Go With the Flow:
THE EFFECT OF PRO- AND COUNTERCYCLICAL ADVERTISING
AND PRICING STRATEGIES ON THE RESILIENCE OF CPG BRAND
1
Go With the Flow: The Effect of Pro- and
Countercyclical Advertising and Pricing Strategies
on the Resilience of CPG Brands
General:
Author: Geert van Greunsven Department: Marketing
Qualification: Master thesis Completion date: January 13th 2020
Contact information of the author:
Address: Gedempte Zuiderdiep 86a, Groningen Phone number: +31 654900698
Email: gvangreunsven@gmail.com Student number: S2374757
Supervisors:
2
MANAGEMENT SUMMARY
The impact of economic downturns on brand can be profound. Gulati, Nohria and Wohlgezogen (2010) report that of the nearly 5000 companies they studied, 17% of the companies did not survive the recession and almost 80% of the companies did not fully recover in the next three years. In recent years, the interest of academics, managers, and politicians towards creating organizations that are more resilient increased. Resilient organizations are able to cope with such adverse and unpredictable events as economic downturns (Van der Vegt et al. 2015). Academic research that guides managers in their search to make their organizations more resilient focusses mainly on measures that are difficult to accomplish or time costly such as changing the management structure or the corporate culture. When the economic downturn kicks in, managers have make decisions fast. Managers are under increasing pressure to improve the effectiveness and accountability of their investments (Van Heerde et al. 2013). Modifying marketing instruments could help managers to make their organizations more resilient toward the adverse impact of economic downturns. Marketing budgets allow for fast adjustments, and most managers already admit to modify their marketing investment in reaction to the overall economy (Shama, 1993). However, academic literature not yet describes how marketing investments can increase organizational resilience.
Therefore, this research starts by exploring the effects of the marketing investments price and advertising on the resilience of organizations. Three different measures were used for the aspects of resilience: downward risk, upward potential and steadiness. The downward risk refers to the magnitude of the impact of the adverse event (Van der Vegt et al. 2015). Organizations with a high downward risk are affected more heavily by economic downturns. Upward potential refers to the ability of the organization to recover from the adverse event (Carmeli et al., 2013). Organizations with a high upward potential restore (a part of) their sales in the subsequent expansion or even outperform their sales prior to the economic downturn. Steady organization do not experience any impact of the contraction at all, so they do not have to cope with the crisis at all. This is also considered as resilience to economic downturns. The sales, advertising and price data of 75 mature CPG brands from the UK is used to gain insights in the way the resilience is affected by advertising and price investments. Several advanced modelling techniques are used, such as spectral filtering techniques to estimate the cyclical components, regressions (with bootstrapping), and meta-analytic tests.
4
Table of Contents
1. Introduction ... 5
2. Theoretical Framework ... 7
2.1 Resilience ... 7
2.2 The impact of business cycle fluctuations ... 8
2.3 The effectiveness of advertising and price over the business cycle... 9
2.4 The contribution of marketing to resilience ... 10
2.5 Conceptual model ... 11
3. Data ... 12
4. Methodology ... 14
4.1 Deriving the series’ cyclical components ... 14
4.2 Determine downward risk and upward potential ... 15
4.3 Determine steadiness of performance ... 16
4.4 Extract co-movement of price and advertising investments with the business cycle ... 16
4.5 Estimate the effect of price and advertising on resilience aspects ... 17
5. Model-Free insights ... 18
5.1 Sales over the business cycle ... 18
5.2 Advertising investments over the business cycle ... 18
5.3 Price over the business cycle ... 19
6. Model-Based Results ... 21
6.1 The cyclical behaviour of CPG sales ... 21
6.2 Steadiness ... 25
6.3 The cyclical behaviour of CPG advertising and price ... 27
6.4 How pricing and advertising strategies affect the resilience aspects ... 30
7. Conclusion & Discussion ... 33
7.1 Summary... 33
7.2 Managerial implications ... 34
7.3 Directions for further research ... 35
5
1. Introduction
Managers find it difficult to deal with the vast impact of economic downturns. They have to overcome immense challenges during contractions, and often not successfully. Gulati, Nohria and Wohlgezogen (2010) studied nearly 5000 companies in different industries over the past three recessions. They report that 17% of the companies did not survive the recession and almost 80% of the companies did not fully recover in the next three years. These findings suggest that managers, in general, could make better choices in the quest to protect their organization from the adverse impact of contractions. Academic research could provide a solution here, by providing managers insights in which strategies can help their company to survive the recession, and possibly to come out of the recession even stronger. Unfortunately, academic research that could advise managers on such normative actions is a growing research area. In marketing research, the interest in this topic grows. In their overview article of marketing related business cycle research over the past two decades, Dekimpe & Deleersnyder (2018) encourage researchers to explore the moderating factors on the effects of economic turbulence on firms further.
According to Van der Vegt et al. (2015) the traditional way of coping with adverse events is to develop approaches and systems to identify risks. Organizations use empirical data and models to analyse past events and predict future events. However, some events are not adequately predictable. Take for example the recent financial crisis in 2008, only a hand full of economists predicted these events. Van der Vegt et al. (2015) also state that for events that are not adequately predictable, a growing number of academics, managers, policy makers and politicians shifted their attention from identifying and mitigating risk to increasing organizational resilience. Resilient organizations are better able to withstand the adverse impact of economic downturns. They possess the ability to cope with (unexpected) adverse events better, and recover faster (Lengnick-Hall & Beck, 2009). They use a more proactive approach regarding adverse events. Resilient organizations focus on getting prepared for adverse events instead of reacting to them (McManus et al. 2008). Therefore, increasing the organizational resilience might be an effective way for managers to deal with the adverse impact of economic downturns.
6 economic downturns. In this paper, modifying advertising investments and price will be proposed as instruments to increase organizational resilience. First, these instruments are easier and faster to modify. Second, advertising and price are two instruments that firms use to compete in the product market during contractions. Therefore, it seems logical to investigate how they affect the organizational resilience.
In practice, many managers seem to already modify their marketing budgets in reaction to business cycle fluctuations (Shama, 1993). Business cycle fluctuations refer to the expansions and contractions in the cyclical pattern that the overall economy follows (Dekimpe & Deleersnyder, 2018). Managers often follow a pro-cyclical pattern with their marketing investments, thereby increasing marketing budgets during expansions and decreasing budgets during contractions. Several academic papers provide evidence that the norm for managers is to cut their marketing investments during contractions and increase them when the economy expands again (e.g. Deleersnyder et al., 2009; Özturan et al., 2014). However, recent academic literature provides evidence that a countercyclical pattern in marketing investments might lead to a better performance in certain situations (e.g. Lamey et al. 2007). The differences in performance between pro- and countercyclical marketing strategies could be caused by effects these strategies have on the organizational resilience.
Therefore, the purpose of this research is to provide managers with insights into how pro- and countercyclical advertising and pricing strategies affect the brand performance in terms of sales with regard to macro-economic recessions. More specifically, the following research question will be addresses:
- To what extent do pro- and countercyclical advertising and pricing investments affect brands’ resilience with regard to macro-economic business cycles in terms of sales?
