Int. Journal of Business Science and Applied Management, Volume 13, Issue 1, 2018

Product recalls: The effects of industry, recall strategy and

hazard, on shareholder wealth

Michael Bernon

Centre for Logistics & Supply Chain Management, Cranfield School of Management

Bedfordshire, MK43 0AL, UK

Tel: +44 1234 751122

Email: m.p.bernon@cranfield.ac.uk

Marko Bastl

Department of Management, College of Business Administration, Marquette University

PO Box 1881, Milwaukee, Wisconsin 53201-1881, USA

Tel: +1 414 288 6866

Email: marko.bastl@marquette.edu

Wenqian Zhang

Cranfield School of Management

Bedfordshire, MK43 0AL, UK

Email: vivienzwq@gmail.com

Mark Johnson

Warwick Business School

Coventry, CV4 7AL, UK

Telephone: +44 2476 523703

Email: mark.johnson@wbs.ac.uk

Abstract

The purpose of this paper is to provide insights into the effects of product recalls on shareholder wealth

of manufacturing firms in different supply chains. Previous research examining this phenomenon is

largely uni-sectorial and/or does not consider the interplay of hazard, recall strategy and sector. By

utilizing the event study method, this study examines investors’ reactions to key product recall

characteristics: industry, recall strategy and hazard level, on a cross-industry sample of 296 product

recall announcements. The results show a significant negative reaction of share values to product

recalls and significant differences between industry type and hazard levels. More regulated and

stringent supply chains, such as the automotive and pharmaceutical, showed statistically significant

losses in share price. The results show that industry sector and level of hazard associated with defective

products are significant factors impacting the shareholder wealth of manufacturing firms. Contrary to

some studies, the impact of recall strategy was not confirmed, although proactive recall strategies led,

in some cases, to an increase in share price. Further research would benefit from more detailed

investigation of recall strategies on the value of companies in specific sectors, particularly ones which

are susceptible to frequent and costly product recalls.

Keywords: product recall, cross-industry sector, shareholder wealth, event study, product hazard,

recall strategy, reverse logistics

Int. Journal of Business Science and Applied Management / Business-and-Management.org

1. INTRODUCTION

A product recall is the act of requesting the return of a batch or an entire production run of a

commercial product, usually because of a defect, safety concern, or efficiency problem. It is an

example of crisis caused by product-harm and is defined by Dawar and Pillutla (2000, p. 215) as a:

“discrete, well-publicised occurrence wherein products are found to be defective or dangerous”.

Product recalls transcend supply chains and can have significant negative effects for firms and

their supply chain partners. The Japanese company Takata, in 2016, recalled 35 to 40 million airbags,

due to the explosion risk they posed to drivers and passengers. The airbags have so far been linked to at

least 10 deaths in the US. The total number of faulty airbags in the US alone exceeds 69 million,

according to The National Highway Traffic Safety Association (NHTSA), a burden big enough to

result in Takata’s bankruptcy (Money CNN, 2017). Further, in complex supply chains, recalls can be

pervasive. In 2013, for example, Europe encountered what was to become known as the “horsemeat

scandal”, when horsemeat was discovered in a wide range of ‘beef’ products. The effect was supply

chain-wide, leading to food manufacturers, retailers and restaurant chains across Britain, Ireland,

France, Spain, Germany, Denmark, Sweden and Norway recalling an ever increasing range of food

products containing traces of horse DNA (BBC News, 2013).

Recalls can adversely affect firm’s performance (Chang et al., 2015), reduce brand equity, damage

its reputation, cause panic among consumers, result in revenue and market share losses (Laufer and

Coombs, 2006; Van Heerde et al., 2007; Chen et al., 2009) and cause an extensive amount of product

returns (Genchev et al., 2011). While the long-term effects of supply chain disruptions, including

product recalls on a firm’s brand, reputation or future revenues, are difficult to estimate (Hendricks and

Singhal, 2005a, Zhao et al., 2013), estimating the short-term impact on shareholder wealth is possible

using event study methodology (cf. Hendricks and Singhal, 2003; Eilert et al., 2017). The key premise

behind this methodology is that an efficient market reacts instantaneously to an event (in our case a

product recall announcement), that could change a stock price evaluation (Brown and Warner, 1985;

MacKinlay, 1997).

To date, the relationship between product recall announcements and shareholder wealth, using

event study methodology, has been examined, but studies are limited in their scope and inconclusive

regarding the levels of impact. Product recalls have been studied: a) in isolation of factors that

influence the magnitude and directionality of investors’ reactions (e.g. Jarrell and Peltzman, 1985;

Govindaraj et al., 2004); and/or b) where influencing factors – such as industry sector, recall strategies

or hazard – have been considered, but have been examined in specific supply chains or product

categories (e.g. retail focus by Ni et al., 2014) or geographical samples (e.g. the Chinese market only

by Zhao et al., 2013).

In order to extend our understanding of the short-term financial impact of product recalls, we

employed the event study methodology on a global, cross-industry sample of 296 product recall

announcements, spanning over a period of ten years. In doing so, we examine both; a) the relationship

between product recall announcements and investors’ reactions and b) the effects of three influencing

factors: supply chain sector, recall strategy and hazard.

This study makes two key contributions: first, it extends our understanding of product recall

announcements on shareholder wealth in a global, cross-supply chain context, which has not been done

to date; second, it provides a more granular insight into the effects that industry, recall strategy and

hazard have on the magnitude and directionality of investors’ reactions.

In the next section we introduce the theoretical background and develop the research hypotheses.

This is followed by the method employed in the research. We then present the results of the study,

followed by the discussion. Finally, we close the paper with conclusions and managerial implications.

2. THEORETICAL BACKGROUND AND HYPOTHESIS DEVELOPMENT

By drawing on inter-disciplinary literature from supply chain management, marketing, economics

and finance, we develop in this section a series of testable hypotheses, beginning with a financial

impact of product recalls.

2.1. Financial Impact of Product Recalls

Studies examining the short-term impact of product recalls on manufacturers’ share price have

been largely uni-sectorial. Manufacturers in automotive supply chains have been the focus of studies by

Jarrell and Peltzman (1985); Hoffer et al. (1987); Bromiley and Marcus (1989); Govindaraj et al.

