A23: Characterizing Asset Pricing Anomalies

Intermediary asset pricing theory has hypothesized that capital constraints on part of financial intermediaries are important ingredients of the cross-sectional variation in risk premia. Corresponding empirical analyses employ stock portfolios as test assets, for example sorted by size, book-to-market ratio, or momentum. While these assets guarantee a pronounced cross-sectional variation in risk premia and are thus sufficient to diagnose the importance of the intermediary channel, the question of whether and which asset pricing phenomena are due to intermediary frictions remains largely unanswered. The goal of our project is to address this question by looking at assets that are traded at two different markets, the stock market and the option market. Some recent papers suggest that stock and option markets are affected by intermediary and market frictions to different degrees. Stock prices are driven by time-variation in the proximity to binding equity capital constraints and also funding liquidity constraints on part of banks and other intermediaries. Moreover, taxes, liquidity constraints, and short sale constraints are also known to have an impact on stock prices. For option prices, margin, liquidity, and short sale constraints are shown to matter. All those channels have in common that they drive prices away from their frictionless counterparts. Market participants are most likely aware of the wedge between the observed asset price and the asset’s true value, but frictions prevent them from trading against it. Thus, frictions can be thought of as limits of arbitrage.The key idea of our project is to investigate if certain asset pricing anomalies are cross-sectionally related to violations of no-arbitrage conditions between the stock and the option market. More precisely, we construct option portfolios that exactly replicate the cash-flows obtained from holding certain stocks. The put-call parity theoretically implies that the prices of the stocks and the replicating option portfolios must be identical. Violation of this parity indicates that either the stock or the options are mispriced, due to capital market frictions. To connect frictions with asset pricing puzzles, we sort stocks into portfolios, according to sorting criteria which have been documented to produce “anomalous” return spreads. Comparing the portfolio returns with the corresponding option portfolio returns allows us to investigate if an asset pricing anomaly can be explained by a friction-based rationale.In a second step, we link the time series and cross-sectional properties of no-arbitrage violations to measures of the different intermediary and capital market frictions and compare the empirical patterns with the implications of an intermediary asset pricing model. This allows us to gain insights into the origins of asset pricing anomalies and provide the missing link between the intermediary asset pricing literature and the literature on asset pricing anomalies.

Project Leader: JProf. Dr. Julian Thimme and Prof. Dr. Marliese Uhrig-Homburg