7
2. Theoretical Framework
2.1 Resilience
Organizational resilience refers to the organization’s ability to cope with adverse events and recover from these events (Van der Vegt, Essens, Wahlstrom & George, 2015). Coping and bouncing back from adverse events is fundamental to human and organizational functioning and may be important for organizational crisis-preparedness, a high reliability, and future growth (Carmeli et al. 2013). In organizational sciences, resilience is considered as a characteristic of a system, rather than a characteristic of the system’s individual parts (Adger, 2000). To understand a system’s resilience, the system’s capability and capacities of important parts should be identified, and their interaction with each other and their environment should be examined (Van der Vegt et al. 2015), suggesting that the organization’s resilience depends on a complex interplay between various characteristics.
The current academic literature on organizational resilience is mainly focused on how (changes in) the organizations’ structure, management or culture affect their resilience (e.g. Carmeli et al. 2013; Van der Vegt et al, 2015; Amann & Jaussaud, 2012). However, these types of changes are difficult and often time costly to realize (Tellis, Prabhu & Chandy, 2009). When an economic crisis hits managers do not have the time to implement such changes. Managers are pressured to make fast decisions in order to cope with the crisis. Since managers already modify their marketing investments in economic downturns, adjusting the investments to create a more resilient organization will be investigated in this paper.
Resilience will be conceptualized along three dimensions: downward risk, upward potential and steadiness. Downward risk refers to the magnitude of the effect the adverse event has on the organization. In other words, the downward risk tells how hard the organization is hit by the adverse event. Therefore, it shows the extent to which the organization is affected by contractions. The upward potential refers to the organization’s ability to recover from adverse events. It represents the extent to which the organization thrives during expansions. Therefore, it shows the organization’s ability to recover or even outperform their original position. It is important to note that strategies to decrease the downward risk do not necessarily contribute (in the same way) to an increase in upward potential.
8 accompanied by higher cost, because organizations have to manage the variability in demand. For example, if there is less demand and the organization has produced too many products, the organization has to store the overproduction. This results in higher costs for the organization (e.g. rent for extra storage space). Additionally, Fisher et al. (2014) find that the volatility of brand performance is affected by the volatility form marketing spending. This provides evidence that the marketing instruments advertising and price affect the steadiness of brand sales. The organization’s steadiness is important, because it shows an aspect of resilience that downward risk and upward potential do not capture. A low upward potential does not necessarily affect the organization’s performance negatively if the organization also has a low downward risk. It then balances each other out. With steadiness, both aspects are incorporated in a combined measure.
2.2 The impact of business cycle fluctuations
Before the impact of business cycles fluctuations is discussed, the business cycle will be conceptualized. An often-used definition of the business cycle is the one of Burns & Mitchell (1946) who define the business cycle as follows: “A cycle consists of expansions occurring at about the same time in many economic activities, followed by similarly general recessions, contractions and revivals which merge into the expansion phase of the next cycle.” The definition of Burns and Mitchel suggests that macro-economy follows a cyclical pattern in which peaks in the overall economic activities (economic growth) are followed by troughs in the overall economic activities (economic decline) at the same time. Although these business cycle fluctuations are not clearly noticeable by themselves, they are visible across many aggregate economic series (Stock & Watson, 1999). Examples of these series are Gross Domestic Product (GDP), employment, and disposable income.
Business cycles fluctuations depend on the whole economy, but this does not mean that the whole economy is affected to the same extent or inthe same direction. For example, Deleersnyder et al. (2004) report that the sales of durables is more heavily affected by contractions and four times more sensitive to business cycle fluctuations than the general economic activity. In their research on the healthcare sector in several countries, Cleeren et al. (2016) reported an overall countercyclical growth in the private healthcare during contractions of more than two times the percentage-growth of the GDP. Therefore, it is important to also provide some insights concerning the extent to which the sales of the CPG brands investigated in this paper move in relation to the business cycle.
9 the 2001 economic crisis, over 20% of the US firms in the top quartile of performance in their industry fell to the bottom quartile of performance, and fewer than 30% that lost ground, were able to regain their position afterwards. From this, we can conclude that business cycle fluctuations (especially economic downturns) can have devastating effects on organizations and that recovery from the hit is very difficult. However, academic research concerning normative guidelines for managers on how to react to business cycle fluctuations is scarce. Therefore, providing managers with concrete and workable insights on how they can cope with such adverse events and how to recover from them should be a valuable contribution to business cycle research.
2.3 The effectiveness of advertising and price over the business cycle
Shama (1993) reports that almost all managers he surveyed modify their marketing strategies in response to contractions. This suggests that managers are aware of the impact such adverse events might have on their organizations and that they recognize the importance of reacting to these events. As response to contractions, many managers cut their marketing budgets, especially advertising. During the financial crisis, 71% of all marketing managers already cut their advertising budgets, and 77% was planning to economize further on media expenditures (Advertising Age, 2009). Academic research repeatedly reports on the managers’ tendencies to cut back their marketing investments during recessions (e.g. Özturan et al., 2014). Several papers provided evidence for the pro-cyclical behaviour of allocating advertising budgets, with decreased advertising budgets during economic contractions and increased advertising budgets during expansions (e.g. Lamey et al. 2007; Deleersnyder et al. 2009). Pro-cyclical investing in marketing instruments seems therefore the most common strategy followed. There are many reasons why organizations adjust to and match their advertising expenditures with the direction of business cycle fluctuations. These include the flexibility and lack of commitment in media contracts, the view of advertising as an expense instead of an investment and the behaviour of competitors who also spend less on advertising during recessions, which justifies the lower advertising spending to achieve the same share-of-voice (Deleersnyder et al. 2009).
10 Managers do not only modify their advertising budgets. In a contraction period, almost all organizations adjust their product prices. Some managers increase their prices to make up for the lost sales caused by the drop in demand and other managers decrease their prices to strengthen their competitive position (Dekimpe & Deleersnyder, 2018). Academic research on pricing in contractions provides no clear recommendations to managers for the pricing direction they should use. Researchers who suggest a price decrease during contraction (countercyclical pricing) argue that when demand is unexpectedly low, organizations should decrease their prices, because the lower demand may be a result their rival’s actions (Green & Porter, 1984). In contrast, it has also been argued that organizations should increase prices during contractions (pro-cyclical pricing), because it is more beneficial for organizations to adapt to competitive (lower) prices during high-demand periods (Rotemberg & Saloner, 1986). The latter suggests that is always a good idea for organizations to lower your prices in order to increase demand. Although this sounds logical, in reality organizations are not able to keep on lowering their prices. They have to make decisions on their pricing strategies. decisions.
2.4 The contribution of marketing to resilience
11
2.5 Conceptual model
Figure 1: Conceptual model
12
3. Data
This study uses a subset of the same dataset as used by Van Heerde et al (2013) provided by M.J. Gijsenberg, who selected the subset of brands. The subset consists of monthly data for the log-transformed volume sales, log-log-transformed price index, and log-log-transformed advertising expenditures for 17 mature CPG product categories in the UK between March 1993 and December 2010. In each CPG product category up to five leading national brands were selected. This led to the inclusion of 76 national brands in the original dataset. However, brand 34 only advertised once during the observed period and it is not possible to show the cyclicality of the advertising investments if they advertised only once. Van Heerde et al (2013) were followed here, and brand 34 is kept out of the selection. This resulted in a final dataset of 75 brands.
