IMPACT OF MARKETING INVESTMENTS ON FIRM VALUE By Malika Chaudhuri A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Business Administration - Marketing - Doctor of Philosophy 2015 ABSTRACT IMPACT OF MARKETING INVESTMENTS ON FIRM VALUE By Malika Chaudhuri Firms engage in marketing communication mix such as sales promotions and advertisements primarily to boost sales, attract potential customers while retaining their ex isting customer base. Marketing communications are therefore critical marketing strategies that are intended to promotions in the marketing literature, there still remains limited insight into the differential impacts of various marketing efforts as well as the conditions under which they are most effective. My first essay seeks to address these gaps by demonstrating the effects of two types of sales promotions (ca sh rebates and financing offers) on consumer perceptions of quality and unit sales across both luxury and mass goods. The results reveal that offering financing incentives can effectively drive sales irrespective of product class, but rebates only impact s ales in the mass market. Interestingly, rebates negatively affect perceptions of quality across both product classes, demonstrating a more complex path to sales than traditional promotion models may suggest. My second dissertation essay examines the downs ide of marketing communication mix by U.S. pharmaceutical firms in the post - marketing activities in the post patent period is a signal that is interpreted differentially by the waves of generic manufacturers who are planning to enter the market. Specifically, the first wave n even after the first wave of generics have entered the market may be interpreted by the second wave of generics as signals of unexplored market potential, thereby attracting competition. rogram on risk and - specific risk. Next, we and sales. In particular, adoption of loyalty program by firms with high market share depletes sales. On the other hand, adoption of loyalty programs by small firms boost sales, thereby iv To my family, who offered me unconditional love and support through out this journey. v ACKNOWLEDGMENTS I am deeply indebted to my dissertation committee - Dr. Roger Calantone ( Dissertation Chair ), Dr. Clay Voorhees (Dissertation Co - chair ), Dr. Tomas Hult , and Dr. Gerry McNamara for their insightful comments, helpful guida nce, and unwavering support. In particular, I owe a huge debt of gratitude to Dr. Roger Calantone for his encouragement, and unflinching support through all phases of the doctoral program. I especially appreciate his guidance in helping me to keep things i n perspective, whether in regards to my dissertation, or my career aspirations. I am indebted, to Dr. Clay Voorhees , who, first as a teacher and later as a coauthor, taught me much of what I know about my field of research, teaching me the art of conductin g research and also that of presenting. Without his constant guidance and support this dissertation would not have been possible. I thank Dr. Tomas Hult for his support and encouragement. I thank Dr. G erry McNamara for his invaluable guidance, constructive criticisms and for being instrumental in my development as an academic. I thank the other professors of the Department of Marketing for their encouragement and advice in the process of completing my dissertation. On behalf of my husband , our children and myself, I would like to express our sincere gratitude for Dr. G. Geoffrey Booth and Elizabeth Booth for treating us like family and for their help and support throughout our stay in East Lansing. I would like to express my sincere appreciation to my broth er, Udipta Mukherji, and my parents, Dr. Raka Mukherji and Manas Mukherji for their unwavering faith and confidence in my abilities. Finally, none of this would have been possible without the patience and encouragement of my family back in India, who have been a constant source of love and concern, over all these years. Support, strength and care from my mother - in - law, Mrs. Renu Chaudhuri and father - in - law, Mr. R.K. Chaudhuri helped me vi overcome setbacks and stay focused on my graduate studies. My deepest gr atitude goes to my husband Ranadeb and my daughter Esha and my son, Rohan for their understanding during all the time spent away from them. Without their sacrifices, this dissertation would not have been possible. Special thanks to Ranadeb for his constant guidance, helpful criticisms, unconditional support and for being there when I needed him most. vii TABLE OF CONTENTS LIST OF TABLES ................................ ................................ ................................ ........................ ix LIST OF FIGURES ................................ ................................ ................................ ....................... x Chapter 1 ................................ ................................ ................................ ................................ ........ 1 Impact of Promotion Mix on Firm Performance: The Mediating Role of Perceived Quality ... 1 1. ABSTRACT ................................ ................................ ................................ ............................ 1 1.1 Introduction ................................ ................................ ................................ ................................ .. 3 1.2 Conceptual Backgr ound ................................ ................................ ................................ ............... 7 1.2.1 Defining Sales Promotions ................................ ................................ ...................... 7 1.2.2 Finance Rates ................................ ................................ ................................ ............ 7 1.2.3 Rebates ................................ ................................ ................................ ....................... 9 1.2.4 Promotions and Price Structure ................................ ................................ ............... 9 1.2.5 Perceiv ed Quality and Firm Sales ................................ ................................ .......... 12 1.2.6 Impact of Promotions on Sales ................................ ................................ ............... 13 1.2.7 Moderating Role of Product Cla ss (Luxury vs. Mass) ................................ .......... 14 1.3 Methodology ................................ ................................ ................................ ................................ 17 1.3.1 MIDAS ................................ ................................ ................................ .................... 17 1.3.2 Empirical Model ................................ ................................ ................................ ..... 19 1.4 Data and Measurement Variables ................................ ................................ ............................. 24 1.5 Results ................................ ................................ ................................ ................................ .......... 25 1.5.1 Descriptive Statistics ................................ ................................ .............................. 25 1.5.2 3SLS Estimation ................................ ................................ ................................ ...... 25 1.6 Post - Hoc Analysis ................................ ................................ ................................ ....................... 29 1.6.1 Contingencies in Sales Promotion Strategies ................................ ....................... 29 1.6.2 Unit Root Test ................................ ................................ ................................ ......... 30 1.6.3 Cointegration Test ................................ ................................ ................................ .. 31 1.6.4 Granger Causality Test ................................ ................................ .......................... 31 1.7 Discussion ................................ ................................ ................................ ................................ .... 34 APPENDIX ................................ ................................ ................................ ................................ .. 39 BIBLIOGRAPHY ................................ ................................ ................................ ......................... 50 Chapter 2 ................................ ................................ ................................ ................................ ...... 57 Does Marketing Communication Mix Attract Generic Competition? ................................ ....... 57 2. ABSTRACT ................................ ................................ ................................ .......................... 57 2.1 Introduction ................................ ................................ ................................ ................................ 58 2.2 Hypothesis Development ................................ ................................ ................................ ............ 61 2.2.1 Market ing Communication Mix Strategy ................................ ............................. 61 2.2.2 Detailing ................................ ................................ ................................ ................. 61 2.2.3 Direct - to - Customer Advertising (DTCA) ................................ .............................. 62 2.2.4 Sample Distribution ................................ ................................ ............................... 63 viii 2.2.5 Journal Advertising ................................ ................................ ............................... 64 2.2.6 Detailing and DTCA Marketing Strategies ................................ ........................... 64 2.2.7 Entry Deterrent Strategies by Prescription Manufacturing Firms ...................... 68 2.3 Methodology ................................ ................................ ................................ ................................ 71 2.3.1 Empirical Model ................................ ................................ ................................ ..... 71 2.3.2 Prentice - Wi lliams - Peterson Gap Time Model ................................ ...................... 72 2.4 Data Collection ................................ ................................ ................................ ............................ 75 2.5 Results ................................ ................................ ................................ ................................ .......... 76 2.6 Discussion ................................ ................................ ................................ ................................ .... 83 APPENDIX ................................ ................................ ................................ ................................ .. 87 BIBLIOGRAPHY ................................ ................................ ................................ ......................... 93 Chapter 3 ................................ ................................ ................................ ................................ ...... 99 The effect of Loyalty Program on firm risk and value ................................ ............................... 99 3. ABSTRACT ................................ ................................ ................................ .......................... 99 3.1 Introduction ................................ ................................ ................................ .............................. 100 3.2 Hypothesis Development ................................ ................................ ................................ .......... 104 3.2.1 Does Adoption of loyalt y programs lowers firm risk? ................................ ........ 104 3.2.2 Loyalty Program and Firm Sales ................................ ................................ ........ 108 3.2.3 Incumbent Effect ................................ ................................ ................................ .. 109 3.2.4 Does market share moderate the relationship? ................................ .................. 110 3.3 Methodology ................................ ................................ ................................ .............................. 112 3.3.1 Measures of Idiosyncratic Risk ................................ ................................ ........... 112 3.3.2 Measures of Firm Performance ................................ ................................ .......... 113 3.4 Data and Measurement Variables ................................ ................................ ........................... 115 3.5 Results ................................ ................................ ................................ ................................ ........ 116 3.5.1 Descriptive Statistics ................................ ................................ ............................ 116 3.5.2 Loyalty programs and Firm risk ................................ ................................ .......... 116 ................................ ... 118 3.6 Post - Hoc Analysis: Response Surface Approach ................................ ................................ ... 120 3.7 Discussion ................................ ................................ ................................ ................................ .. 122 APPENDIX ................................ ................................ ................................ ................................ 125 BIBLIOGRAPHY ................................ ................................ ................................ ....................... 134 ix L IST OF TABLES Table 1: Prior Research on Sales Promotions in Marketing 39 Table 2: 44 Table 3: Impact of Sales Promotions on Perceived Quality and Sales in the U.S. Auto Industry............................................................................... .... 45 Table 4: Impact of Auto Promotions on Firm Value and Perceived Quality: Mass Vs. Luxury Product ...................... ............................ ........ .... . 46 Table 5: Granger Causality Test: Impact of Lagged Sales, and Inventory and Supply on Firm's Promotion al Strategies in the U.S. Automobile Industry (2003 - 2012) .......................................................... ........... 4 7 Table 6 : Distribution of Marketing Expenditure on Prescription Drugs Across Therapeutic Classes ( Sept, 2008 - ..................... 87 Table 7 : Difference in Promotional Expenditure (in Millions $) Pre and Post Patent Expiration (Sept, 2008 - Aug, 2014) 88 Table 8 : ... 89 Table 9 : Effectiveness of M arketing Strategies across Prescription Drugs' Product Life Cycle ..... ................................................ .................... 90 Table 10 : Prescription Drug Manufacturing Firm's Entry Deterrent Strategy U sing PWP - Gap Time Model with Stratum - Specific R egression 91 Table 1 1: Sample Breakdown by Industry 125 Table 12: Descriptive Statistic s ................................. ....... ............ 1 26 Table 13: Firm's Exposure to Risk upon L 127 Table 14: Impact of Adoption of Loyalty Program on Firm Sale 128 Table 15: Analysis of Results Based on Response Surface Approach........ 129 x LIST OF FIGURES Figure 1: Conceptual Model: The Differential Effects of Promotions for Mass versus Luxury Brands 49 Figure 2: Conceptual Model .. 1 33 Figure 3: Ridge of Maximum . 1 34 Figure 4: Rotated Surface Plot 1 35 1 Chapter 1 Impact of Promotion Mix on Firm Performance: The Mediating Role of Perceived Quality 1. ABSTRACT The typical firm invests 20% of its promotional budget on sales promotions in an effo rt to drive short - term sales. Given this heavy investment, academic researchers have modeled the effectiveness of such promotions for decades. Despite the rich body of research on sales promotions in the marketing literature, there still remains limited insight into the differential impacts of various sales promotions as well as the conditions under which they are most effective. This research seeks to address these gaps by demonstrating the effects of two types of sales promotions (cash rebates and fina ncing offers) on consumer perceptions of quality and unit sales across both luxury and mass goods. The authors test these effects by leveraging data across 16 major auto manufacturers operating in the U.S. auto industry between 2003 and 2012. The results reveal that financing incentives positively affect perceptions of quality irrespective of the product class. However, cash rebates have positive impact on consumer perceive value in the mass market with no impact in the luxury market. Moreover, financing incentives limit their effectiveness as a driver of sales in the luxury product market whereas rebates impact sales exclusively in the mass market, demonstrating a more complex path to sales than traditional promotion models may suggest. Based on the find ings, marketing managers in mass markets can effectively leverage rebates to increase sales and improve consumer perception. On the contrary, managers in the luxury market should focus promotional investments solely on financing offers 2 because it not only enhances consumer attitude but also offers significant sales benefit. Next, we conduct post - exogenously determined. Granger causality estimates indicate that promotional strate gies internal factors, such as inventory and sales and history of promotional offerings. Keywords : sales promotion, finance rates, rebates, perceived quality, pr oduct class 3 1.1 Introduction Manufacturers often utilize sales promotion tactics to boost sales and influence Neslin, 2002). These promotions are universally focused on driving purchase behav ior, getting customers out of a holding pattern by offering them incentives to take action before the promotional offers expire (Blattberg, Briesch and Fox, 1995; Nijs et al., 2001). Given evidence of their effectiveness, firms continue to invest heavily i n sales promotions to a tune of $70 billion annually, which accounts for nearly 20 percent of total promotional spending (ZenithOptimedia, 2013), and they have remained an area of focus in the marketing literature. For more than 30 years, scholars have in vestigated the effect of promotions on various aspects of firm performance (see Table 1 for a review), which has provided great insights into how and why promotions drive consumer demand. Despite this progress, less is known about how simultaneous promoti ons may impact consumer demand and firm performance, which is becoming an increasingly important issue for industries like automotive, where firms have large promotional budgets and must allocate this budget across mass and luxury brands. While the desired outcome of promotional investments is invariant across industries, the composition of the promotion mix can vary significantly across industries. For example, consumer packaged goods manufacturers invest heavily in trade promotions as well as in rebates and coupons to drive consumer purchase. In automotive industries, promotions often focus on financing offers from manufacturers or cash rebates. Considerable research has been conducted to understand how promotions can be structured to drive conversion ( Silk and Janiszewski, 2008) and leveraged for success in the presence of price competition and price discrimination (Demirag, Keskinocak and Swann, 2011) as two exemplars. T hroughout these investigations, 4 when scholars focus on analyzing the impact of the promotion mix on firm performance, the level of granularity in the data begins to disappear. With few exceptions, researchers often Pauwels et al. 2004; Leeflang and Parreño - Selva, 2012; Gangwar, Kumar and Rao, 2013). Even though this approach provides some evidence of the impact of promotions, in general, but offers little actionable guidance to managers who need to manage a promotional budget across an array of inve stment areas. One notable exception to this tendency to aggregate promotional types into a single bucket is the study by Lu and Moorthy (2007), which demonstrates the differential effectiveness of coupons and rebates as promotional strategies, conditional reservation price and redemption costs. Failing to disaggregate sales promotions into their respective tactical investment areas results in considerable information loss, and provide us with erroneous conclusions. For example, in industries like automotive, the two most common promotions are cash rebates and financing offers. While both result in cost savings for consumers, they could have differential effects on customer attitudes (i.e., perceptions of quality) and sales. As a result, agg regating actionable guidance for managers and, at worst, lead to incorrect conclusions regarding the effectiveness of promotions in driving attitude change an d firm performance. Building on this issue, most prior research conducts analysis at either the industry level or within a focal product category with little variance in the brands under investigation. This narrow lens limits the ability to assess produc t class contingencies that could alter the nature of the relationship between promotions and sales. One notable factor missing in prior research is product class (luxury 5 versus mass). The very nature of promotions and customer mix for these classes of go ods could result in substantial swings in the effectiveness of sales promotions. The current study seeks to provide advance research on the impact of the promotion mix on firm sales by addressing these two shortcomings of the extant literature. Specific ally, our first contribution focuses on disaggregating promotion incentives into tactical level, operationalizations of finance rates and cash rebate offers in the U.S. automotive industry. As a first step, we focus on the single industry to tease out the effects of the two categories of promotional tactics particularly relevant in the consumer durable industry. In doing so, we provide new insight into the effectiveness of two unique promotional investments in driving firm sales. Second, we examine the effects of these promotions across luxury and mass product classes, thus offering an improved understanding of promotion types that can offer the biggest return for the various product classes. Finally, when testing the se effects, we introduce a new method to the marketing literature to handle the frequency mismatch data issue by applying mixed data sampling regression (MIDAS) as pioneered by Ghysels, Santa - Clara, and Valkanov (2004). Our results demonstrate considerabl e value in disaggregating promotional incentives and modeling their impact separately for luxury and mass goods. For example, our findings reveal of produc t class. However, they are effective demand boosters exclusively in the luxury product market. On the contrary, cash rebates trigger sales increases exclusively for mass brands. cy filing and firm characteristics, cash rebate offerings actually improve consumer perceive value in the mass product market. Empirical estimates suggest that managers may employ perceived quality as a 6 strategic asset that can effectively boost sales, irr espective of product class. As a result, our findings identify critical contingencies regarding the promotions - performance relationship and in doing so has considerable implications for both researchers and practitioners. In the following sections, we int roduce the conceptual basis for our model, describe the MIDAS method, and discuss the results. 