Printer Friendly

Reverse stock splits in the biotechnology industry: an effectuation approach.

INTRODUCTION

ACCORDING TO STANDARD, predictive signaling theory, reverse stock splits send a negative signal to the stock market. This prediction arises from the following logic. Managers are assumed to have superior predictive information about future cash flows. So, when the stock price is below the optimal range, and there are poor prospects about the arrival of good news regarding future cash flows, then the decision to undergo a reverse stock split (RSS) reveals the manager's negative private beliefs (or non-presence of positive beliefs). Studies show that in particular, expert managers in highly uncertain business environments do not use a predictive mental framework; rather, managers think in terms of their ability to effectuate change within their own firm's business environment. (1) Thus, in business environments with a high degree of uncertainty, there is reason to question the explanatory relevance of traditional, predictive signaling theory.

Biotech firms operate in a highly uncertain environment. The sequential progression of products, from pre-clinical and human testing to drug approval requires relatively large sums of capital and multiple rounds of financing in order to progress through critical phases of development. (2) Obtaining financing at each stage of development is crucial for the survival and eventual success of these highly volatile biotech firms. (3-5) Also, valuation of these firms is very difficult. Traditional valuation methods, such as discounted cash flow and relative valuation practices, tend to lead to under-valuation and under-investment in earlier stage drug development projects. (6) Real-option models better capture the stochastic nature of the breakthrough potential and abandonment options for biotech firms, but these models are still very difficult to implement. To place the difficulty of valuing this uncertainty in context, the stock market index for the biotechnology sector (BTK), which has outperformed the overall market, has been 9 times more volatile than the S&P 500; and 5 times more volatile than the NASDAQ (Figure 1).

Because of these features of the biotech industry, we hypothesize that the signaling properties of reverse stock splits for biotech firms will differ from the signaling properties implied by the traditional, predictive model. Investors' ability to predict success among biotech firms does not depend on being able to predict success of specific conceivable scenarios; rather, success depends more on being able to predict how well, and how likely, firms will be successful in "taking effectual action and help[ing] stakeholders make effectual commitments" in a radically uncertain future. (7) Also, with effectual processes, the environment is not exogenous to the firm's transformative actions and, because of this endogenous relationship between stakeholder action and the environment, success depends heavily on endogenous factors, like the ability to obtain stakeholder commitments and the ability to adaptively coordinate and leverage capabilities both within and without the firm. We analyze these differences between the traditional, predictive-signaling model and an effectuation-based signaling model and hypothesize that, for biotech firms, reverse stock splits should comprise a positive signal about future prospects of success.

To empirically test our hypothesis, we utilize event study methodology and find that cumulative abnormal returns (CARs) are positive following a reverse stock split for biotech firms. We also find that this effect is stronger when the split ratio is higher. These results are consistent with our hypothesis that, in accordance with effectuation theory, a manager's commitment to keeping the firm's stock price sufficiently high, in order to avoid the risk of having to delist, is a signal that the manager has positive beliefs about his or her ability to effectively exercise control over endogenous factors important to the firm's ability to succeed in the industry.

In additional cross-sectional regression analysis, we find that abnormal returns are associated with firms that are larger, have greater cash holding, are younger, have a positive amount of long-term debt, and, albeit a less robust finding, have lower market-to-book ratios. These results are consistent with our effectuation-based model in the following ways. With regard to size, larger firms have greater control over their environment, implying that our positive abnormal return hypothesis should be greater for larger firms. With regard to cash holding, firms with more cash have more means to control their environment. With regard to firm age, firms that are older are "past their prime" in the sense of failing to signal their ability to be successful even when given a reasonable amount of time to do so. With regard to long-term debt, the presence of such debt signals an ability to get financial stakeholders to commit to the future of the firm. Finally, with regard to market-to-book ratios, value firms that have more depressed market values, relative to their book value, are able to strengthen any positive impact from the signal to the market and lead to stronger positive abnormal returns. Although not all of these results are uniquely predicted by our effectuation-based framework, these results nevertheless make good sense from an effectuation perspective.

We also study liquidity and find that liquidity measures, such as turnover ratio, the proportion of days with zero returns, and the Amivest liquidity ratio (a measure of the price impact of a trade), all point to a positive impact on liquidity following a reverse stock split by biotech firms. This result is consistent with other studies of reverse splits. (8-9) In light of the positive abnormal returns, the improved liquidity implies the positive signal of the RSS has attracted more participation and trading activity from investors and, consequently, a lower cost of equity in further rounds of financing.

The remainder of our paper is organized as follows. Section 2 analyzes the biotechnology industry and effectuation theory. Section 3 discusses forward and reverse splits, and provides motivation for our empirical hypotheses and predictions. Section 4 describes empirical methodology, data, and sample summary statistics. Section 5 presents our empirical results from analyzing stock returns and liquidity. Section 6 concludes and suggests areas for future research.

AN EFFECTUATION-BASED VIEW OF THE BIOTECH INDUSTRY

The biotech industry is heavily dependent on the research and development of new drugs. Because of the significant uncertainty and long-term nature of biotech research, which typically requires multiple rounds of new financing, the biotech industry has many of the characteristics embedded in effectuation theory. In this section, we first describe the highly uncertain, non-predictive nature of the biotech industry. Then we describe effectuation theory and argue that it provides a framework for understanding the biotech industry that is more suitable than standard predictive frameworks.

THE NON-PREDICTIVE NATURE OF THE BIOTECHNOLOGY INDUSTRY

Biotechnology research is a highly uncertain, long-term affair. Predicting which particular research efforts will be successful is very difficult. Because of this, successful biotechnology firms typically pivot several times, from one area of research to another, before achieving any significant level of success. Moreover, new lines of promising research frequently appear only after initial research in some area is already begun.

This underlying uncertainty of the research process is compounded by a fundamental financial tension that biotechnology firms face: on the one hand, the vast majority of development-stage biotechnology firms have no revenues; on the other hand, these same firms must plan for long product development cycles (12 years on average from initial research to commercialization). Because of this tension, financing occurs in successive incremental rounds that provide resources to the next valuation inflection point (typically 1-3 years). This firm-specific financial risk is compounded by the volatile nature of market-wide "open windows" for subsequent financings. Thus, a firm could be progressing on research goals, but end up being unable to raise capital at accretive terms due either to investor skepticism or a down market. And because biotechnology firms rely so heavily on multiple rounds of financing, setbacks in achieving milestones can be devastating to development-stage biotechnology firms. (10)

The following example of Cytokinetics illustrates the compounding effects associated with the highly uncertain nature of biotechnology research, long-product cycles, and multiple rounds of financing. Cytokinetics was founded in 1998 in San Francisco to pursue therapeutics using a novel technology platform of cytoskeleton and the biology of muscle function to tackle the pursuit of new treatments for multiple disease areas. The company has completed eight different financings totaling $308 million since its IPO in 2004. First, the company experienced multiple setbacks in oncology, notably a Phase 2 trial for SB-715992 (ispinesib) platinum-sensitive and platinum-refractory non-small cell lung cancer (NSCLC) showed that ispinesib led to a disease stabilization rate which was insufficient to proceed to the next stage of the development. After share price declined, Cytokinetics effected a reverse split of 1-for-6, which increased the share price from $2 to $12 with a corresponding decrease in shares outstanding in June 2013. Subsequently, the company went on to complete a financing of $40 million in February 2014. In April 2014, the company announced that tirasemtiv (fast skeletal muscle troponin activator), its lead unpartnered compound, missed the primary endpoint in the Phase 2b trial in 711 patients to treat amyotrophic lateral sclerosis (ALS). On the news, its stock price immediately dropped from $8.40 to $4.59. Despite multiple setbacks, however Cytokinetics continues to move forward with large biotechnology partner Amgen which is evaluating an oral formulation of Cytokinetics' omecamtiv mecarbil in a Phase 2 trial in patients with chronic heart failure (CHF) and left ventricular systolic dysfunction.