(2004); Chen et al. (2009); Zhao et al. (2013), pharmaceutical industry by Jarrell and Peltzman (1985);

Pruitt and Peterson (1986); Ahmed et al. (2002); Zhao et al. (2013), with a few examining other

Michael Bernon, Marko Bastl, Wenqian Zhang and Mark Johnson

industries, such as, electronics, food, consumer goods and toys respectively (e.g. Pruitt and Peterson,

1986; Thomsen and McKenzie, 2001; Chen et al., 2009).

In general terms, firms that experience supply chain glitches report on average 6.92% lower sales

growth, 10.66% higher growth in cost, and 13.88% higher growth in inventories (Hendricks and

Singhal, 2005b). The extant studies show that product recall announcements result in the decline of the

affected firm’s stock price. For example, Jarrell and Peltzman (1985) showed that in both the

pharmaceutical and automotive industries, shareholders experienced financial losses that exceeded the

direct costs of recalling faulty drugs and automobiles. Zhao et al. (2013), showed that the Chinese stock

market also reacts significantly negatively to product recall announcements. A similar conclusion about

the negative impact of product returns on shareholder wealth was drawn by Ni et al. (2014), in their

study of retailers.

However, the extant research is equivocal that the stock market regards a product recall as

negative; moreover, there is a lack of consensus about the magnitude of the markets’ reaction. These

estimates range from -0.4% (Thomsen and Mackenzie, 2001) to -10.57% (Govindaraj et al., 2004).

Given the above, we hypothesize:

H1: A product recall announcement will generate a negative stock market reaction.

2.2. Effect of industry

Beyond the impact of product recalls on shareholders’ wealth, we are interested in more specific

factors that can influence the stock market’s reaction. To date, research has generally examined the

impact of product recalls on shareholder wealth using single industry samples, within which the

automotive supply chains dominate (e.g. Hoffer et al., 1987; Bromiley and Marcus, 1989; Govindaraj

et al., 2004). The few studies that utilized multi-industry samples, did not test the effects of industry

(e.g. Jarrell and Peltzman, 1985; Pruitt and Peterson, 1986), or tested them in a particular geography

such as China (Zhao et al., 2013). We anticipate that investors’ reactions to product recall

announcements will vary based on:

• Product lifecycles, cash-to-cash cycles and Return on Investment (ROI) periods differ between

sectors. For example, in Pharmaceutical supply chains, a substantial upfront Research and

Development (R&D) investment spanning anywhere between ten to fifteen years is needed

before a drug is launched on a market. The R&D processes are stringent and must be agreed

by the regulatory bodies before the drug can be launched (Ahmed et al., 2002; Di Masi et al.,

2016). If successful, the drug will be marketed over a long period of time to recover the

investment, before patent expiration and generics take significant market share. If an issue is

found in a drug, it may be recalled and there could be a lengthy period before the drug re-

enters the market. A drug may be ‘withdrawn’ rather than being simply recalled, if an

associated health hazard is linked with it. As a consequence, a company will not only have

difficulties in recovering the upfront investment, but it may also have to absorb costs related

to law suits in case of significant health hazard for consumers. In the food sector, while health

hazards to consumers may be similar (e.g. food poisoning resulting in severe illness), a food

product can be recalled and replaced with a new product once the problem is discovered,

without significant long-term additional costs. In the automotive supply chains, although a

product recall may be associated with hazard, it can normally be rectified quickly through

dealerships replacing defective parts. Even if an Original Equipment Manufacturer (OEM)

has to redesign a part, this would take significantly less time than changing, for example, a

compound in a medicine, nor would it result in a withdrawal of a particular car from the

market.

• The frequency and value per claim of product recalls vary between industries. According to

Allianz Insurance report, which examined latest trends in product recalls (Allianz Global

Corporate and Speciality, 2017), automotive industry leads in terms of recall frequency - and

it is responsible for 42% of all claims - followed by food/beverage (18%) and domestic

appliances (10%). Automotive industry also leads in average product recall claim value, with

EUR 2.12 million, followed by food/beverage at EUR 1.31 million and IT/Electronics with

EUR 1.1 million. Comparatively, in pharmaceutical sector, the announcements are less

frequent, due to the stringent testing and regulations in place (Narayana et al., 2014). In

consequence, investors will factor both frequency and value of product recalls into their

valuations and react differently to announcements in different industries.

• Industries also vary in terms of the ubiquitousness of their products. For example, food is a

necessity rather than a luxury. Consequently, a significantly larger number of people can be

affected by defective food, than, for example, by a defective toy. Also, as Zhao et al. (2013)

Int. Journal of Business Science and Applied Management / Business-and-Management.org

argued, the effect of bad food on the health of consumers is almost instantaneous, which is

rarely the case with a defective car. Moreover, brand loyalty and availability of substitutes

plays an important role in a market’s reaction to product defects (Rhee and Haunschild,

2008). We posit that brand loyalty is much lower in the food industry compared to the

automotive, and the choice of substitutes is higher. This postulation was evident in the 2013

European horsemeat contaminated beef burgers scandal. Tesco, the UK’s largest supermarket,

recorded a reduction in their overall sales ranging from -5.5% in the UK to -13% in Turkey

(Leach, 2013).

This leads us to the following hypothesis:

H2: The reduction in stock price related to product recalls will have industry specific impacts.

2.3. Effect of Recall Strategy

In line with Zhao et al. (2013), we adopt Signalling theory to examine the impact of recall strategy

and hazard in a product recall. Signalling theory is concerned with reducing information asymmetry

between two parties and the description of behaviours when two parties (e.g. individuals or

organisations) have access to different information (Spence, 2002). The theory hypothesises that a

receiver (e.g. an investor and/or a consumer), interprets a signal from a sender (e.g. a firm that

discovered a product defect) about the quality of a product and the firm’s intent. The availability and

clarity of information affects the decision making process. Individuals make decisions based on

publically available information, as well as private information, which is available only to a subset of

the public (Connelly et al., 2011).