7 1.2 Conceptual Background 1.2.1 Defining Sales Promotions promotional tact ics operationalize short - term techniques to generate almost immediate impact on Belch et al. 2008). In the current study, we focus on finance rates and rebates - the two critical consumer oriented pr omotional strategies frequently employed in the high - value consumer durable goods industries ( Attanasio, Koujianou, and Kyriazidou 2008) . This study focuses on the U.S. automobile industry, a particularly appropriate product category where both types of pr omotional strategies mentioned above are critical demand boosters. In particular, a utomobiles are typical examples of consumer durables where median product price exceeds median household income (Ohta and Griliches, 1986). Consumers may lack the liquid ass ets necessary to make down payments towards the purchase of these consumer durables. They may instead seek loans from banks or other financial institutions to finance their product purchase ( Stango and Zinman 2011) . Additionally, rebates discount product p rice. Thus, consumer - oriented sales promotions, such as finance rate deals and rebates, 1.2.2 Finance Rates These are promotional strategies especially utilized by firms to stimulate purchase of big - ticket items (i.e., automobiles etc.). An auto loan is a contractual agreement between the lender 8 and the borrower where the borrower pledges to repay the l oan at a predetermined rate over a fixed time period. Additionally, it is a secured loan where the financed vehicle is used as the collateral (Forbes 2000). The annual percentage rate (APR), also referred to as finance rate, is a function of the prevailing market interest rates and business environmental conditions Chrysler/GM/Ford cars at an interest rate that is significantly less than the ongoing market interest rate (e. g., 1.9% annual rate) (Varadarajan and Clark 1994). In efforts to boost sales, auto manufacturers typically offer incentives to customers through interest rate reductions from their captive finance subsidiaries (Barron, Chong, and Staten 2008). These fina nce rate deals significantly lower the interest rates on the loans relative to prevailing market interest rate, thereby drastically reducing the monthly payments customers are required to make towards their loan. Such promotional incentives either make the car more income or allow the customer to purchase higher quality product by lowering the monthly loan payments required. Interestingly, auto loans carrying zero p ercent interest rate is not uncommon in the U.S. auto industry. Thus, finance rate deals do not have any explicit discount on the product price. However, they are implicit promotional strategies that decrease the present value f payments made towards the loan repayment. Additionally, Moreover, since the manufacturing firm that sells the product and the financing firm that extends the loan ar e usually independent entities, consumers tend not to associate incidence of attractive finance rates with erosion of quality. 9 1.2.3 Rebates Rebates are monetary inducements in the form of price subsidies offered by manufacturers to potential consumers to stimulate purchase (Blattberg and Neslin 1990; Neslin 2002). Traditionally, this category of inducement involves reducing the sales price of the product equal to the dollar amount of the rebate (Varadarajan and Clark 1994). These are explicit promotiona l tactics such that the price discount can be redeemed after purchase of the product. Interestingly, during purchase of big ticket items, customers are often given the option to apply the rebate towards their down payment or receive cash (Ault et al., 2000 ). Thompson and Noordeweir (1992) analyzes declining impact of continuous incidence of rebates for three successive years in the U.S. automobile industry. Results indicate that these and Moorthy (2007) investigate whether coupons and rebates, two critical promotional incentives, have identical redemption costs since they inherently differ in opportun ity cost of time. Specifically, with coupons, the uncertainty about redemption costs is resolved even before product purchase. However, with rebates, the uncertainty is resolved post product - purchase. Findings also suggest alleviates rebate attractiveness (Lu and Moorthy, 2007). 1.2.4 Promotions and Price Structure these perception s are often driven by brand reputation, price, and advertising efforts (Zeithaml, 1988; Dodds, Monroe and Grewal , 1991; Mitra and Golder, 2006). Quality perceptions serve as 10 - 90) and can serve as a primary driver of purchase intentions (Zeithaml, Berry, and Parasuraman, 1996) and brand preference (Yoo, Donthu, and Lee, 2000). Given the importance of perceived quality, it is widely regarded as a key strategic asset despite its i ntangible nature ( Aaker and Jacobson, 1994) . The l iterature indicates that managers need to complement delivery of quality product with high consumer perceptions regarding ncial value (Aaker, 1991; Aaker and Jacobson, 1994 ). Thus, firms often leverage extrinsic cues to communicate with their customer base and to build positive quality perceptions as consumers interpret these cues when evaluating competing product options and forming quality evaluations (Olson, 1978). This is most commonly done directly through product pricing to the extent that a higher price reflects higher quality (Zeithaml, 1988) or through advertising where higher levels of advertising can result in high er perceptions of product quality (Milgrom and Robers, 1995). Given the rich literature base on these effects, we simply control for these quality drivers in the current study and focus on the potential role of promotions as signals of quality. In a sim ilar vein to price, promotions provide extrinsic cues to customers about the quality of the product. So while promotions are traditionally targeted at changing short - term behavior, they can also be manifested in quality evaluations. In the context of our current research, we consider two types of promotions. At the basic level, cash rebates function as price reduction offers to consumers, thus eroding quality evaluations under the same mechanism as price. However, when evaluating rebates, consumers may also engage in another layer of processing in which they potentially perceive rebate offers as a signal of desperation by manufacturers, which can result in a further reduction in qualit y perceptions (Darke and Chung, 11 2005). As a result, we propose that hi gher rebates erode perceptions of quality. Stated more formally (figure 1): H 1 : The mechanism underlying the effects of cash rebates is relatively straightforward, but the m anner in which financing offers can impact quality evaluations is not explicitly addressed in the literature because a strategies with no direct discount on product price. Instead, thes e promotional tactics decrease - consumers are offered financing incentives, bu t the overall cost to the consumer who finances a vehicle can be greatly reduced. While it has been suggested that consumers account for financing rates in determining the overall cost of a vehicle (Gale, 1994) and class economic investigations have demon strated a relationship between interest rates and demand for durables (e.g., Hamburger, 1967), little empirical evidence has been provided to model the impact of finance rates on perceptions of product quality. However, in line with the same logic on the well documented price - quality relationship, we expect that a higher finance rate, that increases the cost of a product, will result in increased perceptions of quality. Thus, we propose that: H 2 : ed quality to the extent that higher finance rates result in higher perceptions of quality. 12 1.2.5 Perceived Quality and Firm Sales Perceived quality, in association with brand awareness and brand associations, strengthens brand loyalty by increasing cus tomer satisfaction and by providing consumers with reasons to buy the product ( Aaker, 1992) . This gets reflected through increased sales and enhanced firm value in the long run. Aaker and Jacobson (1994) examine the financial information contained in perce ived quality measures and analyze the relationship between firm y and stock returns, thereby impacting firm performance. In particular, improved perceived quality brand loyalty, which translates into higher consumer switching costs. The firm may effectively exploit such high switching costs to in crease its cash flow and revenue generation (Srinivasan et al., 2009). Tellis and Johnson (2007) investigate whether publication of product quality information in The Wall Street Journal generates abnormal return in stock prices. Findings indicate that a f quality generates investor enthusiasm as reflected by abnormal returns in stock prices. Additionally, such signals improve consumer offerings and enhance their willingness to buy ( Oh, 1999) . We H 3 : Perceived quality has positive impact on sales. 13 1.2.6 Impact of Promotions on Sales As we proposed in prior hypotheses, the indirect effect of financing incentives is proposed to by positive; however, the indirect effect of rebates on sales via quality is proposed to be negative. As a result, for cash rebates to offer positive return for firms in either the short or long run, they must have a significant direct effect on the quantit y purchased. In line with this necessity, the marketing literature does provide solid evidence of the impact of price reductions like those offered by rebates on short - term sales spikes (for a review see Blattberg, Briesch, and Fox, 1995). More recent i nvestigations have provided an even more nuanced view of this adoption of promotional strategies consumers significantly increase their purchase quantities in an effort to stockpile. Similarly, Joshi and Hanssens (2010) suggest that rebates reduce sales price and stimulate product demand. As a result, large rebates should trigger a stronger change in demand. Based on classic investigations into the relationship of price promotions and sales and recent empirical and analytical evidence, we suggest: H 4 : Rebates have a positive impact on sales to the extent that higher magnitude cash rebates result in higher sales. In addition to rebates, firms can lower the cos t of products through effective financing 14 loan payment they are required to make, thereby rendering the product more affordable (Attanasio, Goldberg, and Kyriazi dou, 2008). Thus, we propose: H 5 : Finance rates have a negative impact on sales to the extent that higher finance rates result in lower sales. 1.2.7 Moderating Role of Product Class (Luxury vs. Mass) Luxury brands are designed to be deliberately cons picuous and flamboyant, to emit an aura of exclusivity and quality ( Atwal and Williams, 2009; Brown, Kozinets and Sherry, 2003) that distinguishes these brands from mass - market firms by signaling their commitment towards Beverland, 2005 ). Marketing of luxury products has become increasingly multifaceted, being concomitant not only with cuing an aura of to th extant literature indicates that the inescapable desire for social prestige influences consumers to pay a price premium for products that confer status (Shapiro, 1983). Goldsmith, Flynn, and Kim (2010) posit of involvement and brand loyalty make the cons umer less price - sensitive. Consumers associate consumption of luxury goods as signal of status and are willing to pay the price premium (Han, Nunes, and Drèze, 2010). Thus, effective marketing strategies for luxury products are those that convey high quali ty and are less explicit about product pricing structure. The literature on marketing luxury products indicates that luxury is a social marker and 15 a success 2004; Kapferer and Bastien, 2009). Thus, these firms need to adopt strategies that endow the with a halo of superiority with respect to its client. Furthe rmore, these marketing strategies emit signals that emphasize product excellence and perfection, while maintaining an aura of exclusivity. Even though price communicates quality, marketing strategies for luxury brands typically withhold price information f rom being publicly advertised (Kapferer and Bastien, 2009). The role of advertising in the luxury sector is to recreate the dream of exclusivity and not to improve sales growth (Kapferer and Bastien, 2009). Thus, firms whose product offerings target the lu xury market traditionally avoid extension of explicit sales promotions such as rebates and coupons (Kapferer, 2012a; 2012b). Interestingly, signals emitted by the luxury marketing mix are often diametrically different from those of classical marketing emp loyed while promoting mass products (Kapferer and Bastien, 2009). In particular, in the mass market, one observes promotional strategies that offer explicit price discounts and provide consumers with monetary relief that effectively enhance product demand. Additionally, advertisements of products are geared toward accelerating sales growth. Often times, advertisements even provide price information to customers. Thus, considering the characteristics of the target customers in the luxury versus the mass mark et, we posit that finance rates, given their implicit characteristics, have a higher positive impact on perceived quality and sales in the luxury market relative to the mass market. We also theorize that rebates, given their explicit characteristics, erode perceived quality significantly more in the luxury market than in the mass market. Finally, we theorize that rebates are a relatively more effective strategy in boosting sales in the mass market than in the luxury market. We hypothesize: 16 H 6a : The negativ e effect of rebates on perceived quality is stronger for luxury vis - à - vis mass products. H 6b : The positive effect of finance rates on perceived quality is stronger for luxury vis - à - vis mass products. H 7a : The positive effect of rebates on sales is wea ker for luxury vis - à - vis mass products. H 7b : The negative effect of finance rates on sales is stronger for luxury vis - à - vis mass products. 17 1.3 Methodology 1.3.1 MIDAS Our analysis uses data with different sampling frequency. Specifically, informatio n on monthly. Additionally, information on firm performance, dealership and perceived quality data is available annually. Instances when researchers deal with mixed frequency data, they typically have two alternatives: either to align variables downward by aggregating high frequency data to a lower frequency down or to align variables upward by interpolating lower frequency data to high frequency. Both methods suffer from limitations. On one hand, downward adjustments abandon valuable information in the high frequency data, which consequently reduces its estimation and forecast efficiency (Silvestrini and Veredas, 2008). The other alternative which involves upward alig nment based on random mathematical procedures may also be problematic. We address the frequency mismatch data issue by applying mixed data sampling regression (MIDAS) (Ghysels, Santa - Clara, and Valkanov, 2004). MIDAS regression typically projects In particular, MIDAS helps to project the dependent variable onto a history of lagged observations of the independent variables. Suppose the sampling frequency o f variable is between and is unity (say, yearly), whereas that of another variable, say = 12), then MIDAS aids in un and 18 onto a history of lagged observations of script on denotes the higher sampling frequency and its exact timing lag is expressed as a fraction of the unit interval between and - Clara, and Valkanov, 2004). The MIDAS model may be illustrated as: (1) for , where is the regressand, is the regressor, m denotes the frequency of occurren ce of , , is a lag operator and is the disturbance term. The parameter indicates the aggregate impact of lagged on and is the intercept. Following Ghysels, Sinko and Valkanov (2007), we estimate function corresponding to is a vector of parameter with a small dimension. In a MIDAS framework, the coefficients are characterized by . While there are seve ral alternative parametirizations of , in this study we utilize the "Exponential Almon Lag" specification of (Ghysels, Sinko and Valkanov, 2007). (2) 19 1.3.2 Empirical Model We model the relation between promotional strategies, perceived quality, and firm sales as a two - equation simultaneous model (Zellner and Theil, 1962). We use a three - stage least squar e (3SLS) method to estimate the model a method traditionally employed to estimate - While considering cross - equation correlation and potential endogeneity issues, the 3SLS method of estimation yields relatively efficient estimates for simultaneous - equation systems as compared to that of two - stage least squares (2SLS) and ordinary least squares (OLS) (Tamirisa and Igan, 2008). Additionally, the 3SLS method of estimation also does not impose restrictions on the autocovariance matrix of errors. Hence, 3SLS is the preferred estimation method in the current study. The variables used in the estimation are as follows (see Appendix 1): Perceived Quality equation: perceived quality of brand of firm in period ( ) is the dependent variable with offerings of finance rates ( ) and rebate ratio ( ) as the key explanatory variables. Additionally, we include bankruptcy ( oduct offerings based on whether the firm has ever declared bankruptcy in the past. We also include product - class ( ) of brand of firm in period as additional exogenous variables that may impact network ( ) and advertising expenditure ( specific information. 20 Sales equat ion: Logarithmic value of firm total sales of brand in period is the dependent variable ( ) (i.e., equation 4) with perceived quality ( extension of finance rates ( ) and rebate ratio ( ) as key explanatory variables. We also include log of total assets ( ) as a proxy for firm size, inventory ( ) and adjust ed capital expenditure ( - - specific factors that help the firm to adjust its supply function. Additionally, we include product class ( ) of brand of f irm in period as additional exogenous variables that may impact ) and advertising expenditure ( ) as firm level control variables. The unit of analysis is brand. We collected data for brand for the firm at time from 2003 to 2012. However, since some of the brands were discontinued within this time p eriod, we have an unbalanced panel data. 1. Perceived Quality Equation (3) 2. Firm Performance equations: (4) We now provide a definition of the var iables used in the analysis: 21 Perceived Quality ( product quality for brand in period directly from consumer responses regarding brand equity, consumer connection, and brand momentum. Sales ( ) is the total number of brand automobiles sold by firm in time . Finance Rate ( ) is the interest rate extended by banks and financial institutions toward their most creditworthy customers. It is the difference in the interest rate the financing division of manufacturing f irm offers to its customers upon purchase of the brand vehicle and accepting the loan from the firm to finance his/her product purchase in time and the prevailing industry prime interest rate. Thus, the differenc e indicates additional incentives being offered by the financing companies to ensure that customers apply and secure loans from them. Rebate Ratio ( ) is the ratio of dollar value of rebate offered by manufacturi ng firm for brand in time to its customers upon purchase of the automobile to product price ( ). Since luxury cars are prices much higher than mass or economy cars, a $500 rebate offered towards a luxury car has very different implications compared to that towards a mass car. Thus, rebate ratio is a critical factor driving both firm sales and consumer perceived quality. Note that price ( ) is the dollar value of the bran d automobile manufactured by firm in time . Luxury ( ) is a dummy variable that takes the value of 1 if the brand of firm is a luxury product, ot herwise it equals 0. 22 Bankruptcy ( ) is a time varying indicator variable that assumes the value unity when the firm declared bankruptcy and assumes zero when the firm is not under bankruptcy protection. Dealer network ( ) is measured by the number of auto dealers operating in the U.S. for firm in time . They are an important channel of communication between the manufacturer and end customers: the higher the number of aut o dealers, the more intense the supply chain network. Advertising Expenditure ( its total assets. Log of Firm Assets ( ) is the logarithmic value of fir is used in the analysis to control for firm size. Inventory ( ) is the total number of brand vehicles the auto manufacturer has in its reserve at time t. Adjusted Capit al Expenditure ( ) is the ratio of to its total assets at time t . It represents expenditures incurred by firms to upgrade existing physical assets or to acquire assets with the int ention of creating financial benefit for the firm beyond the taxable year. n the sales equation (i.e., equation 4) but not in the perceived quality equation (i.e., equation 3). This makes 23 (Verhoef, Neslin and Vroomen, 2007). The error terms and are potentially correlated with each other for a given firm and across firms. 