The pricing of firms with expected losses/profits: The role of January

• Author: Terry Shevlin,Alexander Nekrasov,Peng‐Chia Chiu
• Published date: 01 May 2018
• Date: 01 May 2018
• DOI: http://doi.org/10.1111/jbfa.12296

DOI: 10.1111/jbfa.12296

The pricing of firms with expected losses/profits:

TheroleofJanuary

Peng-Chia Chiu1Alexander Nekrasov2Terry Shevlin3

1TheChinese University of Hong Kong,

HongKong

2Universityof Illinois at Chicago, Chicago, Illinois,

USA

3Universityof California, Irvine, Irvine, California,

USA

Correspondence

TerryShevlin, PaulMerage School of Business,

Universityof California-Irvine, Irvine, CA 92617,

USA.

Email:tshevlin@uci.edu

Abstract

We examine the role of January in the relation between expected

losses/profits and future stock returns. We predict and find that

the relation between expected losses/profits and future returns

reverses from the usual positive relation in non-January months to

a negative one in January. The reverse January relation is consis-

tent across sample years, is observed in the United States and inter-

national markets, and is incremental to other variables associated

with January returns. At least part of the reverse January relation is

explained bytax-loss selling. Further analysis shows that the reverse

January relation results in a temporary price drift away from funda-

mental value. In other words, we find that abnormal positive (nega-

tive)future returns do not always indicate past under(over)valuation.

Overall, our results illustrate the importance of controlling for the

effect of January when examining how investors price expected

losses/profits.

KEYWORDS

expectedlosses/profits, January effect, return predictability, tax-loss

selling

1INTRODUCTION

Since the seminal paper by Ball and Brown (1968), the relation between earnings and stock prices has been at the

center of capital markets research. While the efficient market hypothesis predicts that stock price fully reflects all

publicly available information about expected future earnings, empirical research finds that stocks with low (high)

expected earnings earn low (high) future risk-adjusted returns.1The return drift is especially large for firms with

past and expected future losses.2Much of the prior research has focused on different measures of firm performance

1Elgers, Lo,and Pfeiffer (2001) show that the ratio of analysts’ annual earnings forecast level to stock price predicts future returns over the next 12 months.

Frankeland Lee (1998) show that the value-to-price ratio based on earnings forecast levels predict future returns over the next three years. Balakrishnan,

Bartov,and Faurel (2010) find that firms with past losses (profits), which can be viewed as expected losses (profits) under a simple random walk model,earn

low(high) future abnormal stock returns. Research finds that expected losses are associated with low future returns in the US (Li, 2011), the UK (Jiang, Soares,

& Stark, 2016), and Australia (Wu, 2017). Cen, Hilary,and Wei (2013) show that firms with forecasted earnings per share lower (higher) than the industry

medianearn abnormally high (low) future stock returns. The magnitude of abnormal returns documented in these studies often exceeds 10% per year.

2Balakrishnanet al. (2010) show that firms with extreme losses (profits) earn an abnormal return of approximately −6% (+4%) over 120 trading daysfollowing

theearnings announcement day.Li (2011) shows that firms with expected persistent losses earn 10.4% lower abnormal one-year returns than loss firms that

544 c

2017 John Wiley & Sons Ltd wileyonlinelibrary.com/journal/jbfa JBus Fin Acc. 2018;45:544–571.

CHIU ET AL.545

to document the return drift. In this study, we argue that another important stock price phenomenon—the January

effect—plays an important role in pricing of expectedearnings, especially expected losses. We predict and find that the

expected loss/profitdrift reverses in January.