A standard discounted-cash-flow framework for analyzing biotechnology firms, like Cytokinetics, has significant weaknesses which lead to under-valuation and under-investment. Although real option techniques can be used to improve valuation accuracy, these models quickly become very complex as the number of development pathways increases. On the contrary, effectuation theory provides an alternative and more suitable way to value and understand the biotechnology industry.

EFFECTUATION THEORY

Effectuation theory refers to "a set of means as given and focus on selecting between possible effects that can be created with that set of means," while predictive models rely on predictable processes that "take a particular effect as given and focus on selecting between means to create that effect." (11) The original effectuation model consists of four dimensions: means, affordable loss, partnership, and expecting the unexpected. (12) In the remainder of this section, we describe how these four dimensions of effectuation theory fit the biotechnology industry.

In previous studies, effectuation has been widely explored in entrepreneurship, (13) but it has also been considered in the context of corporate R&D, (14) management, (15-16) economics, (17) finance (1) and marketing. (12) However, to our knowledge, our research is the first to explore its use specifically within the context of the biopharma industry and to apply it to analyzing reverse stock splits for development stage companies. Also, although the concepts of effectuation theory have been empirically tested at both the individual and the firm level, surveys have been the predominant method of data collection. Dew et al. (11) study individual decision-making in exploring new venture success with data collected from surveys of expert entrepreneurs and compared to MBA student responses. Wiltbank et al. (1) surveyed angel investors and analyze how effectuation framing relates to success. Our research differs from these studies in that we look only at existing firm-level variables, a precedent suggested and supported by Brettel et al (14) who collected their data using surveys of European technology firms rather than adopting archival financial data as proxies. As such, their survey-based results are based on management perceptions.

Means

The "means" construct is a three dimensional variable: "what I know," "who I am," and "who I know." "What I know" tends to be defined as domain specific expertise as well as more general variables such as personality, gender, and management experience. In the biotechnology industry, this dimension is largely comprised of knowledge about the R&D process. "Who I am" is operationalized at both the individual level of analysis (such as propensity for risk and self-efficacy) and the firm level (such as patents, capital, and internal R&D). "Who I know" includes family and friends who are resourceful or well connected, including entrepreneurs, university personnel, scientists, or others experts in the innovation process. (17)

Because pharmaceutical firms enjoy high profit margins, most multinational biopharmaceutical companies have significant financial means or resources to deploy, including large cash balances, borrowing capacity and stock market values. These means allow them to invest heavily in R&D, among other things. However, their decisions on how much to invest and on what segments can differ significantly depending on their degree of diversification and priorities. For example, a diversified biopharma firm like Johnson and Johnson (J&J) gains about 37% of sales from its biopharma segment, but a more focused biopharma firm such as Biogen gains 100% of revenues from drug sales. While both earn about the same profit margins on their biopharma sales (24.4% for J&J and 23.5% for Biogen), in absolute terms, the internally generated cash available to a corporate giant like J&J ($15 billion total, $6.1 billion from biopharma) dwarfs the internally generated cash available to a stand-alone biotechnology firm like Biogen ($1.2 billion). However, this advantage in financial means is a disadvantage when it comes to managing affordable losses, as discussed below.

The relatively diminished means for development-stage biotechnology companies can be crippling. For example, development stage Aveo Oncology completed a Phase 3 trial for ASP4130 (tivozanib) in advanced renal cell carcinoma (RCC), and found a co-promotion partner in Japan-based multinational pharmaceutical Astellas Pharma. However, a FDA advisory committee known as Oncologic Drugs Advisory Committee (ODAC) voted 13-1 to recommend the agency reject tivozanib for RCC in June 2013. The FDA subsequently rejected the company's application which it faulted as uninterpretable and inconclusive, and requested a new trial be conducted in December 2013. Aveo restructured with the layoff of 140 staffers--62% of its workforce--following the advisory committee rejection. Its share price reduced from $7 to $2. Three weeks later, Astellas Pharma informed Aveo it would not pursue European approval for the drug candidate, and would stop funding RCC trials under their collaboration, which ended the company's programs.

Affordable Loss

Rather than using expected return as a criterion for investment, "each effectual stakeholder strives to invest only what he or she can afford to lose." (7) Although large firms have more financial resources than development-stage firms, implementing an affordable-loss approach is easier in smaller biotechnology firms. This is because multiple rounds of financing are frequently needed to keep biotechnology firms afloat, a mechanism that naturally limits losses. The pros and cons of the different ways that large versus small biotechnology firms manage investment decisions can be illustrated in the following examples.

As an example of a large multinational biopharma leveraging its resources to shift from a traditional internal R&D model to biopharmaceutical alliances to further its product pipeline, consider Bristol-Myers Squibb Company (BMY). BMY has been strategically aligning with small and mid-sized drug developers and biotechnology companies by targeting companies whose products and technologies address unmet medical needs and build on BMY's R&D strengths and/or create new areas of expertise. (18) The String of Pearls strategy, formalized in 2007, threads together a library of compounds and portfolio of technologies for the purpose of accelerating the discovery, clinical development and commercialization of new therapies across a broad range of therapeutic areas. However, BMY's acquisition of Inhibitex in Phase 3 clinical development for HCV (hepatitis C virus) for $2.5 billion or 167% premium resulted in a total failure. After only eight months, the lead drug trial was discontinued when a patient death resulted in a $1.8 billion write-off.

While BMY withstood the Inhibitex setback, consider the smaller, development-stage Ziopharm Oncology clinical study for ZIO-201 (palifosfamide) in metastatic soft tissue sarcoma in March 2013. The DNA alkylating agent did not meet its primary endpoint of progressionfree survival (PFS) in a Phase 3 trial, designed to assess the drug as a first-line treatment for metastatic soft tissue sarcoma. The setback resulted in the elimination of the company's entire oncology portfolio, the elimination of 65 positions (leaving approximately 30 employees), and the complete change in strategic focus on its synthetic biology programs being developed with Intrexon. Ziopharm Oncology survived, but the failure resulted in more drastic changes compared to post-setback changes implemented in larger firms like BMY.

Stakeholder Commitments

Because of the greater reliance on multiple rounds of financing, smaller firms depend more than larger firms on commitments from external stakeholders. Effectuation theory frames partnerships as collaborations with stakeholders and organizations willing to make a significant commitment to product and market development. Read et al. (12) distinguish the means-based "who I know" dimension from the stakeholder-commitment-based "partnerships" dimension by determining whether success depends on the firm itself ("means") or the other party ("who I know"--typically as a result of money, equity or a product changing hands).

In the biotechnology industry, the vast majority of the over 600 public and 8,000 private companies worldwide have no revenues or earnings, which means that their investment is funded through grants, public or private equity, and/or through partnerships with larger, better capitalized publicly traded firms. The small percentage of these firms that are successful in moving into later stages of clinical trials or actually receiving FDA approval to market a drug are often acquired by larger biopharma firms in these later stages. (19) Thus, for large, well-established firms, partnering with and/or later acquiring smaller biotechnology companies provides a viable option to committing a firm's R&D investment capital to internal development programs. (20) These partnerships, collaborative agreements, and joint ventures create powerful innovation network effects, (21) as well as allow both firms to learn to work together, providing an option for possible later acquisition.