The original dataset contained a classification of the 17 mature CPG categories into a subset of five product classes: Personal care, dairy food, non-dairy food, drinks and household care products. For the dairy food product class, the issue arises that this class consisted of only a single product category. To provide insights regarding differences between the product classes, it is important that the classes contain at least two categories because potential differences between one product class and the other classes could result from the unusual behaviour in this particular category and not the product class as a whole. This led to combining the dairy food and non-dairy food product classes into a single product class named ‘Food’. Now, the sample consist of 36 brands in the food class, 19 brands in the personal care class, and 10 brands in both the household care and personal care class (see Figure 1).
Figure 1: Distribution of brands over the product classes
36 10 19 10 0 5 10 15 20 25 30 35 40
14
4. Methodology
The method that will be used to answer the research question integrates several modelling techniques. It uses spectral filtering techniques to estimate the cyclical components, regressions (with bootstrapping), and meta-analytic tests. Thereby, the method consists of five steps. In this section these five steps will be discussed.
4.1 Deriving the series’ cyclical components
The first step is to derive the cyclical components of GDP, advertising, pricing, and sales series. This is necessary because these series do not only capture the cyclical variation, but also consist of a long-term trend and short-long-term deviations. Trends often distort cyclical patterns in the data. To remove these trends and smoothen the cyclical component around this trend, spectral filtering techniques for extracting business cycles are used. These are similar techniques to those applied by Deleersnyder et al. (2004; 2009) and van Heerde et al. (2013).
15 Figure 2: Cyclical deviations from the trend in the log-transformed GDP series for the UK economy
4.2 Determine downward risk and upward potential
In the second step the downward risk and the upward potential of the brands are determined. The downward risk and upward potential are defined as co-movement elasticities (see e.g. Deleersnyder et al. 2009): the extent to which the financial performance cycles of the brands are similar to the business cycle. In contrast to the aforementioned studies, which allowed only for symmetric effects, this study allowed for asymmetric effects for downward risk versus upward potential. In other words, where other studies had a single elasticity value for both contractions and expansions, this study allowed for separate values for expansions and contractions.
The elasticities of downward risk and upward potential were obtained by regressing the brands’ financial performance (or sales) cycle on the business cycle. Downward and upward movements in the cycles have been distinguished in this process. The basic model that followed from this process is specified as:
𝑆𝑎𝑙𝑒𝑠
𝑓,𝑡𝑐𝑦𝑐= 𝛼
0,𝑓𝑆𝑎𝑙𝑒𝑠+ 𝛼
𝑑𝑜𝑤𝑛,𝑓𝑆𝑎𝑙𝑒𝑠𝐶𝑜𝑛𝑡𝑟𝑎𝑐𝑡𝑖𝑜𝑛
𝑡𝑐𝑦𝑐+ 𝛼
𝑢𝑝,𝑓𝑆𝑎𝑙𝑒𝑠𝐸𝑥𝑝𝑎𝑛𝑠𝑖𝑜𝑛
𝑡𝑐𝑦𝑐+ 𝜀
𝑓,𝑡𝑆𝑎𝑙𝑒𝑠Based on the BIC, an autoregressive component is added to the model if necessary (See e.g. Deleersnyder et al, 2004). In the model, 𝐶𝑜𝑛𝑡𝑟𝑎𝑐𝑡𝑖𝑜𝑛𝑡𝑐𝑦𝑐 and 𝐸𝑥𝑝𝑎𝑛𝑠𝑖𝑜𝑛𝑡𝑐𝑦𝑐 represent the state of the UK business cycle at time t. 𝐶𝑜𝑛𝑡𝑟𝑎𝑐𝑡𝑖𝑜𝑛𝑡𝑐𝑦𝑐equals the value of the CF-filtered GDP series in case of an economic downturn, otherwise it equals zero. The same condition holds for 𝐸𝑥𝑝𝑎𝑛𝑠𝑖𝑜𝑛𝑡
𝑐𝑦𝑐 . Thus, the -0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04 0.05 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
16 model cannot have a value for both expansion and contraction. 𝛼𝑑𝑜𝑤𝑛,𝑓𝑆𝑎𝑙𝑒𝑠 represents the downward risk
of the brand, and 𝛼𝑢𝑝,𝑓𝑆𝑎𝑙𝑒𝑠 the upward potential. The values of the model’s coefficients can be interpreted as elasticities, because the cycles are defined as percentage deviations from the underlying change of the variables.
To provide overall general insights, a meta-analytic added Z-test is performed. This test allows for the combination of individual results into overall insights (Rosenthal, 1991). While this method provides us with insights on the significance of effects, the uncertainty-weighted average elasticity estimates are reported as meta-values (see Gijsenberg, 2014; 2017).
4.3 Determine steadiness of performance
To determine the third resilience aspect, the steadiness of performance, the (lack of) variability in the cyclical component of the financial performance is considered. The steadiness of the financial performance will be estimated by means of the inverse of the standard deviation of the CF-filtered sales series.
𝑆𝑡𝑒𝑎𝑑
𝑓𝑆𝑎𝑙𝑒𝑠=
1
𝑠𝑡𝑑𝑒𝑣 (𝑆𝑎𝑙𝑒𝑠
𝑓,𝑡=1…𝑇𝑐𝑦𝑐)
As mentioned in step 2, comparability is possible because the series were log-transformed prior to filtering (Deleersnyder et al. 2004). Again, the steadiness is analysed on brand-level. Similar to step 2, an added Z-test is performed for overall general insights.
