Contributions in Merlin

While the comparative statics of asset demand have been studied extensively, surprisingly little work has been done on the behavior of equilibrium asset prices and returns in response to changes in the supplies of securities. This is despite considerable interest in the equity premium and interest rate puzzles. In this paper, we seek to fill this void for the classic case of a representative agent economy with a single risky asset and risk free asset in both one and two period settings. It would seem natural to suppose that in response to an increase in the supply of the risky asset, its price would fall and the gross equity risk premium would increase. We show that in standard settings where preferences are represented by frequently assumed forms of expected utility, one can obtain the opposite result. The necessary and su¢ cient condition for prices (gross equity premium) to increase (decrease) with supply is determined by the sign of the slope of the asset Engel curve. This observation allows us to derive (i) sufficient conditions directly in terms of the representative agent's risk aversion properties for general utility functions and (ii) necessary and su¢ cient conditions for the widely used HARA (hyperbolic absolute risk aversion) class.

This appendix extends the empirical results in Chesney, Crameri, and Mancini (2011). Informed trading activities on put and call options are analyzed for 19 companies in the banking and insurance sectors from January 1996 to September 2009. Our empirical findings suggest that certain events such as the takeovers of AIG and Fannie Mae/Freddie Mac, the collapse of Bear Stearns Corporation and public announcements of large losses/writedowns are preceded by informed trading activities in put and call options. The realized gains amount to several hundreds of millions of dollars. Several cases are discussed in detail.

The paper is the first one outside the high-frequency domain to use sentiment-signed news to directly compare news and no-news stock returns. This is done by estimating whether returns on positive, neutral and negative news days are significantly different from the average daily return for a large sample of US stocks over the period from January 2003 to August 2010. The general results show that positive news days indeed have above-average returns and negative news days returns are below average, while the neutral news days are economically barely distinguishable from the average. The market also proves to be fast and accurate at pricing new information, as there are no signs of drift shortly after news days. On the contrary, a directionally correct and statistically significant movement can be found on the day before the news day. The cross-sectional analysis reveals significant differences in the strength of market reactions between stocks ranked on size, book-to-market or news coverage. The general results however hold across all subsamples and are also not driven by earnings announcements or past stock returns. Moreover, the average news sensitivity is itself a priced source of risk. A portfolio of stocks with high sensitivity to news outperforms a portfolio of stocks with low sensitivity by a statistically and economically significant 0.84% per month. This news premium seems to primarily relate to the high impact of news in situations of general uncertainty.

Most time series models used in econometrics and empirical finance are estimated with maximum likelihood methods, in particular when interest centers on density and Value-at-Risk (VaR) prediction. The standard maximum likelihood principle implicitly places equal weight on each of the observations in the sample, but depending on the extent to which the model and the true data generating process deviate this can be improved upon. For example, in the context of modeling financial time series, weighting schemes which place relatively more weight on observations in the recent past result in improvement of out-of-sample density forecasts, compared to the default of equal weights. Also, if instead of accurate forecasting of the entire density, interest is restricted to just downside risk, placing more weight on the negative observations in the sample improves results further. In this paper, a third and quite general strategy of shifting more weight towards certain observations of the sample is proposed. Weights are derived from external variables that convey additional information about the true DGP, like trading volume, news arrivals or even investor sentiment. As such, those observations are down weighted that bear a high probability of being destructive outliers with no bene¯t of using them when fitting the model. Considerable improvements in forecast accuracy for a variety of data sets and different time series models can be realized.