Indiana University Bloomington, IN East Tenth Street Tel.:

Transcription

1 SHAWN M. O DONOGHUE Department of Finance Mailing Address: 5103 South Rogers Street Bloomington, IN East Tenth Street Tel.: Bloomington, IN sodonogh@indiana.edu AREAS OF INTEREST Research: Empirical and Theoretical Market Microstructure and Empirical Asset Pricing Teaching: Investments and Financial Markets and Institutions EDUCATION Ph.D. Candidate, Finance, Minor: Statistics Committee: Robert H. Jennings (Chair), Craig W. Holden, Zhenyu Wang, Ryan D. Israelsen, and Juan Carlos Escanciano Proposal: August 2013, Expected Final Defense: January 2015 Master of Business, Finance, 2011 University of Denver Master of Science, Mathematics, 2008 Middlebury College Master of Arts, French, Location: Paris, France, 1997 Georgetown University Bachelor of Arts, Major: Economics, Minor: French, 1994 WORKING PAPERS The Effect of Maker-Taker Fees on Investor Order Choice and Execution Quality in U.S. Stock Markets (Job Market Paper) Financial Management Association Doctoral Student Consortium (2014) scheduled Presented at 2013 An Evaluation of Test Portfolios Employed in Tests of Assets Pricing Models Benefit of Being Second: An Event Study of Two Social Media Adoptions, with Jeffrey Cummings and Alan R. Dennis Revise and Resubmit at MIS Quarterly (top tier journal in information systems) HONORS AND AWARDS NASDAQ OMX Group Ph.D. Dissertation Fellowship Research Database Matching Grant 2013 Research Database Matching Grant 2012 American Finance Association Doctoral Student Travel Grant, Chicago, IL 2012 William H. Morris Endowment Scholarship 2011 Assistantship and Fellowship, Graduate Teaching Assistantship, University of Denver Page 1 of 5

2 TEACHING EXPERIENCE Associate instructor with full responsibility Spring 2013 Summer 2012 Security Trading and Market Making (F335) 42 students Dean s 8 Index: Mean 5.23/7.00, Median 5.63/7.00 Security Trading and Market Making (F335) 10 students Dean s 8 Index: Mean 6.49/7.00, Median 6.63/7.00 Working Capital Management (F307) 10 students Dean s 8 Index: Mean 5.23/7.00, Median 5.63/7.00 Summer 2011 Security Trading and Market Making (F335) 31 students Dean s 8 Index: Mean 5.13/7.00, Median 5.75/7.00 University of Denver Teaching Assistant One Quarter Calculus I MATH 1951 Five Quarters Calculus for Business and Social Sciences MATC 1200 RESEARCH EXPERIENCE Research Assistant to Noah Stoffman () Software: Databases: SAS, Stata, R, Eventus, Maple, LaTex, UNIX, MS Excel, MS Powerpoint, MS Access CRSP, Compustat, TAQ, Thomson Reuters, Thomson Financial SDC, Datastream PROFESSIONAL EXPERIENCE VHA Mountain States, L.L.C.! Analyst (Denver, CO) Massachusetts General Hospital! Project Manager (Boston, MA) Oakmont Laboratories, L.L.C.! Bilingual Marketing Assistant (Bedford, MA) 2001 Boston Health Economics, Inc.! Research Assistant (Billerica, MA) PROFESSIONAL AFFILIATIONS American Finance Association and Financial Management Association LANGUAGES French (Advanced) Page 2 of 5

3 REFERENCES Robert H. Jennings (Chair) Craig W. Holden Gregg T. and Judith A. Summerville Boquist-Meyer Faculty Fellow Phone: Phone: Zhenyu Wang Ryan D. Israelsen Assistant Phone: Phone: Page 3 of 5