4.4 Extract co-movement of price and advertising investments with the business
cycle
17 The same approach as in step 2 is adopted here. The brands’ advertising and pricing cycles are regressed on the business cycle. Similar to step 2, the split is made between expansion and contraction periods in the business cycle. This leads to the following models:
𝐴𝐷𝑉
𝑓,𝑡𝑐𝑦𝑐= 𝛼
0,𝑓𝐴𝐷𝑉+ 𝛼
𝑑𝑜𝑤𝑛,𝑓𝐴𝐷𝑉𝐶𝑜𝑛𝑡𝑟𝑎𝑐𝑡𝑖𝑜𝑛
𝑡𝑐𝑦𝑐+ 𝛼
𝑢𝑝,𝑓𝐴𝐷𝑉𝐸𝑥𝑝𝑎𝑛𝑠𝑖𝑜𝑛
𝑡𝑐𝑦𝑐+ 𝜀
𝑓,𝑡𝐴𝐷𝑉𝑃𝑟𝑖𝑐𝑒
𝑓,𝑡𝑐𝑦𝑐= 𝛼
0,𝑓𝑃𝑟𝑖𝑐𝑒+ 𝛼
𝑑𝑜𝑤𝑛,𝑓𝑃𝑟𝑖𝑐𝑒𝐶𝑜𝑛𝑡𝑟𝑎𝑐𝑡𝑖𝑜𝑛
𝑡𝑐𝑦𝑐+ 𝛼
𝑢𝑝,𝑓𝑃𝑟𝑖𝑐𝑒𝐸𝑥𝑝𝑎𝑛𝑠𝑖𝑜𝑛
𝑡𝑐𝑦𝑐+ 𝜀
𝑓,𝑡𝑃𝑟𝑖𝑐𝑒Depending on the BIC, an autoregressive component is added to the model (see e.g. Deleersnyder et al, 2004). In this model, the co-movement elasticities of advertising and pricing investments during an economic contraction are expressed by
𝛼
𝑑𝑜𝑤𝑛,𝑓𝐴𝐷𝑉 and𝛼
𝑑𝑜𝑤𝑛,𝑓𝑃𝑟𝑖𝑐𝑒 . The co-movement elasticities are expressed by𝛼
𝑢𝑝,𝑓𝐴𝐷𝑉 and𝛼
𝑢𝑝,𝑓𝑃𝑟𝑖𝑐𝑒. In line with step 2, an Added Z test is applied and the uncertainty-weighted average elasticities are reported.4.5 Estimate the effect of price and advertising on resilience aspects
The effects of the brands’ pro- or countercyclical advertising and pricing investment strategies on their resilience regarding macro-economic fluctuations will be judged by regressing the downward risk elasticities, upward potential elasticities and steadiness measure on the advertising and pricing co-movements elasticities. This leads to the following model specifications:
(1)
𝛼
𝑑𝑜𝑤𝑛,𝑓𝑃𝑒𝑟𝑓= 𝛽
𝑑𝑜𝑤𝑛,0𝑃𝑒𝑟𝑓+ 𝛽
𝑑𝑜𝑤𝑛,1𝑃𝑒𝑟𝑓𝛼
𝑑𝑜𝑤𝑛,𝑓𝐴𝐷𝑉+ 𝛽
𝑑𝑜𝑤𝑛,2𝑃𝑒𝑟𝑓𝛼
𝑢𝑝,𝑓𝐴𝐷𝑉+ 𝛽
𝑑𝑜𝑤𝑛,3𝑃𝑒𝑟𝑓𝛼
𝑑𝑜𝑤𝑛,𝑓𝑃𝑟𝑖𝑐𝑒+𝛽
𝑑𝑜𝑤𝑛,4𝑃𝑒𝑟𝑓𝛼
𝑢𝑝,𝑓𝑃𝑟𝑖𝑐𝑒+ 𝜀
𝑑𝑜𝑤𝑛,𝑓𝑃𝑒𝑟𝑓(2)
𝛼
𝑢𝑝,𝑓𝑃𝑒𝑟𝑓= 𝛽
𝑢𝑝,0𝑃𝑒𝑟𝑓+ 𝛽
𝑢𝑝,1𝑃𝑒𝑟𝑓𝛼
𝑑𝑜𝑤𝑛,𝑓𝐴𝐷𝑉+ 𝛽
𝑢𝑝,2𝑃𝑒𝑟𝑓𝛼
𝑢𝑝,𝑓𝐴𝐷𝑉+ 𝛽
𝑢𝑝,3𝑃𝑒𝑟𝑓𝛼
𝑑𝑜𝑤𝑛,𝑓𝑃𝑟𝑖𝑐𝑒+ 𝛽
𝑢𝑝,4𝑃𝑒𝑟𝑓𝛼
𝑢𝑝,𝑓𝑃𝑟𝑖𝑐𝑒+ 𝜀
𝑢𝑝,𝑓𝑃𝑒𝑟𝑓(3)
𝑆𝑡𝑒𝑎𝑑
𝑓𝑃𝑒𝑟𝑓= 𝛽
𝑠𝑡𝑒𝑎𝑑,0𝑃𝑒𝑟𝑓+ 𝛽
𝑠𝑡𝑒𝑎𝑑,1𝑃𝑒𝑟𝑓𝛼
𝑑𝑜𝑤𝑛,𝑓𝐴𝐷𝑉+ 𝛽
𝑠𝑡𝑒𝑎𝑑,2𝑃𝑒𝑟𝑓𝛼
𝑢𝑝,𝑓𝐴𝐷𝑉+ 𝛽
𝑠𝑡𝑒𝑎𝑑,3𝑃𝑒𝑟𝑓𝛼
𝑑𝑜𝑤𝑛,𝑓𝑃𝑟𝑖𝑐𝑒+𝛽
𝑠𝑡𝑒𝑎𝑑,4𝑃𝑒𝑟𝑓𝛼
𝑢𝑝,𝑓𝑃𝑟𝑖𝑐𝑒+ 𝜀
𝑠𝑡𝑒𝑎𝑑,𝑓𝑃𝑒𝑟𝑓18
5. Model-Free insights
The purpose of this chapter is to discuss interesting insights found in the data in prior to modeling. These insights provide more context regarding how the business cycle affects the brands and how the brands modify their expanses in reaction to business cycle fluctuations. Additionally, several differences between the product classes of ranks will be discussed.
5.1 Sales over the business cycle
To provide insights into the extent to which the business cycle affects the sales of the CPG brands, several t-tests were performed between the average log-transformed sales in expansions and contractions. Surprisingly, more brands showed a significant increase of their sales during contractions, rather than a significant decrease in sales (See Table 1). Brands who significantly increased their sales during contractions did this with an average of 5.2, while brands who significantly decreases their sales did this with an average of 4.7. Brands without significant differences showed on average a sales change of 0.5, and overall sales increased with 2.1 during contractions. These results suggest that the overall sales of the CPG have a tendency towards countercyclical patterns, showing increased sales during contractions and decreased sales during expansions.
Table 1: The impact of the business cycle on log-transformed brand sales
Number of Brands Avg. change
Increased sales contraction (p < .10) 12 5.2
Decreased sales contraction (p < .10) 9 -4.7
No significant effect 54 0.5
Overall 75 2.1
5.2 Advertising investments over the business cycle
19 results suggest that, although decreasing advertising investments is the more common measure for CPG brands, some brands tend to increase their advertising investments. Additionally, brands that opt for a more countercyclical advertising strategy seem to increase their budgets approximately 15 more compared to brands that follow a pro-cyclical strategy. This finding is surprising, since in contractions advertising becomes less expensive due to decreased competition for the same advertising space.
Table 2: Log-transformed advertising over the business cycle
Number of Brands Avg. Change Increased adv. investments contraction (p < .10) 11 165.4 Decreased adv. investments contraction (p < .10) 21 -149
No significant effect 43 14.9
Overall 75 -8.9
Figure 3 shows the average advertising frequencies based on product class and rank. These figures suggest that, on product class level, brands in the drinks and household care product classes advertise relatively often, while brands in the food and personal care classes advertise relatively little. Based on rank, we see a systematic decrease from the first (148 months) to the fourth ranked brand (55.3 months). However, the fifth ranked brand advertises more frequently than the fourth (67.5 months). This might be caused by the fact that some categories only exist of four brands. In categories with five brands, the competition is fiercer and brands might be forced to advertise more to keep up with the competition. Resulting in a higher advertising frequency for these categories.
