Asset Pricing

April 23, 2010
Nicolae B. Garleanu and Martin Lettau, Organizers

Jules H. Van Binsbergen, Stanford University; Michael W. Brandt, Duke University and NBER; and Ralph S. J. Koijen, University of Chicago
On Timing and Pricing of Cash Flows

van Binsbergen, Brandt, and Koijen study the pricing of short-term assets that are claims to the dividends of the aggregate stock market for a period of up to three years. To compute these prices, they apply put-call parity to a newly constructed, and importantly better synchronized, dataset of liquid, exchange-traded S&P500 index options. They compare the asset pricing properties of the claim to short-term dividends to the pricing of the aggregate stock market, which is the claim to all future dividends. They find that the short-term asset has high expected returns, a beta to the market of 0.5, is excessively volatile, and has returns that are highly predictable. The returns on short-term dividend claims cannot be explained by standard asset pricing models, which makes such claims important candidate test assets. The researchers compare their empirical results to their theoretical equivalents in leading asset pricing models and find that none of them predict the empirical findings that they document.


Gregory R. Duffee, John Hopkins University
Sharpe Ratios in Term Structure Models

Duffee notes that conditional maximum Sharpe ratios implied by fully flexible four-factor and five-factor Gaussian term structure models are astronomically high. Estimation of term structure models subject to a constraint on their Sharpe ratios uncovers properties that hold for a wide range of Sharpe ratios. These robust properties include 1) an inverse relation between a bond's maturity and its average Sharpe ratio; 2) between 15 and 20 percent of annual excess returns to bonds are predictable; and 3) variations in expected excess bond returns are driven by two factors. These factors operate at different frequencies. Non-robust features include the mean level of the term structure. Unconstrained models imply that investors anticipated much of the decline of interest rates in the 1990s. Constrained models disagree.


Tarek A. Hassan, University of Chicago, and Thomas M. Mertens, New York University
The Social Cost Near-Rational Investment: Why We Should Worry About Volatile Stock Markets

Excess volatility in stock returns may arise and drastically reduce welfare even if the stock market appears to be efficient and disconnected from the real economy. Hassan and Mertens solve a macroeconomic model in which information about fundamentals is dispersed and agents make small, correlated errors around their optimal investment policies. As information aggregates in the market, these errors amplify and result in large amounts of excess volatility in stock returns. The increase in volatility makes holding stocks unattractive and distorts the long-run level of capital accumulation. Through its effect on capital accumulation excess volatility causes costly (first-order) distortions in the long-run level of consumption.

Juhani T. Linnainmaa, University of Chicago
Reverse Survivorship Bias

Mutual funds often disappear following poor performance. When this poor performance is partly attributable to negative idiosyncratic shocks, the fund's estimated alpha understates its true alpha. Linnainmaa develops and estimates a structural model to evaluate this bias. He finds that the bias in the mean of the observed alpha distribution is approximately 1 percent per year. After correcting for this bias using historical data, he finds that the majority of fund managers still have negative net alphas, but the average is not nearly as low as what the fund-level estimates suggest. This reverse survivorship bias affects all studies that run fund-level regressions to draw inferences about fund managers' abilities.


Hui Chen, Scott Joslin, and Ngoc-Khanh Tran, MIT
Rare Disasters and Risk Sharing with Heterogeneous Beliefs

Although the threat of rare economic disasters can have large effect on asset prices, diffculty in inferring their likelihood and severity provides the potential for disagreements among investors. Such disagreements lead investors to insure each other against the types of disasters each one fears the most. Because of to the highly non-linear relationship between consumption losses in a disaster and the risk premium, a small amount of risk sharing can significantly attenuate the effect of disaster risk on the equity premium. Chen, Joslin, and Tran show that time variation in the wealth distribution and the amount of disagreement across agents both can lead to significant variation in disaster risk premium. They also highlight the conditions under which disaster risk premium will be large, namely when disagreement across agents is small, or when the wealth distribution is highly concentrated in agents fearful of disasters. Finally, their model predicts an inverse U-shaped relationship between the equity premium and the size of the disaster insurance market.


Maxim Ulrich, Columbia University
Observable Long Run Ambiguity and Long Run Risk

Ulrich derives and estimates a general equilibrium model for the real and nominal term structure of U.S. government bonds with only observable macro variables. The model takes into account that investors are confronted with a set of multiple long-run risk models. He accounts for model misspecification doubts about long-run GDP risk and about long-run inflation risk. He finds that an increase in macro uncertainty leads to a steepening in TIPS and nominal yields. Increased uncertainty about the long-run GDP model generates a steeper slope in TIPS yields than the inflation uncertainty counterpart. On the other hand, he finds that the term premium in TIPS and nominal bond yields is dominated by model uncertainty about long-run inflation. The estimated robustness preference for ambiguity about long-run inflation is 7.6 and 0.3 for long-run GDP ambiguity.