4 ABSTRACTS OF WORKING PAPERS The Effect of Maker-Taker Fees on Investor Order Choice and Execution Quality in U.S. Stock Markets (Job Market Paper) Equity exchanges competing for orders are using new pricing strategies. Typically, liquidity suppliers are compensated and liquidity demanders are charged. This pricing structure is controversial because of its potential effects on investor order choice, market quality, trader welfare, and economic efficiency. I develop a theoretical model of maker-taker fees in the presence of a broker and equity exchanges and test the model empirically using order level data from SEC Rule 605. The broker charges investors a commission and endogenously chooses to route orders to a dealer or equity exchange. The exchanges keep a portion of the taker fee as profit and pass the remaining amount to the broker as a maker rebate when its order providing liquidity executes. The theoretical model predicts that as the taker fee and maker rebate increase, holding constant the amount kept as profit by the exchange: (1) the bid-ask spread declines, (2) the total trading cost increases, (3) the trader participation falls, (4) the proportion of marketable order shares rises, and (5) the non-marketable limit order fill rate increases. These implications are different from how the market is currently expected to operate as understood from the work of Colliard and Foucault (2012), because my model implies that changes in the split of trading fees between liquidity suppliers and demanders affect order choice and thereby execution quality. I find empirical evidence consistent with my model s predictions. In particular, as the taker fee and maker rebate increase, holding constant the amount kept as profit by the exchange: (1) the bid-ask spread declines, (2) the trader participation falls, (3) the proportion of marketable order shares rises, and (4) the non-marketable limit order fill rate increases. An Evaluation of Test Portfolios Employed in Tests of Assets Pricing Models (Second Summer Paper) A large number of new asset pricing models, using economically different risk factors, have been introduced over the past decade and claim to account for size-bm anomalies. Even if these proposed factors do not explain the true expected returns, they will estimate high cross-sectional R 2 s when tested using 25 size-bm portfolios (Fama and French 1993) as long as the factor is correlated with SMB or HML (Lewellen, Nagel, and Shanken 2010). This occurs because the FF test portfolios are created by sorting on size-bm. I test whether cluster (Ahn, Conrad, and Dittmar 2009) or q-theory portfolios (Chen, Novy-Marx, and Zhang 2011) alone or in combination with industry portfolios provide a more rigorous test of asset pricing models than the 25 size-bm portfolios. I randomly generate artificial factors that are a linear combination of FF factors or test portfolios, and report simulated distributions of R 2 s when expected returns of test portfolios are regressed on artificial factor loadings. The artificial factors are created using three different methods: (1) artificial factors are correlated with FF factors but not with idiosyncratic residuals of size-bm portfolios, (2) artificial factors are not completely uncorrelated with residuals of test portfolios, and (3) repeat method two but keep only artificial factors (random combinations of test portfolios) with roughly zero expected returns. After running the simulations I find that cluster portfolios provide a more rigorous test than the FF 25 size-bm portfolios, because cluster portfolios are created using covariance of returns instead of sorts on firm size-bm. Q-theory portfolios are more difficult to explain than the 25 size-bm portfolios, but are a less rigorous test than the cluster portfolios. This occurs because HML and investment-to-asset factors will be correlated as value firms with high BM will invest less relative to growth firms with low BM. Restrictions on cross-sectional intercepts and slopes are less important for cluster portfolios than 25 size-bm portfolios. However, they are important for the q-theory portfolios. Adding 30 industry portfolios to any test portfolio, ensures that smaller R 2 s are generated in regressions with artificial factors. Benefit of Being Second: An Event Study of Two Social Media Adoptions (with Jeffrey Cummings and Alan R. Dennis) New technologies are continually being introduced and changing how firms and customers interact. One of the most challenging questions a firm faces is determining whether and when to adopt technology. Prior research in information systems has linked the business value of IT to overall firm performance but it is not known whether Page 4 of 5

5 social networking sites provide similar benefits. We categorize consumer products firms into types based on the use and timing of technology adopters: innovators (i.e., first movers), early adopters (i.e., fast followers), and early majority. We conduct an event study examining the abnormal return following the date a firm joins Twitter or publishes a Facebook page. When the expected return is estimated using the market model, we find that the cumulative abnormal return in the day following the event is positive and statistically significant at the 1% level or lower for firms using Facebook or Twitter according to the Patell or Portfolio Time-Series tests. Results are similar for firms using Facebook when the FF 3- or Carhart 4-factor model is the benchmark; results are also positive and statistically significant for Twitter adopters but at the 5% level. We find that innovators do not earn a positive abnormal return but early adopters do gain a positive abnormal return independent of the benchmark used to estimate expected returns. In addition, there some evidence that returns of early majority firms increase, but only when adopting Facebook and using the FF 3- and Carhart 4-factor models as benchmarks. Page 5 of 5