Despite the greater information asymmetry associated with early stage novel technologies (e.g., stem cells, checkpoint inhibitors, gene therapy, cancer vaccines, RNAi), signaling mechanisms can help investors discriminate among firms' pipelines. The relevant data in this process includes clinical data (such as announcement of clinical results at medical conferences), publicly announced partnership deals (such as licensing, co-development, co-promotion), and institutional investment by specialist mutual and hedge funds. (22,23) For example, Agios Pharmaceuticals, an early stage drug development company which focused on cancer metabolism with a marquee research partnership with large biopharma Celgene, successfully completed an IPO at $18 which overshot the range of $14-16, raised an additional $106 million, and soared 60% on its first day of trading--sending the market capitalization to over $800 million.

External stakeholder commitments can also be critical in allowing a development stage company to survive a major setback. For example, development stage Rigel partnered R788 (fostamatinib) with multinational pharmaceutical company AstraZeneca (AZ). In June 2013, Rigel announced that R788 (oral spleen tyrosine kinase inhibitor) with methotrexate (MTX) did not show statistically significant improvement compared to placebo in the Phase 3 OSKIRA-3 clinical study. Of note, AZ was solely responsible for all costs and expenses, and subsequently recorded a $136 million pre-tax impairment charge to R&D expense. AZ announced that it would not proceed with regulatory filings, and returned its rights to the compound to Rigel which has since turned its primary focus to other programs.

Expecting the Unexpected

The "expect the unexpected" effectuation principle encourages companies to embrace surprises that arise from uncertain situations, remaining flexible rather than tethered to existing goals. (12) Wiltbank et al. (7) refer to this effectuation dimension as "leveraging contingencies" defined as a willingness to dramatically change goals, products, or strategies.

While all large biopharmaceutical companies have a pressing and ongoing need for new products, they have approached pipeline and product investment and development differentially, in the sense that some rely on internal development and research partnerships, while others rely on purchasing external R&D and/or smaller firms typically in later stages of FDA approval through mergers and acquisitions). Illustrating these different approaches, Pfizer has heavily relied on multibillion dollar acquisitions such as Warner-Lambert in 2000, Pharmacia in 2002, Wyeth in 2009, King Pharmaceuticals in 2010; and Roche has relied on internal development and partnerships (e.g., Genentech partnership to grow a pipeline of blockbuster oncology products such as Herceptin[R] (trastuzumab), Rituxan[R] (rituximab), and Avastin[R] (bezicuzimab) each with greater than $5 billion in 2012 annual revenues. There is a tradeoff between the perceived risk of overpaying for late-stage products, often obtained through mergers and acquisitions, and the uncertainty of valuing internally developed earlier stage products. (24)

An example of an "unexpected" event is the emergence of an unanticipated safety signal, even after extensive clinical studies. For example, Biogen-Idec and Elan's Tysabri (natalizumab) was originally approved for all relapsing forms of multiple sclerosis (relapse-remitting, secondary-progressive, and progressive-relapsing) in 2004. However, four months after its approval in February 2005, the manufacturer withdrew natalizumab voluntarily after two fatal cases of progressive multifocal leukoencephalopathy, and the stock price fell from $66 to $38. The drug was eventually re-approved in June 2006 after an extensive safety review and heavy lobbying by patients, and Tysabri reached $5.5 billion in 2012 revenues.

REVERSE STOCK SPLIT AND EMPIRICAL HYPOTHESES

In this section, we first provide a review of the existing literature on forward and reverse stock splits, paying particular attention to standard predictions based on predictive signaling theory. Then we motivate our empirical hypotheses and predictions by analyzing reverse stock splits from the perspective of effectuation theory.

FORWARD AND REVERSE STOCK SPLITS

There are two types of stock splits, forward splits and reverse splits. A forward split is when one share becomes multiple shares, resulting in more shares but a lower price per share. Between 1933 and 2007 the average share price of major U.S. stocks remained remarkably constant, rarely straying far from $25 to $35. The average forward split was $50 pre-split. Anytime a stock went much higher, the company reduced it back down with a stock split. Conversely, a reverse split occurs when multiple shares are combined into one share, resulting in less shares but a higher price per share. The average reverse split is $1.21 pre-split. (25-27)

Typically, reverse splits are done from a position of weakness such as a setback of some kind (e.g., unanticipated loss of intellectual property protection, loss of market share, natural disaster, adverse regulatory action, significant market correction) which significantly reduces the share price and threatens the company's viability as an exchange traded stock. (28-31) Further, companies must maintain minimum standards to ensure continued compliance and exchange trading.

For example, to maintain a listing on the NASDAQ stock exchange, corporations are required to meet minimum standards for their share price, market value and corporate governance. Generally, stocks must have a share price of at least $1 and a minimum market value of $1 million. In addition, companies listed on the NASDAQ must adhere to federal disclosure requirements for publicly traded securities and pay annual listing fees. The exchange issues a deficiency notice to any company in violation of any of the minimum standards for 30 consecutive days--after which the company has 90 days to regain compliance. For example, if the price were under $1 a company could choose to effect a reverse split to increase its share price. Companies which are delisted from the NASDAQ can continue to trade on the over-the-counter markets and the Pink Sheets, and some can reapply to NASDAQ and regain their listing. Regardless, delisting is often hard on a company, because it can impair its access to capital (e.g., Blue Sky laws which limit retail brokerages to sell stock with a price under $5 per share can reduce the depth and breadth of investors), the ability to borrow and exemptions from various state laws. (32)

There are three main stock split theories: (1) The optimal price/tick theory posits that splits return the stock price and relative tick size to their optimal range in their industry and market; (2) signaling theory posits that splits reveal information about future performance; and (3) the procedure/structure theory explains how a particular feature/structure/rule can cause a certain phenomenon in relation to splits. (27) According to the traditional signaling model, managers have better predictive information about outcome scenarios and so, when a firm is near its delisting boundary, a reverse stock split (RSS) signals negative information about the probability distribution of specific future scenarios. As Rhee and Wu explain:

   A broadly accepted explanation ... is that RSS
   signals to the market that management has
   either lost confidence in future price increases
   or exhausted all other means of maintaining
   the listing. RSS is the last straw before a stock
   is delisted to less liquid and less transparent
   markets, which becomes especially apparent after
   the NASDAQ introduced the one-dollar rule. (33)


EMPIRICAL PREDICTIONS BASED ON EFFECTUATION THEORY

MAIN PREDICTION FOR STOCK RETURNS

Relative to the traditional prediction-based signaling framework, signaling works differently in a non-predictive, "effectual" environment. When there is a large amount of uncertainty and firms have a significant amount of control over their future outcomes, then an RSS signals that the firm's manager is bullish about its own means, its stakeholders' willingness to commit to the future of the company, and that the firm will be able to successfully adapt to unexpected outcomes. Moreover, the RSS is a means by which the firm can, ipso facto, increase stakeholder commitments. However, because multiple rounds of financing are to be expected, this increase in commitment is done in a way that is consistent with the effectual logic of affordable losses.

