Knowledge that Transforms

We address the challenge of dynamic option hedging using deep reinforcement learning (DRL). Unlike traditional model-based approaches, DRL is purely data-driven and does not require explicit modeling of the underlying market dynamics. Leveraging a comprehensive high-frequency dataset of SPX options, we train, validate, and test a DRL agent in a realistic market environment that incorporates actual transaction costs. The study highlights three key contributions. First, we analyze the hedging of 0DTE (zero days-to-expiration) options, which now dominate market trading volume, and show that DRL outperforms Black–Scholes delta hedging at this horizon. Second, we evaluate robustness across regimes, finding that the DRL hedge remains effective in crises such as COVID-19, even when trained only on non-crisis periods. Third, we examine the determinants of the performance gap, the role of alternative reward specifications, and hedging behavior in the presence of price jumps.

We leverage differences in central bank independence and financial stability sentiment across countries to investigate the variability in banks’ stock market reactions to prudential policy announcements during the COVID-19 crisis. Our findings reveal that the relaxation of both macro- and micro-prudential policies leads to negative cumulative abnormal returns (CARs), the reaction being attenuated in countries where the central bank is more independent or communicates deteriorations in financial stability. The CARs around the announcement dates are 0.75 percentage points (pp) and 6.89 pp higher for macro- and micro-prudential policy announcements, respectively, in countries with greater central bank independence compared to those with lesser independence. The difference is approximately 3.73 pp and 5.65 pp between banks located in countries where the central bank communicates a negative sentiment about financial stability, compared to those where a positive sentiment is conveyed. The positive impact of higher degrees of central bank independence and deteriorations in financial stability sentiment on bank market valuation is enhanced for smaller banks, as well as for banks in countries with greater fiscal flexibility and a higher prevalence of privately owned banks. (181 words)

We study whether floods can affect financial stability through a credit risk channel. Our focus is onthe Netherlands, a country situated partly below sea level, where insurance policies exclude property damages caused by some types of floods. Using geocoded data for close to EUR 650 billion in real estate exposures, we consider possible implications of such floods for bank capital. For a set of 38 adverse scenarios, we estimate that flood-related property damages lead to capital declines that mostly range between 30 and 50 basis points. We highlight how starting-point loan-to-value ratios are one important driver of capital impacts. Our estimates focus on property damages as the main transmission channel and are also subject to a number of assumptions. If climate change continues, more frequent floods or flood-related macrofinancial disruptions may have stronger implications for financial stability than our estimates so far indicate.

This paper studies systemic-risk connectedness in the European insurance sector at three levels of granularity: across major segments of financial markets, across insurance subsectors, and across individual insurance companies. Using a common connectedness framework applied to returns, volatility, value-at-risk, and expected shortfall, we document that insurers are an important component of systemic-risk connectedness, especially during stress episodes. We also provide reduced-form evidence on economically relevant channels in the European institutional setting: aggregate insurer spillovers co-move with term spreads, sovereign spreads, and funding stress, and firm-level insurer-to-bank spillovers vary with sovereign risk and domestic sovereign-bond home bias in a way consistent with a balance-sheet channel. The analysis further reveals substantial heterogeneity across subsectors and identifies a stable core of systemically central insurers in firm-level networks.