Working Paper: CEPR ID: DP7188
Authors: Jun Liu; Allan G. Timmermann
Abstract: We define risky arbitrages as self-financing trading strategies that have a strictly positive market price but a zero expected cumulative payoff. A continuous time cointegrated system is used to model risky arbitrages as arising from a mean-reverting mispricing component. We derive the optimal trading strategy in closed-form and show that the standard textbook arbitrage strategy is not optimal. In a calibration exercise, we show that the optimal strategy makes a sizeable difference in economic terms.
Keywords: Cointegrated Asset Prices; Optimal Portfolio Choice; Risky Arbitrage
JEL Codes: G11
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| mispricing (D49) | traditional arbitrage strategies are not optimal (G19) |
| mispricing (D49) | optimal portfolio weights (G11) |
| optimal portfolio weights (G11) | utility losses associated with suboptimal strategies (L97) |
| degree of mispricing (G19) | optimal strategy (L21) |
| correlation between asset prices (G19) | optimal strategy (L21) |
| asymmetric mispricing (G19) | significant economic benefits compared to traditional approaches (O22) |