Figure 3: Average advertising frequency per product class and per rank
5.3 Price over the business cycle
To provide insights into how the brands modify their price over the business cycle, several t-tests were performed to test for significant differences for each brand between expansions and contractions. Again, the log-transformation allows for interpreting the differences as percentage changes. The
130 92.7 123 73.6 0 20 40 60 80 100 120 140
Drinks Food Household
20 results show (Table 3) that 12 brands increased their prices significantly during a contraction, while 16 brands significantly decreased their prices in contractions. The brands that increased their prices significantly during contractions did this on average with 3.8. Brands that significantly decreased their price did this with an average of 5.2. These results suggest that CPG brands both increase and decrease their prices during contractions, thereby following both pro-cyclical and counter cyclical strategies. This implies that there is no dominant pricing strategy for the CPG brands.
Table 3: Log-transformed price over the business cycle
Number of Brands Avg. Change
Increased price contraction (p < .10) 12 3.8
Decreased price contraction (p < .10) 16 -5.2
No significant effect 47 1.7
21
6. Model-Based Results
Before discussing the results of the model, some more general modelling issues will be discussed. Although the equations from step 2 and step 4 of the method allow for direct influences of business cycle fluctuations on these elements, they may still be influenced by fluctuations in the previous month(s). Therefore, up to three lags were added to the models. The models with one lag are considered best, based on the Bayesian Information Criterion (BIC) and the 𝑅2’s of these models. Table 4 shows the BIC-scores for the lagged model. Although the BIC scores for models with two or three lags are lower, the one-lag models is considered best since the average 𝑅2‘s of the two- and three-lagged models tell that these explain respectively 99.9% and 100% of the variability in the sample. To avoid the risk of overfitting the model, I chose not to use the model with the lowest BIC. The direct effects (and standard errors) of the models with one lag were taken to calculate the parameters (and standard errors) for the final model.
Table 4: Average BIC scores
No lag One lag Two lags Three lags
Uw. pot/Dw. risk -972.9 -1699 -2474 -3148
Advertising 778.4 90.4 -711.3 -1386
Pricing -845.5 -1565.2 -2350 -3016
The adjusted 𝑅2 of the regression to extract the upward potential and the downward risk tells us that the business cycle predicts between the 94% and 99% of the variability in the cyclical sales series with an average of 97%. The model fit of the models for the advertising and pricing co-movements show similar results. The adjusted 𝑅2‘s of the models in step 4 tell us that business cycle fluctuations explained between 94% and 98% of the variability in the cyclical advertising series with an average of 96% and between 94% and 99% with an average of 97% of the variability in the cyclical pricing series. From these results, we can conclude that the sales (or marketing investments) of last month are good predictors for the sales (or marketing investments) for this month.
6.1 The cyclical behaviour of CPG sales
22 significantly positive effect of the business cycle component on the overall sales (β = .15, Z = 5.52, p < .01). For the interpretation of the result, it is important to note that the cyclical log-transformed GDP series in contractions only decreases. Therefore, the result suggest that when the economy as a whole decreases with 1%, the sales across the CPG market decreases with 0.15%. The size of the overall effects show that the CPG market reacts inelastic to the business cycle. This could be explained by the findings of Gicheva et al. (2008) that consumers substitute away from out-of-home consumption and back towards groceries to compensate for the reduced discretionary income during contractions. The switch consumers make possibly attenuate the growth of the CPG market.
Another interesting aspect of the overall effect is the magnitude of the effect the business cycle components have on the sales. An effect size of <|1| implies that compared to the general economic activity, the general business cycle fluctuations are attenuated in the context of sales for the CPG category as a whole. Further investigation of the strength of the relationships for the brands separately confirms this believe (See Table 5). For all brands in the sample, business cycle fluctuations are attenuated in the context of sales for both expansions and contraction. Additionally, the extent to which the business cycle affects the sales of the brand with the most pro-cyclical sales pattern in a expansion is almost two times the extent to which the sales of the brand with the most countercyclical brand is affected. In contractions, extent to which the sales of brand with the most pro-cyclical sales pattern is affected three times the extent to which the most countercyclical brand is affected. These findings suggest that the sales of the CPG brands are less affected by business cycle fluctuations compared to the overall economic activity, that pro-cyclical sales patterns are stronger, and that the difference between the extent that brands with pro- vs. countercyclical sales patterns are affected by the business cycle increases in contractions.
23 Table 5: Overall results on the extent of cyclical sensitivity
Number of brand with Weigthed β Range Q1-Q3 Effect Size |1|> Pro-cyclical sales (p < .10) Countercyclical sales (p < .10) Upward potential .15*** -.05 – .09 0 24 17 Downward risk .15*** -.04 – .08 0 25 16 ***p<0.01
Differences between product classes
24 The results on the individual brands provide insights concerning the amount and distribution of brands for which a significant pro- or countercyclical sales pattern is found within each product class. In line with the Added-Z Test, the product classes food, drinks, and personal care contain more brands with pro-cyclical sales, whereas the product class household care contains more countercyclical brand. Additionally, the number of significant pro- or countercyclical brands decreases as the effect decreases (e.g. for the food class’ upward potential (13) vs. (10) downward risk). Therefore, the number of significant brands seems a viable representation of the sign of the effect and changes between expansions and contractions.
The results on the product classes imply that companies with multiple CPG brands can expect pro-cyclical development of the sales for their food, drinks, and personal care brands, and an overall countercyclical development of the sales in household care brands. Based on these implications, managers are recommended to focus on capturing the potential growth of food, drink, and personal care brands in expansions and focus on avoiding to lose ground in contractions, whereas they should focus on avoiding to lose ground in expansions and they should try to take this ground back (or expand it) in contractions.
Table 6: Results on the extent of cyclical sensitivity per product class
Number of brand with
N Weigthed β Pro-cyclical sales (p < .10) Countercyclical sales (p < .10) Upward potential Food 36 .031*** 13 7 Drinks 10 .020*** 2 1 Personal care 19 .011*** 8 5 Household care 10 -.070*** 1 4 Downward risk Food 36 .015*** 10 7 Drinks 10 .036*** 4 2 Personal care 19 .010*** 9 4 Household care 10 -.011 2 3 ***p<0.01
Differences in market position
25 in both expansions (β = .038, Z = 4.14, p <.01) and contractions (β = .056, Z =5.94, p <.01). The brands in the second position in the product category shows more variability in how they are affected by the business cycle, resulting in no significant effects on sales for both expansions and contractions. The sales of the brands in the third position in the market are, like the market leaders, significantly positive affected by expansions (β = .026, Z = 4.18, p <.01) and contractions (β = .038, Z = 6.01, p <.01). The brands occupying the fourth position in the market show no overall significant effects of the business cycle on their sales in both expansion and contractions. The fifth ranked brands in the market show also a significantly positive effect of the business cycle in expansions (β = .013, Z = 3.45, p <.01), however no significant effects for contractions. The decreasing size of the coefficients suggests that the possible impact the business cycle has on the upward potential and the downward risk of the brand decreases with their position. Therefore, market leaders seem to be more vulnerable to business cycle fluctuations as lower ranked brands in the same category. Additionally, brands in the second and fourth position are differently affected by business cycle fluctuations, then brands in the uneven market positions, which suggests that contractions provide them the opportunity to overtake the position of their closest competitors.
