The value of the ‘swap’ feature in equity default swaps

When equity default swap (EDS) contracts were first included in a rated collateralized debt obligation (CDO) deal, some critics doubted the originality of the product. In fact, EDSs are equivalent to already existing binary barrier options on equity, except the premium is not paid upfront, but over time, and conditional on the trigger event not having occurred. Therefore, as opposed to existing options, the buyer of an EDS: (1) postpones payment for protection, and (2) purchases not only protection against a sharp drop in the price of equity, but also the right to cease payments in case the barrier is hit. This paper derives the closed-form pricing formula for equity default swap spreads under the Black–Scholes assumptions, and then quantifies the fraction of the EDS spread actually due to the ‘swap’ feature of the contract for plausible parameter values. It is found that the extra spread due to the swap nature of EDSs is economically significant only for high volatility, high trigger levels, and long time-to-maturity. The impact of interest rates on the value of the ‘swap’ feature is almost exclusively due to the postponement of payments.     

Market timing: A global endeavor

This paper employs daily fund and index data, the classical Treynor and Mazuy timing model, and two multi-factor extensions to measure the market timing ability of global asset allocation funds. These funds differ from traditional global or international funds in that they face fewer investment constraints and are known to actively shift funds across a wide variety of asset classes. When using the classical Treynor and Mazuy timing models, I find evidence of poor market timing ability. However, this evidence disappears when timing ability is examined using two multi-factor models. The results from Treynor and Mazuy are spurious since both multi-factor extensions do a much better job in explaining the variation in average fund returns.     

Flexible distribution functions,higher-order preferences and optimal portfolio allocation

In this paper we show that flexible probability distribution functions, in addition to being able to capture stylized facts of financial returns, can be used to identify pure higher-order effects of investors' optimizing behavior. We employ the five-parameter weighted generalized beta of the second kind distribution—and other density functions nested within it—to determine the conditions under which risk averse, prudent and temperate agents are diversifiers in the standard portfolio choice theory. Within this framework, we illustrate through comparative statics the economic significance of higher-order moments in return distributions.     

The cost of equity of internet stocks: a downside risk approach

Beta as a measure of risk has been under fire for many years. Although practitioners still widely use the CAPM to estimate the cost of equity of companies, they are aware of its problems and are looking for alternatives. A possible alternative is to estimate the cost of equity based on the semideviation, a well-known and intuitively plausible measure of downside risk. Complementing evidence reported elsewhere about the ability of the semideviation to explain the cross-section of returns in emerging markets and that of industries in emerging markets, this article reports results showing that the semideviation also explains the cross-section of Internet stock returns.     

Beyond Single-Factor Affine Term Structure Models

This article proposes a new approach to testing for the hypothesisof a single priced risk factor driving the term structure ofinterest rates. The method does not rely on any parametric specificationof the state variable dynamics or the market price of risk.It simply exploits the constraint imposed by the no-arbitragecondition on instantaneous expected bond returns. In order toachieve our goal, we develop a Kolmogorov-Smirnov test and applyit to data on Treasury bills and bonds for both the United Statesand Spain. We find that the single risk factor hypothesis cannotbe rejected for either dataset.