PhD Candidate

Marshall School of Business - USC

Bridge Hall 308

3670 Trousdale Parkway

Los Angeles, CA 90089

Phone: 310.309.0845

Email: duchin@usc.edu   

Cash Holdings and Corporate Diversification, September 2007

Journal of Finance, Revise and Resubmit

Job Market Paper

This paper documents a strong connection between corporate cash holdings and organization form, and explores why such a relation exists. Over the period 1990-2004, U.S. multi-division firms held 39 percent less cash as a fraction of assets than stand-alone firms, and the difference persists after controlling for firm size, capital structure, cash flows, growth opportunities and investments. To explain the difference, I study the determinants of cash holdings in a sample of 10,380 firms over 1990-2004 and find that (a) correlations between divisional cash flows or investment opportunities completely explain the relation between cash holdings and organization form and (b) their effect is big: an increase of one standard deviation in the correlations between cash flows (investment opportunities) implies an increase of 13 percent (10 percent) in cash holdings of the average firm. This suggests that corporate diversification can serve as a way to economize on cash. The paper also documents a substantial increase in average cross-divisional correlations from 1990 to 2004, which might explain why diversified firms hold more cash than they used to.

When Are Outside Directors Effective? (with John G. Matsusaka and Oguzhan Ozbas), October 2007

Presented at the NBER Summer Institute, 2007

American Economic Review, Under Review

New regulations and corporate governance activists have called for increased numbers of outside directors on boards and committees, yet existing research has failed to find convincing evidence that outside directors improve firm performance. This paper estimates the effect of outside directors using a new empirical strategy that controls for the well known endogeneity problem in board composition by focusing on firms that were required to increase the number of outside directors as a result of the Sarbanes-Oxley Act. We find that the effect of outside directors on performance depends on their cost of acquiring information: outside directors are effective when the cost of acquiring information is low and are ineffective when the cost of acquiring information is high. We also find that firms compose their boards as if they understand that outsider effectiveness varies with information costs.

Riding the Merger Wave (with Breno Schmidt), November 2007

This paper proposes that self-serving, empire building managers strategically initiate inefficient acquisitions during periods of intense merger activity (“merger waves”). We document that corporate acquisitions during waves lead to worse long-term performance relative to non-wave mergers. However, managers that acquire during waves are less likely to be fired following unsuccessful mergers. We examine reasons for this and argue that merger waves provide a mechanism through which empire builders can conceal their true motives. Consistent with our story, we also find evidence associating in-wave mergers with poor governance. Overall, our results bring forth a possible link, unexplored in the literature, between agency theory and merger waves.

Disagreement, Portfolio Optimization and Excess Volatility (with Moshe Levy), May 2007

Econometrica, Under Review

A central task facing investors who believe in market efficiency is that of portfolio optimization. As it is far from obvious how to best estimate the ex-ante expected returns and covariances, it is quite plausible that investors would hold different beliefs regarding these parameters, and that the degree of disagreement about the parameters may change over time. Levy, Levy and Benita (2006) have shown that in the portfolio context disagreement regarding the expected returns does not affect asset prices. In this paper we study the pricing effects of disagreement regarding return variances. We show that disagreement about variances has systematic and significant pricing effects. Even if the average belief about the variance is constant, tiny fluctuations in the disagreement about the variance lead to substantial price fluctuations. This result may offer an explanation for the excess volatility puzzle: small changes in the degree of disagreement are very likely to occur, and they induce relatively large price changes. Yet, the changes in disagreement may be hard to directly detect empirically, leading to apparent “excess volatility”. 

Portfolio Optimization and the Distribution of Firm Size (with Moshe Levy), September 2007

In the context of portfolio optimization, a firm’s market capitalization reflects the optimal portfolio weight of the firm, and is determined by the return parameters. The empirical distribution of firms’ market capitalizations is in excellent agreement with the lognormal distribution. This distribution is very skewed: the largest firms are about 1000 times larger than the median firm. The empirical distribution of average returns is not nearly as skewed: the maximal average return is only about 6 times larger than the median average return. Can the empirical firm size distribution be consistent with mean-variance portfolio optimization with realistic return parameters? We show that the expected returns implied by the empirical firm size distribution and portfolio optimization are actually in very good agreement with the empirical average returns. Moreover, the portfolio optimization framework can provide a constructive explanation for the exact lognormal functional form empirically observed. Thus, portfolio optimization is not only consistent with the empirical lognormal size distribution, it can actually explain it.

Yes, You Can Assume Normality (with Haim Levy), January 2007

Most theoretical economic models under uncertainty assume that asset returns follow a specific parametric distribution (e.g.: Normal). This study develops a methodology to examine the economic loss, in terms of expected utility, induced by approximating the historical distribution of asset returns with various parametric distributions. We compare the performance of different parametric approximations relative to the empirical distribution, using a power utility function. Our results reveal that while the statistical goodness-of-fit tests strongly reject the Normal distribution, assuming Normality induces a negligible economic loss. Thus, practically one can employ the Mean-Variance rule even without Normality. We also conduct a horse race between various parametric distributions for various risk aversion levels. We find that the Skew Normal and the Multivariate Skew t distributions induce the smallest economic loss.

Publications

Asset Return Distributions and the Investment Horizon (with Haim Levy), Spring 2004

                Journal of Portfolio Management, 47-61.

The optimal investment decision rule and asset equilibrium prices depend on the assumed distribution of rates of return. And empirical distributions vary with the assumed time interval (investment horizon). We test the goodness of fit of 11 theoretical distributions including the normal distribution, fat-tailed distributions, and skewed distributions for investment horizons ranging from one day to four years. The normal distribution performs poorly, and never provides the best fit for any time interval. In the 330 goodness of fit tests reported, at least one distribution of the 11 always fits the data better than the normal distribution. As the logistic distribution fits the data best for investment horizons of up to one year, analysis focuses on this distribution and its implications for equilibrium asset prices.