Table 6: Results on the extent of cyclical sensitivity per rank
Number of brand with
N Weigthed β Pro-cyclical sales (p < .10) Countercyclical sales (p < .10) Upward potential Rank 1 17 .038*** 6 2 Rank 2 16 .003 4 6 Rank 3 16 .026*** 6 3 Rank 4 15 .004 3 4 Rank 5 11 .013*** 5 2 Downward risk Rank 1 17 .056*** 9 5 Rank 2 16 .012 6 3 Rank 3 16 .038*** 6 1 Rank 4 15 -.004 2 4 Rank 5 11 -.003 2 3 ***p<0.01
6.2 Steadiness
26 compared to check for differences between the product classes. Table 8 shows the mean steadiness scores per product class. In line with the outcomes on the effect of the business cycle, household care is the least steady product class with a mean of 21.44. Personal care is the steadiest product class with a mean of 112.99. These findings are in line with the findings on the downward risk and upward potential. From this we can conclude that the sales of personal care brands are the least volatile and the sales of household care products most volatile. Therefore, managers of multiple CPG brand companies who want to invest more defensive, better invest in personal care brands (over e.g. household care brands), because an increase in the sales of these brands will probably last longer. Manager who want to invest more offensive, better invest in household care brands (over e.g. personal care brands), because these brands give are more high risk – high reward oriented.
Table 8: Steadiness by product class
Category Mean
Food 44.48
Drinks 73.30
Household Care 21.44
Personal care 112.99
The means for the market positions of the brands also show interesting insights. The results in Table 9 show that brands ranked first in their category have on average the lowest steadiness score with 37.05 this slightly increases with the ranks and brands ranked fifth have on average the highest steadiness score with 114.15. This suggests that the standard deviation of the brand sales becomes higher when the brand gains market share. However, research on this phenomenon by Jung, Gruca & Rego (2010) points in the opposite direction. They find that for CPG brand, a higher market share is associated with customers who are more loyal and buy more. Therefore, one would expect the steadiness to be decreasing with position. However, Jung Gruca & Rego (2010) used annual data instead of the monthly data used in this paper. Therefore, a possible explanation for this discrepancy is that the steadiness of higher ranked brands is better over a longer period, but more volatile on smaller aggregation levels.
27
6.3 The cyclical behaviour of CPG advertising and price
Table 10 shows the results of the Added-Z test which is computed to provide some overall insights regarding the brands’ advertising and pricing strategies. The results show that advertising is significantly negative affected by business cycle fluctuations. The effects are almost symmetrical for expansions (β = -.87, Z = -4.33, p<0.01) and contractions (β = -.86, Z = -4.16, p<0.01). These findings suggest that a 1% growth (decline) of the overall economy in expansions (contractions) results in a 0.87% decrease (0.86% increase) of advertising investments across all CPG products. Implying that the CPG market invests less in advertising in expansions and more in contractions. Additionally, the size of the effect suggest that across CPG brands adverting investments are inelastic relative to business cycle fluctuations. These implications contradict the findings of Lamey et al. (2007, Table 5) that the majority of CPG brands cut advertising investments during contractions. They compared the growth of advertising investments of 1997 – 1998 (expansion) to these of 2001 – 2002 (contraction) for ten “Top Brands” in the UK. The contradiction between the results might be caused by the multiple expansions and contractions and the larger variety of brands this paper accounted for. Further research is recommended to provide clear insight concerning advertising investments of CPG brands.
Investigation of the coefficients and significance pro- and countercyclical advertising investments shows that more brands advertise countercyclical than pro-cyclical, and that the advertising investment for many brands are elastic with respect to business cycle fluctuation. Of the 75 brands, 58 report elastic co-movements of advertising investments relative to economic downturns, whereas 65 brands report elastic co-movements of advertising investment in expansions. Additionally, more brands report a significant countercyclical co-movement in expansions and contractions. However, in contractions the number of brands with significantly pro-cyclical advertising investments increases, and the number of brands with significantly countercyclical advertising investments decreases. The interquartile range shows a similar pattern and moves from -5.54 – 2.57 in expansions to -4.43 – 2.67 in contractions. The results imply that managers of CPG brands use both pro- and countercyclical advertising strategies in expansion and contractions, but they seem modify the strategy in contractions slightly toward the business cycle. Thereby they attenuate the counter cyclicality of their strategy or amplify the pro-cyclicality.
28 prices more in contractions than they do in expansions. This could be explained by the increased popularity of price promotions during contractions among managers (De Chernatony et al. 1991). Managers might compensate for these price promotions by increasing the price of their products and because our model ignores the fluctuations of such temporary promotional activities, only the long-term deviations are visible.
Investigation of the cyclicality of the individual brands shows that almost twice the number of brands that prices pro-cyclical in expansions (20) does this in contraction (38). This means that approximately half of the CPG brands in the sample increases their prices in contractions. Similar to the overall effects, the effects of temporary price promotions are not captured by the cyclical component used in this model. As the overall effect already indicates, price reacts very inelastic to business cycle fluctuations with only a single brand reporting an elastic price co-movement in contractions and none during expansions. Surprisingly, the interquartile range shows that the brand on the third quartile in a contraction reports a value of three times its equivalent in expansions. This provides evidence that the pricing elasticities contain more variability in contractions relative to expansion, which implies that managers react stronger to economic downturns than economic growth.
Table 10: Overall results on the advertising and pricing co-movements
Number of brand with Weighted β Range Q1-Q3 β >|1| Pro-cyclical
investments (p < .10) Countercyclical investments (p < .10) Advertising co-movement Expansions -.873*** -5.536 – 2.571 58 15 24 Contraction -.859*** -4.430 – 2.671 65 18 21 Pricing co-movement Expansions .014*** -0.105 - 0.109 0 25 20 Contraction -.061*** -0.179 - 0.375 1 15 38 ***p<0.01
Differences on product class level
meta-29 analytic co-movement coefficient for the drinks class is smaller, although still significant and negative (β = -3.34, Z = -5.17, p<0.01). Resulting in an increase in advertising investments during contractions for this product class. Additionally, the food class reports significantly negative co-movements in contractions (β = -.72, Z = -3.24, p<0.05), suggesting that the advertising investments increase during contractions across this product class. The results imply that in expansions “the battle for the consumers’ attention” becomes less cluttered for brands in the drinks product class. This leads to opportunities for brands in this class, since it is easier to get consumers’ attention in a less cluttered environment (Assael, 1998). However, although the advertising clutter for drinks reduces, the overall advertising clutter becomes more extensive in expansions (Deleersnyder et al., 2009). There is less competition from inside the product class, but the overall “competition for the consumers’ attention” becomes fiercer. In contractions, the results imply that across both the drinks and food product classes advertise investments are increased. This leads to more advertising clutter from inside these classes, but since the overall advertising clutter reduces in contractions this might not be problematic. Future research is recommended to investigate how these classes are differently affected by the extent of advertising clutter in contractions and expansions.