24 1.4 Data and Measurement Variables In this study, we considered 16 major auto manufacturers that were operating in the U.S . auto industry between 2003 and 2012 and offering either luxury or economy or both brands of products to the customers. Perceived quality information by brand was obtained from Harris Interactive . We obtained weekly brand specific promotional information (i.e. cash rebate and finance rate), monthly sales transaction, inventory and supply information by brand and firm specific dealership network information from Automotive News. Brand specific price information was sourced from Kelly Blue Book and warranty information from Gillis (2007). We obtained advertisement and capital expenditure data from COMPUSTAT and product age 25 1.5 Results 1.5.1 Descriptive Statistics Table 2 p rovides the correlation coefficient estimates of the variables used in the analysis. The estimates indicate that perceived quality has negative correlation with finance rate and rebate ratio, significant at 1 percent level of significance. This indicates t hat lower is the finance quality. On the contrary, perceived quality has positive correlation with price, significant at 1 percent level of significance. This implies that higher price is associated with higher perceived quality. Rest of the estimates may be interpreted accordingly. 1.5.2 3SLS Estimation Table 3 provides the 3SLS estimates of the model. In the first column, the dependent variable is pe rceived quality and the independent variables are sales promotions (i.e., finance rate and rebate ratio), vehicle characteristics (i.e., price, luxury). We include dealer network and fic information. We also include warranty information to control for brand specific information (Erdem and Swait 1998). quality increases by .0646 units (p<.001). Fin dings suggest that consumers perceive incidence of rate being offered by the firm, higher is the value of perceived quality. This confirms hypothesis H2 t hat incidence of finance rates have positive and significant impact on perceived quality. Estimates also indicate that for every 1 unit increase in the rebate ratio increases consumer 26 perceived value by .0200 units (p<.001). Thus findings contradict hypoth esis H1 that cash Findings also suggest that when rebates are offered to promote luxury products, it erodes consumer perceived value by .0092 units (p<.1). This confirms hypothesis H6a that inc idence of rebates erodes perceived quality significantly more of luxury products relative to that of mass products. However, no such differential impact of incidence of finance rates on perceived quality has been observed across luxury and mass product mar kets. Thus, hypothesis H6b is not supported. In the second column, we have logarithmic value of sales of brand for the firm as the dependent variable and perceived quality, sales promotions (i.e., finance rate an d rebate ratio), and product class as the independent variables. Consistent with the above analysis, we access to firm related information. We also include firm size measured by logarithmic value of firm assets, inventory and adjusted capital expenditure as firm level controls. Results indicate that 1 unit improvement in perceived quality increases log of sales by .5964 units (p<.001). Consistent with the existing li terature, estimates confirm hypothesis H3 that perceived quality has positive impact on firm sales. Results also suggest that one unit increase in rebate ratio boosts log of sales by .0232 units (p<.05). This confirms hypothesis H4 that promotional strateg ies such as cash rebates tend to have positive and significant impact on firm sales. Interestingly, we do not observe any significant impact of finance rates on sales. Thus, findings do not validate hypothesis H5 that incidence of finance rates boosts firm sales. Findings also indicate that for every one unit increase in dealership network, perceived quality and sales improve by .0007 units (p<.001) and .0005 units (p<.001) respectively. 27 Dealership network provides effective communication channel between th e manufacturer and the consumers and are able to provide authentic information regarding product quality along with information on lucrative promotional deals to their customers. Additionally, higher dealership network corresponds to higher competition amo ngst the dealers. In such a competitive environment, as survival strategies, dealers would strive to provide better service and offer better deals to customers. This eventually improves perceived quality and enhances sales. To farther unravel the differen tial impact of promotional strategies on perceived quality and firm sales across product class, we estimate the model for two subsectors (i.e., luxury and mass automobiles) (Table 4). First two columns provide us with estimates for the mass product whereas third and fourth column provide us with estimates of the luxury product. Additionally, we have perceived quality (log of sales) as the dependent variable in the first and third (second and forth) columns. Consistent with the previous sections, we have pro motional strategies (i.e., finance rates and rebate ratio) as the independent variables. Furthermore, we control for Estimates suggest that in the mass market, one unit increase in finance rates im proves consumer perceived value by .0663 units (p<.001). Similarly, in the luxury market, one unit improvement in finance rates drives up perceived quality by .028 units (p<.05). Interestingly, the differential impact of finance rate across the two markets is not statistically significant. Thus, results do not support hypothesis H6b that the positive effect of finance rates on perceived quality is stronger for luxury vis - à - vis mass products. Post - hoc, we conducted one - way analysis of variance to test for d ifference in means in perceived quality across product class. Findings suggest that mean (standard deviation) of perceived quality of luxury products is higher (lower) 28 than that of mass products. Hence, consumers targeting the luxury market are less sensit ive to perceived quality. Empirical findings suggest that in the mass market, one unit increase in rebate ratio increased perceived quality by .0203 units (p<.001). Contrary to the findings in the literature, estimates indicate that after controlling for firm's filing for bankruptcy protection and other firm characteristics, rebates have a positive impact on perceived quality. However, rebate ratios may not have any impact on perceived quality in the luxury market. Empirical estimates indicate that one un it increase in perceived quality in the mass (luxury) market enhances log of sales by .5972 units (.8145 units) both at p<.001. Thus, findings support hypothesis H3 that perceived quality boosts sales, irrespective of the product market characteristics. Re sults also indicate that one unit drop in finance rates in luxury market improves sales by .0761 units (p<.05), with no significant impact in the mass market. Findings support hypothesis H7b that the negative effect of finance rates on sales is strong in t he luxury market compared to the mass market. On the other hand, estimates indicate that one unit increase in rebate ratio in the mass market increases log of sales by .0279 unit (p<.05) with no significant impact on sales in the luxury market. This confir ms hypothesis H7a that cash rebates may be an important driver of sales in the mass market relative to that in the luxury market. Thus, findings suggest that attractive finance rates are effective promotional strategies to boost sales in the luxury market whereas rebates drive sales e xclusively in the mass market. 29 1.6 Post - Hoc Analysis 1.6.1 Contingencie s in Sales Promotion Strategies Having established the relation between promotional strategies, perceived quality and firm performance, post - hoc we a by lagged values of perceived quality, along with firm characteristics such as sales, inventory and supply functions. Contingency theory is a strand of behavioral theory that proposes that str ategic decisions adopted by managers are contingent on its internal needs and the environmental circumstances that the firms needs to adjust to ( Morgan 1986 ). Even though strategy is not a universal concept, its structure must be fitted into its context to enhance organizational performance (Schoonhoven 1981; Mohr 1982). This fit is a vital concept, as it and develop critical resources and capabilities, which may endow it with competitive edge (Drazin and Van de Ven, 1985). managers to comprehend whether consumers perceive their product offerings of high quality relative to th eir competitors. We posit that while crafting market - mix plans, managers internalize perceived quality information. Specifically, if perceived qualities of product offerings are high, firms would continue to adopt marketing strategies that would convey sim ilar and consistent cue to their consumer base. On the contrary, if firms have a history of low perceived quality, regarding product quality towards better. Thus, we theor strategies is contingent on lagged value of perceived quality of its product offerings. 30 Extant literature indicates that two common features of consumer durable goods markets are inventory pile up relative to sales a nd declining product prices over its life cycle (Copeland, Dunn, and Hall (2005). Thus, managers are constantly required to synchronize their promotional strategies, inventory and supply management to maximize firm valuation and sale. Even though depleting inventory restricts inventory carrying costs, running too low on inventory may prove to be detrimental for firm reputation (Hendricks and Singhal 2003). Additionally, supply uncertainties due to external factors often require managers to adjust promotiona l strategies to adjust with projected product demand. For example, due to natural calamities, manufacturing firms may experience unplanned manufacturing parts or product supply disruptions (Chopra and erings are exposed to seasonal variations. Consistent with contingency theory, we posit that promotional incentives are often endogenously determined, adopted by managers to improve sales, deplete inventory and adjust variations in supply. Specifically, th ey provide the firm with a strategic fit between its internal requirements and the environmental conditions it is exposed to, thereby improving its valuation. 1.6.2 Unit Root Test First, we conduct unit - root tests to determine whether the variables use d in the study (i.e., sales, promotions, inventory, etc. ) are stationary or evolving over time. A unit root test helps us to determine whether the variables should enter the Granger Causality Model in level or difference form. We applied the augmented Dick ey Fuller (ADF) test to examine the stationarity of each individual series. Following is the general form of the test equation: (5) 31 where is the variable of interest; is a trend variable; where is the lag operator; and is a white noise term. Additionally, is the intercept term that accounts for the fact that at (i.e., ) need not necessarily be equal to zero (Nijs et al., 2001). The null hypothesis is and is said to possess the unit root property if one fail s to reject (Dickey and Fuller, 1979). 1.6.3 Cointegration Test Evolving variables are said to be cointegrated when a linear combination of the variables exists and results in stable residuals (Dekimpe and Hanssens, 2003). Various factors may drive such long - run equilibria. For example, a boost in sales may translate into higher marketing addition, competitive decision rules may restrict s kewed distribution of the marketing budget. This may ensure that budget allocation across marketing mix variables does not deviate substantially. test for possible pairwise cointegra tion of the five time series (i.e., finance rates, rebates, inventory, supply and sales functions) (Johansen, 1995). 1.6.4 Grange r Causality Test We perform the Granger causality test to analyze whether an incidence of promotions is triggered by the fir 32 perceived quality of its product offerings. This is a well - established test for bivariate causality, which involves estimating a linear reduced - form vector autoregression (VAR) (Gran ger, 1988): (6) where is the coefficient on the lagged values, is the coefficient on the lagged values, and and are assumed to be independent and identically distributed (i.e., ). To examine Granger causality between and , the fol lowing null hypotheses were tested: and . If neither set of null hypotheses can be rejected, then and are an independent series. If both are rejected, then there is " feedback" between and . If the hypothesis is rejected but the other is not, there is unidirectional causality running from and . Conversely, if hypothesis is rejected but the other is not, then the reverse is true (Hiemstra and Jones, 1994; Granger, Huang and Yang, 2000) . Further, to avoid model misspecification, appropriate lag structure must be identified based on statisti cal criterion ( Thornton and Batten, 1985) . Results from the Dickey Fuller test reveal that cash rebates, finance rate, sales, inventory, and supply series are stationary. Next, we conduct a Johansen cointegration trace test to examine if pairwise series sh are a common stochastic drift. Cointegration test results reveal that each of the pairwise series has two cointegrated processes . Thus, we introduced the error correction term while conducting the Granger Causality test ( Granger, Huang and Yang, 2000 ). 33 Es timates in the Granger Causality test reveal that for every 1 unit increase in sales in period , finance rate offered by the manufacturing firm against brand in period increases by .00001 percentage points, significant at a 10 percent level of significance. Additionally, for every 1unit drop in inventory in period , finance rate offered by the manufacturing firm against brand in period increases by .00001 percentage points, significant at a 5 percent level of significance. Improvement in sales and depletion in inventory gement. Thus, managers respond to such positive signals by increasing finance rates, thereby adjusting the marketing - mix strategies by reducing their promotion expenditure against finance rate deals. Estimates also reveal that for every 1 unit drop in sal significance. Thus, managers often extend cash rebates incentives to consumers to make product purchases, thereby improvi ng sales and depleting excess inventory. Interestingly, we observe a two period lag between the dip in sales and an increase in inventory and the execution of the ion to lagged sales, inventory, and supply functions. However, findings fail to confirm our hypothesis quality regarding its product offerings. This may be due to the intangibility characteristic of 34 Estimates also indicate that incidence of finance rate in per iod exhibits a negative correlation with finance rates and cash rebates offered in periods and , significant at a 10 percent or higher level of significance. This suggests that firms are les s likely to offer attractive finance rates as promotional strategies in consecutive periods. Additionally, if the firm offers attractive finance rates in the current period, it is less likely that the firm would have offered cash rebates in the last two pe riods. This may be indicative of the fact that managers are less inclined to opt for alternative promotional tactics in consecutive periods. Results also indicate that an incidence of cash rebates in period exhibits a negative correla or , significant at a 5 percent level of significance or higher. This may indicate that firms are less likely to extend cash rebates in consecutive period. a cash rebate is not contingent on its decision to extend an attractive finance rate in previous periods. 1.7 Discussion Key Findings : Threats of market share erosion, mass customizat ion, and product commoditization are some critical factors that motivate firms to differentiate their product offerings through various promotional incentives (Neslin, 2002, Busse, Simester and Zettelmeyer, 2010 ). U.S. auto manufacturers are no exceptions. They frequently extend sales promotions as rebates, attractive finance rates, or a combination of the two to enhance product Thompson and Noordewier, 1992). Historically, while the big - 3 automaker s typically offer a combination of cash rebate and 35 attractive finance rate deals to attract consumers, Honda and Volkswagen exclusively offer attractive financing rates cut to attract potential consumers (Automotive News). Thus, the intriguing question is why firms, even in the same industry, differ in their extension of promotional incentives. Do firms use promotional strategies to signal product quality? In the current study, we investigate if incidence of sales promotions has a direct and an indirect imp act on sales. Results confirm that in addition to the direct relationship between sales ristics, incidence of both categories of promotional tactics (i.e., finance rates and cash rebates) improves perceived quality. Moreover, perceived quality has a positive and significant impact on firm sales irrespective of the product market characteristi cs. Next, we examine whether the impact of promotion mix on sales is moderated by product class. Results indicate that an extension of attractive finance rates boosts sales exclusively in the luxury market where as effectiveness of rebates as a driver of sales is limited in the mass market. Estimates indicate that incidence of finance rates improves consumer perceived value in both categories of product classes. Interestingly, positive relation between incidence of cash rebates and perceived quality is obs erved exclusively in the mass market. regarding product quality and sales. This may indicate that a dense dealer network increases competition among the dealers and in duces them to provide better value propositions to their influence product sales. 36 Ceteris paribus, managers may strive to adjust their promotion - mix tactics to maximize firm sales while dynamically optimizing their inventory holdings as well as adjusting for variations in product demand and uncertainty in supply. Thus, to comple te the analysis, our post - hoc analysis investigates firm specific criterion, such as lagged sales, inventory, and supply functions as well as . Grange r causality estimates indicate that promotional tactics adopted by firms are not ad - hoc decisions imposed by the top management team. Rather, inventory and sales. I nterestingly, estimates suggest that managers may not internalize - mix decisions. This to quantify the concept. sales and inventory information is not instantaneous; rather one observes the adjustment process with few lag periods. In particular, with a dip in sales or inventory pile up in period , managers typically extend rebates to boost sales and deplete inventory in period . On the contrary, a boost in sales and the depletion of inventory in period encourages managers to increase their finance rate offerings in period , thereby reducing budget allocation targeted towards this specific promotion strategy. Managerial Implications : Auto firms tend to invest heavily to advertise and promote their product offerings. For example, according to Kantar Media's 2011 index of top advertisers, General Motors ranks second in marketing budget, with an overall budget approximately equal to $3.1 billion, 2.081 percent of its revenu e. 37 adopt optimal marketing - mix strategies that would not only cover the marketing cost but also ensure increasing returns. Estimates indicate that effectiveness of finance rates as a demand booster is rest ricted in the luxury product market whereas that of cash rebates is limited to the mass market. - mix decisions are contingent on the product class of their offerings. In particular, incidence of rebates in the luxury market may not exchequer without any significant improvement in sales but also may have no implication s on consumer perceived value. Limitations and Future Research : In this section, we address some of the limitations of the current study and list pote ntial future research. Even though we have access to promotional tactics, sales, and inventory information at the brand level, information on firm performance (i.e., revenue, net income, earnings per share etc.) is only available at the firm level. Thus, d ue to data limitations, analysis of the impact of promotions on firm performance in the luxury and mass product classes is restricted to variation in firm sales. The current study assumes that at any given time period, firms offer identical finance rates to their entire consumer base. However, in reality, finance lenders segment their finance rate credit information and information regarding the credit segmentation p rocess employed by the lenders, we were unable to incorporate segmentation analysis in our study. It may be interesting to analyze how incidence of varying finance rates across segments of consumers with differing credit scores affects perceived quality an d sales. Furthermore, it may be interesting to investigate the moderating effect of market dynamism on the relationship between promotional tactics and firm performance. Specifically, in a highly competitive industry, firms are expected to aggressively o ffer price promotions to attract 38 consumers and to maintain their market share in the industry. On the other hand, with less competition and few key players in the market, firms may be less threatened by erosion of market share. Thus, one may observe manage rs allocating significantly less budget resources toward promotions in low to mildly competitive markets. 