In the context of product recalls, Chen et al. (2009), show that during a product recall crisis,

information asymmetry (i.e. difference in the access to the same information) between firms,

consumers and stock market investors, increases. Normally, through their traceability systems (Dai et

al., 2015), firms possess significantly more information about a product recall than the stock market or

consumers. Conversely, the stock market relies on multiple external sources of information, such as

corporate or governmental announcements and the business press, to analyse a firm’s actions and

strategies, and to interpret those signals in terms of future earnings and firm value (Ross, 1977).

Firms’ responses to product recalls differ. The extant literature (Dawar and Pillutla, 2000; Laufer

and Coombs, 2006) classifies firms’ responses to product recalls into four groups: denial, forced

compliance (involuntary recall), voluntary recall, and ‘super-effort’. Broadly, denial and involuntary

recall fall under the category of passive responses, while voluntary recall and super-effort are proactive

responses.

In a proactive response, a firm that discovers a product defect, either through internal inspections

or external sources, releases a voluntary product recall. Kumar and Schmitz (2011) state that in 2009,

the US Consumer Product Safety Commission recorded 465 voluntary product recalls involving 229.6

million product units. Voluntary recalls are released prior to any safety incidents or concerns being

reported by consumers. In 2016, for example, Apple announced a voluntary recall of millions of its

two prong AC wall plugs, after it became aware of their potential to break, causing an electric shock

(BBC News, 2016).

A passive response, conversely, involves delaying the recall process or shifting the blame to other

supply chain members, such as distributors, wholesalers and/or suppliers. Firms are particularly

motivated to adopt a passive response when they discover a serious and pervasive product defect where

they cannot profitably repair or replace all defective products. In these situations, companies may

attempt to deceive consumers about unobserved quality attributes, hoping that the issues remain

undetectable (Zhao et al., 2013). Consequently, firms that adopt a passive recall strategy tend to issue

recall announcements much later, compared to proactive companies, often after serious complaints,

injuries or even death to consumers (Chen et al., 2009).

The current evidence as to which product recall strategy is penalized with a higher negative

abnormal stock return is mixed. Zhao et al. (2013) found stock markets in China reacted significantly

more negatively to passive recalls. Their findings suggest that in China, investors perceive companies

who adopt a proactive approach to product recalls as socially more responsible. This is in line with

Margolis et al. (2007), who posit that a proactive approach enhances consumers’ confidence in the

firms’ products and helps firms to recover sooner. Siegel and Vitaliano (2007) argue that a proactive

approach is seen by consumers and investors as a signal of corporate responsibility, even when that

event leads to reduced cash flows that may devalue a firm. In contrast, Chen et al. (2009) found that

proactive recall strategies have a more negative effect on a firm’s financial value than passive

strategies. While the extant literature is mixed, most research suggests taking a proactive approach will

be less punitive. Thus, we hypothesize:

Michael Bernon, Marko Bastl, Wenqian Zhang and Mark Johnson

H3: A passive product recall strategy will result in more negative abnormal returns than a

proactive product recall strategy.

2.4. Effect of hazard

The hazard that recalled products pose to consumers varies between different recalls (e.g. Ahmed

et al., 2002; Ni et al., 2014). Federal departments and previous research classify product hazard in

various ways. The US Food and Drug Administration (FDA) applies a three-level classification to

recalls based on product hazard (Food and Drug Administration, 2014). Class I refers to dangerous or

defective products that could cause serious health problems or death; Class II refers to recalls where

products might cause a temporary health problem, or pose only a slight threat of a serious nature; and

Class III refers to recalls where products are unlikely to cause any adverse health reaction, but violate

FDA labelling or manufacturing laws. Jarrell and Peltzman (1985) applied this classification to

distinguish between severities of drug-related recalls.

For the assessment of hazard in automotive sector, Crafton et al. (1981) and Reilly and Hoffer

(1983) developed recall classifications for the severity level of vehicle recalls, when investigating the

impact of automotive recalls on consumer demand. Their classification labelled minor problems, such

as mislabelled or missing placards and tyre-related difficulties, as a Type I recall; Type II recalls

considered intermediate problems such as defective windshield wipers or problems with carburettor

brackets; while Type III were recalls due to severe safety hazards, such as loss of steering and braking

functions or early failure of an axle shaft.

Empirically, the effect of hazard on manufacturers, using cross-sectorial studies is yet to be tested.

Given the variation in hazard levels, investors may react differently to recalls where faulty products

pose a risk of severe impact to the health or even the death of consumers, as opposed to those where

there are unlikely to be any adverse effects. We offer three arguments for this: First, firms would

normally send a signal to the investors and public about the level of hazard posed by the product to be

recalled. We anticipate that, irrespective of industry, investors would associate increased levels of

hazard, leading to serious personal injury or death, as a signal of more serious, costly, and difficult to

resolve problems. Such problems may result in the loss of market share, decreased profitability,

damaged reputation and lengthy lawsuits (Zhao et al., 2013).

Second, due to loss aversion, investors’ interpretation of a product recall signal that is linked to the

greater hazard, results in greater stock depreciation. This reaction is rooted in the notion of loss

aversion, which suggests that change for the worse is perceived in people’s minds larger than an

equivalent change for better (Novemsky and Kahneman 2005).

Third, as posited by Ni et al. (2014), product recall announcements linked to serious hazard are

likely to receive substantially more media coverage. The FDA’s and the Consumer Product Safety

Commission (CPSC) websites are regularly updated with a plethora of product recalls with low levels

of hazard, which never find their way into mainstream media. However, the newsworthy nature of

recalls linked to serious injuries or death, is more likely to be reported by the media (Barber and Odean,

2008). Therefore, as argued by Ni et al. (2014, p. 314): “product recall announcements for a more

severe product safety issue will receive greater media attention, leading to greater disutility by

stakeholders.” Thus, we hypothesize that:

H4: A higher level of hazard will lead to a greater reduction in stock price.

3. METHODOLOGY

In line with other research that examined the impact of product recalls on shareholder wealth (cf.

