State Dependence Can Explain the Risk Aversion Puzzle

Risk aversion functions extracted from observed stock and optionprices can be negative, as shown by Aït-Sahalia and Lo(2000), Journal of Econometrics 94: 9–51; and Jackwerth(2000), The Review of Financial Studies 13(2), 433–51.We rationalize this puzzle by a lack of conditioning on latentstate variables. Once properly conditioned, risk aversion functionsand pricing kernels are consistent with economic theory. Todifferentiate between the various theoretical explanations interms of heterogeneity of beliefs or preferences, market sentiment,state-dependent utility, or regimes in fundamentals, we calibrateseveral consumption-based asset pricing models to match theempirical pricing kernel and risk aversion functions at differentdates and over several years.     

Empirical pricing kernels

This paper investigates the empirical characteristics of investor risk aversion over equity return states by estimating a time-varying pricing kernel, which we call the empirical pricing kernel (EPK). We estimate the EPK on a monthly basis from 1991 to 1995, using S&P 500 index option data and a stochastic volatility model for the S&P 500 return process. We find that the EPK exhibits counter cyclical risk aversion over S&P 500 return states. We also find that hedging performance is significantly improved when we use hedge ratios based the EPK rather than a time-invariant pricing kernel.     

Option Pricing Bounds and the Elasticity of the Pricing Kernel

In this paper we use power functions as pricing kernels to derive option-pricing bounds. We derive option pricing bounds given the bounds of the elasticity of the true pricing kernel. The bounds of the elasticity of the true pricing kernel are closely related to the bounds of the representative investor's coefficient of relative risk aversion. This methodology produces a tighter upper call option bound than traditional approaches. As a special case we show how to use the Black–Scholes formula to obtain option pricing bounds under the assumption of lognormality.     

The Recovery Theorem

We can only estimate the distribution of stock returns, but from option prices we observe the distribution of state prices. State prices are the product of risk aversion—the pricing kernel—and the natural probability distribution. The Recovery Theorem enables us to separate these to determine the market's forecast of returns and risk aversion from state prices alone. Among other things, this allows us to recover the pricing kernel, market risk premium, and probability of a catastrophe and to construct model‐free tests of the efficient market hypothesis.     

On the Distributional Conditions for a Consumption-oriented Three Moment CAPM

In this paper, we develop sufficient conditions on probability distributions for a three moment (mean, variance, and skewness) consumption-oriented capital asset pricing model (CAPM) to price correctly a subset of assets. The assumptions that individuals in an allocationally efficient capital market have identical probability beliefs and monotone increasing strictly concave utility functions displaying nonincreasing absolute risk aversion imply an aggregate preference function that exhibits preference for expected return, aversion to variance of return, and preference for positive skewness. For otherwise arbitrary preferences, we show that quadratic characteristic lines are sufficient for a subset of assets to be priced according to a three moment consumption-oriented CAPM.     

Stock performance by utility indifference pricing and the Sharpe ratio

We compare stock performance based on utility indifference pricing and the Sharpe ratio assuming that stock returns follow the class of discrete normal mixture distributions. The utility indifference price with an exponential utility function satisfies several desirable properties that a suitable value measure should satisfy. For utility indifference pricing, we employ the inner rate of risk aversion proposed by Miyahara [Evaluation of the scale risk. RIMS Kokyuroku, No. 1886, Financial Modeling and Analysis (2013/11/20-2013/11/22), 181–188, 2014], which is the degree of risk aversion that makes the utility indifference price with the exponential utility function zero in order to evaluate stock performance. Using a selection of U.S. stocks, the results show that the evaluation of stock performance based on the inner rate of risk aversion is more relevant for risk-averse investors than that based on the Sharpe ratio, which represents performance by the first two moments.     

Generalized Disappointment Aversion and Asset Prices

We characterize generalized disappointment aversion (GDA) risk preferences that can overweight lower‐tail outcomes relative to expected utility. We show in an endowment economy that recursive utility with GDA risk preferences generates effective risk aversion that is countercyclical. This feature comes from endogenous variation in the probability of disappointment in the representative agent's intertemporal consumption‐saving problem that underlies the asset pricing model. The variation in effective risk aversion produces a large equity premium and a risk‐free rate that is procyclical and has low volatility in an economy with a simple autoregressive endowment‐growth process.     