39 A PPENDIX 40 Table 1 : Prior Research on Sales Promotions in Marketing Article Assess Various Promotions Independently? Consider Contingency Effects? Consider Attitudinal Outcome? Consider Behavioral Outcome? Address Data Frequency Mismatch Issues? Current Paper Rebates and finance rate offers Moderating Effects of Product Class Perceived Quality Unit Sales Mixed Data Sampling Regression (MIDAS) Gangwar, Kumar and Rao (2013) Shallow versus deep price promotions with varying depth and frequency Consumer stockpiling Leeflang and Parreño - Selva (2012) Cross category demand Martín - Herrán, Sigué and Zac cour (2010) Rebates, trade deals Moderating effect of consumer sensitivity to promotions Unit sales Busse, Simester and Zettelmeyer (2010) Consumers' perception of price changes Purchase acceleration 41 Silva - Ris so and Ionova (2008) Cash discounts, finance rates, and lease payment discounts sensitivity towards pricing instruments, transaction type and brand choice Silk and Janiszewski (2008) Mail - in rebates Consumers' price sensitivity pattern Attanasio, Koujianou, and Kyriazidou (2008) Finance rate High versus low income households sensitivity to maturity of loans and interest rate changes Barron, Chong and Staten (2008) Finance rate Ba nks versus captive financing institutions likelihood of loan repayment Manning and Sprott (2007) Magnitude of quantity specified in the promotion offer accessing anchor - consistent knowledge Lu and Moorthy (2007) C oupons, rebates Redemption costs riskaversity and redemption periods of rebates 42 Chen, Moorthy and Zhang (2005) Coupons, rebates willingness to pay Pauwels, Silva - Risso, Srinivasan and Hanssens (2004) Stock market performance, top and bottom line financial metrics Pauwels, Hanssens and Siddarth (2002) Category - incidence, brand - choice and purchase - quantity Nijs, Dekimpe, Steenkamp and Hanssens (2001) Marketing intensity and competition Category demand Zhang, Krishna and Dhar (2000) Front - loaded versus rear - loaded coupons Variety - seeking, inertia Sales, profit Yoo, Donthu and Lee (2000) Frequency of price promotions Brand equity Chandon, Wansi nk, and Laurent (2000) Band equity Hedonic benefits Utilitarian benefits Jedidi, Mela, and Gupta (1999) Brand equity Brand choice, sales Krishna and Zhang (1999) Short versus long - duration coupons share, coupon durat ion Coupon profitability and redemption 43 Dekimpe, Hanssens and Silva - Risso (1998) National and private - label brands Category and brand sales Dhar and Raju (1998) Cross - ruff coupons Demand complements or substitutes Target and carrier brand sales and profit Mela, Gupta, and Lehmann (1997) Consumers' brand choice behavior Narasimhan, Neslin, and Sen (1996) Featured price cuts, displayed price cuts, and pure price cuts Number of brands offered, category penetration, interpurchase times, and consumer propensity to stockpile Impulse buying, private label market share Dhar and Hoch (1996) In - store coupons, off - the - shelf price discounts Unit category sales, retailer profit Greenleaf (1995) Price promotions, trade deals to reference price effects Profit Raju, Dhar and Morrison (1994) Package coupons Market share 44 Thompson and Noordewier (1992) Financing rate, reb ates Sales Grover and Srinivasan (1992) Price, feature, coupon or combination Brand loyal versus switching segments Purchase acceleration, stockpiling activities Campbell and Diamond (1990) Monetary versus non - monetary promotions Customers' suspiciousness 45 Table 2 : Summary Statistics 46 Table 3 : Impact of Sales Promotions on Perceived Quality and Sales in the U.S. Auto Industry 47 Table 4 : Impact of Auto Promotions on Firm Value and Perceived Quality: Mass Vs. Luxury Product 48 Table 5 : Granger Causality Test: Impact of Lagged Sales, and Inventory and Supply on Firm's Promotional Strategies in the U.S. Automobile Industry (2003 - 20 12) 49 Figure 1 : Conceptual Model: The Differential Effects of Promotions for Mass v ersus Luxury Brands Perceived Quality Sales Cash Rebates Finance Rate MASS BRANDS Perceived Quality Cas h Rebates Finance Rate LUXURYBRANDS + ¯ ¯ + + ¯ + + + ¯ ¯ ¯ ¯ + + Sales 50 BIBLIOGRAPHY 51 BIBLIOGRAPHY Aaker, David A. (1991), Managing Brand Equity. New York:The Free Press. Aaker, David A. "The value of brand equity." Journal of business strategy 13.4 (1992): 27 - 32. Management(May, 1994), pp. 191 - 201. Arellano, Manuel. "Testing for autocorrela tion in dynamic random effects models." The review of economic studies 57.1 (1990): 127 - 134. Attanasio, Orazio P., Pinelopi Koujianou Goldberg, and Ekaterini Kyriazidou. "Credit Constraints In The Market For Consumer Durables: Evidence From Micro Data On Car Loans*." International Economic Review 49.2 (2008): 401 - 436. Atwal, Glyn, and Alistair Williams. "Luxury brand marketing the experience is everything!." Journal of Brand Management 16.5 (2009): 338 - 346. Ault, Richard W., T. Randolph Beard, David N. Economic Inquiry, 38 (4), 570 8. Babakus, Emin, Carol C. Bienstock, and James R. Van Scotter. "Linking Perceived Quality and Customer Satisfaction to Store Traffic and Revenue Growth*." Decision Sciences 35.4 (2004): 713 - 737. Barron, John M., Byung Uk Chong, and Michael E. Staten. "Emergence of captive finance companies and risk segmentation in loan markets: theory and evidence." Journal of Money, Credit and Banking 40.1 (2008): 173 - 192. Belch, G. E., Belch, M. A., Kerr, G. F., & Powell, I. (2008). Advertising and promotion: An integrated marketing communications perspective . mcgraw - hill. Bell, David R., and Xavier Drèze. "Changing the channel: A better wa y to do trade promotions." Image (2012). Beverland, Michael B. "Crafting brand authenticity: the case of luxury wines*." Journal of Management Studies 42.5 (2005): 1003 - 1029. Blattberg, Robert C., and Scott A. Neslin. "Sales promotion: the long and the s hort of it." Marketing Letters 1, no. 1 (1989): 81 - 97. 52 Blattberg, Robert C., and Scott A. Neslin. Sales promotion: Concepts, methods, and strategies . Englewood Cliffs, NJ: Prentice Hall, 1990. Blattberg, Robert C., Richard Briesch, and Edward J. Fox. "Ho w promotions work." Marketing Science 14, no. 3_supplement (1995): G122 - G132. Busse, Meghan, Jorge Silva - Risso, and Florian Zettelmeyer. "$1,000 cash back: The pass - through of auto manufacturer promotions." The American Economic Review (2006): 1253 - 1270. Marketing Science 29(2), pp. 268 290. Copeland, Adam, Wendy Dunn, and Ge orge Hall. Prices, production and inventories over the automotive model year . No. w11257. National Bureau of Economic Research, 2005. Chopra, Sunil, and ManMohan S. Sodhi. "Supply - chain breakdown." MIT Sloan management review (2004). Connelly, B. L., Cer to, S. T., Ireland, R. D., & Reutzel, C. R. (2011). Signaling theory: A review and assessment. Journal of Management , 37 (1), 39 - 67. Darke, Peter R., and Cindy MY Chung. "Effects of pricing and promotion on consumer perceptions: it depends on how you fram e it." Journal of Retailing 81.1 (2005): 35 - 47. Dawar, Niraj, and Philip Parker. "Marketing universals: consumers' use of brand name, price, physical appearance, and retailer reputation as signals of product quality." The Journal of Marketing (1994): 81 - 9 5. Dekimpe, Marnik G., Dominique M. Hanssens and Jorge M. Silva - - run effects - 291. Dekimpe, M. G., & Hanssens, D. M. (2004). Persistence modeling for assessing mark eting strategy performance. In D. Lehmann & C. Moorman (Eds.), Assessing marketing strategy performance, Marketing Science Demirag, Ozgun Caliskan, Pinar Keskinocak, and Julie Swann. "Customer rebates and retailer incentives in the presence of competition and price discrimination." European Journal of Operational Research 215.1 (2011): 268 - 280. Dickey , David A., and Wayne A. Fuller. "Distribution of the estimators for autoregressive time series with a unit root." Journal of the American statistical association 74.366a (1979): 427 - 431. Drazin, Robert, and Andrew H. Van de Ven. "Alternative forms of fit in contingency theory." Administrative science quarterly (1985): 514 - 539. 53 Erdem, Tülin, and Joffre Swait. "Brand equity as a signaling phenomenon." Journal of consumer Psychology 7.2 (1998): 131 - 157. Forbes, Mark P. "Method for retrieving vehicular colla teral." U.S. Patent No. 6,025,774. 15 Feb. 2000. Gambacorta, Leonardo. "How do banks set interest rates?." European Economic Review 52.5 (2008): 792 - 819. Gangwar, Manish, Nanda Kumar, and Ram C. Rao. "Consumer Stockpiling and Competitive Promotional Stra tegies." Marketing Science (2013). Sloan management review 26.1 (1984). Ghysels, Eric, Pedro Santa - Clara, and Rossen Valkanov. "The MIDAS touch: Mixed data sampling regression models." (2004). Ghysels, Eric, Arthur Sinko, and Rossen Valkanov. "MIDAS regressions: Further results and new directions." Econometric Reviews 26.1 (2007): 53 - 90. Gillis, Jack. The Car Book, 2007 . Gillis Publishing Group. Goldsmith, Ronald E., Leisa R. Flynn, and Daekwa n Kim. "Status consumption and price sensitivity." The Journal of Marketing Theory and Practice 18.4 (2010): 323 - 338. Granger, Clive WJ. "Some recent development in a concept of causality." Journal of econometrics 39.1 (1988): 199 - 211. Granger, Clive WJ, Bwo - Nung Huang, and Chin - Wei Yang. "A bivariate causality between stock prices and exchange rates: evidence from recent Asianflu ." The Quarterly Review of Economics and Finance 40.3 (2000): 337 - 354. Han, Young Jee, Joseph C. Nunes, and Xavier Drèze. "S ignaling status with luxury goods: the role of brand prominence." Journal of Marketing 74.4 (2010): 15 - 30. Hendricks, Kevin B., and Vinod R. Singhal. "The effect of supply chain glitches on shareholder wealth." Journal of Operations Management 21.5 (2003) : 501 - 522. Hiemstra, Craig, and Jonathan D. Jones. "Testing for Linear and Nonlinear Granger Causality in the Stock Price Volume Relation." The Journal of Finance 49.5 (1994): 1639 - 1664. Jedidi, Kamel, Carl F. Mela, and Sunil Gupta. "Managing advertising and promotion for long - run profitability." Marketing science 18, no. 1 (1999): 1 - 22. 54 Jensen, T., Kees, J., Burton, S., & Turnipseed, F. L. (2003). Advertised reference prices in an internet environment: effects on consumer price perceptions and channel search intentions. Journal of Interactive Marketing , 17 (2), 20 - 33. Johansen, Søren. "Identifying restrictions of linear eq uations with applications to simultaneous equations and cointegration." Journal of econometrics 69.1 (1995): 111 - 132. Joshi, Amit, and Dominique M. Hanssens. "The direct and indirect effects of advertising spending on firm value." Journal of Marketing 74. 1 (2010): 20 - 33. Kapferer, Jean - Noël, and Vincent Bastien. "The specificity of luxury management: Turning marketing upside down." Journal of Brand Management 16.5 (2009): 311 - 322. Kapferer, Jean - Noël a. The luxury strategy: break the rules of marketing to build luxury brands . Kogan Page Publishers, 2012. Kapferer, Jean - Noël b. "Why Luxury Should not Delocalize." European Business Review March (2012): 58 - 62. Kirmani, Amna, and Akshay R. Rao. "No pain, no gain: A critical review of the literature on signali ng unobservable product quality." Journal of marketing 64.2 (2000): 66 - 79. Leeflang, Peter SH, and Josefa Parreño - Selva. "Cross - category demand effects of price promotions." Journal of the Academy of Marketing Science 40, no. 4 (2012): 572 - 586. Low, Geor ge S., and Charles W. Lamb Jr. "The measurement and dimensionality of brand associations." Journal of Product & Brand Management 9.6 (2000): 350 - 370. Lu, Qiang, and Sridhar Moorthy. "Coupons versus rebates." Marketing Science 26.1 (2007): 67 - 82. Milgrom, Paul, and John Roberts. "Complementarities and fit strategy, structure, and organizational change in manufacturing." Journal of accounting and economics 19.2 (1995): 179 - 208. Miyazaki, Anthony D., Dhruv Grewal, and Ronald C. Goodstein. "The effect of mul tiple extrinsic cues on quality perceptions: a matter of consistency." Journal of Consumer Research 32.1 (2005): 146 - 153. Mohr, Lawrence B. (1982) Explaining Organizational Behavior. Jossey - Bass Inc., San Francisco, CA. Morgan, G. (1986), Images of Organ ization, Beverly Hills, CA: Sage Publications. Neslin, Scott. (2002). Sales Promotion. Cambridge: Marketing Science Institute Relevant Knowledge Series 55 Nijs, Vincent, Marnik G. Dekimpe, Jan - Benedict E. M. Steenkamp, Dominique M. Hanssens. 2001. The categ ory demand effects of price promotions. Marketing Sci. 21(1) 1 - 22. Oh, Haemoon. "Service quality, customer satisfaction, and customer value: A holistic perspective." International Journal of Hospitality Management 18.1 (1999): 67 - 82. Ohta, Makoto and Zv 4, No. 2 (Apr., 1986), pp. 187 - 198. Pauwels, Koen, Dominique M. Hanssens, and Sivaram akrishnan Siddarth. "The long - term effects of price promotions on category incidence, brand choice, and purchase quantity." Journal of marketing research 39.4 (2002): 421 - 439. Pauwels, Koen, Jorge Silva - Risso, Shuba Srinivasan and Dominique M. Hanssens (2 Marketing, Vol. 68, No. 4 (Oct., 2004), pp. 142 - 156. Phau, Ian, and Prendergast, Gerard (2001). Offensive advertising: A view from Singapore. Journal of Promotion Management , 7 (1/2), 71 - 90. Qian, Hang. "Vector autoregression with mixed frequency data." (2013). Raju, Jagmohan S., Sanjay K. Dhar, and Donald G. Morrison. "The effect of package coupons on brand choice." Marketing Science 13, no. 2 (1994): 1 45 - 164. Sarmah, S. P., Acharya, D., & Goyal, S. K. (2006). Buyer vendor coordination models in supply chain management. European journal of operational research , 175 (1), 1 - 15. Schoonhoven, Claudia Bird. "Problems with contingency theory: testing assumpti ons hidden within the language of contingency" theory"." Administrative Science Quarterly (1981): 349 - 377. Shapiro, Carl. "Premiums for high quality products as returns to reputations." The quarterly journal of economics 98.4 (1983): 659 - 679. Silk, Tim, and Chris Janiszewski. Managing mail - in rebate promotions . Working paper, 2008. Silva - Risso, Jorge, and Irina Ionova. "Practice Prize Winner - A Nested Logit Model of Product and Transaction - Type Choice for Planning Automakers' Pricing and Promotions." Mark eting Science 27, no. 4 (2008): 545 - 566. Silverstein, Michael J., and Neil Fiske. "Luxury for the masses." Harvard Business Review 81.4 (2003): 48 - 59. Silvestrini, Andrea, and David Veredas. "Temporal aggregation of univariate and multivariate time series models: a survey." Journal of Economic Surveys 22.3 (2008): 458 - 497. 56 Snoj, Boris, Aleksandra Pisnik Korda, and Damijan Mumel. "The relationships among perceived quality, perceived risk and perceived product value." Journal of Product & Brand Management 1 3.3 (2004): 156 - 167. Outcomes: Evidence from Truth - in - The Review of Financial Studies / v 24 n 2 2011, pp: 507 534. Srinivasan, Shuba, Koen Pauwel s, Jorge Silva - Product Innovations, Advertising, and Stock Returns Journal of Marketing, Vol. 73 (January 2009), 24 43 Tamirisa, Natalia T., and Deniz O. Igan. "Are Weak Banks Leading Credit Booms? Evidence f rom Emerging Europe." Comparative economic studies 50.4 (2008): 599 - 619. No. 6 (Nov. - Dec., 2007), pp. 758 - 773. Thompson, Patrick A., and Thomas Noordewier. "Estimating the effects of consumer incentive programs on domestic automobile sales." Journal of Business & Economic Statistics 10.4 (1992): 409 - 417. Thornton, Daniel L., and Dallas S. Batten. "Lag - length selection and tests of Granger causality between m oney and income." Journal of Money, credit and Banking (1985): 164 - 178. Varadarajan, P. Rajan, and Terry Clark. "Delineating the scope of corporate, business, and marketing strategy." Journal of Business Research 31.2 (1994): 93 - 105. Varki, Sajeev, and M ark Colgate. "The role of price perceptions in an integrated model of behavioral intentions." Journal of Service Research 3.3 (2001): 232 - 240. Verhoef, Peter C., Scott A. Neslin, and Björn Vroomen. "Multichannel customer management: Understanding the rese arch - shopper phenomenon." International Journal of Research in Marketing 24, no. 2 (2007): 129 - 148. Vigneron, Franck, and Lester W. Johnson. "Measuring perceptions of brand luxury." The Journal of Brand Management 11.6 (2004): 484 - 506. Yoo, Boonghee, Nave en Donthu, and Sungho Lee. "An examination of selected marketing mix elements and brand equity." Journal of the Academy of Marketing Science 28.2 (2000): 195 - 211. Zeithaml, Valarie A. "Consumer perceptions of price, quality, and value: a means - end model a nd synthesis of evidence." The Journal of Marketing (1988): 2 - 22. Zellner, Arnold, and Henri Theil. "Three - stage least squares: simultaneous estimation of simultaneous equations." Econometrica: Journal of the Econometric Society (1962): 54 - 78. 57 Chapter 2 Does Marketing Communication Mix Attract Generic Competition? 2. ABSTRACT Once the prescription drug patent expires, generic manufacturing firms enter the industry hare is not immediate, thereby extending patent life beyond the patent expiration date. Current study utilizes diffusion theory to analyze effectiveness of marketing communications to ensure diffusion of prescription drugs and enhance customer responsivene ss across product life cycle (PLC). Next, we utilize signaling theory to analyze whether in the post - patent era, marketing efforts undertaken by incumbents discourage generic competition or do they signal unexplored market potential and thereby lure compet ition. Current study utilizes sales, revenue and marketing expenditure data across 11 therapeutic classes from September, 2008 to October, 2014. Estimates indicate that ) and DTCA (i.e., direct - to - consumer advertising prescription drugs in the post - patent period acts as an entry dete rrent strategy for the first and second waves of generic entry whereas high brand price induces competition. Keywords : prescription drugs, detailing, direct - to - customer - advertising, generic manufacturers 58 2.1 Introduction Firms in the pharmaceutical indu stry adopt aggressive marketing mix strategies to aid in the rapid diffusion of new products (Leffler, 1981; Mackowiak and Gagnon, 1985; Vogel, Ramachandran and Zachry, 2002 ). Detailing and Direct - to - customer - advertisement (DTCA) are two critical component s of the marketing mix plan. Detailing educates physicians regarding the new product characteristics whereas DTCA exposes existing and new consumers to product information (Narayanan, Desiraju and Chintagunta, 2004). Firms also promote their products throu gh distribution of free samples and product advertisement in medical journals. Interestingly, Fischer and Albers, 2010) . For example, prescription drugs protected by the patent windo w are among the most heavily promoted drugs in the U.S. economy. During this period, incumbents adopt marketing mix strategies that help them to build a loyal customer base and strong brand equity (Ladha, 2007) However, once the drug patent expires, incumb ents may encounter competition for generic manufacturers conditional on economic and financial factors ( Grabowski, Ridley and Schulman, 2007 ). Interestingly, Hudson (2000) indicates that even with competition from generics, is not immediate, implying that value of patents extend beyond their expiration period (Hudson, 2000). Thus, during the post - patent period, incumbents tend to e off - patent prescription drugs and generics (Agrawal and Thakkar, 1997). However, for generic manufacturers, decision to enter a new market is exponentially more challenging and risky. Königbauer (2007 ) suggests that generic manufacturers may consider en tering the drug market if expected stream of revenue income significantly outweighs the cost 59 and risk associated with the corporate strategy. Literature is, however, equivocal regarding the - patent window o n generic entry decision. Hurwitz and Caves (1988) and Rizzo (1999) demonstrate that brand - name marketing activities expenditure in the pre - expiration period has no imp On the other hand, Königbauer (2007) uses a two - period Bertrand model of competition to demonstrate that product differentiation through advertising induces generic entry. The current study identifies some of the critical gaps in the pharmaceutical promotion literature and addresses them. product diffusion tend to concentrate on product life cycle within the patent protection window ( Sr idhar, Mantrala, and Albers, 2014) . We utilize diffusion theory to analyze the effectiveness of customer responsiveness across product life cycle, with particular emphasis in the post patent on generic entry decision. Current study solves the puzzle by demonstrating that certain factors (i.e. aggressive marketing strategies suc h as detailing and DTCA) may serve as entry deterrent strategies where as certain economic factors (i.e. high prescription drug prices) may actually lure generic competition. Moreover, marketing s market simultaneously. In reality, once the patent window expires, generic manufacturers tend to enter the market in sequential waves ( Grabowski and Vernon, 1992 ). We uti lize signaling theory to of sequential generic entries. 60 current e conomic and marketing conditions, independent of the decisions taken by the previous waves of generics that have already entered the market. In particular, in the post - patent period, (direct - to - consumer advertising) help them to keep their consumers and physicians well informed regarding effectiveness of the prescription drug and emphasize on its comparative advantage. Thus, ry deterrent strategy for the first and second waves of generic entry. Interestingly, high prescription drug prices even in the off - patent period may signal revenue - generation potential, which may consequently induce generic entry. The article is organiz ed as follows. Section 2 deals with literature review and hypothesis development. Section 3 provides details of the Prentice - Williams - Peterson Gap time (PWP gap - time) conditional model, methodology and variables used in the study and section 4 reports the data collection. Section 5 reports the results and section 6 provides a brief summary and discussion. 61 2.2 Hypothesis Development 2.2.1 Marketing Communication Mix Strategy e efforts directed at physicians encompass detailing (i.e., personal selling through sales representatives ), sampling distribution (i.e., distribution of free sampl es of drugs), physician meetings and events, and advertisements in medical journals. Even though promotion expenditure - to - consumer advertising (DTCA) has gained prominence ove r the last decade (Ma et al., 2003). 