Jarrell and Peltzman, 1985; Zhao et al., 2013; Ni et al., 2014), we used the event study methodology. In

the following section we describe the data collection and analysis procedures.

3.1. Data

The process of identifying product recall announcements started with a free text search using the

Factiva database focused on the Wall Street Journal (WSJ) - All Sources. This included its US,

European and Asian printed editions, plus the online edition. Use of the WSJ as a source for recall

announcements is in line with other studies that used the event study methodology (e.g. Hendricks and

Singhal, 2003).

The search covered a ten year period from 2005 until 2014. The key search word used was

“recall*”, which yielded articles containing words such as “recall”, “recalls”, “recalled” or “recalling”.

Each article/announcement was screened to identify the nature of the recall and whether it was germane

to product recalls. In this process we eliminated articles/announcements with the following

characteristics:

Int. Journal of Business Science and Applied Management / Business-and-Management.org

• Articles/announcements which reported the same information made in an earlier

announcement, unless new information was disclosed. Thus, only the first announcement of a

product recall was taken into consideration;

• Articles/announcements related to privately owned companies, as there would be no publicly

traded shares; and

• Articles/announcements and earnings pre-announcements where product recalls were

mentioned. Product recalls can influence earnings expectations (Bowen et al., 1992), as they

contain factors that can bias the perceptions of a firm’s performance.

The final sample consisted of 296, non-contaminated (i.e. no other recalls within the event

window), product recall announcements, comprised of 150 recalls (50.7%) from automotive supply

chain manufacturers; followed by 72 recalls from manufacturers in pharmaceutical supply chains

(24.3%), 21 in food (7.1%), 20 in electronics (6.8%), 9 from toy (3.0%) and 24 (8.1%) from other

supply chains. Following the sample selection, we developed a cross-industrial classification of hazard

levels. This was necessary for testing H4, given that the selected announcements spanned multiple

industries. We started the process by reviewing the classifications of three main US government

regulation agencies:

• The Consumer Product Safety Commission (CPSC) responsible for consumer products;

• The Food and Drug Administration (FDA) which regulates consumer products encompassing

food, pharmaceuticals, medical devices and cosmetics; and

• The National Highway Traffic Safety Association (NHTSA), responsible for regulating the

automotive industry.

We further reviewed regulatory bodies in other countries and industries and consolidated these in

the following way (see Table 1):

• Class I involves recalls that could lead to death or severe injury;

• Class II comprises recalls that may cause a temporary health problem or a minor to moderate

injury; and

• Class III is composed of defects that are not likely to have health or safety threats, but breach

the legislation.

Table 1: Hazard level classification across industry

Automotive

Pharmaceutical & Food

Toy & Electronics

Class I

Death or severe injury caused by

road accidents or fire, or risks that

can be devastating

Dangerous or defective products

that predictably could cause serious

health problems or death

Death or severe injury caused by

choking or exploding, etc., or

predictably could cause serious

health problems

Class II

Minor or moderate injury caused by

road accidents or fire

Products that might cause a

temporary health problem, or pose

only a slight threat of a serious

nature

Minor or moderate injury or adverse

health reaction

Class III

Defects that are not causing any

road accidents or fires

Products that are unlikely to cause

any adverse health reaction, but that

violate FDA labelling or

manufacturing laws

Defects that are unlikely to cause

any adverse health reaction or

safety-related risks

3.2. Event Study Method

In order to determine the impact of product recalls on shareholder wealth, we utilized the event

study methodology (McWilliams and Siegel, 1997; Hendricks and Singhal, 2003; Eilert et al., 2017).

This methodology utilizes daily stock returns to estimate the abnormal share price changes due to

product recall announcements. It takes into consideration both industry and systematic risks, while also

estimating investors’ reactions to specific events (Brown and Warner, 1980, 1985; MacKinlay, 1997).

It assumes that an efficient market immediately reacts when an event that could change a stock price

value is announced. The method has been previously applied in the fields of operations and supply

chain management, marketing, information technology and accounting, to examine phenomena such as

the increase in capital expenditure (McConnell and Muscarella, 1985), new product introduction

Michael Bernon, Marko Bastl, Wenqian Zhang and Mark Johnson

(Chaney et al., 1991), strategic alliance or acquisitions (Chan et al., 1997), supply chain glitches

(Hendricks and Singhal, 2003), and innovative IT investments (Dos Santos et al., 1993).

We adopted this method in line with Hendricks and Singhal (2003), Zhao et al. (2013) and Ni et

al. (2014), who used a modified version of Fama et al. (1969), to evaluate the daily stock returns.

Appendix A provides an overview of the method used. The following section contains the results.

4. RESULTS

4.1. The Financial Impact of Product Recalls

The results from the event study are summarized in Tables 2 and 3. Table 2 shows the daily

abnormal return during the event window of 296 product recall announcements. It demonstrates that on

announcement day (Day 0), the mean abnormal return is -1.34%, while the cumulative mean abnormal

return over the 3-day event period is -1.87%. As shown in Table 2, there are statistically significant

results for mean, median and percentage of negative abnormal return on Day -1 and Day 0, while the

abnormal return on Day 1 is not significant at the 5% level. The negative abnormal return on days -1

and 0 is in line with the results of other product recall studies using the methodology (e.g. Ahmed et al.,

2002; Chu et al., 2005; Zhao et al., 2013). However, the magnitude of the negative abnormal return in

our study (i.e. -1.87%), is significantly higher than in previous studies that examined more than one

industry (e.g. Jarrell and Peltzman, 1985; Pruitt and Peterson, 1986; Chu et al., 2005).

Table 2: Abnormal returns for 296 product recalls announcements - event period from day -1 to

day 1

Day -1

Day 0

Day 1

Event Period

(Day -1 to Day 1)

Mean abnormal return

-0.30%

-1.34%

-0.23%

-1.87%

t-statistic

-3.04

-2.96

-1.44

-4.21

Median abnormal return

-0.11%

-0.25%

-0.01

-0.222%

Wilcoxon signed-rank test Z-

statistic

-1.726

-2.049

-.694

-2.554

Abnormal returns negative (%)

55.74%

58.78%

52.03%

61.82%

Table 2 indicates that the results support H1 that investors react negatively to product recall

announcements. This results in a significant (t=4.21) cumulative abnormal return during the event

window, and a negative impact on shareholder wealth. Table 3 shows the Dollar change in the stock

price for the product recall announcements in this study.

