Refine
Year of publication
Document Type
- Working Paper (16)
- Report (2)
Language
- English (18)
Has Fulltext
- yes (18)
Is part of the Bibliography
- no (18)
Keywords
- General Equilibrium (7)
- Asset Pricing (4)
- Contagion (4)
- stochastic volatility (4)
- Asset Allocation (3)
- Recursive Preferences (3)
- Discretization Error (2)
- Hedging (2)
- Model Error (2)
- Stochastic Volatility (2)
Institute
- Wirtschaftswissenschaften (18) (remove)
n this paper we analyze an economy with two heterogeneous investors who both exhibit misspecified filtering models for the unobservable expected growth rate of the aggregated dividend. A key result of our analysis with respect to long-run investor survival is that there are degrees of model misspecification on the part of one investor for which there is no compensation by the other investor's deficiency. The main finding with respect to the asset pricing properties of our model is that the two dimensions of asset pricing and survival are basically independent. In scenarios when the investors are more similar with respect to their expected consumption shares, return volatilities can nevertheless be higher than in cases when they are very different.
This paper provides an in-depth analysis of the properties of popular tests for the existence and the sign of the market price of volatility risk. These tests are frequently based on the fact that for some option pricing models under continuous hedging the sign of the market price of volatility risk coincides with the sign of the mean hedging error. Empirically, however, these tests suffer from both discretization error and model mis-specification. We show that these two problems may cause the test to be either no longer able to detect additional priced risk factors or to be unable to identify the sign of their market prices of risk correctly. Our analysis is performed for the model of Black and Scholes (1973) (BS) and the stochastic volatility (SV) model of Heston (1993). In the model of BS, the expected hedging error for a discrete hedge is positive, leading to the wrong conclusion that the stock is not the only priced risk factor. In the model of Heston, the expected hedging error for a hedge in discrete time is positive when the true market price of volatility risk is zero, leading to the wrong conclusion that the market price of volatility risk is positive. If we further introduce model mis-specification by using the BS delta in a Heston world we find that the mean hedging error also depends on the slope of the implied volatility curve and on the equity risk premium. Under parameter scenarios which are similar to those reported in many empirical studies the test statistics tend to be biased upwards. The test often does not detect negative volatility risk premia, or it signals a positive risk premium when it is truly zero. The properties of this test furthermore strongly depend on the location of current volatility relative to its long-term mean, and on the degree of moneyness of the option. As a consequence tests reported in the literature may suffer from the problem that in a time-series framework the researcher cannot draw the hedging errors from the same distribution repeatedly. This implies that there is no guarantee that the empirically computed t-statistic has the assumed distribution. JEL: G12, G13 Keywords: Stochastic Volatility, Volatility Risk Premium, Discretization Error, Model Error
Tests for the existence and the sign of the volatility risk premium are often based on expected option hedging errors. When the hedge is performed under the ideal conditions of continuous trading and correct model specification, the sign of the premium is the same as the sign of the mean hedging error for a large class of stochastic volatility option pricing models. We show, however, that the problems of discrete trading and model mis-specification, which are necessarily present in any empirical study, may cause the standard test to yield unreliable results.
When options are traded, one can use their prices and price changes to draw inference about the set of risk factors and their risk premia. We analyze tests for the existence and the sign of the market prices of jump risk that are based on option hedging errors. We derive a closed-form solution for the option hedging error and its expectation in a stochastic jump model under continuous trading and correct model specification. Jump risk is structurally different from, e.g., stochastic volatility: there is one market price of risk for each jump size (and not just \emph{the} market price of jump risk). Thus, the expected hedging error cannot identify the exact structure of the compensation for jump risk. Furthermore, we derive closed form solutions for the expected option hedging error under discrete trading and model mis-specification. Compared to the ideal case, the sign of the expected hedging error can change, so that empirical tests based on simplifying assumptions about trading frequency and the model may lead to incorrect conclusions.
