<p>Diversification potential refers to the ability of an investment to enhance the diversification of a portfolio. Based on Euler allocation, we introduce a diversification potential index termed the marginal diversification quotient (MDQ), and a diversification index termed the marginal diversification quotient mean (MDQM). The two indices satisfy non-negativity, location invariance, scale invariance, riskless invariance and replication consistency for common risk measures. We derive the asymptotic distributions of the proposed nonparametric estimators for MDQ and MDQM based on expected shortfall (ES) for dependent data under a mixing sequence. To assess the performance of the estimators, we model each investment within a portfolio using an AR-GARCH model. Furthermore, we employ a residual-based bootstrap method to quantify the estimation uncertainty. The proposed indices are applied to real data sets and compared with existing indices. A simulation study is conducted to evaluate the performance of the proposed nonparametric estimators in finite samples.</p>