Thus, in contrast to the predictive framework of traditional signaling theory where the manager and the stakeholders of the firm have relatively little control over outcomes, in an effectual environment this relationship is reversed: the firm operates in a non-predictive environment and the manager and stakeholders of firm have a relatively significant amount of control over the firm's outcomes. Because of this, an RSS strengthens commitments to the firm's future and signals the manger's confidence that the firm will be able to continue its operations in a propitious way. And because it is not possible in an effectual environment to exhaustively specify these possible scenarios, the signaling effect about the firm's confidence in its ability to control its own fate has a greater effect than any effect based on predictions about any specific future scenarios. Based on this logic and our previous argument that biotechnology firms are, in fact, in an effectual environment, we hypothesize the following:

[H.sub.1]: Biotechnology firms who conduct a reverse stock split will experience a positive abnormal return.

EXPLANATORY VARIABLES FOR STOCK RETURNS

Our key hypothesis, that RSS-firms will have positive ex post abnormal stock returns, is rooted in effectuation theory. Effectuation theory can also be used to predict the sign of the coefficient for various explanatory variables. These predictions cannot be cleanly contrasted with predictions obtained using a predictive signaling model. Nevertheless, to better understand the empirical implications of effectuation theory in the context of biotechnology RSSs, below we discuss the predicted sign on the coefficient for various variables in a regression where abnormal stock returns are the dependent variable.

If a firm chooses a larger split ratio, then this comprises a stronger signal and effects a greater commitment. Thus, the effect underlying [H.sub.1] will be more significant and we predict that the estimated coefficient for the split ratio will be positive. Because of economies of scale, larger firms tend to have greater control of their own destiny. This is because larger firms have more means and resources to survive and adapt when setbacks occur. All else equal, size is also an indicator of commitment by internal and external stakeholders. Thus, in an effectual environment, measures of size should have a positive coefficient. Similarly, if a firm has a great deal of cash, the cash can be used as a means of increasing the firm's ability to prolong projects and successfully navigate or adapt in the face of setbacks. Cash thus comprises an alternate form of control and implies a positive predicted coefficient.

In a slightly different vein, research and development (R&D) spending represents an alternative indicator of means, commitment, and adaptability. This is because firms with larger R&D spending will, all else equal, have greater resources to spur innovation, a larger network of synergistic partners and potential partners, and a larger number of options to adapt in the face of setbacks. Thus, the estimated coefficient for R&D should have a positive sign. In a similar vein, but with a stronger emphasis on commitments, long-term debt signals that a firm has committed financial partners (debt holders). This comprises a positive signal with respect to the firm's commitments from existing financial investors and prospects for successfully navigating future rounds of financing. Additionally, the structure of debt more strongly parallels the logic of affordable losses than the structure of equity. Thus, the estimated coefficient for an indicator of long-term debt should be positive.

If the market is bearish about a firm's future prospects, this will lead to a lower market-to-book ratio, all else equal. If a reverse stock split sends a positive signal to investors, the reversal in investors' expectations is apt to be greater for these firms with low market-to-book ratios. This implies that the market-to-book ratio should have a negative coefficient estimate. Finally, with regard to firm age, older biotechnology firms can be understood as being less likely to face setbacks, since they should have more controls and means compared to younger firms, all else equal. So, when an older firm does experience a setback, as indicated by the need to undergo an RSS, then this is likely to comprise a negative signal about the firm's ability to successfully control its environment. Thus, we predict that the coefficient for firm age will be negative.

Liquidity

With regard to stock liquidity, reverse stock splits are known to improve liquidity due to reduced effective (percentage) bid-ask spread that captures round-trip trade execution costs. (8,9) In an effectual environment, an RSS draws attention from investors. This, in turn, leads to a stronger signaling effect. Also, because the signaling effect is positive, as we have argued above, the greater attention also leads to improved commitments from investors and other stakeholders. Also, inasmuch as an RSS increases the firm's stock price, this leads to a positive feedback effect in the form of a lower cost of capital, thus improving the firm's ability to adapt to unexpected setbacks. We thus hypothesize the following:

[H.sub.2]: Biotechnology firms who conduct a reverse split following a setback will experience an improvement in liquidity.

METHODOLOGY AND DATA

In this section, we first explain our empirical methodology. We then discuss our sample, and provide basic summary statistics.

METHODOLOGY

We utilize the methodology of event study to test our hypotheses. An event study attempts to measure the valuation effects of a corporate event, such as a reverse stock split announcement, by looking at the response of the stock price around the announcement of the event and determine whether there is an abnormal return or not. One underlying assumption is that the market processes information about the event in an efficient and unbiased manner. To alleviate this assumption, we consider a number of lengths of event windows from one day to one year.

To estimate the normal return of a stock, we first use market model with CRSP value-weighted index as the proxy for the market return. We then use a four-factor model with Fama-French three factors (market, size, and book-to-market) and momentum factor. (34,35) The four-factor model has the advantage to control for risk premiums associated with

size, growth, and market momentum. It is important to control for size and growth when estimating the normal return as our sample firms are relatively small and still in their early growth stage. We then calculate the abnormal return and the cumulative abnormal return (CAR) based on the estimated normal return and test the average of CAR using the methodology in Brown and Warner. (36,37)

In addition to the CAR approach, we follow Barber and Lyon (38) and analyze buy-and-hold returns using matched control firms. Barber and Lyon point out a potential bias induced by cumulating short-term abnormal returns, such as CARs, over long periods (see also Conrad and Kaul (39) and conclude that the matched control firm approach leads to unbiased test statistics. Because of this potential bias, we measure stock performance by computing holding-period returns (HPRs) for each adopting firm and its matched control firm over one-month, six-month, and one-year periods following RSS. The holding periods start on the RSS announcement day.

Following Spiess and Affleck-Graves, (40) we first choose our control firms on the basis of industry, size (market capitalization), and book-to-market ratio. We avoid look-ahead bias by using only the information available at the time of RSS announcement. For each RSS sample firm, we identify all public firms in CRSP that have not undergone a RSS in the previous three years and belong to the biotechnology industry as defined by their 2-digit NAICS. We select the first matched firm from the set of potential matches such that the sum of absolute percentage difference between the size and book-to-market ratio of the sample firm and the control firm is minimized. If the first-best matched firm is delisted, we substitute returns from the second-best matched firm, starting at the close of trading on the date of the delisting payment and including the delisting return. If the first-best matched firm subsequently undergoes a RSS, we substitute the second-best matched firm on the next trading day.

In our liquidity analysis, we use four measures of Chordia et al. (41) First, we construct a share turnover ratio, Turnover, by dividing the total number of shares traded by the number of shares outstanding for a trading day and then average the daily ratios over a sample period to have the mean share turnover ratio:

Turnover = [1/T][T.summation over (t=0)][number of shares traded on day t/number of shares outstanding on day t]

Lesmond, Ogden (42) consider the proportion of days with zero returns as a proxy for liquidity. There are two key arguments that support this measure. First, stocks with lower liquidity are more likely to have days with little to no trading activity, and thus zero volume and zero return on these days. Second, stocks with higher transaction costs have less private information acquisition because of the higher transaction costs which gives traders a low incentive to obtain private information. Thus, even on positive volume days, these illiquid stocks can experience no-information-revelation and therefore zero return on these days. Thus:

Zeros = number of days with zero returns/total number of days in the subsample

The Amivest liquidity ratio is a measure of price impact which can be interpreted as the dollar volume of trading associated with a 1-percent change in the price of a security:

Liquidity = [1/T][T.summation over (t=0)][[volume.sub.t]/[absolute value of [r.sub.t]]]

where [volume.sub.t] is the dollar volume on day t and [r.sub.t] is the return on day t. The average is calculated over all nonzero-return days since the ratio is undefined for zero-return days. A larger value of Liquidity implies a lower price impact. This measure has been used by Amihud, Mendelson, (43) Berkman and Eleswarapu, (44) and others.