2.2.2 Detailing In the pharmaceutical industry, detailing, i.e. personal selling by pharmaceutical firms sales representatives to hospital and office - based physicians has been a critical component of drug promotions fo r decades (Donohue et al., 2004; Gagnon and Lexchin, 2008; Sridhar, Mantrala, and Albers, 2014 ). It includes direct contact by sales representatives at drug fair or a brief mention of the drug in hospital clinics, meeting rooms etc. Firms also provide phys icians information regarding the drug over the phone or through educational press. Cegedim - SK&A ( 2011 ) reports that in 2009 2010, U.S. Pharma Companies spent about $28 billion promoting drugs to prescribers, with detailing accounting for about $15.3 bill ion, or about 54 % of total annual promotion spending. Meta - analysis estimates by Sridhar, Mantrala, and Albers (2014) indicate that current - period detailing elasticity is 0.21. Additionally, elasticity estimates are higher for products 62 that are offered in early life cycle stages and differ across countries. Even though detailing is a dominant marketing strategy, yet firms selectively employ this strategy to aid product diffusion (Donohue, Cevasco and Rosenthal, 2007). Additionally, studies suggest that det ailing impacts (Venkataraman and Stremersch, 2007). 2.2.3 Direct - to - Customer Advertising (DTCA) Drug manufacturing firms often spend millions of dollars to promote t heir products directly to the customers through multiple media channels (ie, internet, television, newspapers, magazine, radio) ( Bell, Kravitz and Wilkes, 1999; 2000a; 2000b ). For example, in 2001, the US pharmaceutical industry spent an aggregate of US$2. 7 billion in DTCA campaigns (Young, 2003). Proponents of DTCA argue that this marketing channel is an opportunity to enhance health care by having patients identify symptoms of a curable medical condition and seek medical attention and also treat more broa dly diseases that are currently underdiagnosed or undertreated, and improve communication between the health care system and their patients ( Pines, 2000 ).Such massive advertising efforts are geared towards improved consumer awareness of advertised drugs th at may eventually open up dialog between physicians and patients. That conversation is most likely to induce the physician to prescribe the recommended drug, thereby generating demand for prescription drugs (Mintzes et al., 2003; Frosch et al., 2007 ). Oppo nents of DTCA argue that marketing effort by pharma firms directed at the customers is motivated by profit making incentives rather than concern for the public health ( Gellad and Lyles, 2007 ). They also argue that DTCA often results in wasting causes physi patients and also encourages the consumption of expensive and often times, unnecessary 63 medications (Rosenthal et al., 2002). Additionally, they argue that manufacturing firms tend to promote expensive prescr iption drugs that are newer with incomplete safety information (Lexchin 1999; Bradford et al., 2006). Therapeutic classes such as allergies, obstetrical/gynecological, dermatological, Cardiovascular, tobacco addiction are those that are advertised most fre quently (Wilkes, Bell and Kravitz, 2000). There is ample evidence in the literature supporting the positive Donohue, Cevasco and Rosenthal, 2007; Dave and Saffer, 201 0). 2.2.4 Sample Distribution Sales representatives of the manufacturing firms typically distribute samples either in person or during service visit or through mail (Dong, Li, and Xie, 2014). A drug sample in the prescription drug industry is defined as pharmaceutical product sufficient to evaluate clinical response, distributed to authorized health Dispensing free samples by pharmaceutical companies' sales representatives is one of the competitive marketing practices in the prescription drug industry ( Gönül et al, 2001 ). Moreover, distributing free samples to patients may be indicative of care and involvement that may eventual ly improve the physician - patient relationship (Groves, Sketris and Tett, 2003). Groves, Sketris and Tett (2003) indicates that expenditure on free sampling distribution accounts for more than half of the total marketing expenditure incurred by the U.S. pha rmaceutical industry. Specifically, in aggregate, pharmaceutical firms delivered an estimated $18.4 billion worth (in retail value) of free drug samples to doctors in year 2005 alone more than all other marketing expenses combined (Dong, Li and Xie, 2014 ). Findings indicate that distribution of free drug samples and detailing as the two 64 critical pharmaceutical marketing practices with significant positive impact on demand for prescription drugs (Mizik and Jacobson, 2002). 2.2.5 Journal Advertising Phys icians derive valuable information regarding latest drugs and devices from medical journals. In addition, these journals are also the source of scholarly articles. Simultaneously, medical journals often contain advertisements of drugs, thereby explicitly p romoting sales of the consideration as well as the advert isements. Moreover, by printing advertisements for drugs and devices, these medical journals are indirectly recommending these drugs and complement pharmaceutical firms that advertise in those reputed journals but also for the medical journals and the physician organizations that publish in those journals ( Fugh - Berman, Alladin and Chow , 2006 ). 2.2.6 Detailing and DTCA Marketing Strategies Rogers (1976; 1995) defin adopted by the mass consumers over time. Products have a life span. Long established products eventually loose consumer demand, while in contracts, demand for new product or idea increases dramatically after they are launched. Literature categorizes product life cycle (PLC) by four distinctive stages, namely introduction, growth, maturity and decline ( Qualls, Olshavsky, and Michaels, 1981) . 65 In line with the above categorization of PLC, we de prescription drugs that have received FDA approval over the last five years and are within the patent protection window ( Sridhar, Mantrala and Albers, 2014 ). These drugs are in their initiation stage in the diffusion process when inf ormation regarding the product becomes available to potential customers (i.e. physicians and consumers) . The manufacturing firms typically adopt market expansion strategies that target and provide product related information to physicians and consumers so given the monopoly market structure, incumbents typically allocate marketing budgets to educate the physicians (i.e., Detailing) and the consumers (i.e. DTCA) regarding the new drug and build C when the manufacturing firms already have a loyal consumer base and benefit from economies of scale in production. Consistent information asymmetry among consumers a nd physicians, improve market share and maximize sales. With patent expiration and generics entering the market, the competitive nature of the market segment changes. Off - patent prescription drugs experiences price competition and threat of market erosion. However, given that these drugs still have positive revenue generating potential, managers may need to optimally distribute marketing budget across the communication channels so as to maximize returns. Specifically, they may tend to adopt defensive market ing strategies geared towards retaining their existing customer base and fight competition (Berndt et al., 1996). Thus, one may conclude that the post - 66 stage in PLC. In the current study, we d - patent protection over the last five years. marketing efforts geared towards educating physician s help to address this information asymmetry problem, thereby aiding new product diffusion process. However, as new drug move along the PLC, physicians access to drug related information increases. Thus towards the later stages in PLC, increased detailing expenditure may not necessarily translate into enhanced product demand. However, its ability to generate high product demand and revenue drops as the product pro gresses Once the patent expires, the incumbent experiences increased competition from generic drug difference (Agrawal and Thakkar, 1997). In particular, the marketing communications not only emphasize the off - generics that have flooded the market. Generics are typically priced much lower than the prescrip tion drugs. However, physicians may not have full information regarding chemical side effects (Borgheini, 2003) Moreover, these cheaper alternatives may not ne cessarily have gone through all the steps in clinical trials and physicians may not be very comfortable prescribing the generics over the off - mix plan may help the incumbents to ret ain their consumer base and prevent market erosion. We 67 H 1 : Similar line of reasoning applies for effectiveness of DTCA as the marketing strategy that Towards the early stage of PLC, consumers have incomplete information regarding effectiveness of the drug as well as the possible side effects. Thus advertising strategies that aim at educating the customers help in the product diffusion process (Lexchin 1999; Bradford et al., 2006). As the product progresses towards its maturity stage, information regarding effectiveness of the drug and towards DTCA may not t ranslate into higher sales. However, once the patent window expires, the manufacturing firm may have to defend its market share. It may do so by adopting marketing strategies that effectively communicates with its loyal customer base and reconfirms drug ef drug progresses along its PLC. H 2 e. mood, behavior and thinking pattern (Mayo Clinic). Depression, schizophrenia, addictive behaviors, anxiety disorders are some examples of mental illness. On the o ther hand, disease is a 68 pathological condition of a body part, an organ, or a system caused by infection, inflammation, environmental factors, or genetic defect ( Tikkinen et al., 2012) . Diagnosis and treatment of mental illness is relatively more dependent rather than on concrete laboratory results. One may posit that mental illness is relatively more subjective than diseases; demand for drugs that treat mental illness is more driven by consumers t han by physicians. Firms employ DTCA (detailing) as the primary communication channel to promote their products directly to the consumers (physicians). Thus, we hypothesize that compared to detailing, DTCA is more effective in promoting drugs that treat me ntal illness. H 3 mental illness relative to those that treat diseases. Diagnosis and treatment of diseases are relatively more driven by laboratory examinations r objective and organic than mental illness. Since detailing primarily educates the physicians, we hypothesize that relative to DTCA, detailing is more effective in promoting drugs that treat diseases rather than those that treat mental illness. H 4 diseases relative to those that treat mental illness. 2.2.7 Entry Det errent Strategies by Prescription Manufacturing Firms 69 Patent window ensures market exclusivity to the inventor(s) of the prescription drugs for a fixed period of time (Königbau er, 2007) and to ensure diffusion of new products (Eisenberg, 2003). Within this nt on several economic (Caves, Whinston & Hurwitz, 1991; Hudson, 2000) and marketing (Königbauer, 2007) factors. One observes entry of generic manufacturing firms with a time lag ranging from few days to several years (Hudson, 2002). Consequently, incumben ts experience significant erosion of market share and loss in revenue (Grabowski and Vernon, 1992). There is as extensive body of literature that captures the relation between the marketing efforts and generic market entry. Interestingly, studies are equi vocal with regards to the relationship. Hurwitz and Caves (1988) demonstrates that in the pharmaceutical industry, consumers help the firm to retain its market share. However, generic price discounts erode market share. Rizzo (1999) suggests that brand - name advertising enhances brand loyalty, thereby decreasing price - elasticity of demand. Thus, both these studies conclude that brand - name marketing inhibits generic mark et entry. Scott Morton (2000) investigates the influence of - - name adverti sing. Königbauer (2007) uses a two - period Bertrand model of competition to demonstrate that product differentiation through advertising induces generic entry. The author argues that market entry is costly. Thus, the expected profit that the generic manufac turers may earn once they 70 through advertising activities induces generic market entry. Despite many studies documenting the incidence of generic entries in the pharmaceutical industry, there is scarcity of research that recognizes the sequential nature of generic entry. In particular, generic manufacturing firms enter the drug market in waves, contingent on FDA approval for their bioequivalent drug. Addition ally, entry of generics range from the date of patent expiration of the prescription drug to few years, depending on the revenue generation capability of the latter (Hudson, 2000). We posit that market communication mix and market saturation are two critic al factors that influence generic entry decision. Entry into a new market is costly. Hence, generic manufacturing firms are likely to enter the market only if they expect the payoff to be greater the cost and risk of entering the market. In particular, we hypothesize that in the off - patent market dominance. H 5 : Inc - patent drugs deter generic competition. 71 2.3 Methodology 2.3.1 Empirical Model to boost sales us ing the following regre ssion analysis (i.e. equation, 7 ). (7) We now provide a definition of the variables used in the analysis: Dependent Variable : Total sales of the drug th at belongs to therapeutic class in time . Explanatory Variables : therapeutic class treats non - fatal m edical conditions else is assumes the value of unity. : therapeutic class treats mental health else is assumes the value of unity. Detail ( drug by engaging in one - on - one conversations. DTCA ( ing effort to promote drug directly to its customers. 72 2.3.2 Prentice - Williams - Peterson Gap Time Model Cox Proportional Hazard (CoxPH) model analyzes time to event outcomes (Fox, 2002) . It takes into consideration of the time to relapse and does not assume a constant hazard rate. Alternatively, it assumes that the ratio of risk for generic competition between two off - patent prescription drugs is constant over time. Cox proportional hazard function may be expressed as: (8) where are values of covariates and is a vector of regression parameters. is t he baseline hazard function that describes how risk of an event is a function of time at baseline levels of covariates and t he effect parameters describe response of hazard to changes in explanatory covariates. Since the generic manufacturers enter the off - patent drug market in sequential waves, each wave is expected to alter the competitive environment. In particular, with every wave of generics entering the market, the off - patent prescription drug and the existing bio - equivalent generics compete for marke t share. Thus, this is a typical example where the baseline hazard function varies by strata, thereby not satisfying the basic assumption of proportional hazard model. Additionally, the sequence of occurrence of events is critical, specifically, the second wave of generics will not enter the market until the first wave of generic have already entered the market. Our study employs Prentice - Williams - Peterson Gap time (PWP - GT) conditional model, an extension of the COX proportional hazard model that analyzes r ecurrent events with stratifications. It is the conditional model that conserves the order of sequential entry of generic manufacturing firms in the creation of the risk set, thereby allowing for entry dependence. In particular, PWP - GT model incorporates f or 73 events correspond to different baseline hazards. We estimate the PWP - GT model with the data organized in gap time (ie, time since last wave of generic entry) (Ullah, Gabbett, and Fi nch, 2014). PWP - GT specified that the hazard function at time t as a function of earlier generic entries and firm characteristics of , as given by equation (9 ) (9) Where are completely artibrary baseline intensity functions; stratification variable may vary as a function of time for a specific wave of entry, is a colum n vector of regression coefficients specific for the strata. is the time of the failure and represents the gap time between occurrences of two sequential events. In th e PWP - GT model, dependent variable is the hazard rate, i.e., the likelihood of generic closed. In the current study, we control for type of medical condition (i.e . mental health vs diseases) and criticality of the medical condition (i.e. non - fatal vs fatal). We also control for the . Following are the independent variables used in the analysis: Sample: It is the quantity of the drugs distributed in the form of samples by manufacturing Percentage of Marketing Expenditure on RVOS: It represents the percentage of pharmaceutical the distribution of free samples among consumers. Percentage of Marketing Expenditure on Detailing: It represents the percentage of 74 Percentage of Marketing Expenditu re on DTCA: It represents the percentage of pharmaceutical asset. Market S total sales to industry sales in that period. Prescription Drug Price: Price of the prescription drug charged by the manufacturer. 75 2.4 Data Collection We obtained mo nthly marketing expenditure data for eleven therapeutic classes for the month September, 2008 through August, 2014 from IMS Health. Marketing expenditure directed to physicians is composed of four components: detailing (i.e. providing drug related informat ion in a face - to - face meeting to the office and hospital - based physicians, providing free samples to physicians, and advertising in medical journals (Rosenthal et al., 2002). Drug patent applicant and supplier information, patent expiration dates are obtai ned from drugpatentwatch.com. Additional information on generic manufacturing firms entering the respective drug market are obtained from drugs.com and WebMD .com. We obtained information on firm performance (i.e., total sales, firm size, R&D intens ity) f rom COMPUSTAT . 76 2.5 Results Table (6) provides the summary statistics of promotional expenditure by prescription drug manufacturers across 11 therapeutic classes. Estimates indicate that drugs that treat fatal diseases (i.e., Pyrimidine, Antineo Monoclon al Antibody, Tyrosine Kinase Inhibator etc.) channelize substantially less marketing expenditure towards DTCA vis - à - vis drugs that treat mental illness (i.e. Serotonin) and non - fatal diseases (Beta Blockers, HMG - CoA Reductase etc.). Similar dichotomy is ob served in pharma marketing expenditure towards sampling distribution and journal advertising. Specifically, incumbents that manufacture drugs that treat mental illness (i.e. Serotonin, SSRI and SNRI) and those that treat non - fatal diseases tend to invest i n sampling distribution. However, there is a significant drop in marketing expenditure geared towards distribution of free samples by firms that manufacture cancer treatment drugs (i.e. Pyrimidine, Antineo Monoclonal Antibody, Tyrosine Kinase Inhibator etc .). One may infer that high value of sales towards detailing are approximately consistent across all the therapeutic classes, ranging between .0 01 to .1 percen tage of sales. In this section we classify therapeutic classes against three broad categories, specifically, non - fatal mental illness (category 1), non - fatal disease (category 2) and fatal disease (category 3). Table (7) presents the difference in marketin g expenditure by manufacturing firms before and after prescription drug patent expiration across these three categories. Estimates indicate that the difference in mean expenditure before and after patent window closes for drugs that treat category 1, categ ory 2, and category 3 are 14.8349 (p<.1), 18.5372 (p<.05), and 13.9029 (p<.05) million dollars respectively. Thus, findings suggest that manufacturing firms tend to significantly reduce 77 their promotional expenditure in RVOS once the patent window expires. Interestingly, findings also suggest that pharmaceutical firms that produce drugs to treat category 2 (i.e., non - fatal disease) actually enhance their DTCA expenditure once the patent window closes by 3.2047 million dollars (p<.05). Table 8 provides the co rrelation coefficient matrix of the variables used in the current study. expenditure across all the marketing channels of communication. In particular, percent age of expenditure for distribution of free samples ( ) is negatively correlated with that of detailing ( ) (p<.001), journal advertisement ( ) (p<.001) and direct - to - customer advertising ( ) (p<.001). Additionally, estimates indicate negative correlation between brand price ( ) and brand sales ( ) (p<.001) and percentage of marketing expenditure across the different channels (p<.001 ). Interestingly, brand sales ( ) has positive association with percentage of expenditure for distribution of free samples ( ) (p<.001) and direct - to - customer advertising ( ) (p<.001), whereas it is negatively correlated with percentage of expenditure geared towards detailing ( ) (p<.001) and that of journal advertisement ( ) (p<.001). The elasticity estimates of the marketing effort are presented in tab le (9). In panel (A), dependent variable is logarithmic value of total revenue generated by the drug, whereas in panel (B), the dependent variable is log of total sales of the product. Exogenous variables used in the analysis are logarithmic values of the building awareness among the consumers (i.e. Direct to Customer Advertising (DTCA)) and the physicians (Detailing). Panel A (Panel B) provides us with the estimates of responsiveness of revenue (total sales) to changes in marketing expenditure. 