The results of the Added-Z test for the price co-movements in expansions shows that economic growth has a significantly positive effect on the price of products across the food (β = .014, Z = 2.14, p<0.05) and drinks (β = .069, Z = 5.53, p<0.01). The business cycle has a significantly negative effect on the price of products of the household care class (β = -.034, Z = -3.31, p<0.01). These results suggest that in expansions food and drinks become overall more expansive, whereas household care products become less expansive. Whether the brands increase or decrease their prices seems to be related to how the brands are affected by the business cycle, since the directions of the effects are similar, although their size may vary. Thereby, brands’ price decisions may be a function of the cyclical demand.
30 Table 11: Overall results on the advertising and pricing co-movements per product class
Number of brand with
N Weigthed β β >|1| Pro-cyclical
investments (p < .10)
Countercyclical investments (p < .10) Advertising co-movement expansions
Food 36 -.292 27 9 10
Drinks 10 -4.822*** 9 0 6
Personal Care 19 -.713 15 4 5
Household Care 10 -.245 8 2 3
Advertising co-movement contractions
Food 36 -.720** 31 8 12
Drinks 10 -3.337*** 8 1 3
Personal Care 19 -.359 17 5 3
Household Care 10 .596 10 4 3
Price co-movement expansions
Food 36 .014** 0 13 9
Drinks 10 .069*** 0 6 0
Personal Care 19 .021 0 5 6
Household Care 10 -.034*** 0 1 5
Price co-movement contractions
Food 36 -.053*** 0 10 20
Drinks 10 -.090*** 0 0 6
Personal Care 19 -.052*** 1 4 6
Household Care 10 -.074*** 0 1 6
***p<.01; **p<.05
6.4 How pricing and advertising strategies affect the resilience aspects
31 Table 12: Results on how price and advertising affect resilience aspects
Estimates Standard errors Upward potential
Intercept 0.024 0.020
Co-movement Adv. in contraction -0.004** 0.002 Co-movement Adv. in expansion -0.003 0.003 Co-movement Pricing in contraction -0.035 0.088 Co-movement Pricing in expansion -0.029 0.098
Downward risk
Intercept 0.020 0.025
Co-movement Adv. in contraction -0.004* 0.003 Co-movement Adv. in expansion -0.003 0.003 Co-movement Pricing in contraction -0.215** 0.109 Co-movement Pricing in expansion 0.140 0.120 Steadiness
Intercept 64.379*** 9.525
Co-movement Adv. in contraction 1.216 1.062 Co-movement Adv. in expansion -0.799 1.242 Co-movement Pricing in contraction 10.114 41.390 Co-movement Pricing in expansion 1.409 45.810 *p < .10; **p < .05; ***p < .01
Notes: Standard errors are the robust standard errors. The t-statistic and significance are estimated based on the robust standard errors.
33
7. Conclusion & Discussion
7.1 Summary
In this paper, the extent to which pro- and countercyclical advertising and pricing investments affects brands’ resilience with regard to the macro-economic business cycle is investigated. For this purpose, three different measures of resilience were used: upward potential, downward risk and steadiness. Downward risk and upward potential refer to the ability of the brands to mitigate the adverse impact of the economic downturn and to recover from it respectively. However, a brand does not have to deal with a potential crisis if the brand is not affected by the economic downturn at all. Therefore, steadiness is introduced as a measure.
The results provide us with two main findings: 1) CPG brands should cut less or increase their advertising investment in contractions to strengthen their upward potential, 2) brands should decrease their price (or increase the price less) in contractions to decrease their downward risk. The results show that cutting advertising investments less or even increase them in contractions positively affects the brands’ upward potential. This assumes that when the economy is in heavy weather, increased advertising investments strengthen the brands ability to recover from the sustained damage afterwards. These findings provide evidence that getting in the consumers’ minds before the expansion begins and the advertising clutter intensifies, give brands a lead compared to their competitors. For the second finding, the results provide evidence that brands can mitigate the adverse impact of economic downturns by decreasing (or less increasing) their prices. Common sense tells us that this effect is obvious, because lowering prices positively affect sales in most situations. Consumers often search for less expensive alternatives or better deals. The price-sensitivity of consumers seems to increase during contractions (e.g. Estelami et al, 2001; van Heerde et al., 2013). My findings provide additional evidence for this notion.
34
7.2 Managerial implications
When economic downturns hit, the result on firms can be disastrous. Managers realize this and most managers admit that they modify their marketing budgets in reaction to economic downturns. Creating a more resilient organization might be the solution to cope with these adverse events, because resilient organizations are better capable to mitigate the impact of the economic downturn and to recover faster. However, research on organizational resilience mainly focuses on the organization’s structure and culture, which are time costly and difficult to change. In the crisis situation that economic downturn can cause, managers do not have the time to implement these changes. Additionally, managers are under increasing pressure to improve the effectiveness of their marketing investments and account for their decisions. Increasing the organizational resilience with their marketing investments would kill two birds with one stone. Therefore, this research investigated how modifying advertising investments and price affect organizational resilience.
The findings discussed in this paper can provide manager of CPG brands with evidence for normative actions they can take to make their organization more resilient to the adverse impact of economic downturns. First of all, the results point out that in contractions more advertising can help brands to recover faster (in terms of sales) from the economic downturn. By being in the consumers’ mind when the economy starts growing again, brands can provide themselves with a head start to their competitors. This results in enhanced sales during the subsequent expansion. The second key finding of this research is that lowering (or less increasing) prices during contractions mitigates the adverse impact the economic downturns. This seems logically, since in most scenarios lowering prices increase sales. However, lowering prices becomes even a more important measure during contractions due to consumers’ increased price-sensitivity in these periods (e.g. Van Heerde et al., 2013). Based on these two key findings, managers of CPG brands are recommended to increase (or at least not decrease) advertising investments during contractions and to decrease (or at least not increase) prices. Furthermore, this research provides evidence that the brands’ marketing actions in expansions to not affect the resilience of their brands at all. However, this does not imply that these actions are not effective at all. Only their influence on the resilience of the brand is discussed here.
35 than the general economic activity. However, the investigation of the differences between the different product classes shows major differences. The sales of food, drinks and personal care classes are positively affected by economic growth, whereas the household care class is affected negatively, thus the sales in this class decrease when the overall economy grows. The effect size shows great variability, not only between the product classes, but also between expansions and contractions. The drinks class for example, shows that the effects of economic growth on the overall sales in this product class is approximately half the size of the effect economic decline has on the overall sales in this category. For a more extensive discussion of the overall differences and differences between product classes, I want to refer the interested reader to the results section of this paper. Conclusively, the variability implies that the product classes do not react similarly to business cycle fluctuations, therefor managers should be aware of these differences to make concise decisions regarding their marketing investments.
