Browsing by Author "Kumar, Alok"
Now showing 1 - 4 of 4
Results Per Page
Sort Options
Item Early career experience and optimism spillover(2012-05) Law, Kai Fung; Starks, Laura T; Clement, Michael B; Han, Bing; Kumar, Alok; Sialm, ClemensUsing a long panel on employment history, I exploit a novel setting to examine if sell-side analysts carry over their early career experience into their future professional careers. I find that analysts' early mentorship experience has a long-lasting impact on their professional styles. Analysts are more optimistic if they work with optimistic mentors in their first jobs as junior analysts: they issue more strong buy recommendations and upgrade jumps, and they are also more optimistic in earnings forecasts and price targets. While it is easy to pick up their mentors' styles, I show that it is apparently harder for them to learn their mentors' skills, as indicated by the lack of spillover in forecast accuracy. Only talented superstar mentors can unwind this pattern, passing their skills and reputation to their proteges. The market—especially sophisticated institutional investors—is smart in identifying the apprentices of optimistic mentors as short-run market reactions to their forecast revisions are weaker. Collectively, these results have important implications for financial economists and regulators (on a new source of optimism), for analyst profession (on talent management and portability), and for market participants (on information dissemination and optimism debias).Item Essays on participation dynamics and informational value of virtual communities(2010-08) Park, Jaehong, 1973-; Gu, Bin, Ph. D.; Konana, Prabhudev C.; Raghunathan, Rajagopal; Kumar, Alok; Whinston, Andrew B.In recent years, virtual communities have become increasingly popular among online users and businesses. Public press has shown a strong interest in virtual communities and announced their informational value. The explosive growth of virtual communities has aroused tremendous interest among academic researchers as well. Academic researchers have noticed that participation dynamics of virtual communities among online users create valuable information that influences subsequent economic outcome. However, relatively little research has explored the underlying motives of passive and active participation in virtual communities. In particular, research explicating how the information created by user participation influences users’ economic decisions has not been investigated. To investigate user participation dynamics and the resulting informational value of virtual communities, I explore three major issues in my dissertation. First, I investigate the determinants of passive and active users of virtual communities using survey data from 502 online investors. The results indicate that social, psychological, and community factors influence two different behavioral intentions – the intention to share and the intention to seek. For instance, social factors such as reputation seeking increase the intention to share in virtual communities, whereas psychological factors such as perceived knowledge deteriorate the intention to seek. Second, I explore how an online investor processes information posted on virtual communities and its subsequent economic outcomes by conducting a field experiment. I find that psychological bias (e.g., confirmation bias) influences investors’ information processing behaviors, which ultimately leads to a low return of investment as compared to economic rationales. Following this line of exploration, I empirically examine the relative informational impact of virtual communities on users’ decisions and market performance in the context of electronic markets. Using customer review data from a well-known online retailer and from three third-party customer review websites, I find that consumers obtain product information from external information sources during the information search stage for high involvement products and are thus less influenced by retailer-hosted information. All in all, my dissertation contributes to the understanding of user participation dynamics and informational value of virtual communities by investigating users’ information processing behavior and the subsequent economic outcomes and performance.Item Essays on social values in finance(2011-05) Page, Jeremy Kenneth; Kumar, Alok; Titman, Sheridan; Parrino, Robert; Sialm, Clemens; Spalt, OliverThis dissertation consists of three essays on the role of social values in financial markets. Chapter 1 uses geographic variation in religious concentration to identify the effect of people's gambling behavior in financial market settings. We argue that religious background predicts people's gambling propensity, and that gambling propensity carries over into their behavior in financial markets. We test this conjecture in various financial market settings and find that the predominant local religion predicts variation in investors' propensity to hold stocks with lottery features, in the prevalence of broad-based employee stock option plans, in first-day returns to initial public offerings, and in the magnitude of the negative lottery-stock return premium. Collectively, our findings indicate that religious beliefs regarding the acceptability of gambling impact investors' portfolio choices, corporate decisions, and stock returns. In Chapter 2 I examine the impact of social norms against holding certain types of stocks (e.g. "sin stocks", or stocks with lottery features) on trading decisions and portfolio performance. I argue that trades which deviate from social norms are likely to reflect stronger information. Consistent with this hypothesis, I find that the most gambling-averse institutions earn high abnormal returns on their holdings of lottery stocks, outperforming the holdings of the most gambling-tolerant institutions. An analysis of institutions' sin stock holdings provides complementary evidence using another dimension of social norms, supporting the hypothesis that trades which deviate from norms reflect stronger information. In the third essay, we conjecture that people feel more optimistic about the economy and stock market when their own political party is in power. We find supporting evidence from Gallup survey data and analyze brokerage account data to confirm the impact of time-varying optimism on investors' portfolio choices. When the political climate is aligned with their political preferences, investors maintain higher systematic risk exposure while trading less frequently. When the opposite party is in power, investors exhibit stronger behavioral biases and make worse investment decisions. Investors improve their raw portfolio performance when their own party is in power, but the risk-adjusted improvement is economically small.Item Fuzzy-based wind damage model for single-family housing(Texas Tech University, 2001-05) Kumar, AlokResidential construction is generally non-engineered and sustains the most damage of all the construction types in the event of a windstorm. With extensive development along the hurricane prone coastal areas, more and more residential property is exposed to hurricane risk. As a result, a probabilistic estimate of damage to these structures is crucial to the insurance industry. In the present work of developing a wind damage model, the authors have proposed a methodology for damage prediction that takes into account both expert opinion and strength of components and connections in a building, as provided in the literature. A building is considered to be an aggregation of components and connections that are assembled together according to standard construction practices. The strength of each component and connection is represented by a suitable probability distribution. To account for the uncertainty in the strength, a probability distribution is represented by fuzzy membership functions. The effect of damage to the building envelope is included in the analysis by updating the wind load acting on components when the building envelope is breached. Expert opinion is used to establish the fraction of interior damaged given there is damage to any of the exterior components. A component exposure parameter, based on expert opinion, defines the maximum fraction of damage to a component at any given wind speed. These expert supplied parameters are represented by fuzzy membership functions to account for uncertainty in the expert judgment. A fuzzy model of building damage as a function of wind speed is then developed. .