Table 3: Description of Dollar change in stock price for the 296 product recall announcements

Mean

-US$362.6m

Median

-US$55.7m

S.D.

US$2,564.4m

Maximum

US$5,840.4m

Minimum

-US$37,250.4m

From Table 3 it can be determined that the average Dollar loss over the event period (day -1 > day

1) is US$ 362.6 million.

4.2. The Effects of Industry, Recall Strategy and Hazard on Abnormal Stock Returns

4.2.1. Descriptive Statistics

Table 4 shows the descriptive, cross-tabulated results for both industry and hazard level. The

figure in parenthesis is the proportion of hazard level results within the industry.

Table 4: Descriptive results of industry and hazard classification

Industry

Class I

Class II

Class III

Unclassified

Total

Automotive

4 (2.67%)

12 (8.00%)

125 (83.33%)

9 (6.00%)

150

Int. Journal of Business Science and Applied Management / Business-and-Management.org

Pharmaceutical

24 (33.33%)

25 (34.73%)

15 (20.83%)

8 (11.11%)

Food

4 (19.05%)

7 (33.33%)

6 (28.57%)

4 (19.05%)

Electronics

5 (25.00%)

13 (65.00%)

2 (10.00%)

Toys

6 (66.67%)

3 (33.33%)

Miscellaneous

8 (33.33%)

6 (25.00%)

8 (33.33%)

2 (8.34%)

Table 4 shows that recalls from manufacturers in automotive supply chains are the largest

proportion of recalls within the sample. Pharmaceutical recalls are the next largest number (n=72) and

they comprise one third of recalls for the most severe form (Class I).

4.2.2. Effects of Industry

The statistical results for Cumulative Average Abnormal Return (CAAR) by industry are

presented in Table 5.

Table 5: Statistical results of cumulative average abnormal return by industry

Industry

Mean

Std.

Deviation

Std. Error

p-value

(t-value)

Automotive

150

-0.0078

0.0210

0.0017

0.000 (-4.536)

Pharmaceutical

-0.0460

0.1408

0.0166

0.007 (-2.774)

Food

-0.0016

0.0185

0.0040

0.692 (-0.402)

Electronics

-0.0024

0.0182

0.0041

0.565 (-0.586)

Toy

-0.0258

0.0713

0.0238

0.309 (-1.085)

Miscellaneous

-0.0395

0.0788

0.0191

0.055 (-2.066)

From Table 5 it can be seen from the mean results that the manufacturers in pharmaceutical supply

chains have the highest CAAR during the event period (day -1 > day 1), followed by miscellaneous

(mean= -0.0395) and the toy supply chains (mean= -0.0258). However, only the automotive (p= 0.000,

t= -4.536) and pharmaceutical industries (p= 0.007, t= -2.774) underwent significant changes to share

prices. In order to check for statistically significant differences between groups (cf. Hendricks and

Singhal, 2003), we conducted an analysis of variance (ANOVA). The ANOVA indicated that there was

a statistically significant difference between groups (p= 0.014), indicating that the CAAR was different

between sectors. Thus, H2 (The reduction in stock price related to product recalls will have industry-

specific impacts) is supported.

4.2.3. Effects of Recall Strategy

The impact of product recall strategy (i.e. proactive or passive), on CAAR is captured in Table 6.

From the table, it can be determined that passive recalls have a CAAR that is statistically significantly

different compared to the market. This indicates that manufacturers that adopted a passive strategy

suffered a greater loss in share price compared to those that adopted a proactive strategy. The results of

the ANOVA (p= 0.372) indicate there is no statistically significant difference between recall strategy

types, potentially indicating there is some overlap in the distributions for each strategy.

Table 6: Statistical results of cumulative average abnormal return (CAAR) by recall strategy

Mean

Std.

Deviation

Std. Error

p-value

(t-value)

Proactive

-0.011

0.068

0.007

0.098 (-1.670)

Passive

136

-0.025

0.091

0.008

0.002 (-3.222)

N/A

-0.016

0.047

0.006

0.011 (-2.636)

Total

296

-0.019

0.076

0.004

0.000 (-4.214)

These results suggest that H3 is not supported. This may be due to the wider spread of samples for

proactive recalls (both positive and negative share price changes), while passive recall samples are

more concentrated in the negative direction. Thus, it can be posited that only passive recalls will be

punished by the stock market. Conversely, for proactive recalls, the financial consequences can be both

positive and negative.

Michael Bernon, Marko Bastl, Wenqian Zhang and Mark Johnson

4.2.4. Effects of Hazard Level

Table 7 shows the results of the CAAR by hazard level.

Table 7: Statistical results of cumulative average abnormal return by hazard level

Mean

Std. Deviation

Std. Error

p-value (t-value)

Class I

-0.0672

0.1536

0.0226

0.005 (-2.965)

Class II

-0.0107

0.0260

0.0035

0.004 (-3.053)

Class III

167

-0.0054

0.0220

0.0017

0.002 (-3.173)

N/A

-0.0341

0.1208

0.0228

0.147 (-1.494)

Total

296

-0.0187

0.0764

0.0228

0.000 (-4.214)

As is evident from Table 7, the CAAR of the stock will become more negative as the hazard level

increases – i.e. from a Class III recall through to a Class I recall. Overall, there is a significant

difference according to hazard level (p= 0.000).

We conducted a multiple comparison analysis, which determined there is no significant difference

(p= 0.643) between Class II and Class III recalls; a statistically significant difference is detected

between Class I and Class II recalls (p =0.000), and between Class I and Class III recalls (p= 0.000).

Therefore, it can be posited that the stock market reacts differently to hazard level, with more

hazardous events leading to greater reductions in share price. The results of the analysis provide

support to H4 (A higher level of hazard will lead to a greater reduction in stock price).