The January or turn-of-the-year effect refers to abnormally highreturns earned by stocks, especially stocks of small

firms, during the month of January (Blume & Stambaugh, 1983; Keim, 1983; Reinganum, 1983; Rozeff & Kinney,1976;

Thaler, 1987). The most common explanation of the January effect is tax-loss selling. According to this explanation,

in order to minimize tax liability, tax-sensitive investors—suchas individual investors—sell losing stocks before year-

end and then buy them back in January.3The positive price pressure in January leads to positive abnormal returns

in that month. Several studies find evidence consistent with the tax-loss selling explanation.Poterba and Weisbenner

(2001) show that changes in the tax rules for capital gains explain changes in January returns. Kang, Pekkala, Polk,

and Ribeiro (2015) show that the turn-of-the-year tax-selling pressure is stronger when interest rates are high and

therefore the cost of delaying tax-selling benefits is high. Sias and Starks (1997) find that stocks with greater individ-

ual investor interest earn higher (lower) returns in January (December) relative to stocks with greater institutional

investor interest. Badrinath and Lewellen (1991), Ritter (1988), and Dyl and Maberly (1992) show that tax-motivated

trading by individual investors is associated with January returns. Sikes (2014) shows that in addition to individual

investors, institutional investors with strong tax incentives also exhibit tax-loss selling trading behavior, which con-

tributes to the January effect. Starks, Yong,and Zheng (2006) find that the trading behavior of tax-sensitive investors

explains the January effect in municipal bond closed-end funds.4

Tax-lossselling, and therefore the January effect, should be important for firms with poor financial performance

such as stocks with expected negative earnings. Prior studies argue that the January effect is strong for small firms

because they are more likely to haveperiods of poor operating and stock return performance and therefore are more

likely to bring capital losses to some investors (Reinganum, 1983; Roll, 1983). Since expected negative earnings are

also likely to lead to poor stock performance and capital losses during the current year, tax-loss selling should also

be evident for such firms.5Further, since the price pressure induced by tax-loss selling behavior is large and con-

centrated in one month, we expect the January effect to dominate (during that month) the documented expected

loss/profit drift.6Wetherefore predict that firms with expected losses will earn higher January returns than will firms

with expectedprofits.7As a result, the usual positive relation between expected losses/profits and future stock returns

will reverse to a negative relation in January. The predicted reversalis markedly different from a usual combination

of two anomalies—such as the size effect and the January effect—wherein the abnormal returns are enhanced, not

reversed.

An important implication of the January reversal is that it can result in underestimation of the loss/profitdrift dur-

ing the whole year or even make it appear insignificant. Consequently,in a study that uses cumulative returns that

aremore likely to return to profitability. Jiang et al. (2016) show that future abnormal returns associated with expected losses are concentrated in months of

subsequentearnings announcements and greater for stocks with higher trading costs.

3The time between a stock sale and repurchase must be greater than 30 days to avoid a wash sale. Losses from tradesof securities in a wash sale cannot

be deducted under Internal Revenue Service rules. Followingprior studies, we refer to this tax-motivated selling (buying) before (after) year-end as tax-loss

selling.

4TheJanuary effect has also been observed in countries with a non-December tax year-end and in countries with no taxes on capital gains, which may suggest

that tax-loss selling is not the sole explanation(Berges, McConnell, & Schlarbaum, 1984; Gultekin & Gultekin, 1983; Kato & Schallheim, 1985; Ko, 1998; Lee,

1992; Thaler,1987). At the same time, as Thaler (1987) notes, ‘Still, returns are high in April in Great Britain, and in July in Australia, so taxes do seem to be

part of the story.’Also, tax-loss selling by investors from the US and other countries with taxes on capital gains and a December tax year-end could result in

theJanuary effect in such countries (Berges et al., 1984; Kato & Schallheim, 1985).

5Although capital losses refer to investors’losses in the financial market, it is difficult to measure these losses directly based on past stock returns because

anygiven stock was purchased and sold at various points in time by different investors (Roll, 1983). Therefore, past returns will not fully explain the effects of

sizeand negative earnings.

6The typical increase in monthly stock return due to the January effect exceeds 3%, while the typical monthly hedge return due to loss/profit drift is

around1%.

7Although,firms with expected profits will have lower tax-loss selling than firms with expected losses, the likelihood of tax-loss selling for these stocks is still

positive. Therefore, even firms with expectedprofits will have somewhat higher returns in January than in non-January months. Our empirical findings are

consistentwith this prediction.