78 - PLC if it has been launched in the market less than 5 years. ars for patent to expire. - - patent over the last 5 years. In panel (A), the coefficients are marketing elasticity of revenue (MER) that measures responsiveness in total revenue to a change in the expenditure on the marketing effort for a specific product (% change revenue / % change in marketing expenditure). Similarly, panel (B) provides estimates of marketing elasticity of sales (MES) that measures responsiveness in total sales (Sa les) to a change in the expenditure on the marketing effort for a specific product (% change in sales / % change in marketing expenditure). - fatal medical condition; else is assum - non - fatal diseases has higher effect on revenue that those that treat fatal d iseases. Interestingly, we observe a switching pattern in the growth PLC stage improvement in marketing expenditure on fatal drugs has greater effect on revenue generation vis - à - vis non - fatal drugs. if the drug treats mental illness; else is drugs that treat diseases are more responsive to changes in marketing expenditure vis - à - vis those that treats me ntal health. However, as the drugs moves through their stages in PLC (i.e. growth and 79 post - patent stage) , changes in revenue generation to changes in marketing expenditure is more for those drugs that treat mental health than those that treat diseases. Fi ndings in Panel A indicate that a 10% increase in the detailing expenditure increases revenue generated by the product by a 16.54% (p<.001) in its infant/early stage. However, a similar increase in marketing expenditure improves revenue by 2.82 (p<.001) an d 7.98 (p<.001) percentage points in the growth and post - patent expiration era respectively. Additionally, estimates also suggest that revenue responsiveness to changes in DTCA expenditure is inelastic. In particular, one percentage change in manufacturing revenue generated by the product by .860 (p<.001), .129 (p<.001) and .465 (p<.001) percentage points in the infant, growth and post - patent expiration era respectively. The fourth column provides us with difference estimates between the infant and growth stages whereas the fifth column provides us with difference estimates between the mature and growth stages. Estimates indicate that effectiveness of Detailing as a marketing strategy to improve revenue (sales) is hi gher in the infant stage than the growth stage by 1.372 (p<.001) and .841 (p<.001) percentage points respectively. Additionally, its effectiveness in improving revenue (sales) is higher in the mature stage than the growth stage by .561 (p<.01) and (.313) ( p<.05) percentage points respectively. Thus, one may conclude that effectiveness of detailing as a Furthermore, results suggest that effectiveness of DTCA as a m arketing strategy to improve revenue and sales is higher in the infant stage than the growth stage by .731 (p<.001) and .284 (p<.05) percentage points respectively. Additionally, its effectiveness in improving sales is higher in the mature stage than the g rowth stage by .079 ( p<.1) percentage points respectively. Thus, one may conclude that effectiveness of detailing and DTCA as a marketing communication to 80 age in PLC. Thus, findings are consistent with hypotheses H1 and partially satisfy hypotheses H2. Empirical findings in panel (A) indicate that for prescription drugs that treat diseases, for one towards detailing enhances revenue generation by the product by .267 (p<.001), .330 (p<.001) and .530 (p<.001) percentage points in the infant, growth and post - patent expiration stages respectively. Similarly, for prescription drugs that treat diseases, on DTCA depresses revenue generated by the product by .344 (p<.001) and 0.141 (p<.05) in the infant and post - patent expiration stages respectively. The difference in estimates between responsiveness of detailing and DTCA in generating revenue is .611 (p<.001), .256 (p<.001) and .671 (p<.001) in the infant, growth and mature stages respectively. Similarly, according to the estimates in panel (B), the difference in estimates between responsiveness of detai ling and DTCA in generating sales is .117 (p<.01) and .290 (p<.001) in the infant and mature stages respectively. revenue and sales when the drug under conside post - patent period of the drug. revenue varies across PLC and is contingent of the characteristics of the medical condition it treats. In particular, DTCA is an effective strategy for those drugs that treat mental health and are either in their 81 - ge of their PLC. However, it is an effective strategy to enhance revenue if the drug treats diseases and is in its mature stage. Table (5) provides us with the estimates of PWP - Gap time model. The dependent variable is the hazard rate, i.e., the likelihoo d of generic manufacturing firms entering the market, given that Column (A) provides us with the estimates of n (B) provides us with the estimates targeted at the second wave of generic manufacturers who enter the drug market. The standard errors have been reported in parenthesis. We report the corresponding hazard ratio directly below the standard errors. If the hazards ratio an independent variable is less than 1, an improvement in the variable decreases the hazard rate. According to HR estimates in column A, HR for marketing strategy targeted at educating the physicians (Detailing) and customers (DTCA) is 0.017 (p<.05) and .001 (p<.01) respectively. Results indicate that in the post - expenditure incurred by the prescription drug manufacturing firm helps to them to continue building awareness among physicians and customers regardi ng the effectiveness of the drug. Specifically, it improves the likelihood of physicians continue to prescribe the drug and customers continue to consume it. These entry deterrent strategies employed by the incumbent consequently discourage generic manufac turers from entering the market. If the hazards ratio (HR) of an independent variable is larger than 1, an increment in the variable increases the hazard rate. Estimates of HR for the prescription drug price is 1.005 (p<.05). This may indicate that high pr escription drug prices are likely to attract competition from generics. In the post - patent era, incumbents typically experience competition from generic manufacturing firms. Drop in prescription drug prices may have been an effective strategy adopted by in cumbents 82 to prevent market share erosion. However, the very fact that incumbents continue to maintain high prescription drug prices even in the post - patent period is a strong signal to potential competitors regarding unexplored market potential. This prici ng strategy may however attract the first wave of generic entry. Column (B) provides us with the estimates for the second wave of generic entry. HR for Retail value of samples (RVOS) and Detailing are .018 (p<.01) and .011 (p<.05) respectively. Distributi on of free samples to consumers and continuance of physician education are expensive affairs. Findings suggest that when incumbents continue to allocate significant marketing budget towards these marketing strategies, they successfully deter even the secon d wave of generic entry. Interestingly, firm size of the generic manufacturer is a critical factor influencing its decision to enter the market. Finally, HR for firm size of the generic manufacturer is .190 (p<.05). Thus, estimates suggest that prescriptio n drug manufacturing firm is less likely to encounter competition 83 2.6 Discussion Key Findings and Theoretical Implications : T he U.S. drug manufacturers promote their products heavily to ensure accelerat ed adoption of new drugs and retain market share of existing drugs (Neslin, 2002; Rosenthal et al., 2002; Donohue, Cevasco, and Rosenthal, 2007). Although the patent window of a prescription drug closes on a specific date, the drug's trademark continues to live on as the vehicle for maintaining the goodwill and possibly delaying or impeding subsequent generic competition (Caves, Whinston and Hurwitz, 1992). Thus, the incumbent often continues to promote its prescription drug even after the latter goes off patent and faces competition from generic manufacturers (Aitken, Berndt, and Cutler, 2009). These marketing communications by the incumbent serve two primary purposes. First, they continue to build brand loyalty and re - establish t he relative effectiveness of the brand drug compared to the bio - equivalent generics that may be available in the market (Grabowski and Vernon, 1992). Second, they serve as an entry deterrent strategy (Ellison and Ellison, 2007). In the current study we ana stages of PLC. We consider detailing, direct - to - customer advertising (DTCA), sample distribution and journal advertising as the four broad categories of marketing strategies that are typical ly adopted by the U.S. pharmaceutical firms to promote their drugs. Our results indicate that he product life cycle stages, with minimum effectiveness at the growth stage. One may conclude that drugs in their growth stage have an established market with known effectiveness and possible side effects. Thus, additional marketing expenditure to increas e consumer awareness may not necessarily yield high 84 returns. However, when a drug is in its infant stage, physicians and consumers have limited information regarding its effectiveness in treating the medical condition and possible side effects. Thus, incum of these uncertainties and decrease risk. strategy also varies across therapeutic - a - conditions that are relatively organic and subject to observable pathology vis - à - vis mental health tha t are more descriptive and not observable readily. Additionally, diagnosis of the former requires expert clinical eye of physicians whereas the latter subject to interpretation of the patient. Thus, increased budget allocation towards detailing (i.e. marke ting effort geared towards educating the physicians) is expected to generate sales and revenue for drugs treating diseases relative to those treating mental health. Our findings confirm our hypothesis. Interestingly, DTCA is an effective strategy to impro ve sales and revenue in the infant and post - patent period. In both these stages, the incumbent utilizes DTCA to establish product credibility and emphasize on its comparative advantage. Since mental illness is a subjective medical condition and depends lar resolve some of the uncertainties by providing information directly to the customers. Finally, results indicate that continuance of marketing communications mix by the incumbent in the post - expiration period emits mixed signal to the wave of generics entering the market. Specifically, estimates indicate that in the post - marketing communications through detailing and DTCA (direct - to - consumer advertising) ac t as an entry deterrent strategy for the first and second waves of generic entry. Given that these strategies 85 - patent period may deter generic. Interestingly, continuance of high bran d price even in the off - patent period lures competition. In particular, when the generic manufacturers observe that the incumbent continues to charge high price for its off - generation poten tial, which may consequently induce them to enter the market. Managerial Implications : Current study indicates that effectiveness of marketing strategies in promoting the product and generating sales varies across product lifecycle. Thus, in order to maxim ize returns and firm valuation, managers of pharmaceutical firms may need to adjust their marketing expenditure and effort contingent on whether the drug is in its infant/ mature or post - patent stage. Additionally, findings suggest that once the prescript ion drug goes off - patent and generic expenditure as signals of market potential. Thus, managers of the incumbent firm may need to be aware of the downside of con tinuance in marketing efforts even in the post - patent stage. They may need to adjust their m arketing effort accordingly. Limitations and Future Research : The current study establishes that effectiveness of marketing communications vary across drug life cy cle. An interesting extension of the current across marketing channels and across product life cycle that maximizes total sales/ revenue generation. Incumbents adopt detaining and DTCA as primary promotional vehicles to diffuse their product. It may be interesting to analyze in a game theoretic setup how these strategies may impact - bioequivalent or quasi - bioequival ent drugs 86 in the same therapeutic class. An interesting analysis may be examining the cross - marketing effect if the rival firm introduces the close - substitute drug with a time lag. 87 A PPENDIX 88 Table 6 : Distribution of Marketing Expe nditure on Prescription Drugs Across Therapeutic Classes (Sept, 2008 - Nov, 2014) 89 Table 7 : Difference in Promotional Expenditure (in Millions $) Pre and Post Prescription Drugs' Patent Expiration (Sept, 2008 - Aug , 2014) 90 Table 8 : Summary Statistics 91 Table 9 : Effectiveness of Marketing Strategies Across Prescription Drugs' Product Life Cycle 92 Table 10 : Prescription Drug Manufacturing Firm's Entry Deterrant Strategy Using PWP - Gap time Model with Stratum - Speci fic Regression Coefficients 93 BIBLIOGRAPHY 94 BIBLIOGRAPHY Agrawal, Madhu, and Nimish Thakkar. "Surviving patent expiration: strategies for marketing pharmaceutical products." Journal of product & brand management 6.5 (1997): 305 - 314. Aitken, Murray, Ernst R. Berndt, and David M. Cutler. "Prescription drug spending trends in the United States: looking beyond the turning point." Health Affairs 28.1 (2009): w151 - w160. Aronsson, Thomas, Mats A. Bergman, and Niklas Rudholm. "The impact of generic drug competition on brand name market shares Evidence from micro data." Review of Industrial Organization 19.4 (2001): 423 - 433. Basara LR. The impact of a direct - to - consumer prescription medication advertising campaign on new prescription volume. Drug InfJ. 1996;30:715 - 729. Bell, Robert A., Richard L. Kravitz, and Michael S. Wilkes. "Direct to consumer prescription drug advertising and the public." Journal of General Internal Medicine 14.11 (1999): 651 - 657. Bell, Robert A., Michael S. Wilkes, and Richard L. Kravitz (2000a) "The educational value of consumer - targeted prescription drug print advertising." Journal of Family Practice 49.12: 1092 - 1098. Bell, Robert A., Richard L. Kravitz, and Michael S. Wilkes (2000b)"Direct - to - consumer prescription drug advertising, 1989 - 1998. A content analysis of conditions, targets, inducements, and appeals. " J Fam Pract 49.4: 329 - 335. Berndt, Ernst R., Linda T. Bui, David H. Lucking - of Marketing, Product Quality, and Price Competition in the Growth and Composition of the mics of New Goods, Timothy F. Bresnahan and Robert J. Gordon, editors, University of Chicago Press, pp: 277 328. Blattberg, R. C., & Neslin, S. A. (1990). Sales promotion: Concepts, methods, and strategies (pp. 349 - 350). Englewood Cliffs, NJ: Prentice H all. Borgheini, Giuseppe. "The bioequivalence and therapeutic efficacy of generic versus brand - name psychoative drugs." Clinical therapeutics 25.6 (2003): 1578 - 1592. Bradford, W. D., Kleit, A. N., Nietert, P. J., Steyer, T., McIlwain, T., & Ornstein, S. (2006). How direct - to - behavior. Health Affairs , 25 (5), 1371 - 1377. and Co 95 Cegedim - SK&A (2011) 2010 U.S. Pharma Company promotion Spending, January. Available at www.forums.pharma - mkting.com/attac hment.php?attachmentid=154&d Dave, Dhaval, and Henry Saffer. "Impact of direct - to - consumer advertising on pharmaceutical prices and demand." Southern Economic Journal 79.1 (2012): 97 - 126. Dong, Xiaojing, Michael Li, and Ying Xie. "Understanding Sample Usage and Sampling as a Promotion Tool: State of Industry Practice and Current Research." Innovation and Marketing in the Pharmaceutical Industry . Springer New York, 2014. 507 - 530. Donohue, Julie M., Ernst R. Berndt, Meredith Rosenthal, Arnold M. Epstein, and Richard G. Medical care 42, no. 12 (2004): 1176 - 1185. Donohue, Julie M., Marisa Cevasco, and Meredith B. Rosenthal. "A decade of direct - to - consumer advertising of prescription drugs." New England Journal of Medicine 357.7 (2007): 673 - 681. Duflos, Gautier, and Frank R. Lichtenberg. "Does competition stimulate drug utilization? The impact of changes in market structure on US drug prices, marketing and utilization." International Review of Law and Economics 32.1 (2012): 95 - 109. 72 , issue 3, pp 477 491. Ellison, Glenn, and Sara Fisher Ellison. Strategic entry deterrence and the behavior of pharmaceutical incumbents prior to patent expiration . No. w13069. National Bureau of Economic Research, 2007. Fischer, Marc, and Sönke Albers . "Patient - or physician - oriented marketing: what drives primary demand for prescription drugs?." Journal of Marketing Research 47.1 (2010): 103 - 121. Fox, John. "Cox proportional - hazards regression for survival data." An R and S - PLUS companion to applied r egression (2002): 1 - 18. Frosch, D. L., Krueger, P. M., Hornik, R. C., Cronholm, P. F., & Barg, F. K. (2007). Creating demand for prescription drugs: a content analysis of television direct - to - consumer advertising. The Annals of Family Medicine , 5 (1), 6 - 13 . Fugh - Berman, Adriane, Karen Alladin, and Jarva Chow. "Advertising in medical journals: should current practices change?." PLoS Medicine 3.6 (2006): e130. Gagnon, Marc - pharmaceutical PLoS medicine 5.1 (2008): e1. 96 - to - The American journal of medicine 120.6 (2007): 475 - 480. Gönül, Füsun F., Franklin Car ter, Elina Petrova, and Kannan Srinivasan. "Promotion of Journal of Marketing 65, no. 3 (2001): 79 - 90. ition (Oct., 1992), pp. 331 - 350. Grabowski, Henry G., David B. Ridley, and Kevin A. Schulman. "Entry and competition in generic biologics." Managerial and Decision E conomics 28.4 5 (2007): 439 - 451. Groves, K. E. M., I. Sketris, and S. E. Tett. "Prescription drug samples does this marketing strategy counteract policies for quality use of medicines?." Journal of clinical pharmacy and therapeutics 28.4 (2003): 259 - 271. - up in the pharmaceutical market following patent expiry A multi - 221 ares of brand name and generic pharmaceuticals , Journal of Law and Economics, Vol. 31, No. 2 (Oct., 1988), pp. 299 - 320. Kadiyali, Vrinda. "Entry, its deterrence, and its accommodation: A study of the US photographic film industry." The Rand Journal of Ec onomics (1996): 452 - 478. Königbauer, Ingrid. "Advertising and generic market entry." Journal of health economics 26.2 (2007): 286 - 305. Kvesic, Dennis Z. "Product lifecycle management: marketing strategies for the pharmaceutical industry." Journal of Medi cal Marketing: Device, Diagnostic and Pharmaceutical Marketing 8.4 (2008): 293 - 301. Ladha, Z. (2007). Marketing strategy: are consumers really influenced by brands when purchasing pharmaceutical products?. Journal of Medical Marketing: Device, Diagnostic and Pharmaceutical Marketing , 7 (2), 146 - 151. Leffler KB. Persuasion or information? The economics of prescription drugs advertising, J Law Econ. 1981;24:45 - 74. - to - Disease Management and Health Outcomes 5.5 ( 1999): 273 - 283. 97 Luo, Xueming, and Chitra Bhanu Bhattacharya. "The debate over doing good: Corporate social performance, strategic marketing levers, and firm - idiosyncratic risk." Journal of Marketing 73.6 (2009): 198 - 213. Ma, Jun ; Stafford, Randall S. ; Co ckburn, Iain M. ; Finkelstein, Stan N. A statistical analysis of the magnitude and composition of drug promotion in the United States in 1998, Clinical Therapeutics , 25 (5) (May 2003): 1503 - 1517. Mackowiak, John I., and Jean Paul Gagnon. "Effects of promotion on pharmaceutical demand." So cial Science & Medicine 20.11 (1985): 1191 - 1197. Masood, I., Ibrahim, M. I., Hassali, M. A., & Ahmed, M. (2009). Evolution of marketing techniques, adoption in pharmaceutical industry and related issues: a review. J Clin Diagnostic Res , 3 , 1942 - 52. Mintz es, Barbara, Morris L. Barer, Richard L. Kravitz, Ken Bassett, Joel Lexchin, Arminée - to - consumer advertising (DTCA) affect prescribing? A survey in primary care environments w ith Canadian Medical Association Journal 169, no. 5 (2003): 405 - 412. detailing and sampling on new prescriptions." Management Science 50 .12 (2004): 1704 - 1715. Neslin, Scott. (2002). Sales Promotion. Cambridge: Marketing Science Institute Relevant Knowledge Series. Rogers, Everett M. "New product adoption and diffusion." Journal of consumer Research (1976): 290 - 301. Rogers Everett, M. "Di ffusion of innovations." New York (1995). Scott Morton, Fiona M. "Barriers to entry, brand advertising, and generic entry in the US pharmaceutical industry." International Journal of Industrial Organization 18.7 (2000): 1085 - 1104. Narayanan, Sridhar, Ram arao Desiraju, and Pradeep K. Chintagunta. "Return on investment implications for pharmaceutical promotional expenditures: The role of marketing - mix interactions." Journal of marketing 68.4 (2004): 90 - 105. - to - The Annals of pharmacotherapy 34, no. 11 (2000): 1341 - 1344. Qualls, William, Richard W. Olshavsky, and Ronald E. Michaels. "Shortening of the PLC: An empirical test." The Journal of Marketing (1981): 76 - 80. Rizzo, John A. "Advertising and Competition i n the Ethical Pharmaceutical Industry: The Case of Antihypertensive Drugs*." The Journal of Law and Economics 42.1 (1999): 89 - 116. 98 Rosenthal, M. B., Berndt, E. R., Donohue, J. M., Frank, R. G., & Epstein, A. M. (2002). Promotion of prescription drugs to c onsumers. New England Journal of Medicine , 346 (7), 498 - 505. A meta - Innovation and Marketing in the Pharmaceutical Industry . Springer New York, 2014 . 531 - 556. Ullah, Shahid, Tim J. Gabbett, and Caroline F. Finch. "Statistical modelling for recurrent events: an application to sports injuries." British journal of sports medicine (2012): bjsports - 2011. Venkataraman, Sriram, and Stefan Stremersch. "The debate on influencing doctors' decisions: Are drug characteristics the missing link?." Management Science 53.11 (2007): 1688 - 1701. Vogel, Ronald J., Sulabha Ramachandran, and Woodie M. Zachry. "A 3 - stage model for assessing the probable economic effects o f direct - to - consumer advertising of pharmaceuticals." Clinical therapeutics 25.1 (2003): 309 - 329. Warrier, R., Monaghan, M. S., Maio, A., Huggett, K., & Rich, E. (2010). Effect of drug sample availability on physician prescribing behavior: a systematic re view. - to - consumer prescription Health Affairs 19.2 (2000): 110 - 128. - to - CMAJ 169.5 (2003 ): 385. 