Finally, we define two volatility variables: [Volatility.sub.d] as the standard deviation of daily returns, annualized by multiplying by the square root of 252; [Volatility.sub.m] as the standard deviation of monthly returns, annualized by multiplying by the square root of 12. A reverse split reduces the relative bid-ask spread due to an increase in share price. This change in market microstructure alone may cause volatility to decrease. (26) As monthly returns are less impacted by bid-ask bounce, [Volatility.sub.m] can reflect the level of volatility due to trading activities, which we intend to measure.

SAMPLE AND SUMMARY STATISTICS

We use the Biocentury database to identify 40 biotechnology firms with RSS and collect split-related information. All 40 biotechnology firms were listed on NASDAQ and announced their reverse stock split during the 2011-2013 period. Table 1 summarizes the 40 reverse stock splits by split ratio and by their announcement year. Company financial data and stock return data are collected from COMPUSTAT (active and research) and CRSP tapes, respectively. The COMPUSTAT data includes "research" firms that have failed or been acquired eventually and CRSP stock data includes delisting returns if a firm's stock is delisted. Imposing that firms need to have data in all three sources leaves us with a total of 35 RSS firms. The choice of the sample period is governed by the availability of data.

We show summary statistics for our sample in Table 2. As shown in Panel A, the average (median) split ratio is 14.38 (7) with a range from 2 to 125 and an interquartile range from 6 to 15. The average (median) 30-day closing price for RSS firms prior to their reverse split event is 0.62 (0.52), with a range from 0.16 to 1.96 and interquartile range from 0.37 to 0.66. Thus, the majority of our biotechnology RSS firms have a prior price below $1. Average (median) market capitalization three days prior to the RSS event is 40.06 (28.58) million dollars.

Panel B of Table 2 shows summary statistics of our key variables. Panel C shows correlations between the explanatory variables that we intend to use in subsequent analysis. The variables with absolute correlation greater than 0.40 are as follows: LogEmp, LogSales, and LogTA are all highly correlated, with correlations ranging from 0.45 to 0.77. All three variables are proxies of size. As half of our sample firms do not have any sales, we use LogTA to measure size in our regression analysis. Also, there is a high degree of negative correlation between LogSales and LogSplitRatio, LogSales and Cash/TA, and LogEmp and Cash/TA, ranging from 0.42 to 0.54.

RESULTS

ANALYSIS OF RETURNS

Table 3 shows cumulative abnormal returns (CARs) for RSS biotechnology firms over the following six time windows, relative to each firm's RSS event: (1) 30 days before to 1 day before [-30, -1]; (2) the day of [0, 0]; (3) the day after [+1, +1]; (4) two days after to one month after [+2, +30]; (5) 1 month after to 6 months after [+31, +180]; (6) the day after to one year after [+1, +365]. Panel A shows CAR results using the market model with CRSP value-weighted index as the proxy for the market return whereas Panel B shows CAR results using a 4-factor model with Fama-French 3 factors (market, size, and B/M) and the momentum factor.

The results in Table 3 tell a fairly clear empirical story: biotechnology RSS firms experience positive abnormal returns prior to the RSS event, negative returns on the day of and the day after, and positive returns in 1-, 6-, and 12-month periods following the RSS event. These results are generally statistically significant, although if the Z-statistic is adjusted for both time-series and cross-sectional dependence, following Mikkelson and Partch, (37) then the day-of and month-after results are not significant. The economic significance of these results is, on average, quite large: 16% for the one-month prior; about 2.5% on the event day; 6% for the day-after; 33% for the month after; an additional 61% for the next 5 months; an additional 59% for the next 6 months, or 120% for the 1-year post-RSS window. The stock market initially reacted negatively to the announcement as shown by the negative CARs on the event day and the day after (albeit a less robust finding), and quickly reversed to strong positive returns in longer event windows.

CROSS-SECTIONAL CAR REGRESSIONS

Table 4 shows the results of cross-sectional regressions with CARs for various event windows as the dependent variable. In accordance with our effectuation-based prediction, we find that the coefficient on LogSplitRatio is positive and significant for each event window, and that the magnitude of the effect is larger for longer horizons.

LogTA, our size measure, has a positive and significant coefficient for the 6-month and 1-year post-RSS returns. This result is consistent with our effectuation-based prediction. For M-B, the coefficient is negative, in line with our prediction, but it is only significant for the one-month prior, event-day, and day-after returns. With regard to cash holding, we find that the coefficient on Cash/TA is significant only for the 1-year-after event window. The coefficient on Cash/TA is positive, in accordance with our prediction.

The coefficient on R&D/TA is positive and significant, as predicted, for the 1-month prior, 6-month after, and 1-year-after event windows. The coefficient on Age is, as predicted, negative and significant for the 1-month prior, day-after, 6-month-after, and 1-year-after event windows. The coefficient on IndLTDebt is positive and significant, as predicted, for the 6-month-after and 1-year-after event windows.

MATCHED RETURNS

Table 5 shows holding-period returns (HPRs) for RSS biotechnology firms compared to a matched sample of non-RSS firms on industry, size, and book-to-market. The returns for our biotechnology RSS firms are significantly higher than our control sample. This difference is 7.5% at 1 month, 20.7% at 6 months, and 27.0% at 1 year. These results corroborate our CAR findings reported in Table 3 and strengthen our [H.sub.1] hypothesis.

ANALYSIS OF LIQUIDITY

Table 6 reports mean and median values for each liquidity measure and the corresponding difference of the measure of the same firm in the windows of 180 days before and after the effective day. We conduct the paired sample t-tests and Wilcoxon signed-rank tests of differences in means and medians respectively, and report corresponding p-values. The number of firms in this table is 34.

In accordance with [H.sub.2], we find that the share turnover ratio and the Amivest liquidity ratio are higher and Zeros is lower after the RSS by the mean and median. Both volatility measures indicate a slight increase in return volatility after the RSS despite insignificant p-values. When combining with the positive abnormal returns, these data support the idea that the reverse split draws positive attention and trading activities from investors. The improved liquidity can enhance the ability of biotechnology firms to raise capital in subsequent rounds of financing.

CONCLUSION

The highly volatile nature of the biotechnology industry possesses several features that make it an ideal fit to evaluate effectuation theory. In particular, there is significant uncertainty in developing specific product development scenarios which makes it confounding to predict results, as firm success depends on their internal means and ability to procure stakeholder commitments, limit losses, and being prepared to adapt to unexpected results (i.e., expecting the unexpected). Because this environment differs substantially from the presumed predictable environment of traditional stakeholder theory, the usual negative-signal predictions regarding reverse stock splits are not appropriate. We conjecture, instead, that reverse stock splits following a setback comprise a positive signal for biotechnology firms regarding their own competencies and commitments pertaining to operations and future rounds of financing.

In our empirical analysis, we find that biotechnology firms who conduct a reverse split following a setback experience positive abnormal returns over 1-, 6-, and 12-month periods. We also find, in accordance with the effectuation-theory perspective, that the abnormal returns are positively related to the reverse split ratio, size, cash holding and long-term debt, and negatively related to the market-to-book ratio and firm age. Moreover, we find that liquidity improves after reverse stock splits.