7.3 Directions for further research
My study shows several limitations that offers opportunities for further research. First, the dataset consists of relatively mature CPG categories. Compared to less mature CPG categories, mature CPG categories are characterized by a lower advertising sensitivity (Allenby & Hanssens, 2005). The advertising sensitivity for less mature CPG categories might be differently affected by business cycle fluctuations. Because this paper focuses on mature categories, the question remains whether more recently introduced categories are affected similarly by the business cycle. Further research is necessary to provide better insights into this issue.
Second, the products investigated in this paper are for everyday use. Unlike to other products and services (e.g. durables, construction work), purchases of CPG products cannot be delayed until the contraction is over. Categories for which sales can be postponed, might be more sensitive to business cycle fluctuations. Durables for example, are more prone to business cycle fluctuations (Deleersnyder et al. 2004). This offers an interesting research opportunity.
36 Fourth, as previously stated, the advertising and pricing co-movements do not affect the steadiness in sales at all. The S-shaped function of advertising and price responses might be an influential factor in these (lack of) findings regarding the steadiness, because this research focused on advertising and price movements instead of the lack of movement. Further investigation concerning the lack of movement in advertising and price could provide answers for this remaining question.
37
References
Allenby, G., & Hanssens, D. (2005). Advertising response. Marketing Science Institute, Special Report, (05-200), 8.
Amann, B., & Jaussaud, J. (2012). Family and non-family business resilience in an economic downturn. Asia Pacific business review, 18(2), 203-223.
Assael, H. (1998). Consumer Behavior and Marketing Action, 6th Edition. Cincinatti, OH. SouthWestern College Publishing.
Baveja, S., Postma, G., & Pritzl, R. (2002). How to bounce higher out of recession. European Business Journal, 14(4), 193-193.
Carmeli, A., Friedman, Y., & Tishler, A. (2013). Cultivating a resilient top management team: The importance of relational connections and strategic decision comprehensiveness. Safety Science, 51(1), 148-159.
Christiano, L. J., & Fitzgerald, T. J. (2003). The band pass filter. international economic review, 44(2), 435-465.
Cleeren, K., Dekimpe, M. G., & van Heerde, H. J. (2017). Marketing research on product-harm crises: a review, managerial implications, and an agenda for future research. Journal of the Academy of Marketing Science, 45(5), 593-615.
Danaher, P. J., Bonfrer, A., & Dhar, S. (2008). The effect of competitive advertising interference on sales for packaged goods. Journal of Marketing Research, 45(2), 211-225.
De Chernatony, L., Knox, S. & Chedgey, M. (1991). Brand Pricing in a Recession. European Journal of Marketing, Vol 26 (2). pp 5-14.
Dekimpe, M. G., & Deleersnyder, B. (2018). Business cycle research in marketing: a review and research agenda. Journal of the Academy of Marketing Science, 46(1), 31-58.
Deleersnyder, B., Dekimpe, M. G., Sarvary, M., & Parker, P. M. (2004). Weathering tight economic times: The sales evolution of consumer durables over the business cycle. Quantitative Marketing and Economics, 2(4), 347-383.
Deleersnyder, B., Dekimpe, M. G., Steenkamp, J. B. E., & Leeflang, P. S. (2009). The role of national culture in advertising's sensitivity to business cycles: An investigation across continents. Journal of Marketing Research, 46(5), 623-636.
Du, R. Y., & Kamakura, W. A. (2008). Where did all that money go? Understanding how consumers allocate their consumption budget. Journal of Marketing, 72(6), 109-131.
38 Fischer, M., Shin, H., & Hanssens, D. M. (2007). Marketing Spending and the Volatility of Revenues and Cash Flows. Diskussionspapier der Anderson School of Management. University of California, Los Angeles.
Gicheva, D., Hastings, J., & Villas-Boas, S. (2010). Investigating income effects in scanner data: do gasoline prices affect grocery purchases?. American Economic Review, 100(2), 480-84.
Gordon, B. R., Goldfarb, A., & Li, Y. (2013). Does price elasticity vary with economic growth? A cross-category analysis. Journal of Marketing Research, 50(1), 4-23.
Graham, R. C., & Frankenberger, K. D. (2011). The earnings effects of marketing communication expenditures during recessions. Journal of Advertising, 40(2), 5-24.
Green, E. J., & Porter, R. H. (1984). Noncooperative collusion under imperfect price information. Econometrica: Journal of the Econometric Society, 87-100.
Gulati, R., Nohria, N., & Wohlgezogen, F. (2010). Roaring out of recession. Harvard Business Review, 88(3), 62–69.
Jung, S. U., Gruca, T. S., & Rego, L. L. (2010). Excess loyalty in CPG markets: a comprehensive examination. Journal of Empirical Generalisations in Marketing Science, 13(1).
Kashmiri, S., & Mahajan, V. (2014). Beating the recession blues: Exploring the link between family ownership, strategic marketing behavior and firm performance during recessions. International Journal of Research in Marketing, 31(1), 78-93.
Lamey, L., Deleersnyder, B., Steenkamp, J. B. E., & Dekimpe, M. G. (2012). The effect of business-cycle fluctuations on private-label share: what has marketing conduct got to do with it?. Journal of Marketing, 76(1), 1-19.
Lengnick-Hall, C. A., & Beck, T. E. (2009). Resilience capacity and strategic agility: Prerequisites for thriving in a dynamic environment: UTSA. College of Business.
Marketing Science Institute. (2002). MSI/AAAA Special Request for Proposals: Marketing in Turbulent Times. Cambridge, MA.
McManus, S., Seville, E., Vargo, J., & Brunsdon, D. (2008). Facilitated process for improving organizational resilience. Natural Hazards Review, 9(2), 81-90.
Monroe, K. B. (1973). Buyers’ subjective perceptions of price. Journal of marketing research, 10(1), 70-80.
Özturan, P., Özsomer, A., & Pieters, R. (2014). The role of market orientation in advertising spending during economic collapse: The case of Turkey in 2001. Journal of Marketing Research, 51(2), 139-152.
39 Shama, A. (1993). Marketing strategies during recession: A comparison of small and large firms. Journal
of small business management, 31(3), 62.
Steenkamp, J. B. E., & Fang, E. (2011). The impact of economic contractions on the effectiveness of R&D and advertising: evidence from US companies spanning three decades. Marketing Science, 30(4), 628-645.
Stock, J. H., & Watson, M. W. (1999). Business cycle fluctuations in US macroeconomic time series. Handbook of macroeconomics, 1, 3-64.
Tellis, G. J., Prabhu, J. C., & Chandy, R. K. (2009). Radical innovation across nations: The preeminence of corporate culture. Journal of marketing, 73(1), 3-23.
The Economist (2009). The game has changed. The recession has spelt disaster for most brands of packaged goods, but not all. August 20. (accessed September 24, 2019), [available at
https://www.economist.com/business/2009/08/20/the-game-has-changed].
Vakratsas, D., Ambler, T. & Naik, P.A. (2007). Essentials of Planning Media Schedules. In: Tellis, G.J & Ambler, T., eds. The SAGE Handbook of Advertising. Thousand Oaks, CA. Sage Publications. Van Der Vegt, G. S., Essens, P., Wahlström, M., & George, G. (2015). Managing risk and resilience.
Academy of Management Journal, 58 (4), 971-80.