5. DISCUSSION AND CONCLUSIONS

In this study we investigate the key characteristics of manufacturer firms’ product recalls on the

stock market reaction. In Table 8 we present the hypotheses within this study and whether they were

supported or unsupported.

Table 8: Summary of tested hypothesis and their empirical support

No.

Hypothesis

Supported?

A product recall announcement will generate a negative stock market reaction.

Supported

The reduction in stock price related to product recalls will have industry-specific

impacts.

Supported

A passive product recall strategy will result in more negative abnormal returns than

a proactive product recall strategy

Unsupported

A higher level of hazard will be associated with more severe market penalty.

Supported

The contributions of this study are twofold. First, the results of the study revealed a significant

negative stock market reaction to product recalls, confirming H1. The direction of the stock market

reaction on days -1 and 0 is in line with previous literature. However, the magnitude of the reaction (i.e.

-1.87%) in our study is significantly higher than in previous studies, where more than one industry was

studied, for example -0.81% on day -1 (Jarrell and Peltzman, 1985) and -0.76% on days -1 and 0 (Pruitt

and Peterson, 1986) based on US data, and lower when comparing to Zhao et al.’s (2013) findings of -

2.21% based on data from China. As this study is based on global data, the result of -1.87% is within

this range and possibly reflects the impact of both developed and developing countries.

Second, we provide further granularity in the relationship between product recalls and

shareholder’s wealth by examining the effects of industry, recall strategy and hazard. Our findings

indicated that there were differences in the impact between different sectors. Both, manufacturers in

automotive (t= -4.536) and pharmaceutical supply chains (t= -2.774) had significant negative abnormal

returns. This difference is also significant (p= 0.001) with pharmaceutical showing more variability

when compared to automotive (SD= 14.174% vs. 2.101%). We suggest there are two distinct reasons

behind this. For manufacturers in pharmaceutical supply chains, one third of the total recalls were for

the most severe type of recall, i.e. those likely to cause death. They often result not only in the need for

a complete drug withdrawal and consequent loss of market, but also in very high litigation costs. For

example, Merck’s legal costs related to Vioxx withdrawal were estimated to exceed $7.7bn

(Bloomberg News, 2010). For the recalls in automotive supply chains, investors may perceive that the

recall is indicative of a more costly and pervasive problem, with the recalled product being shared

Int. Journal of Business Science and Applied Management / Business-and-Management.org

between different automotive firms due to supply base over-rationalisation (cf. Choi and Linton, 2011)

and sharing of the same supplier (Yan et al., 2015). For example, Takata was a shared supplier of

multiple automotive OEM’s whose airbag recall impacted more than 20 manufacturers. By comparison,

toy recalls were not found to be statistically significant. We suggest this was due to only having four

firms in our sample. Whilst it is logical to expect food recalls to have a significant impact, we suggest

that due to diversification of brands and range of stock keeping units within food manufacturers’

portfolios, the results are to be expected.

Furthermore, rooted in signalling theory, we also examined the effects of recall strategy. The

results of the analysis showed no statistically significant difference between the abnormal returns of

passive and proactive recalls. Moreover, the effects of proactive recall strategies can have mixed

financial impact. This is an important finding and it is in contrast to the extant literature (e.g. and

Davidson and Worrell, 1992; Zhao et al., 2013), which suggests significant differences between passive

and proactive recalls as well as passive recall strategies being more punitive than proactive. Our finding

suggests that manufacturers who adopt proactive recall strategy may not always elicit positive

responses from the investors. This suggests that investors may not always see the proactive recall

strategy in the same light as consumers, potentially interpreting them as a signal of more severe product

hazard, potential financial damage or, as shown in the study of Hora et al. (2011) - longer recall times.

The effects of hazard have been largely left out of examinations of product recall announcements

on cross-sectorial samples. We have shown that higher hazards lead to greater negative abnormal

returns, and such plays an important role in the directionality and the magnitude of stock price

movement for three key reasons; a) severe hazards, such as consumers’ death, are indicative of

products that require costly rectification or even a complete withdrawal from an entire supply chain; b)

the more severe the hazard, the higher the expected litigation costs and; c) extensive media coverage of

recalls associated with severe hazard is resulting in greater disutility by stakeholders. Our findings

extend the findings of Ni et al. (2014), Ahmed et al. (2002) and Thomsen and McKenzie (2001),

showing that irrespective of the structural position of a firm in a supply chain, investors will react more

strongly to recalls that are associated with the higher risk of injury or death.

5.1. Managerial implications

Given the global nature of supply chains, supply base rationalization, tougher regulations and

economic pressures, product recalls will remain ubiquitous. It is unlikely that a firm could change

sector to one that was less sensitive to recall announcements, unlike the pharmaceutical and automotive

sectors. Firms do, however, need to be transparent and proactive when recalling products, as this can

result in minimization of short-term as well as long-term financial consequences. Consumers and

shareholders are interested in whether a firm cares. The results around hazard and sector also suggest

that firms need to be incredibly rigorous in ensuring the safety and quality of their products in various

stages, from product development, sourcing, manufacturing and distribution. The manufacturers should

also be motivated to play an active role, not only in ensuring internal quality standards, but also to

proactively manage their supply network partners.

5.2. Future research

This work examined four sectors; future research could examine a broader range to unveil further

sectoral differences and seek to examine whether the source of the recall (e.g. focal firm, partner,

supply chain) leads to a greater or lesser impact upon share price. Further, considering the impact of

severe hazard on shareholder value, research is needed to improve strategies and quality procedures to

mitigate the occurrence of hazard.

In our research, we focussed on major recall announcements published by news agencies. While

this is important, the vast majority of recalls do not gain such press attention. However, agencies such

as the US CPSC and NHTSA provide data on all recalls reported within their area of responsibility.

Further research could evaluate the effects of the full range of recall announcements, to establish

further granularity of the financial implications on shareholder values.

Finally, the research could be extended to include other potential mediating factors, such as, the

effects of branded and non branded products. For example: do firms with strong brand loyalty suffer

lower – or perhaps higher – decreases in share price?