We propose a long-run risk model with stochastic volatility, a time-varying mean reversion level of volatility, and jumps in the state variables. The special feature of our model is that the jump intensity is not affine in the conditional variance but driven by a separate process. We show that this separation of jump risk from volatility risk is needed to match the empirically weak link between the level and the slope of the implied volatility smile for S&P 500 options.
Tractable hedging - an implementation of robust hedging strategies : [This Version: March 30, 2004]
(2004)
This paper provides a theoretical and numerical analysis of robust hedging strategies in diffusion–type models including stochastic volatility models. A robust hedging strategy avoids any losses as long as the realised volatility stays within a given interval. We focus on the effects of restricting the set of admissible strategies to tractable strategies which are defined as the sum over Gaussian strategies. Although a trivial Gaussian hedge is either not robust or prohibitively expensive, this is not the case for the cheapest tractable robust hedge which consists of two Gaussian hedges for one long and one short position in convex claims which have to be chosen optimally.
Stocks are exposed to the risk of sudden downward jumps. Additionally, a crash in one stock (or index) can increase the risk of crashes in other stocks (or indices). Our paper explicitly takes this contagion risk into account and studies its impact on the portfolio decision of a CRRA investor both in complete and in incomplete market settings. We find that the investor significantly adjusts his portfolio when contagion is more likely to occur. Capturing the time dimension of contagion, i.e. the time span between jumps in two stocks or stock indices, is thus of first-order importance when analyzing portfolio decisions. Investors ignoring contagion completely or accounting for contagion while ignoring its time dimension suffer large and economically significant utility losses. These losses are larger in complete than in incomplete markets, and the investor might be better off if he does not trade derivatives. Furthermore, we emphasize that the risk of contagion has a crucial impact on investors' security demands, since it reduces their ability to diversify their portfolios.
This paper compares two classes of models that allow for additional channels of correlation between asset returns: regime switching models with jumps and models with contagious jumps. Both classes of models involve a hidden Markov chain that captures good and bad economic states. The distinctive feature of a model with contagious jumps is that large negative returns and unobservable transitions of the economy into a bad state can occur simultaneously. We show that in this framework the filtered loss intensities have dynamics similar to self-exciting processes. Besides, we study the impact of unobservable contagious jumps on optimal portfolio strategies and filtering.
This paper analyzes the equilibrium pricing implications of contagion risk in a Lucas-tree economy with recursive preferences and jumps. We introduce a new economic channel allowing for the possibility that endowment shocks simultaneously trigger a regime shift to a bad economic state. We document that these contagious jumps have far-reaching asset pricing implications. The risk premium for such shocks is superadditive, i.e. it is 2.5\% larger than the sum of the risk premia for pure endowment shocks and regime switches. Moreover, contagion risk reduces the risk-free rate by around 0.5\%. We also derive semiclosed-form solutions for the wealth-consumption ratio and the price-dividend ratios in an economy with two Lucas trees and analyze cross-sectional effects of contagion risk qualitatively. We find that heterogeneity among the assets with respect to contagion risk can increase risk premia disproportionately. In particular, big assets with a large exposure to contagious shocks carry significantly higher risk premia.
This paper analyzes the equilibrium pricing implications of contagion risk in a two-tree Lucas economy with CRRA preferences. The dividends of both trees are subject to downward jumps. Some of these jumps are contagious and increase the risk of subsequent jumps in both trees for some time interval. We show that contagion risk leads to large price-dividend ratios for small assets, a joint movement of prices in the case of a regime change from the calm to the contagion state, significantly positive correlations between assets, and large positive betas for small assets. Whereas disparities between the assets with respect to their propensity to trigger contagion barely matter for pricing, the prices of robust assets that are hardly affected by contagion and excitable assets that are severely hit by contagion differ significantly. Both in absolute terms and relatively to the market, the price of a small safe haven increases if the economy reaches the contagion state. On the contrary, the price of a small, contagion-sensitive asset exhibits a pronounced downward jump.