In sum, we believe this study contributes to the research literature by expanding and extending the use of effectuation theory as an integrative and highly relevant framework for assessing biotechnology firms, especially with regard to financial decisions. More specifically, our analysis suggests that the concept of effectuation theory is better suited to analyzing reverse stock splits in the biotechnology industry. To the best of our knowledge, this is the first effectuation research to use archival financial data as opposed to relying on surveys of management perceptions, as in prior research. Further, our integration of effectuation and stock split theories provides a lens from which to explore emerging approaches for breakthrough innovation and technology development. Future research, particularly in the biotechnology industry, should therefore, pay more careful attention to the distinct aspects of effectuation theory.

REFERENCES

(1.) Wiltbank, R., Read, S., Dew, N. and Sarasvathy, S.D. (2009) Prediction and control under uncertainty: Outcomes in angel investing. Journal of Business Venturing 24(2): 116-133.

(2.) Wierenga, D.E. and Eaton, C.R. (2006) Drug Development and Approval Process (USA): Office of Research and Development--Pharmaceutical Manufacturers Association. (http:// ageconsearch.umn. edu), accessed on December 7, 2013.

(3.) Ahn, M.J., Couch, R.B. and Wu, W. (2008) Financing Development Stage Biotechnology Companies: RMs vs. IPOs. Journal of Health Care Finance 38(1): 32.

(4.) Black, B.S. and Gilson, R.J. (1998) Venture capital and the structure of capital markets: banks versus stock markets. Journal of Financial Economics 47(3): 243-77.

(5.) Allen, F. and Gale, D. (1999) Diversity of opinion and financing of new technologies. Journal of Financial Intermediation 8(1): 68-89.

(6.) Newton, D.P., Paxson, D.A. and Widdicks M. (2004) Real R&D options. International Journal of Management Reviews 5(2): 113-30.

(7.) Wiltbank, R., Dew, N., Read, S. and Sarasvathy, S.D. (2006) What to do next? The case for non-predictive strategy. Strategic Management Journal 27(10): 981-98.

(8.) Han, K.C. (1995) The effects of reverse splits on the liquidity of the stock. Journal of Financial and Quantitative Analysis 30(01): 159-69.

(9.) Mehta, C., Yadav, S.S. and Jain, P. (2011) Managerial motives for stock splits: survey based evidence from India. Journal of Applied Finance 21(1): 103-17.

(10.) Huggett, B., Hodgson, J. and Lahteenmaki, R. (2009) Public biotech 2008--the Numbers. Nature Biotechnology 27(8): 710.

(11.) Dew, N., Sarasvathy, S.D. and Venkataraman, S. (2004) The economic implications of exaptation. Journal of Evolutionary Economics 14(1): 69-84.

(12.) Read, S., Dew, N., Sarasvathy, S.D., Song, M. and Wiltbank, R. (2009) Marketing under uncertainty: The logic of an effectual approach. Journal of Marketing 73(3): 1-18.

(13.) Sarasvathy, S.D. (2008) Effectuation: Elements of entrepreneurial expertise: Edward Elgar Publishing.

(14.) Brettel, M., Mauer, R., Engelen, A. and Kupper, D. (2012) Corporate effectuation: Entrepreneurial action and its impact on R&D project performance. Journal of Business Venturing 27(2): 167-84.

(15.) Augier, M. and Sarasvathy, S.D. (2004) Integrating evolution, cognition and design: Extending Simonian perspectives to strategic organization. Strategic Organization 2(2): 169-204.

(16.) Hitt, M.A. and Duane, R. (2002) The essence of strategic leadership: Managing human and social capital. Journal of Leadership & Organizational Studies 9(1): 3-14.

(17.) Hemlin, S. (2009) Creative Knowledge Environments: An Interview Study with Group Members and Group Leaders of University and Industry R&D Groups in Biotechnology. Creativity and Innovation Management 18(4): 278-85.

(18.) Ahn, M.J., Wu, W. and Rahman, M. (2010) Medarex: Realizing its Potential? In: Ahn MJ, Alvarez MA, Meyers AD, York AS, editors. Building the Case for Biotechnology: Logos Press.

(19.) Kirchhoff, M. and Schiereck, D. (2011) Determinants of M&A Success in the Pharmaceutical and Biotechnological Industry. Journal of Business Strategy 8(1): 25-50.

(20.) Ahn, M.J., Meeks, M., Davenport, S. and Bednarek, R. (2010) Exploring technology agglomeration patterns for multinational pharmaceutical and biotechnology firms. Journal of Commercial Biotechnology 16(1): 17-32.

(21.) Hemphala, J. and Magnusson, M. (2012) Networks for Innovation--But What Networks and What Innovation? Creativity and Innovation Management 21(1): 3-16.

(22.) Gulati, R. and Higgins, M.C. (2003) Which ties matter when? The contingent effects of interorganizational partnerships on IPO success. Strategic Management Journal 24(2): 127-44.

(23.) Nicholson, S., Danzon, P.M. and McCullough, J.S. (2002) Biotech-pharmaceutical alliances as a signal of asset and firm quality. National Bureau of Economic Research.

(24.) Ford, C.M., Sharfman, M.P. and Dean, J.W. (2008) Factors Associated with Creative Strategic Decisions. Creativity and Innovation Management 17(3): 171-85.

(25.) Angel, J.J. (1997) Tick size, share prices, and stock splits. The Journal of Finance 52(2): 655-681.

(26.) Koski, J.L. (2007) Does volatility decrease after reverse stock splits? Journal of Financial Research 30(2): 217-235.

(27.) Wang, J. (2012) Stock Split Decisions: A Synthesis of Theory and Evidence. Journal of Applied Finance 22(2): 124-142.

(28.) Desai, H. and Jain, P.C. (1997) Long-Run Common Stock Returns following Stock Splits and Reverse Splits. The Journal of Business 70(3): 409-433.

(29.) Garcia de Andoain, C. and Bacon, F.W., eds. (2009) The Impact of Stock Split Announcements on Stock Price: A Test of Market Efficiency. Proceeding of American Society of Business and Behavioral Sciences Annual Conference, Las Vegas.

(30.) Martell, T.F. and Webb, G.P. (2008) The performance of stocks that are reverse split. Review of Quantitative Finance and Accounting 30(3): 253-79.

(31.) Woolridge, JR and Chambers, D.R. (1983) Reverse splits and shareholder wealth. Financial Management 5-15.

(32.) NASDAQ. Initial Listing Guide 2014 (https:// listingcenter.nasdaqomx.com), accessed on February 28, 2014.

(33.) Rhee, S.G. and Wu, F. (2012) Anything wrong with breaking a buck? An empirical evaluation of NASDAQ's $1 minimum bid price maintenance criterion. Journal of Financial Markets 15(2): 258-285.

(34.) Carhart, M.M. (1997) On persistence in mutual fund performance. The Journal of Finance 52(1): 57-82.

(35.) Fama, E.F. and French, K.R. (1993) Common risk factors in the returns on stocks and bonds. Journal of Financial Economics 33(1): 3-56.

(36.) Brown, S.J. and Warner, J.B. (1980). Measuring security price performance. Journal of financial Economics 8(3):205-58.

(37.) Mikkelson, W.H. and Partch, M.M. (1986) Valuation effects of security offerings and the issuance process. Journal of Financial Economics 15(1): 31-60.

