Long and Short of It
Crash Narratives
William Goetzmann, Dasol Kim & Robert Shiller
NBER Working Paper, July 2022
Abstract:
The financial press is a conduit for popular narratives that reflect collective memory about historical events. Some collective memories relate to major stock market crashes, and investors may rely on associated narratives, or "crash narratives," to inform current beliefs and choices. Using recent advances in computational linguistics, we develop a higher-order measure of narrativity based on newspaper articles that appear following major crashes. We provide evidence that crash narratives propagate broadly once they appear in news articles, and significantly explain predictive variation in market volatility. We exploit investor heterogeneity using survey data to distinguish the effects of narrativity and fundamental conditions and find consistent evidence. Finally, we develop a measure of pure narrativity to examine when the financial press is more likely to employ narratives.
The Value of Differing Points of View: Evidence from Financial Analysts' Geographic Diversity
William Gerken & Marcus Painter
Review of Financial Studies, forthcoming
Abstract:
Using satellite imagery of retail firms parking lots to measure time-varying local firm-specific performance, we document that analysts incorporate local information into their forecasts. Analysts rely more on local signals when less firm-wide information is available. This incorporation of noisy local firm information has firm-level implications. Examining across industries, we find causal evidence that geographic concentration of analysts increases consensus forecast errors and decreases firm liquidity. These effects are stronger for harder-to-value stocks. The market values geographic firm information, as the abnormal return around forecast revisions is higher for analysts who cover a firm from a unique location.
The Role of Social Media in the Corporate Bond Market: Evidence from Twitter
Eli Bartov, Lucile Faurel & Partha Mohanram
NYU Working Paper, April 2022
Abstract:
Prior studies document the role social media information plays in the stock market as well as the important dissimilarities between the bond and stock markets. Bridging these two literatures, we examine the role of social media information in the corporate bond market. Analyzing a broad sample of messages by Twitter individual users, posted just prior to earnings announcements, containing bond, credit risk, and fundamental information, we find that aggregate Twitter opinion (OPI) predicts upcoming announcement bond returns and changes in CDS spreads, and is associated with future changes in bond yield spreads and credit ratings, thereby providing economically important information to the bond market. This interpretation is bolstered by results from a variety of cross-sectional analyses. Finally, we document an association between OPI and future changes in default risk, which casts light on the nature of the Twitter information underlying our findings. Overall, our findings demonstrate that Twitter appears to disseminate potentially economically important information to even the presumably sophisticated bond and CDS investors, as well as information intermediaries.
Contagion of Investor Sentiment in Online Investment Communities: Evidence from Dynamic Visuals on Stocktwits
Ming Gu, Siew Hong Teoh & Shijia Wu
University of California Working Paper, April 2022
Abstract:
Psychological research suggests that communication using dynamic visuals attract higher attention and induce higher activation response than text and static images. Stocktwits is a social media platform where postings about a stock can include GIF images (henceforth GIFs) and declare their sentiment (bullish or bearish). We study the associations of GIFs with user-declared stock sentiment, trading behavior, and future stock returns at the firm-day level. On firm-days with GIF postings, we find higher net bullish sentiment, higher retail order imbalance, and short sale volume; and initial positive abnormal returns followed by a longer-term reversal. The GIF associations with sentiment, return overreaction, and short sale volume are stronger on firm-days with popular GIFs, with GIFs with declared sentiment, with GIFs from influencers, and in firms with higher retail ownership. Retail buys and retail sells separately, and not order imbalance, are stronger in firms with higher retail ownership. These findings suggest that dynamic visuals in social media postings are associated with greater investor attention and higher bullish sentiment, resulting in net buying and overpricing of the stock, especially among retail investors.
Prior Experience of Managers and Maladaptive Responses to Performance Feedback: Evidence from Mutual Funds
Vibha Gaba et al.
