“…The coefficient on Gender i,t is positive and highly statistically significant in every regression specification. As more leverage increases portfolio risk, this result is consistent with prior studies (e.g.,Barber & Odean, 2001;Davydov et al, 2017), showingTA B L E 3 Determinants of leverage usageThe table presents the estimates of the analysis of the determinants of leverage product use and trading margin account. The logistic model of the analysis is as follows: Leverage i,t = α i + Demographic i,t + Activity i,t + ε i,t , where Leverage i,t is one of the three dummy variables for the use of leverage: Margin i,t is a dummy variable for trading on margin account by investor i during a year, EmbedLev i,t is a dummy variable for the use of products with embedded leverage by investor i during a year, or AnyLev i,t , a dual variable indicating both trading on margin and/or use of products with embedded leverage by investor i during a year.…”