(38.) Barber, B.M. and Lyon, J.D. (1997) Firm Size, Book-to-Market Ratio, and Security Returns: A Holdout Sample of Financial Firms. The Journal of Finance 52(2): 875-883.

(39.) Conrad, J. and Kaul, G. (1993) Long-Term Market Overreaction or Biases in Computed Returns? The Journal of Finance 48(1): 39-63.

(40.) Spiess, D.K. and Affleck-Graves, J. (1999) The long-run performance of stock returns following debt offerings. Journal of Financial Economics 54(1): 45-73.

(41.) Chordia, T., Huh, S-W. and Subrahmanyam, A. (2009) Theory-based illiquidity and asset pricing. Review of Financial Studies 22(9): 3629-3668.

(42.) Lesmond, D.A., Ogden, J.P. and Trzcinka, C.A. (1999) A new estimate of transaction costs. Review of Financial Studies 12(5):1113-1141.

(43.) Amihud, Y., Mendelson, H. and Lauterbach, B. (1997) Market microstructure and securities values: Evidence from the Tel Aviv Stock Exchange. Journal of Financial Economics 45(3): 365-390.

(44.) Berkman, H. and Eleswarapu, V.R. (1998) Short-term traders and liquidity: a test using Bombay Stock Exchange data. Journal of Financial Economics 47(3): 339-355.

Wei Wu is Assistant Professor of Finance at the Atkinson Graduate School of Management at Willamette University.

Robert Couch is Assistant Professor of Finance at the Atkinson Graduate School of Management at Willamette University.

Yulianto Suharto is pharmacist and PhD candidate at Portland State University.

Mark J. Ahn is Principal at Pukana Partners and Professor (adjunct) at Portland State University.

Correspondence:

Mark J. Ahn, Pukana Partners Ltd., US. Email: mark@ pukanapartners.com

Table 1: Biotech Reverse Splits by Split Ratio
and Announcement Year

Split Ratio         Number of Splits

              2011   2012   2013   Total

1:2            1                     1
1:3                   1              1
1:4                   3              3
1:5                   1      1       2
1:6            1      6      2       9
1:7                   3              3
1:10           2      4      1       7
1:12                  1      1       2
1:15                  1              1
1:16                  1              1
1:20                  2      2       4
1:25                         1       1
1:30                  1              1
1:40           1                     1
1:50                         1       1
1:56                  1              1
1:125                        1       1
Total          5      25     10     40

This table summarizes the 40 reverse stock splits
by split ratio for biotechnology firms during the
sample period 2011-2013.

Table 2: Summary Statistics and Correlations

Panel A: Share Characteristics

Variable                   N    Mean    StDev    min    25th %

Split Ratio                35   14.38   21.78   2.00     6.00
Prior Avg (30-day) Price   35   0.62    0.43    0.16     0.37
Prior Mkt Cap              35   40.06   55.94   2.89     7.62

Panel B: Key Variables

Variable                   N    mean    StDev    min    25th %

Total Assets (TA)          35   30.26   38.37   3.18     8.66
LogTA                      35   2.90    0.99    1.16     2.16
Employees (Emp)            35   0.04    0.07    0.00     0.01
LogEmp                     35   -3.77   1.02    -5.81   -4.61
Sales                      35   5.93    12.83   0.00     0.00
LogSales                   25   -0.28   2.69    -4.61   -2.43
M-B                        32   6.65    10.29   0.17     1.89
R&D                        35   10.47   12.17   0.73     3.52
Age                        35   12.01   8.32    1.93     5.21
LogAge                     35   2.24    0.77    0.66     1.65
Cash                       35   18.21   31.15   0.59     6.20
LT Debt                    35   2.21    6.13    0.00     0.00

Variable                   Median   75th %    Max

Split Ratio                 7.00    15.00    125.00
Prior Avg (30-day) Price    0.52     0.66     1.96
Prior Mkt Cap              28.58    57.89    263.05

Panel B: Key Variables

Variable                   Median   75th %    Max

Total Assets (TA)          16.62    40.23    204.99
LogTA                       2.81     3.69     5.32
Employees (Emp)             0.02     0.05     0.44
LogEmp                     -4.02    -3.00    -0.82
Sales                       0.10     4.07    49.32
LogSales                   -0.78     1.68     3.90
M-B                         2.79     6.51    49.45
R&D                         5.62    12.69    52.40
Age                         9.95    16.22    34.96
LogAge                      2.30     2.79     3.55
Cash                       10.82    17.72    187.66
LT Debt                     0.00     1.67    32.73

Panel C: Correlations

                LogTA   logEmp   logSales    m-b    logSplitRatio

LogTA           1.00
LogEmp          0.68     1.00
LogSales        0.45     0.77      1.00
M-B             -0.06   -0.08      0.10     1.00
LogSplitRatio   -0.19   -0.39     -0.42     -0.12       1.00
R&D/TA          -0.33   -0.24     -0.03     0.00        0.11
Age             0.02    -0.19     -0.12     -0.16       0.10
Cash/TA         -0.19   -0.54     -0.44     0.12        0.14
IndLTDebt       -0.03    0.21      0.17     -0.14       0.06

Panel C: Correlations

                r&d/ta    Age    Cash/TA   IndlTDebt

LogTA
LogEmp
LogSales
M-B
LogSplitRatio
R&D/TA           1.00
Age              0.04    1.00
Cash/TA          0.09    -0.03    1.00
IndLTDebt        0.29    0.05     -0.32      1.00

Panel A reports summary statistics of equity share
characteristics for our sample of biotechnology firms that
undergo a reverse stock split between 2011 and 2013. Panel B
reports summary statistics of key variables. For variables with
high degree of right skewness, we also show the logged version.
Panel C shows correlation coefficients. Split Ratio is the
reverse stock split ratio, the ratio between the number of new
and old shares. Prior Avg (30-day) Price is the average closing
price over 30 days prior to the announcement. Prior Mkt Cap is
the market capitalization three days prior to the announcement.
Total Assets is the total value of assets. Employees is the
number of company workers as reported to shareholders (measured
in thousands). Sales is the total sales of the firm. M-B is the
market value of equity divided by the book value of equity. R&D
is the research and development expenses, including all costs
incurred during the year that relate to the development of new
products or services. Age is the number of years between the
split announcement day and the IPO day. Cash is the total amount
of cash. LT Debt is the long-term debt. IndLTDebt is a dummy
variable equal to one if the firm has a positive amount of long-term
debt and zero otherwise. All variables, except split ratios,
share prices, employees, and age, are in millions of dollars.