99 Chapter 3 The effect of Loyalty Program on firm risk and value 3. ABSTRACT Loyalty programs (LPs) are dynamic incentive programs where consumers are benefitted from cumulative purchase over time. Studies indicate that not all LP are equally suc cessful and some fail to generate the expected stream of revenue for the firm, leading to volatility in expected stream of revenue. A report published by Colloquy reiterates the fact by indicating that American businesses distribute approximately $48 billi on worth of perceived value in reward points and miles annually; surprisingly only two - third of these points are redeemed by consumers. The current study utilizes a sample of 336 U.S. firms inclusive of retail, hospitality, telecommunication and entertai nment sectors idiosyncratic risk after it adopts the program following the Fama - French four - factor model. - specific risk. Next, we demonstrate that In particular, adoption of loyalty program by firms with high market share depletes sales. On the other hand, adoption of loyalty programs by small firms boost sales market share. Keywords : loyalty program, firm - specific risk, market share, sales 100 3 .1 Introduction Loyalty programs are designed to offer accumulated economic benefits to customers who purchase the product in the near future (U ncles, Dowling and Hammond, 2003). Customers typically accumulate points over a period of time, which he/she may consequently exchange for free products or rewards such as air miles ( Dowling and Uncles, 1997; Sharp and Sharp, 1997) . Consequently, these pro grams encourage consumers to shift their purchase decisions to a multi - period framework rather than focus on single - period decisions (Lewis, 2004). Over the last decade, loyalty programs have assumed a critical role in customer relationship management (CRM ), thereby rendering these strategies critical for firm management in initiating and maintaining relationships, motivating product and service usage, and retaining customer base (Musalem and Joshi, 2009). Acquiring a new customer base is few folds more ex pensive than customer retention ( Blattberg, Getz and Thomas, 2001; Griffin and Lowenstein, 2001; Thomas, Blattberg and Fox, 2004 ) . Moreover, on average, existing customers spend significantly more than a new customer ( Zeithaml, Rust, Lemon, 2001). Thus, on e of the important factors driving marketing strategists to implement loyalty programs is retaining existing customer base ( Lee, Lee, Feick, 2001) . A report published by Colloquy reiterates the fact by indicating that American businesses distribute approxi mately $48 billion worth of perceived value in reward points and miles annually; surprisingly only two - third of these points are redeemed by consumers. Thus, significant portion stomers do not get any additional benefit from buying business to which they are loyal ( Keh and Lee, 2006) . Another report by Colloquy indicates that the average U.S. household has joined 14.1 loyalty and rewards programs; however, they actively operate on ly 6.2 of them. Thus, market strategists 101 realize that loyalty and rewards programs have the potential to spark business growth. However, the business that needs to be addressed to increase the effectiveness of loyalty programs ( McCall and Voorhees, 2010) . Customers express the need for loyalty programs that are relevant and customized based on individual consumer preference structure ( Kivetz and Simonson, 2003). Fir m management and consumer enthusiasm for loyalty programs (LP) has been echoed in the marketing literature. In particular, scholars have examined effectiveness of loyalty programs in changing consumer purchase pattern. - gradient hy Kivetz, Urminsky, and Zheng (2006) demonstrates consumers expend more effort as they approach a reward. Moreover, consumers are more likely to have high perception regarding the LP if they experience an idiosyncratic fit with the program offeri ngs (Kivetz and Simonson, 2003). Additionally, LPs with high requirements tend to shift consumer preference towards luxury rewards as compared to necessity rewards (Kivetz and Simonson, 2002). Studies demonstrate that loyalty program in conjunction with m arketing instruments such as shipping fees , e - mail coupons etc. aids in customer retention ( Verhoef, 2003; Lewis, 2004). Studies have also focused on the economic aspects of loyalty programs (Shugan, 2005). Retail firms with high assortment homogeneity and product offerings characterized by high purchase frequency are ; Leenheer and Bijmolt, 2008). Adoption of successful loyalty program helps the firm to build a s trong customer base that tends to discount negative evaluations of the company relative to its competitors (Bolton, Kannan and Bramlett, 2000). Kim, Shi and Srinivasan (2001) demonstrates ffering the incentives for 102 customers. Moreover, since firms gain less from undercutting their prices, equilibrium prices go Adoption of LP is a critical compo nent of customer relation management that not only 1997). Moreover, LPs are designed to decrease consumer defection rate and build a loyal customer base ( Zhang e t al., 2000 market - based assets may dampen uncertainty in future cash flow that may lead to a decrease in firm risk (Rego, Billett and Morgan, 2009) and improvement in firm valuation (Srivastava, R eibstein, and Joshi, 2006). Thus, it may be interesting to of loyalty program on firm risk. - customers into the LP (Keiningham et al. , 2005), market entry position. adoption of loyalty program and first mover advantage. An important question is whether the pioneerin g firm who is first to adopt loyalty program in the industry has relative advantage vis - à - vis other competitors in the industry who are yet to adopt similar corporate strategy. An underlying assumption to an effective loyalty program is that the offerings match challenging for a firm with high market share since it typically has a diverse customer base with differentiated preference structure. On the other hand, a firm with low market share usually has a niche customer base. Thus, tailoring its loyalty program to satisfy the requirements of its clientele may not be an impossible task. A critical question is whether the relation between 103 Our study makes the following contributions to the extant literature. We demonstrate that adoption of next three years vis - à - vis its exposure to idiosyncratic risk prior to the launch of the program. Firms with high market share often adopt loyalty programs to retain th eir current customer base and prevent market share erosion. We empirically illustrate that such a defensive marketing strategy is likely to hurt firm performance, as indicated by drop in sales. Interestingly, when a low market share firm adopts loyalty pro gram as an offensive strategy to improve upon its customer base, it boots sales. The article is organized as follows. Section 2 deals with literature review and hypothesis development. Section 3 provides details of the measures of idiosyncratic risk and firm performance. Section 4 reports the data collection and sample selection procedure used in the analysis. Finally, section 5 provides the results of the analysis followed by a discussion section. 104 3 .2 Hypothesis Development 3.2.1 Does Adoption of lo yalty programs lowers firm risk? Valuation of a firm is determined by the present value of expected future cash flows (Kaplan and Ruback, 1995). Financial managers may improve firm valuation either by increasing expected future cash flows or by reducing uncertainty of the cash flows, which translates into a lower discount rate and firm risk. If one compares two firms with identical cash flows but differ th e firm with higher risk structure. Thus, former has higher firm valuation than the later, even though the level of cash flow is identical for both the firms (Rego, Billett and Morgan, 2009). - based asset s such as brand, patents, trademarks etc., and risk is central to the relation between marketing and firm performance (Madden, Fehle and Fournier, 2006). This is because if such an investment helps to alleviate risk and reduce uncertainty in future cash fl ow, it improves firm value. Fornell et al. (2006) posits that by investing in superior market based assets, managers may be able to simultaneously investment on con to risk. In particular, the authors demonstrate that even though CBBE has significant risk - reducing effect on both idiosyncratic as well as systematic risk, its impact on alleviatin g the former is stronger than shielding the firm from economy - level shocks. Low, 2009) . While systematic risk o 105 information regarding broad market changes (e.g., unemployment, natural disaster etc.) that are common to all stocks, idiosyncratic risk - systematic risk has been accounted for (Fu, 2009). Sinc e, by definition, idiosyncratic risk is unique to a specific firm, it is a diversifiable risk (Lee and Faff, 2009), whereas systematic risk involves uncertainty inherent to the entire market and hence is non - diversifiable. - Clara, 2003). Studies indicate that firm specific idiosyncratic risk has profound influence in stock market performance, rendering i t as an important factor for the managers as well as the investors. Thus, in the presence of transaction traditionally lay greater emphasis in managing unsystemat ic risk (Brown and Kapadia, 2007). idiosyncratic risk and economic and financial factors such as firm profitability and investment decisions (Wei and Zhang 2006; Panousi an d Papanikolaou, 2012 ), corporate governance (Ferreira and Laux 2007), institutional holding (Xu and Malkiel 2003), consumer word of mouth and its impact on stoc k market returns, generation of systematic and unsystematic risk. The study focuses on direct consumer (i.e., direct - to - consumer advertising, DTCA) and physician (i.e., direct - to - physician, DTP) advertising expenditure incurred by pharmaceutical firms. Fin dings indicate that investors regard expenditure on DTCA as value enhancing as reflected in upward movement of stock prices. Interestingly, results also indicate that such marketing activities generate higher idiosyncratic risk. In contrast, DTP marketing activities have relatively modest impact on stock returns and idiosyncratic risk. Luo and Bhattacharya (2009) examines whether 106 Empirical results indicate that even though there is a n CSP and enhance its idiosyncratic risk. oneered by the airline industry in the 1980s, LP has penetrated virtually all industries, ranging from retail to hospitality, department stores to specialty stores, and entertainment to communications. With growth of the internet, LP has also captured the online shopping market ( Keegan 2010 , Wong 2011 ). Firms embrace LPs and invest billions of dollars in their implementation and maintenance for multitude of reasons (Nunes & Dréze, 2006). First, LPs helps to reduce customer defection (Keiningham et al., 2005 - cost fallacy locks them into the LP and ensures continued interaction in the future even though th e t may even entice customers to buy more than they originally intended. Thus, LPs help firms to win a greater share of Furthermore, with strong customer patronage, LPs may decrease variability in ironmental shocks, such as negative press coverage, product recall etc. Thus, establishing a strong customer base helps the firm to reduce uncertainty in cash flow and alleviate risk (Kumar and Shah, 2015). Firms often utilize LP as a tool to obtain custo mer specific data which might yield valuable insight into customer behavior and purchase pattern. In particular, marketing managers 107 volume and frequency of purch ase. Consumer specific information may be utilized to determine customer segmentation and relative effectiveness of marketing strategies across segment. Additionally, insider information may help the firm to reduce overall marketing and promotional costs, and maximize effectiveness of marketing communication mix. Furthermore, marketing managers may even use insider information to establish special bond with the customers, which goes beyond offering just economic or functional value of the product or servic e. It may help the firm to establish a relationship of trust and commitment with its customer base. Using customer relation management ( CRM), one may observe friendshi Sheth and Parvatiyar, 1995 ). cks specific to the firm (i.e. massive product recall, disruption in supply chain etc.) or the industry. firm may also be exposed to shock targeted at the broa der market due to unexpected events (i.e., hurricane Katrina, tsunamis, earthquakes etc. ), impact of which is captured by systematic risk. We hypothesize that adoption of a successful loyalty program and building a strong customer base shields the firm fro m both categories of shocks and helps the manager to manage risk better. H 1a H 1b 108 3.2.2 Loyalty Program and Firm Sales Loyalty pro grams are dynamic incentives designed to benefit consumers from cumulative purchase over time and helps the firm to retain its current customer base while attracting potential customers ( Liu, 2007 ). In other words, an optimally designed loyalty program is expected to benefit both the parties involved in the transaction (i.e. the consumers as well as the firm). Thus, adopting a successful loyalty program is an effective marketing strategy by the firm, especially in a competitive environment. Bolton, Kannan a nd Bramlett (2000) demonstrates that an effective reward program not only make customers happy and makes them believe that the program provides them good value for their money but also makes them less sensitive to any negative evaluations of the firm vis - à - vis its competitors. Thus, it helps the firm to build a loyal customer base with positive evaluations and repeat purchase intentions. In particular, these loyalty program members provide firms with a consistent source of revenue (repeat and increased purc hases) and helps in cost reduction through less promotional expenses, thereby elevating profit. However, building a loyal customer base takes time and resources ( Taylor and Neslin, 2005). Furthermore it is a learning process for the firm to be able to off er a loyalty program that matches customer requirements as well as satisfies their corporate goals. There are numerous instances of firms re - launching loyalty programs, each time with minor revisions that better suits customer needs and helps them to fulfi ll their organizational goals (Nunes and Drèze, 2006). For - launching their loyalty programs, each time with incremental changes in their program offering packet. Thus we posit that launching of loyalty pr ogram may not have a significant impact on firm sales in the immediate future . However, it helps the firm to improve sales in the long run (Figure 1). 109 H 2 : Launch of LP has positive impact on sales in the long run. 3.2.3 Incumbent Effect In this section we investigate whether firms who are the pioneers in adopting loyalty programs in their respective industry enjoy first mover advantage. Extant literature indicates that first - mover advantage depends on certain demand - related inertial advantage and supply - related efficiency advantages (Mueller, 1997). Specifically, first - mover advantage is significant in industries where products are associated with high set - up and switching costs, product with high network externalities or high dollar value of transaction s ( Kerin, Varadarajan and Peterson, 1992 ) costs, or seller - and value of a credit card is directly proportional to the number of stores, restaurants, etc., which accept it, which in turn is a function of the number of possible customers who also use similar uncertainty regarding product quality is a demand - related factor th at may actually prove to be disadvantageous for the first mover in the industry. On the other hand, network externalities, economies of scale, set - up and sunk costs are some of the supply related factors typically enjoyed by the pioneering firm in the indu stry. For example, a firm that develops a new product may be able to establish a contractual relationship with suppliers of important inputs. Literature indicates that pioneers firms have different skill sets and resources at their disposal relative to th e early adopters and late entrants. Robinson, Fornell and Sullivan (1992) investigates whether successful market pioneers necessarily have access to superior skills and 110 resources. Findings suggest that market pioneers are not necessarily stronger and have access to superior skills. However, skills and resource profiles of market pioneers vary significantly from that of early followers and late entrants ( Lieberman and Montgomery, 1988 ). However, studies indicate that first mover advantage is contingent on in dustry and product characteristics ( Lieberman and Montgomery, 1998) . In similar vein, studies indicate that the pioneer firm offering loyalty program enjoys distinct advantage over the other firms in the industry (Van Osselaer, Alba and Manchanda , 2003). H owever, the relation may be moderated by loyalty program offerings and its pricing structure. We hypothesize that the pioneer firm who is the first to launch a loyalty program in the industry enjoys a significant improvement in sales vis - à - vis its competit ors who are yet to adopt such a marketing strategy. H 3 : Pioneering firms who are first to launch loyalty programs in their industry may experience improvement in sales relative to other firms in the industry. 3.2.4 Does market share moderate the relatio nship? Fornell (1992) and Griffin and Hauser (1993) indicated the possibility of a negative association between customer satisfaction and market share. Authors posit that a firm with small market - share may target niche customers and address their needs, thereby resulting in high preferences further away from the firm's targe t market, the overall level of customer satisfaction is likely to fall. 111 programs. A small market share firm serving a niche customer base may be able to tailor its loyalty program offerings to address the needs of its customer. Loyalty program members may believe that the program provides them good value for their money. This in turn may translate into higher customer satisfaction (Bolton, Kannan and Bramlett, 2000) and cus tomer retention (Lewis, 2004). Thus, an effective loyalty program not only makes the customers happy but also helps the firm to improve its sales and performance through customer retention and by attracting potential customers. Thus, one may conclude that small market share firm utilizes announcements and adopting launching loyalty programs as an offensive marketing strategy to improve its customer base and gain market share in the long run (Hauser and Shugan, 1983). On the contrary, a high - market share fi rm with large and diverse customer base may offer a generic loyalty program that addresses overall customer needs. However, "one size fits all" The firm may offer multiple loyalty programs targeted at its multiple customer segments. However, such a differentiated marketing approach may not necessarily translate into higher egment Lehmann, 1994). Additionally, such a differentiated marketing strategy may not be very cost effective. Thus, we posit that for firms with high market share, either strategy (one generic loyalty program vs. multiple loyalty program targeted at multiple customer segments) may not necessarily translate into high firm performance. 112 H 4 programs on its sales to the extent that firms with high market share will experience loss in sales from launching of the loyalty program. 3 .3 Methodology 3.3.1 Measures of Idiosyncratic Risk total risk may be measured by standard deviation of r eturns. In particular, it is given by where is the on day and is the risk free rate based the Fama and French four facto risk may be decomposed into Systematic Risk (or market risk) and Firm - Specific Risk (or idiosyncratic risk). We measure idiosyncratic risk of an individual stock using the Fama and French (1993) three - factor mo del expanded with the Carhart (1997) momentum factor: (10) where is the subscript for the day and is the subscript for the month, and , , and are factor sensitivities or loadings. is the day, is the return in month on a value - weighted market proxy, is the risk free return and is measured by in month of a one - month treasury bill. Daily stock returns are obtained from the Center for Research in Security Prices (CRS P). We regress daily excess returns of individual stocks on the following four factors: (i) the excess return based on a market portfolio , (ii) the difference in return between a portfolio of small stocks and t hat of a 113 portfolio of large stocks , (iii) the difference in return between a portfolio of high book - to - market stocks and that of a portfolio of low book - to - market stocks , and (iv) the difference in return betw een a portfolio of long on past one - year winners and that of short on past one - year losers . The residual ( ) of the model is a measure of firm - idiosyncratic excess return (Ang et al. 2006; Cao, Simin, and Zhao 2 008). Following Lou and Bhattacharya (2009), we assume that , where is a normal random variable with mean and variance . Thus, presence of serial correlation is eviden measured by the variance of the residuals over the time period under consideration. Specifically, in the year, idiosyncratic risk may be expressed as where denotes the number of days (i.e., 252) over which the model is estimated. 3.3.2 Measures of Firm Performance We employed multivariate regression analysis with performance indicator as the dependent vari ables and launching of loyalty programs as the independent variable. We used sales ( ) as indicators of firm performance . We used firm size and leverage as control variables (equation 2). Consistent with Leenheer and Bijmolt (2008), we include customer (11) 114 Loyalty Program ( ) is an indicator variable that takes the value of 1 if the firm has adopted a loyalty program in period else equals. Incumbent ( ) is an indicator variable that takes the value of 1 if the firm is the pioneering firm in the industry to launch the loyalty program els e equals 0. Market Share ( ): It is the ratio of to total industry sales in that period. It is an indicator of relative competitiveness of the firm in the industry. Consumer Satisfaction index ( ): It is economic indicator that measures the satisfaction of the U.S. consumers for product and/or service offerings of the firm in period . Size: We use log was obtained from the Compustat database. Leverage ( asset information was obtained from the Compustat database. Retail dummy ( ): this is an indicator variable that takes the value of unity if the firm under consideration belongs to the retail industry; else it takes the value of zero. Hospitality dummy ( ): this is an indicator variable that takes the value of unity if the firm under consideration belongs to the hospitality industry; else it takes the value of zero. 115 3 .4 Data and Measurement Variables Current analysis includes four sectors, na mely: retail, hospitality, telecommunication and information, and entertainment. Consistent with ASCI convention, we include department and discount stores (SIC: 5651, 5311), specialty retail stores (SIC: 5700, 5940), drug stores (SIC: 5912), and super mar 7011, 6794), restaurants (SIC: 5812, 6794), airlines (SIC: 4512), internet travel (SIC: 4700) as ar phone (SIC: 4812), and subsciption TV/ Cable (SIC: 4841). Finally, we include amusement and theme parks consolidated list of publicly traded firms corresponding to the list of SIC given above, we acquired loyalty program related information from company websites, COLLOQUY (Colloquy.com) and LexisNexis ( LexisNexis .com). Information on firm performance measures (i.e., sales, debt - to - asset ratio, total asset) are obtained f rom COMPUSTAT. Finally, we obtain information on customer satisfaction index from American Customer Satisfaction Index ( ACSI ) . Thus, we constructed a panel data set that contains financial as well as loyalty program information of the publicly traded firms in the four sectors identified in the study (i.e., retail, hospitality, telecommunication and information, and entertainment) from 1980 to 2013. 