This paper studies feedback effects and endogenous risk in financial markets. In order to model those effects in a non parsimonious manner, we propose a general framework of a financial market with multiple assets which takes into account feedback effects from systematic trading by large financial institutions and which is flexible enough to incorporate the impact of any type of trading strategy that can be source of feedback. The model yields tractable formulas linking realized volatilities and correlations to the strategies followed by large financial institutions and the asset liquidities and shows that, in the presence of feedback effects, asset dynamics may deviate significantly from fundamentals and be driven more by the market capitalizations and strategies of large financial institutions. We quantify the price-mediated contagion to other investors generated by feedback effects and give a decomposition of endogenous risk between a volatility component and a correlation component. The results developed in this paper are useful in a risk-management perspective as they provide a flexible framework to better tackle and anticipate liquidity events caused by large trades and also in a systemic risk-management perspective as they enable to quantify price-mediated contagion.
Renaud BEAUPAIN, Stephanie HECK
In this paper we use a repeat-sales index methodology to construct US corporate bond price indices. Using several performance tests, we show that this methodology provides superior index estimates. In particular when assets trade at infrequent and irregular intervals the repeat-sales index is superior to taking an arithmetic price average. The methodology can readily be applied to any sub-sample of bonds based on a particular characteristic, such as the rating or the maturity. We further study the sensitivity of individual bond returns to systematic market risk as measured by a repeat-sales price index. Results indicate that variations in the price index are an important determinant of the time series and of the cross-sectional variation of corporate bond returns.
In recent years firms have been shifting their executive compensation packages from plain stock and option grants to grants with accounting-based performance vesting provisions and awards that benchmark firm performance against those of a designated peer group. One potential explanation for this trend is that firms want to improve the signal-noise ratio of performance measures. This paper directly tests this hypothesis. It estimates the statistical relationship between various performance measures and computes the optimal combination of awards. It finds that the observed trend is consistent with compensation optimization. It also finds that firms with higher potential gains from switching to alternative performance characteristics have increased their usage more than other firms.