Market pricing of executive stock options and implied risk preferences

When managers get to trade in options received as compensation, their trading prices reveal several aspects of subjective option pricing and risk preferences. Two subjective pricing models are fitted to show that executive stock option prices incorporate a subjective discount. It depends positively on implied volatility and negatively on option moneyness. Further, risk preferences are estimated using the semiparametric model of Aït-Sahalia and Lo (2000). The results suggest that relative risk aversion is just above 1 for a certain stock price range. This level of risk aversion is low but reasonable, and it may be explained by the typical manager being wealthy and having low marginal utility. Related to risk aversion, it is found that marginal rate of substitution increases considerably in states with low stock prices.     

Market incompleteness and the equity premium puzzle: Evidence from state-level data

This paper investigates the importance of market incompleteness by comparing the rates of risk aversion estimated from complete and incomplete markets environments. For the incomplete-markets case, we use consumption data for the 50 US states. We find that the rate of risk aversion under the incomplete-markets setup is much lower. Furthermore, including the second and third moments of the cross-sectional distribution of consumption growth in the pricing kernel lowers the estimate of risk aversion. These findings suggest that market incompleteness ought to be seen as an important component of solutions to the equity premium puzzle.     

First steps towards an equilibrium theory for Lévy financial markets

For a continuous-time financial market with a single agent, we establish equilibrium pricing formulae under the assumption that the dividends follow an exponential Lévy process. The agent is allowed to consume a lump at the terminal date; before that, only flow consumption is allowed. The agent’s utility function is assumed to be additive, defined via strictly increasing, strictly concave smooth felicity functions which are bounded below (thus, many CRRA and CARA utility functions are included). For technical reasons we require for our equilibrium existence result that only pathwise continuous trading strategies are permitted in the demand set. The resulting equilibrium asset price processes depend on the agent’s risk aversion (through the felicity functions). Even in our simple, straightforward economy, the equilibrium asset price processes will essentially only be (stochastic) exponential Lévy processes when they are already geometric Brownian motions. Our equilibrium asset pricing formulae can also be modified to obtain explicit equilibrium derivative pricing formulae.     

Option pricing under time-varying risk-aversion with applications to risk forecasting

We present a two-factor option-pricing model, which parsimoniously captures the difference in volatility persistences under the historical and risk-neutral probabilities. The model generates an S-shaped pricing kernel that exhibits time-varying risk aversion. We apply our model for two purposes. First, we analyze the risk preference implied by S&P500 index options during 2001–2009 and find that risk-aversion level strongly increases during stressed market conditions. Second, we apply our model for Value-at-Risk (VaR) forecasts during the subprime crisis period and find that it outperforms several leading VaR models.     

Option-Implied Risk Aversion Estimates

Using a utility function to adjust the risk‐neutral PDF embedded in cross sections of options, we obtain measures of the risk aversion implied in option prices. Using FTSE 100 and S&P 500 options, and both power and exponential‐utility functions, we estimate the representative agent's relative risk aversion (RRA) at different horizons. The estimated coefficients of RRA are all reasonable. The RRA estimates are remarkably consistent across utility functions and across markets for given horizons. The degree of RRA declines broadly with the forecast horizon and is lower during periods of high market volatility.     

Heterogeneity and Option Pricing

An economy with agents having constant yetheterogeneous degrees of relative risk aversion prices assetsas though there were a single decreasing relative risk aversion``pricing representative' agent. The pricing kernel has fattails, and option prices do not conform to the Black-Scholesformula. Implied volatility exhibits a ``smile.' Heterogeneityas the source of non-stationary pricing fits Rubenstein's (1994)interpretation of the ``over-pricing' as an indication of ``crash-o-phobia'.Rubinstein's term suggests that those who hold out-of-the moneyput options have relatively high risk aversion (or relativelyhigh subjective probability assessments of low market outcomes).The essence of this explanation is investor heterogeneity.     

Decision-making under uncertainty – A field study of cumulative prospect theory

The presented research tests cumulative prospect theory (CPT, [Kahneman, D., Tversky, A., 1979. Prospect theory: An analysis of decision under risk. Econometrica 47, 263–291; Tversky, A., Kahneman, D., 1981. The framing of decisions and the psychology of choice. Science 211, 453–480]) in the financial market, using US stock option data. Option prices possess information about actual investors’ preferences in such a way that an exploitation of conventional option analysis, along with theoretical relationships, makes it possible to elicit investor preferences. The option data in this study serve for estimating the two essential elements of the CPT, namely, the value function and the probability weighting function. The main part of the work focuses on the functions’ simultaneous estimation under CPT original parametric specification. The shape of the estimated functions is found to be in line with theory. Comparing to results of laboratory experiments, the estimated functions are closer to linearity and loss aversion is less pronounced.     