116 3 .5 Results 3.5.1 Descriptive Statistics Table 11 gives the summary statistics of firms included in our stu dy. number of firms in the respective sector that has launched loyalty programs since 1980. In particular, there are approximately 228, 75, 23 and 10 firms that have adopted loyalty programs in the retail, hospitality, telecommunicatio ns and information and entertainment sectors provides us with information regarding the number of firms in of these four sectors who are yet to adopt loyalty programs in the corresponding time period . Table 12 provides the descriptive statistics of the financial variables used in the current study. launching loyalty program and its market share (p>.001), sales to asse t ratio (p>.05) and size (p>.001). Estimates also indicate that firms with high market share tend to have positive association with consumer satisfaction (p>.05), sales to asset ratio (p>.001), and size (p>.001). 3.5.2 Loyalty programs and Firm risk In t to risk. We constructed a panel data set that contains financial information of the firm three years prior to and post adoption of the loyalty program. In particular, if firm adopts loyalty program in period , we consolidated a data set with firm financial information for period to . First two columns of table (13) pr ovide estimates for all three components of risk (i.e., 117 total firm risk, systematic risk and firm - specific risk) for one year before and after the launch of the program. The third column provides the difference in estimates. We measure total risk by standa (p<.001) and .3810 (p<.001) one year before and after the adoption of the loyalty program respectively. Findings also suggest that over this time period, systematic risk has been 1.0135 (p<.001) and 1.0084 (p<.001) respectively. Finally, estimates indicate that firm - specific risk has been .3521 (p<.001), .3285 (p<.001) one year pre and post adoption of the loyalty program respectively. Next, empirical estimates indic ate that the difference in overall firm - risk and firm - specific risk over the time period is - .0210 (p<.1) and - .0236 (p<.001) respectively. Columns four and five of table (13) provide estimates of firm risk for three years before and after the launch of th e program. The sixth column provides the difference in estimates. years before and after the adoption of the loyalty program respectively. Findings also suggest t hat over this time period, systematic risk has been 1.0590 (p<.001) and 1.0339 (p<.001) respectively. Finally, estimates indicate that firm - specific risk has been .3950 (p<.001), .3388 (p<.001) three years pre and post adoption of the loyalty program respe ctively. Finally, empirical estimates indicate that the difference in overall firm - risk and firm - specific risk over the time period is - .0636 (p<.001) and - .0561 (p<.001) respectively. eviates firm risk in both the one year and three years interval. Interestingly, we do not observe any significant difference both the time intervals. Thus, findi 118 of loyalty program reduces idiosyncratic risk. However, findings do not support hypothesis program and impact on sales year, three years and five years after the launch date. We use sales as the measures of firm performance. As mentioned earlier, loyalty program is an indicator variable that takes the value of unity (i.e., ) if the firm has launched a loyalty program; else it takes the value of 0 (i.e., ). We include prior year change in sales as a control in the regression analysis since firms may launch LPs in light of declining sales. We also include firm size measured by log of total assets and consumer satisfaction index to control for firm characteristics. Effectiveness of LP varies across sectors. In parti cular, even though the emphasis on LP in the retail sector is minimal, it is exponentially significant in the hospitality industry. Thus, we included sector dummy in the analysis to control for the imbalance in emphasis on LP across sectors. Parameter est improvement in sales. Interestingly, it registers a boost in sales by 4893.66 (p<.1) and 9905.70 (p<.05) units after three and five years of launching the program respectively. Thus, consistent positive impact on sales. Estimates indicate that pioneering firms who were among the first to launch loyalty programs in the respective industry experience improvement in sales by 6373.09 units (p<.1) three years after the launch date. Interestingly, we do not observe any significant impact of loyalty programs 119 on sales one year and five years after the launch date. Thus, findings partially confirm hypothesis program. (p<.001) and 76837.00 (p<.001) one year, three years and f ive years after the launch date respectively. Consistent with the literature, findings indicate a positive association between market share and sales (Szymanski, Bharadwaj, and Vara - darajan, 1993). Interestingly, market share of firms who have launched loy alty programs diminishes sales by 32933.00 (p<.05), 44187.00 (p<.001) and 52270.00 (p<.001) one year, three years and five years after the launch date respectively. This suggests that adoption of loyalty program hurts firms with high market share in the sh ort and long run. Thus, findings confirm hypothesis H4 that market share the three periods under consideration. Findings suggest that in the hospitality sector, loyalty programs hurts sales by 9417.10 units (p<.05), 8068.89 units (p<.05) and 11008.00 units (p<.05) one year, three and five years after the launch year respectively. However, we do not observe any significant impact on sales after the of LP in the retail sector. 120 3 .6 Post - Hoc Analysis : Response Surface Approach Analysis in the previous sections indicates the presence of possible non - linear relation between firm characteristics and adoption of loyalty program. Thus, post hoc, we condu ct an is a portfolio of mathematical methods that helps to develop, improve, and optimize processes in which a response of interest is contingent on several independ ent factors and the objective is to design, development and formulation of new products, as well as in the improvement of existing Mittal and Kamakur a , 2001; Kim and Hsieh, 2003). In particular, it characterizes the impact of the independent variables, alone or in combination, on the response of interest. is given in e quation (3) : (12) where is the response, is the unknown function of response, , , and are the independent variabl es and finally is the statistical error that represents other sources of variability, such as measurement error, that has not been accounted for in the analysis. It is generally assumed that follows normal dist ribution with mean zero and variance. Panel A of table (15) indicates that the quadratic model fits the data very well, suggesting a non - 121 Additionally, R - square is 0.13 59 , which indicates 13.59% of variability explained by the fitted model. Panel B of table (15) provides the estimates of the coefficient of response surface analysis. Estimates indicate that market share is not significant in the analysis of variance for the model. Findings also indicate that quadratic and interaction terms are significantly important, with the exception of interaction between consumer satisfaction index and market share. 122 3 .7 Discussion Key Findings and Theoretical Implications : Loyalt y programs (LPs) are dynamic incentive programs designed to benefit consumers from cumulative purchase over time. Despite is known about the its valuation in the long run. Studies indicate that not all LP are equally successful and some fail to generate the expected stream of revenue for the firm (Reinartz and V. Kumar, 2002; Shugan, 2005). Thus, from firm management perspective, it is important to analyze whether launching a idiosyncratic risk after the firm adopts the program. We empirically demonstra adoption of loyalty program depletes risk. In particular, estimates indicate that upon adoption of the loyalty program, firm specific risk drops steadily and significantly. Firm management realizes that loyalty programs or frequent shopper p rograms involve generated due to discounts may be negated by overwhelming increase in sales to new and existing customers over a period of time (Lal and Bell, 2 003). A critical question is whether firms experience a drop in performance within a year of adopting the LP and whether it is able to recuperate its losses and is able to successfully enhance firm value in the long run. Third, we analyze if firms enjoy f irst mover advantage by being the pioneer in the industry to adopt the loyalty program. In particular, we examine whether pioneer firms enjoy a distinct advantage over the rival firms in the industry by examining short and long term firm performance of the 123 loyalty program depletes sales in the short run. However, it is able to recuperate its loss in the long run. We observe improvement in sales three years after launching the program . Results also and sales. In particular, adoption of loyalty programs by firms with high market share hurts sales. Additionally, estimates suggest that first - m over advantage is limited to the first year after adopting the loyalty program. Finally, we conduct a post - hoc analysis to investigate probable non - linear relationship lly, we sales to asset ratio, consumer satisfaction index and market share that makes launching of loyalty program a feasible solution for the firm. Findings c onfirm existence of non - linear relationship. Managerial Implications : It has been well established in the literature that loyalty programs are risky marketing strategies. A report published by Colloquy reiterates the fact by indicating that Ame rican businesses distribute approximately $48 billion worth of perceived value in reward points and miles annually; surprisingly only two - third of these points are redeemed by consumers. Thus, one may conclude that either the consumers are unaware of the b enefits offered by the loyalty program or their requirements are not addressed by the program offerings. Current study reiterates the statistics through empirical findings. Results suggest that successful adoption of loyalty programs require managers to c raft programs whose offerings matches closely with that of the expectations and requirements of the consumers. However, this may be a challenging requirement for managers, especially for those of big firms. In particular, firms with high market share typic ally cater to a consumer base with diverse preference structure. Thus, tweaking program offerings to meet the requirements of each consumer segment 124 may be a daunting task almost impossible to achieve. On the other hand, managers of small firms may be able to customize program offerings to satisfy the requirements of their niche customer base, and in the pr ocess yield positive returns. Limitations and Future Research : Heerde and Bijmolt (2005) investigates the differential impact of communication mode of loy alty programs (i.e. direct mail to loyalty program members only vs. door - to - door flyers to its entire customer base) across its customer base. We would like to extend the study by analyzing wthere effectiveness of loyalty programs is contingent on communic ation channel and characteristics of the loyalty program offerings. In particular, we would like to examine the effectiveness of corporate websites, direct contact with customers through emails, word - of - mouth, point - of - sale information, direct mail, dedica ted club sites, SMS text messages and social network as effective marketing channels impacting success of the launching program. Furthermore, firms differ in their loyalty program offerings. Some programs are built on tier system to reward initial loyalty, where as some charge an initial fee to receive benefits ( Zeithaml, Rust and Lemon, 2001 ). Some firms even structure non - monetary programs around their customer's values while some opt to partner with another company to provide all - inclusive offers to its customers. Additionally, we would like to investigate whether reward program characteristics are critical factors driving loyalty program success. 1 25 A PPENDIX 126 Table 11 : Sample Breakdown by Industry 127 Table 12 : Descriptive Statistics 128 Table 13 : Firm's Exposure to Risk upon Launching of Loyalty Programs 129 Table 14 : Impact of Adoption of Loyalty Program on Firm Sales 130 Table 15 : Analysis of Results Based on Response Surface Approach 131 Figure 2 : Conceptual Model 132 Figure 3 : Ridge of Maximum 133 Figure 4 : Rotated Surface Plot 134 BIBLIOGRAPHY 135 BIBLIOGRAPHY Anderson, Eugene W., Claes Fornell, and Donald R. Lehmann. "Customer satisfaction, market share, and profitability: Findings from Sweden ." The Journal of Marketing (1994): 53 - 66. Ang, Andrew, section of - 299. methodology." Journal of Food Engineering 78.3 (2007): 836 - 845 . Bijmolt, Tammo HA, Harald J. van Heerde, and Rik GM Pieters. "New empirical generalizations on the determinants of price elasticity." Journal of marketing research 42.2 (2005): 141 - 156. Blattberg, Robert C., Gary Getz, and Jacquelyn S. Thomas (2001), C ustomer Equity: Building and Managing Relationships as Valuable Assets . Boston: Harvard Business School Press. Implications of Loyalty Program Membership and Service Experiences for Customer Re tention and Value, Journal of the Academy of Marketing Science ; Winter2000, Vol. 28 Issue 1, pp 95 - 108. - specific risk and equity market 388. Cao, Charles, Ti mothy Simin, and Jing Zhao. "Can growth options explain the trend in idiosyncratic risk?." Review of financial studies 21.6 (2008): 2599 - 2633. Campbell, J. Y., Lettau, M., Malkiel, B. G., & Xu, Y. (2001). Have individual stocks become more volatile? An em pirical exploration of idiosyncratic risk. The Journal of Finance , 56 (1), 1 - 43. Carhart, M., 1997. On persistence in mutual fund performance. Journal of Finance 52, 57 82. Dowling, Grahame R. and Mark Uncles (1997), "Do Customer Loyalty Programs Really W ork?" Sloan Management Review, 38 (Summer), 71 - 82. Fama, Eugene F., and Kenneth R. French. "Common risk factors in the returns on stocks and bonds." Journal of financial economics 33.1 (1993): 3 - 56. Ferreira, Miguel A. and Paul A. Corporate Governance, Idiosyncratic Risk, and Information Flow, The Journal of Finance, Vol. 62, No. 2, 951 - 989. Fornell, Claes. "A national customer satisfaction barometer: The Swedish experience." the Journal of Marketing (1992): 6 - 21. 136 Griffin, Abbie, and Joh n R. Hauser. "The voice of the customer." Marketing science 12.1 (1993): 1 - 27. Goyal, Amit and Santa - Idiosyncratic Risk Matters!, Finance, Vol. 58, No. 3 (Jun., 2003), pp. 975 - 1007. Griffin, Jill, and Michael W. Low enstein. "Customer winback." San Francisco (2001). Hauser, John R., and Steven M. Shugan. "Defensive marketing strategies." Marketing Science 2.4 (1983): 319 - 360. Kaplan, Steven N., and Richard S. Ruback. "The valuation of cash flow forecasts: An empiric al analysis." The Journal of Finance 50.4 (1995): 1059 - 1093. Keh , Hean Tat and Lee, Yih Hwai ( 2006 Do reward programs build loyalty for services? The moderating effect of satisfaction on type and timing of rewards Journal of Retailing 82 (2), 127 136 . Keiningham, T. L., Vavra, T. G., Aksoy, L., & Wallard, H. (2005). Loyalty myths: hyped strategies that will put you out of business -- and proven tactics that really work . John Wiley & Sons. Kerin, Roger A., P. Rajan Varadarajan and Robert A. Peterson (1 - Mover Advantage: No. 4 (Oct., 1992), pp. 33 - 52. Kim, Byung - ing Science Kim, Stephen Keysuk, and Ping - Hung Hsieh. "Interdependence and its consequences in distributor - supplier relationships: a distributor perspective through response surface approach." Journal of Marketing Research 40.1 (2003): 101 - 112. Kivetz, Ran and Itamar Research, Vol. 39, No. 2 (May, 2002), pp. 155 - 170. Kivetz, Ran and Itamar Simonson (2003 The Idiosyncratic Fit Heuristic: Effort Advantage as a Determinant of Consumer Response to Loyalty Programs Vol. 40, No. 4 (May, 2002), pp. 454 - 467. - Gradient Hypothesis Journal of Marketing Research, Vol. 43 , No. 1 , pp. 39 - 58 . 137 Klemperer, Paul. "Markets with consumer switching costs." The quarterly journal of economics (198 7): 375 - 394. profitable customer loyalty for the 330 Handbook of Research on Customer Equity in Marketing , Busi ness & Economics Lee, Jonathan, Janghyuk Lee, Lawrence Feick, (2001) "The impact of switching costs on the customer satisfaction - loyalty link: mobile phone service in France", Journal of Services Marketing, Vol. 15 Iss: 1, pp.35 48 Lee, Darren D., and Robert W. Faff. "Corporate sustainability performance and idiosyncratic risk: A global perspective." Financial Review 44.2 (2009): 213 - 237. programs really enhan ce behavioral loyalty? An empirical analysis accounting for self - selecting Leenheer, Jorna, and Tammo HA Bijmolt (2008) "Which retailers adopt a loyalty program? An empirical study." Journal of Reta iling and Consumer Services 15.6: 429 - 442. - Term Promotions on - 292. Lieberman, Marvin B. and David First - mover (dis) advantages : Retrospective and link with the resource - based view, Strategic Management Journal. Dec1998, Vol. 19 Issue 12, pp 1111 1125. - Term Impact of Loyalty Programs on Consumer Purc hase - 35 Liu, Yuping and Competing Loyalty Programs: Impact of Market Saturation, Market Share, and Category Expandability, Journal of Marketing, Vol. 73 (Ja nuary 2009), 93 108. - taking behavior and equity - of Financial Economics, Volume 92, Issue 3, June 2009, Pages 470 490 Luo, Xueming. "Consumer negative voice and firm - idiosyncratic stock retur ns." Journal of Marketing 71.3 (2007): 75 - 88. The Debate over Doing Good: Corporate Social Performance, Strategic Marketing Levers, and Firm - Idiosyncratic Risk, Journal of Marketing, 73 (6), 198 213. 138 Mägi, Ann e W. "Share of wallet in retailing: the effects of customer satisfaction, loyalty cards and shopper characteristics." Journal of Retailing 79.2 (2003): 97 - 106. Building Brand Community, Journal of Marketing, Vol. 66, No. 1 (Jan., 2002), pp. 38 - 54 Organizing Framework and Research Agenda, Cornell Hospitality Quarterly February 2010 vol. 51 no. 1 35 - 52. Mittal, Vikas, and Wagner A. Kamakura. "Satisfaction, repurchase intent, and repurchase behavior: Investigating the moderating effect of customer characteristics." Journal of marketing research 38.1 (2001): 131 - 142. Mueller, Dennis C. "First - mover advant ages and path dependence." International Journal of Industrial Organization 15.6 (1997): 827 - 850. Research Note How Much Should You Invest in Each Customer Relationship? A Competitive Strategic Approach, Market ing Science May/June 2009 vol. 28 no. 3 555 - 565 Nunes, Joseph C., and Xavier Drèze. "Your loyalty program is betraying you." Harvard Business Review 84.4 (2006): 124. Do Rewards Really Create Loyalty?, Harvard B usiness Review, 73 (May/June), pp 75 - 82. Why Do Firms Invest in Consumer Advertising with Limited Sales Response? A Shareholder Perspective, Journal of Marketing, Vol. 75 (January 109 2011), 109 124. Panousi, Vasia, and Dimitris Papanikolaou. "Investment, idiosyncratic risk, and ownership." The Journal of Finance 67.3 (2012): 1113 - 1148. Reinartz, Werner and V. Kumar The Mismanagent of Customer Loyalty, Harvard Business Review. Jul2002, Vol. 80 Issue 7, p86 - 94. - 624. Sharp, Byron and An ne Loyalty programs and their impact on repeat - purchase loyalty Patterns, Intern. J. of Research in Marketing 14 (1997) 473 - 486. 139 Sheth, Jagdish N., and Atul Parvatiyar. "The evolution of relationship marketing." International Business Revie w 4.4 (1995): 397 - 418. Shugan, Steven M. Brand Loyalty Programs: Are They Shams?, Marketing Science Spring 2005 vol. 24 no. 2 185 - 193 Sirdeshmukh, Deepak, Jagdip Singh and Barry Sabol, (2002), Consumer Trust, Value, and Loyalty in Relational Exchanges, Journal of Marketing, Stauss, Bernd, Maxie Schmidt and Andreas Schoeler (2005) Customer frustration in loyalty programs, International Journal of Service Industry Management, Vol. 16 No. 3, pp. 229 - 252. Szymanski, David M., Sundar G. Bharadwaj, and P. Rajan Varadarajan. "An analysis of the market share - profitability relationship." The Journal of Marketing (1993): 1 - 18. Thomas, Jacque - 45 Uncles, Mark D., Grahame R. Dowling, and Kathy Hammond. "Customer loyalty and customer loyalty program s." Journal of consumer marketing 20.4 (2003): 294 - 316. Van Osselaer, Stijn M. J. , Joseph W. Alba and Puneet Manchanda - 274. Venkatesh, Viswan ath, and Sandeep Goyal. "Expectation disconfirmation and technology adoption: polynomial modeling and response surface analysis." MIS quarterly 34.2 (2010): 281 - 303. Verhoef, Ef 67, No. 4, 30 - 45. Wagner, Tillmann, Thorsten Hennig - Thurau, Thomas Rudolph (2009). Does Customer Demotion Jeopardize Loyalty? Journal of Marketing: Vol. 73, No. 3, pp . 69 - 85. Wei, Steven X., and Chu Zhang. "Why Did Individual Stocks Become More Volatile?*." The Journal of Business 79.1 (2006): 259 - 292. Management Review, Vol. 43, No . 4, 89 - 105. Xu, Yexiao, and Burton G. Malkiel. "Investigating the behavior of idiosyncratic volatility*." The Journal of Business 76.4 (2003): 613 - 645. Yi, Youjae and Hoseong Jeon (2003) Effects of Loyalty Programs on Value Perception, Program Loyalty, and Brand Loyalty, Journal of the Academy of Marketing Science 140 The Customer Pyramid: Creating and Serving Profitable Customers, 43 Issue 4, p118