Working Paper: NBER ID: w22146
Authors: Itzhak Bendavid; Justin Birru; Viktor Prokopenya
Abstract: We document evidence consistent with retail day traders in the Forex market attributing random success to their own skill and, as a consequence, increasing risk taking. Although past performance does not predict future success for these traders, traders increase trade sizes, trade size variability, and number of trades with gains, and less with losses. There is a large discontinuity in all of these trading variables around zero past week returns: e.g., traders increase their trade size dramatically following winning weeks, relative to losing weeks. The effects are stronger for novice traders, consistent with more intense “learning” in early trading periods.
Keywords: risk taking; forex trading; self-attribution; overconfidence; behavioral finance
JEL Codes: G02; G11; G32
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| past gains (J17) | increase in risk-taking (G41) |
| past gains (J17) | increase in perceived skill (D29) |
| past losses (G33) | minimal adjustment in risk-taking (D81) |
| past losses (G33) | minimal adjustment in perceived skill (D29) |
| small gains (D61) | significant increase in trading activity (G15) |
| small losses (G33) | no increase in trading activity (F49) |
| discontinuity at zero past returns (D52) | increase in trading behavior (F19) |