REFERENCES

Ahmed, P., Gardella, J., & Nanda, S. (2002). Wealth Effect of Drug Withdrawals on Firms and their

Competitors. Financial Management, 31, 3, 21–41.

Michael Bernon, Marko Bastl, Wenqian Zhang and Mark Johnson

Allianz Global Croporate and Speciality. (2017). Product Recall - Managing the impact of the new risk

landscape.URL:http://www.agcs.allianz.com/assets/PDFs/Reports/AGCS_ProductRecallReport.pd

Barber, B.M., & Odean, T. (2012). All that Glitters: The effect of Attention and news on the Buying

Behavior of Individual and Institutional Investors. The Handbook of News Analytics in Finance,

21, 2, 173–210.

BBC News. (2016). Apple plug recall over safety concerns. URL:

http://www.bbc.co.uk/news/technology-35432682.

BBC News. (2013). QandA: horsemeat scandal. URL: http://www.bbc.co.uk/news/uk-21335872.

Bloomberg News. (2010). Merck’s Reserves $950 Million for Vioxx Crime Probe.

URL:http://www.bloomberg.com/news/articles/2010-10-29/merck-legal-costs-for-vioxx-reduce-

profit-6-years-after-recall.

Bowen, R.M., Johnson, M.F., Shevlin, T., & Shores, D. (1992). Determinants of the Timing of

Quarterly Earnings Announcements. Journal of Accounting, Auditing & Finance.

Bromiley, P., & Marcus, A. (1989). The Deterrent to Dubious Corporate Behavior: Profitability,

Probability and Safety Recalls. Strategic Management Journal, 10, 3, 233-250.

Brown, S.J., & Warner, J.B. (1985). Using daily stock returns. The case of event studies. Journal of

Financial Economics, 4, 1, 3–31.

Chan, S.H., Kensinger, J.W., Keown, A.J., & Martin, J.D. (1997). Do strategic alliances create value?

Journal of Financial Economics, 46 2, 199–221.

Chaney, P.K., Devinney, T.M., & Winer, R.S. (1991). The Impact of New Product Introductions on the

Market Value of the Firm. Journal of Business, 64, 4, 573–610.

Chang, W., Ellinger, A.E., & Blackhurst, J. (2015). A contextual approach to supply chain risk

mitigation. The International Journal of Logistics Management, 26, 3, 642–656.

Chen, Y., Ganesan, S., & Liu, Y. (2009). Does a Firm’s Product-Recall Strategy Affect Its Financial

Value? An Examination of Strategic Alternatives During Product-Harm Crises. Journal of

Marketing, 73, 6, 214–226.

Choi, T., & Linton, T. (2011). Don’t let your supply chain control your business. Harvard Business

Review.

Chu, T.-H., Lin, C.-C., & Prather, L.J. (2005). An Extension of Security Price Reactions around

Product Recall Announcements. Quarterly Journal of Business and Economics, 44, 3, 33–48.

Connelly, B.L., Certo, S.T., Ireland, R.D., & Reutzel, C.R. (2011). Signalling theory: A review and

assessment. Journal of Management, 37, 1, 39–67.

Dai, H., Tseng, M.M., & Zipkin, P.H. (2015). Design of traceability systems for product recall.

International Journal of Production Research, 53, 2, 511–531.

Davidson, W.N., & Worrell, D.L. (1992). Research notes and communications: The effect of product

recall announcements on shareholder wealth. Strategic Management Journal, 13, 6, 467–473.

Dawar, N., & Pillutla, M.M. (2000). Impact of Product-Harm Crises on Brand Equity: The Moderating

Role of Consumer Expectations. Journal of Marketing Research, 37, 215–226.

DiMasi, J.A., Grabowski, H.G., & Hansen, R.W. (2016). Innovation in the pharmaceutical industry:

New estimates of R&D costs. Journal of Health Economics, 47, 20–33.

Dos Santos, B.L., Peffers, K., & Mauer, D.C. (1993). The impact of information technology investment

announcements on the market value of the firm. Information Systems Research, 4, 1–23.

Eilert, M., Jayachandran, S., Kalaignanam, K., & Swartz, T.A. (2017). Does It Pay to Recall Your

Product Early? An Empirical Investigation in the Automobile Industry. Journal of Marketing, 81,

3, 111–129.

Fama, E., Fisher, L., Jensen, M., & Roll, R. (1969). The Adjustment Of Stock Prices To New

Information. International Economic Review, 10, 1, 1–21.

Food and Drug Administration. (2014). Background and Definitions, URL:

http://www.fda.gov/Safety/Recalls/ucm165546.htm.

Genchev, S.E., Glenn Richey, R., & Gabler, C.B. (2011). Evaluating reverse logistics programs: a

suggested process formalization. The International Journal of Logistics Management, 22, 2, 242–

263.

Int. Journal of Business Science and Applied Management / Business-and-Management.org

Govindaraj, S., Jaggi, B., & Lin, B. (2004). Market overreaction to product recall revisited - The case

of firestone tires and the ford explorer. Review of Quantitative Finance and Accounting, 23, 1, 31–

54.

Grafton, S.M., Hoffer, G.E., & Reilly, R.J. (1981). Testing the impact of recalls on the demand for

automobiles. Economic Inquiry, 19, 4, 694–703.

Hendricks, K.B., & Singhal, V.R. (2003). The effect of supply chain glitches on shareholder wealth.

Journal of Operations Management, 1,5, 501–522.

Hendricks, K.B., & Singhal, V.R. (2005a). An empirical analysis of the effect of supply chain

disruptions on long-run stock price performance and equity risk of the firm. Production and

Operations Management, 14, 1, 35–52.

Hendricks, K.B., and Singhal, V.R, (2005b). Association between supply chain glitches and operating

performance. Management Science, 51, 5, 695–711.

Hoffer, G.E., Pruitt, S.W., & Reilly, R.J. (1987). Automotive recalls and informational efficiency.

Financial Review, 2, 4, 433–442.

Hora, M., Bapuji, H., & Roth, A. V. (2011). Safety hazard and time to recall: The role of recall

strategy, product defect type, and supply chain player in the U.S. toy industry. Journal of

Operations Management, 29, 7–8, 766–777.