A productivity analysis of Central and Eastern European banking taking into account risk decomposition and environmental variables

This paper develops a Luenberger productivity index that is applied to a technology where desirable and undesirable outputs are jointly produced and are possibly negative. The components of this Luenberger productivity index - the efficiency change and the components of the technological shift - are then decomposed into factors determined by the technology, adjusted and then for ‘risk and environment’, ‘risk management’ and ‘environmental effects’. The method is applied to Central and Eastern European banks operating during 1998-2003 utilising three alternative input/output methodologies (intermediation, production and profit/revenue). Additionally, the comparative analysis of the sensitivity of the productivity indices in the choice of the methodologies is undertaken using statistical and kernel density tests. It is found that the main driver of productivity change in Central and Eastern European banks is technological improvement. That is, in the beginning of the analysed period, the results hinged on the banks ability to capitalise on advanced technology and successfully take into account ‘risk and environmental’ factors. Whereas, in later periods, one of the most important factors of technological improvement/decline was ‘risk management’. Finally, the tests employed confirm previous findings, such as Pasiouras (2008)in this journal, that different input/output methodologies produce statistically different productivity results. Finally, we find that external factors, such as ‘risk in the economy’ and banking production, and a ‘corruption perception’ affect the productivity of banks.     

Time-varying risk preferences and emerging market co-movements

This paper examines how shocks can transmit across international stock markets through the channel of time-varying investor risk preferences. We highlight the effects of this channel by comparing the conventional constant relative risk aversion utility function with the habit-formation utility function of Campbell and Cochrane (J. Pol. Econ. 107 (1999) 205). Calibrating our model with data from Argentina, Korea and Mexico, we find that in the presence of time-varying investor risk preferences, market integration generates a substantial increase in cross-country co-movements of stock returns.     

Risk Premia and Variance Bounds

If a pricing kernel assigns a premium to a risk variable that differs from the one assigned by the minimum-variance admissible kernel, then the pricing kernel must exhibit more variability than the minimum-variance kernel. Based on this intuition, we derive a variance bound that is more stringent than that of Hansen and Jagannathan (1991) . When we apply our bound to the kernel of a representative consumer with power utility, we find that the consumption risk premium increases the severity of the “equity-premium puzzle” of Mehra and Prescott (1985) .     

Equilibrium asset pricing: with non-Gaussian factors and exponential utilities

We analyse the equilibrium asset pricing implications for an economy with single period return exposures to explicit non-Gaussian systematic factors, that may be both skewed and long-tailed, and Gaussian idiosyncratic components. Investors maximize expected exponential utility and equilibrium factor prices are shown to reflect exponentially tilted prices for non-Gaussian factor risk exposures. It is shown that these prices may be directly estimated from the univariate probability law of the factor exposure, given an estimate of average risk aversion in the economy. In addition, a residual form of the capital asset pricing model continues to hold and prices the idiosyncratic or Gaussian risks. The theory is illustrated on data for the US economy using independent components analysis to identify the factors and the variance gamma model to describe the probability law of the non-Gaussian factors. It is shown that the residual CAPM accounts for no more than 1% of the pricing of risky assets, while the exponentially tilted systematic factor risk exposures account for the bulk of risky asset pricing.     

The shape of small sample biases in pricing kernel estimations

Numerous empirical studies find pricing kernels that are not-monotonically decreasing; the findings are at odds with the pricing kernel being marginal utility of a risk-averse, so-called representative agent. We study in detail the common procedure which estimates the pricing kernel as the ratio of two separate density estimations. In the first step, we analyse theoretically the functional dependence for the ratio of a density to its estimated density; this cautions the reader regarding potential computational issues coupled with statistical techniques. In the second step, we study this quantitatively; we show that small sample biases shape the estimated pricing kernel, and that estimated pricing kernels typically violate the commonly believed monotonicity at the centre even when the true pricing kernel fulfils these. This contributes to an alternative, statistical explanation for the puzzling shape in pricing kernel estimations.