Organization Science, forthcoming
Abstract:
In this study, we examine how the prior experiences of decision makers systematically influence their assessment of and responses to negative performance feedback. We posit that, although greater and more specialized experiences enable managers to build relevant knowledge and expertise in specific domains, they also make them overconfident in their abilities and strategies. Such experience-induced overconfidence further leads to distortions in the performance assessment process, hindering a firm's ability to recognize and respond to poor performance. We empirically test these arguments in the context of U.S. mutual fund managers making investment decisions in response to fund performance below aspirations. As hypothesized, we find that more experienced and more specialized fund managers change their investment decisions less when faced with negative performance feedback than managers who are less experienced and less specialized. In additional analyses, we further show that the lower responsiveness of more experienced (specialized) managers is associated with the fund's lower future performance, supporting our proposed theoretical mechanism (overconfidence). This study augments existing performance feedback research by showing how decision makers' prior experience can impede problem-solving behavior in organizations. It also contributes to the literature on human capital and organizational learning by documenting an unintended consequence of accumulated human capital on firm adaptive behavior.
Unpacking the Rise in Alternatives
Juliane Begenau, Emil Siriwardane & Pauline Liang
Stanford Working Paper, May 2022
Abstract:
We document a large and heterogeneous shift by public pensions into alternative investments (hedge funds, private equity, and real estate) since 2006. Interpreting the data through the lens of the two-fund separation theorem we find that pensions' attitude towards taking risk cannot account for the size and heterogeneity of the shift. Pension characteristics like size and underfunding explain almost none of the cross-sectional heterogeneity in the shift to alternatives but some of the shift out of fixed income. In contrast, our results are consistent with a sizable shift in beliefs changing the composition of the risky portfolio. Consistent with this hypothesis, a simple variance decomposition shows that investment consultants explain nearly 25% of the variation of which pensions shifts into alternatives. After controlling for consultants, we also document evidence of peer effects in the sense that pensions are more likely to invest in alternatives if their neighbors do.
Fully Closed: Individual Responses to Realized Gains and Losses
Steffen Meyer & Michaela Pagel
Journal of Finance, June 2022, Pages 1529-1585
Abstract:
We analyze how individuals reinvest realized capital gains and losses exploiting plausibly exogenous sales due to mutual fund liquidations. Individuals reinvest 83% if a forced sale results in a gain relative to the initial investment; but reinvest only 40% in the event of a loss. This difference is statistically significant for more than six months and arises because many individuals forced to realize a loss choose not to reinvest anything and some even exit the stock market altogether. Individuals treat realized losses differently from paper losses and are discouraged from investing more and participating in the stock market.
Idiosyncrasy as a Leading Indicator
Randall Morck, Bernard Yeung & Lu Zhang
NBER Working Paper, May 2022
Abstract:
Disequilibrating macro shocks affect different firms' prospects differently, increasing idiosyncratic variation in forward-looking stock returns before affecting economic growth. Consistent with most such shocks from 1947 to 2020 enhancing productivity, increased idiosyncratic stock return variation forecasts next-quarter real GDP growth, industrial production growth, and consumption growth both in-sample and out-of-sample. These effects persist after controlling for other leading economic indicators.
The causal effect of improved readability of financial reporting on stock price crash risk: Evidence from the Plain Writing Act of 2010
Shiyan Yin, Thanaset Chevapatrakul & Kai Yao
Economics Letters, July 2022
Abstract:
This paper shows that obfuscating financial reports leads to an increase in the risk of stock price crash. Exploiting the Plain Writing Act of 2010 (PWA) as the exogenous source of variation, the results of the difference-in-differences (DID) estimation show that improved readability of 10-Ks, as a result of the PWA, caused the stock price crash risk to fall. Our results survive the falsification check and are robust under different measures of readability and crash risk.
Learning by Owning in a Lemons Market
Jordan Martel, Kenneth Mirkin & Brian Waters
Journal of Finance, June 2022, Pages 1737-1785
Abstract:
We study market dynamics when an owner learns about the quality of her asset over time. Since this information is private, the owner sells strategically to a less informed buyer following sufficient negative information. In response, market prices feature a "U-shape" and trading probabilities a "hump-shape" with respect to the time to sale. As the owner initially acquires greater information, buyers suffer greater adverse selection, and prices fall accordingly. Eventually, the probability of an informed sale shrinks, and prices rebound. We provide evidence consistent with our model in markets for residential real estate, venture capital investments, and construction equipment.