In this paper, we provide a thorough study on the relevance of liquidity-adjusted value-at-risk (LVaR) and expected shortfall (LES) forecasts. We measure additional liquidity of an asset via the difference between its respective bid and ask prices and we assess the non-normality of bid-ask spreads, especially in turbulent market times. The empirical assessment comprises German stocks in both calm and turmoil market times, and our results provide evidence that liquidity risk turns out to be crucial for the quality of regulatory risk assessment in turmoil market times. We find that a Cornish-Fisher approximation describes a sensible choice for LVaR forecasts whereas an extreme value approach results in adequate LES forecasts.