{"id":"arxiv/2611.00578","url":"https://arxiv.org/abs/2611.00578","tags":["q-fin.PM","math.OC","math.PR","math.ST","q-fin.MF"],"score":0,"title":"Non-asymptotic analysis of SGLD with discontinuous stochastic gradients: a new route to portfolio CVaR minimization","domain":"q-fin.PM","source":"arxiv","authors":["Yan Ding","Sotirios Sabanis","Ying Zhang"],"reviews":0,"subject":"q-fin.PM","summary":"Conditional-Value-at-Risk (CVaR) optimization is a fundamental tool in portfolio selection, yet its nonsmooth loss structure poses a challenge for the existing theory of stochastic gradient methods including associated Langevin algorithms. In particular, the positive-part loss and the softmax portfolio parametrization lead to stochastic gradients that need not be globally Lipschitz continuous, but rather discontinuous, where available non-convex stochastic gradient Langevin dynamics (SGLD) guarantees often exhibit an unfavorable exponential dependence on the problem dimension. In this…","category":"research","comments":0,"disputed":false,"division":"econ","mentions":0,"discussed":0,"published":"2026-10-31","reactions":{}}