Working Paper: CEPR ID: DP6074
Authors: Qinglei Dai; Kristian Rydqvist
Abstract: We estimate the costly-arbitrage model of Boyd and Jagannathan (1994) using Norwegian stock market data. Taxable distributions take place at two separate dates, one that entails the distribution of an imputation-tax credit and another the distribution of the cash dividend. We find that the costly-arbitrage model is consistent with observed stock returns around the ex-dividend day, but the model cannot explain the return patterns around the distribution of the tax credit. We relate the difference in price formation to uncertainty.
Keywords: costly arbitrage model; estimation; risk; ex-dividend day; imputation tax credit; legal risk; withholding tax
JEL Codes: C78; D40; G10; H26
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| estimation risk (C13) | trading decisions (G11) |
| trading decisions (G11) | stock prices (G12) |
| legal risk (K13) | price formation (L11) |
| costly arbitrage model (G19) | observed stock returns around the ex-dividend day (G35) |
| tax credit distribution (H23) | stock prices (G12) |
| costly arbitrage model (G19) | return patterns linked to tax credit distribution (H23) |
| tax audit (H26) | changes in trading volume (G15) |
| tax audit (H26) | price behavior (D40) |