Jarrell, G., & Peltzman, S. (1985). The Impact of Product Recalls on the Wealth of Sellers. The Journal

of Political Economy, 93, 3, 512-536.

Kumar, S., & Schmitz, S. (2011). Managing recalls in a consumer product supply chain - Root cause

analysis and measures to mitigate risks. International Journal of Production Research, 49, 1,

235–253.

Laufer, D., & Coombs, W.T. (2006). How should a company respond to a product harm crisis? The

role of corporate reputation and consumer-based cues. Business Horizons, 49, 5, 379–385.

Leach, A. (2013). Tesco sales hurt by horse meat scandal.

URL: http://www.supplymanagement.com/news/2013/tesco-sales-hurt-by-horse-meat-scandal.

MacKinlay, A.C. (1997). Event Studies in Economics and Finance. Journal of Economic Literature,

35,. 1, 13–39.

Margolis, J., Elfenbein, H., & Walsh, J. (2009). Does it pay to be good... and does it matter? A meta-

analysis of the relationship between corporate social and financial performance. SSRN Electronic

Journal, 1–68.

McConnell, J.J., & Muscarella, C.J. (1985). Corporate capital expenditure decisions and the market

value of the firm. Journal of Financial Economics, 14, 3, 399–422.

McWilliams, A., & Siegel, D. (1997). Event Studies in Management Research: Theoretical and

Empirical Issues. Academy of Management Journal, 40, 3, 626–657.

Money CNN. (2017). Takata, brought down by airbag crisis, files for bankruptcy.

URL:http://money.cnn.com/2017/06/25/news/companies/takata-bankruptcy/index.html.

Narayana, A.S., Elias, A.A., & Pati, K.R. (2014). Reverse logistics in the pharmaceuticals industry: a

systemic analysis. The International Journal of Logistics Management, 25, 2, 379–398.

Ni, J.Z., Flynn, B.B., & Jacobs, F.R. (2014). Impact of product recall announcements on retailers

financial value. International Journal of Production Economics, 153, 309–322.

Pruitt, S.W., & Peterson, D.R. (1986). Security price reactions around product recall announcements.

Journal of Financial Research, 9, 2, 113–122.

Reilley, R.J., & Hoffer, G.E. (1983). Will retarding the information flow on automobile recalls affect

consumer demand? Economic Inquiry, 21, 3, 444–447.

Rhee, M., & Haunschild, P.R. (2006). The Liability of Good Reputation: A Study of Product Recalls in

the U.S. Automobile Industry. Organization Science, 17, 1, pp. 101–117.

Ross, S.A. (1977). The Determination of Financial Structure: The Incentive-Signalling Approach. The

Bell Journal of Economics, 8, 1, 23.

Siegel, D.S., & Vitaliano, D.F. (2007). An Empirical Analysis of the Strategic Use of Corporate Social

Responsibility. Journal of Economics & Management Strategy, 16, 3,. 773–792.

Spence, M. (2002). Signaling in retrospect and the informational structure of markets. American

Economic Review.

Michael Bernon, Marko Bastl, Wenqian Zhang and Mark Johnson

Thomsen, M.R., & McKenzie, A.M. (2001). Market Incentives for Safe Foods: An Examination of

Shareholder Losses from Meat and Poultry Recalls. American Journal of Agricultural Economics,

83, 3, 526–537.

Van Heerde, H., Helsen, K., & Dekimpe, M. (2007). The impact of a product-harm crisis on marketing

effectiveness. Marketing Science, 26, 2, 230-245.

Yan, T., Choi, T.Y., Kim, Y., & Yang, Y. (2015). A Theory of the Nexus Supplier: A Critical Supplier

From A Network Perspective. Journal of Supply Chain Management, 51, 1, 52–66.

Zhao, X., Li, Y., & Flynn, B.B. (2013). The financial impact of product recall announcements in China.

International Journal of Production Economics, 142, 115–123.

APPENDIX A: OVERVIEW OF THE EVENT STUDY METHOD

In line with Hendricks and Singhal (2003), we used a modified version of the Fama et al. (1969)

market model to estimate the daily share price returns:

(1)

Where R

is the return of stock i on day t, R

is the market return on day t (calculated using the

local market index of each stock i). α

is the intercept of the relationship for stock i, β

is the slope of the

relationship for stock i with the market return R

, and ε

is the error term for stock i on day t which

captures the part of R

that cannot be explained by market movements and captures the effect of firm-

specific information. For each company, the change in the intercept , the change in the slope

of the relationship, and the variance of the error term ε

were estimated using Ordinary Least Squares

(OLS) regression over a 190-day estimation period. The event window spanned three days to include

one day prior to the announcement date (-1) and one day following the announcement date (+1). This

was done to both effectively measure the market reaction and control for confounding effects

(McWilliams and Siegel 1997). The estimation period (day -201, -12) was ended 10 trading days

before the event window to avoid any potential bias caused by the data used to estimate the parameters

of the market model (Hendricks and Singhal 2003). The abnormal return for stock i on day t is the

difference between the actual price of stock i on day t (R

) and the expected return of stock i on day t

. It is defined as:

(2)

From Fama et al. (1969), the average abnormal return across N sample observations of the sample

of firms at day t is described as:

(3)

Where N is the number of sample companies on day t. In this study, N=296. The cumulative

average abnormal return (CAAR) over the time period (t

,…,t

) is the sum of AAR

and is expressed as:

(4)

In order to determine whether the abnormal return is different to zero at a statistically significant

level, we first standardize the abnormal return by dividing the abnormal return (AAR

) by , its

estimated standard deviation:

(5)

The test statistic (TS) employed to test the statistical significance of the average abnormal return

for day t is thus defined as:

(6)

Int. Journal of Business Science and Applied Management / Business-and-Management.org

To evaluate the multiple day test statistics, assume that abnormal returns are independent and

identically distributed across time (Hendricks and Singhal 2003). Thus, the t-test over multiple days

,…t

), TS

, is presented as:

(7)