Table 3: Cumulative Abnormal Returns

Panel A: Market Model

Event Window   Average (%)   T-Statistic    Z-statistic

[-30, -1]         16.04         2.275 *       1.913 *
[0, 0]            -2.48         1.929 *      -0.648
[+1, +1]          -6.25        -4.855 ***    -1.928 *
[+2, +30]         33.04         4.765 ***    -0.648
[+31, +180]       61.01         3.869 ***     2.340 **
[+1, +365]       120.34         4.892 ***     3.620 ***

Panel B: Fama-French-momentum Four-Factor model

Event Window     CAR (%)     T-Statistic    Z-statistic

[-30, -1]         16.70         2.372 **      1.917 *
[0, 0]            -2.62        -2.036 *      -0.644
[+1, +1]          -6.09        -4.739 ***    -1.925 *
[+2, +30]         34.03         4.917 ***    -0.644
[+31, +180]       62.40         3.965 ***     2.770 **
[+1, +365]       120.86         4.923 ***     3.624 ***

Event Window        25th             75th
               Percentile (%)   Percentile (%)

[-30, -1]          -4.74             28.94
[0, 0]              0.04              1.70
[+1, +1]            0.90              2.55
[+2, +30]          10.10             47.34
[+31, +180]        16.63            289.35
[+1, +365]         95.43            581.90

Panel B: Fama-French-momentum Four-Factor mode

Event Window         25th             75th
               Percentile (%)   Percentile (%)

[-30, -1]           -9.53            47.59
[0, 0]              -8.52             3.82
[+1, +1]           -13.56             3.02
[+2, +30]          -20.20            39.11
[+31, +180]         -2.59           112.15
[+1, +365]          33.28           199.96

This table reports certain measures of the distribution
of cumulative abnormal returns (CARs) for various event
windows surrounding the announcement day of the reverse
stock split of the biotechnology firms in our sample.
Panel A shows the results based on market-model
adjusted stock returns. Panel B shows the results based
on Fama-French-Momentum four-factor model adjusted
stock returns. The T-statistics of average CARs are
based on the time-series standard deviation test in
Brown and Warner (1980). The Z-statistics of average
CARs are computed using the methodology of Mikkelson
and Partch (1986), which considers both time-series and
cross-sectional dependence, as well as unequal
variances in returns. The 25th percentile and the 75th
percentile of the distribution of CARs are also
reported. The symbols *, **, and *** denote statistical
significance at the 0.05,0.01, and 0.001 levels,
respectively, using a two-tailed test.

Table 4: Cross-Sectional CAR Regressions

                    [-30, -1]          [0, 0]

Variable        Param    P-value   Param   P-value

Intercept       53.17    0.258     0.83    0.730
LogSplitRatio   27.95    0.072     1.47    0.014
LogTA           2.10     0.721     -0.06   0.882
M-B             -1.12    0.013     -0.08   0.011
R&D/TA (%)      0.35     0.040     0.00    0.880
LogAge          -13.90   0.067     -0.26   0.213
Cash/TA (%)     0.02     0.902     0.00    0.602
IndLTDebt       -8.60    0.431     -0.32   0.547
[R.sup.2]             0.41              0.50

                   [+1, +1]           [+2, +30]

Variable        Param   P-value   Param    P-value

Intercept       9.86    0.039     17.30    0.845
LogSplitRatio   1.71    0.003     61.59    0.043
LogTA           -0.09   0.877     21.33    0.040
M-B             -0.06   0.046     -1.45    0.121
R&D/TA (%)      0.00    0.882     0.32     0.234
LogAge          -1.30   0.001     -21.93   0.120
Cash/TA (%)     0.00    0.657     -0.09    0.663
IndLTDebt       0.76    0.464     20.87    0.241
[R.sup.2]            0.42               0.45

                    [+31, +180]         [+1, +365]

Variable         Param    P-value    Param    P-value

Intercept       227.49    0.477     -49.31    0.925
LogSplitRatio   233.12    0.020     410.26    0.001
LogTA           63.12     0.077     117.90    0.066
M-B             -1.90     0.434     -3.84     0.219
R&D/TA (%)      2.76      0.025     4.52      0.036
LogAge          -114.99   0.019     -163.32   0.027
Cash/TA (%)     0.57      0.621     3.40      0.042
IndLTDebt       221.60    0.012     502.93    0.004
[R.sup.2]              0.49                0.68

This table reports coefficient estimates from the
cross-sectional OLS regressions where the dependent
variable is the cumulative abnormal return (CAR) in
percentage for various event windows surrounding the
announcement day of the reverse stock split of the
biotechnology firms in our sample. Variable
descriptions are provided in Table 1. P-values are
adjusted for heteroskedasticity. We report the
adjusted [R.sub.2] in the last row.

Table 5: Holding-Period Returns Following Reverse Split
Matched on Industry, Size, and Book-to-Market

                               1-month

         Sample Firms   Matched Firms   Difference   P-Value

Mean     17.16          9.61            7.54         0.040
Median   9.05           1.93            6.12         0.054

                               6-Month

         Sample Firms   Matched Firms   Difference   P-Value

Mean     30.64          9.98            20.67        0.016
Median   27.39          10.51           18.81        0.024

                                1-Year

         Sample Firms   Matched Firms   Difference   P-Value

Mean     50.21          23.16           27.04        0.008
Median   39.65          16.95           24.19        0.013

This table shows mean and median holding-period returns (HPRs) for
sample firms, matched control firms, along with paired differences
over 1-month, 6-month, and 1-year horizons. HPRs are calculated as
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] where Rit is
the return on stock i on the t-th day after the announcement day
and Ti is the number of days from the announcement to the end of
the holding period. Matched firms are chosen based on industry
(2-digit NAICS code), size, and book-to-market ratio. If the
matched firm is delisted, or undergoes a reverse split
subsequently, we use the next closest matched firm's return. There
are 32 sample firms that can be paired to control firms. We conduct
the paired sample t-tests and Wilcoxon signed-rank tests of
differences in means and medians respectively, and report
corresponding p-values.

Table 6: Liquidity Measures: Before and After the Reverse Split
Announcement

               Stat     [-180, 0]   [0, 180]   Difference   P-value

Turnover (%)   Mean        0.80        4.73       -3.94      0.018
               Median      0.60        0.96       -1.52      0.023

Zeros (%)      Mean        6.26        4.21        2.05      0.010
               Median      5.88        3.36        2.46      0.002

Liquidity      Mean       30.11      172.27     -142.17      0.070
               Median      7.88       38.40      -13.30      0.002

[Volatility.   Mean      110.92      125.89      -14.97      0.044
sub.d] (%)     Median     98.00       89.69      -20.31      0.207

[Volatility.   Mean      100.55      116.86      -16.31      0.156
sub.m] (%)     Median     68.51      101.48      -32.39      0.207

This table compares liquidity of stock trading activity around the
reverse split announcement day for the biotech firms in our sample.
The first period is from 180 days before the announcement to the
announcement day [-180,0]. The second period is from the
announcement day to 180 days after the announcement [0, 180].
Turnover is defined as the daily number of share-trading volume
divided by the number of common shares outstanding, and then the
daily ratios are averaged over time. Zeros is the proportion of
trading days with a zero price change from the previous day over a
specified time period. Liquidity is the Amivest liquidity ratio to
measure price impact defined by dividing daily dollar trading volume
by the absolute daily return, and then averaging over time.
[Volatility.sub.d] is defined as the standard deviation of daily
returns, multiplied by the square root of 252. [Volatility.sub.m] is
defined as the standard deviation of monthly returns, multiplied by
the square root of 12. The table reports mean and median values for
each measure and the corresponding difference of the measure of the
same firm before and after the announcement day. We conduct the
paired sample t-tests and Wilcoxon signed-rank tests of differences
in means and medians respectively, and report corresponding
p-values. The number of firms in this table is 34.
COPYRIGHT 2015 thinkBiotech LLC
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2015 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Wu, Wei; Couch, Robert; Suharto, Yulianto; Ahn, Mark J.
Publication:Journal of Commercial Biotechnology
Article Type:Report
Date:Jan 1, 2015
Words:10591
Previous Article:Differentiating public policy for technology startups--essential for biotech?
Next Article:Using biotechnology, CSI, and zombies to promote science education in one of America's most challenging regions.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters