Contributions in Merlin

The first equity future was launced in 1982 on the S&P 500 index. The future market has grown enormously since then and in more liquid than the cash market in many countries. The analysis here proves that the future market is highly efficient in many perspectives. Transaction costs in futures are on average one third of those in the cash market. Arbitrage is hardly ever possible. As futures contain information about future dividends, they can be used to analyze the market estimation for dividends. Even this proves to be quite efficient, with an exception in 2008, where the market has either underestimated the subsequent dividends or increased risk aversion caused too low market values of future dividends. Withholding taxes cause a split between the future and cash market. For investors who cannot reclaim foreign withholding taxes, it is beneficiary for them to only invest in futures, as dividend earnings implied in the future price are not taxed.

This paper examines how liquidity affects market efficiency in a market environment where securities' fundamental values are revealed at a predetermined point in time. We employ differences in minimum tick sizes at the betting exchange Betfair which induce exogenous variation in liquidity. The results show that liquidity significantly decreases market efficiency for bets on weekend matches but not for bets on weekday matches. As uninformed noise bettors are more likely to bet on weekends than on weekdays, the type of liquidity seems to matter for market efficiency.

We generalize the Kou (2002) double exponential jump-diusion model in two directions. First, we independently displace the two tails of the jump size distribution away from the origin. Second, we allow for each of the displaced tails to follow a gamma distribution with an integer-valued shape parameter. Both extensions introduce additional exibility in the tails of the corresponding return distribution. Our model is supported by an equilibrium economy and we obtain closed-form solutions for European plain vanilla options. Our valuation function is computationally fast to evaluate and robust across the full parameter space. We estimate the physical model parameters through maximum likelihood and for a diverse sample of equities, commodities and exchange rates. For all assets under consideration, the original Kou (2002) model can be rejected in favor of our newly introduced asymmetrically displaced double gamma dynamics.

An auction is said to be externality-robust if unilateral deviations from equilibrium leave the remaining bidders’ payoffs unaffected. The equilibrium and its outcome will then persist even if externalities between bidders arise from certain types of interdependent preferences. One important example are externalities due to spiteful preferences, which are frequently used to explain overbidding in auctions. Another important example are cross-shareholdings between firms that compete in an auction. For the independent private values model, we derive an auction that coincides with the second-price auction (SPA) in terms of efficiency and expected revenue, and, in contrast to the second-price auction, is also externality-robust. The externality-robust auction (ERA) is a first-price auction in which truthful bidding is encouraged by bonus payments. We test the robustness property experimentally by comparing the outcomes of the SPA and the ERA. We replicate the earlier finding of significant average overbidding in the SPA, while we find no overbidding on average in the ERA. We also conduct additional treatments where bidders play against the computer, and we use controls for cognitive skills and joy of winning to further pin down the reasons behind the subjects’ bidding behavior and to understand the sources of individual heterogeneity.

We introduce an adaptive importance sampling method for the loss distribution of credit portfolios based on the Robbins-Monro stochastic approximation procedure. After presenting the subtle construction of the algorithm, we apply our adaptive scheme for calculating the risk figures of a typical medium-sized credit risk portfolio with 2000 obligors. Simulating the tail of the loss distribution, we can improve significantly the variance reduction and outperform other recently proposed importance sampling approaches that are based on deterministic methods providing asymptotically optimal importance sampling distributions. Furthermore, the simple structure of the algorithm not only allows a straightforward implementation, but also offers a lot of flexibility for extensions to more complex models. Therefore, our numerical results motivate interesting future research paths for the application of stochastic approximation methods in risk management.

Belief that a single number can capture the degree of risk being taken within a bank or an investment is mistaken - especially when that number is value at risk. Markus Leippold explains why the measure is flawed, points to the dangers of its widespread acceptance by regulators and investors, and suggests an alternative.

In this paper, we develop a method to efficiently estimate dynamic discrete choice models with AR(n) type serial correlation of the errors. First, to approximate the expected value function of the underlying dynamic problem, we use Gaussian quadrature, interpolation over an adaptively refined grid, and solve a potentially large non-linear system of equations. Second, to evaluate the likelihood function, we decompose the integral over the unobserved state variables in the likelihood function into a series of lower dimensional integrals, and successively approximate them using Gaussian quadrature rules. Finally, we solve the maximum likelihood problem using a nested fixed point algorithm. We then apply this method to obtain point estimates of the parameters of the bus engine replacement model of Rust [Econometrica, 55 (5): 999–1033, (1987)]: First, we verify the algorithm's ability to recover the parameters of an artificial data set, and second, we estimate the model using the original data, finding significant serial correlation for some subsamples.