Conference Agenda

At year-end 2022, U.S. banks reported $620 billion in unrealized losses on their investment securities portfolios, as the Federal Reserve Board raised its target interest rate by 400 basis points to combat inflation. These losses are strikingly similar in character to the losses on residential mortgages experienced by savings & loan (S&L) institutions in the early 1980s when the Federal Reserve Board raised interest rates to combat inflation—despite subsequent regulatory reforms that were ostensibly put into place to prevent such crises. In this study, we analyze the role of interest rate risk in losses in bank securities investments. We show that banks used RMBS to “reach for yield” during 2020–2022, and we show the losses that ensued. We find that the equities market for publicly traded bank holding companies failed to price this risk. We use publicly available data to show these results, raising the question of why neither bank shareholders nor regulators responded to the interest rate threat that was “hiding in plain sight.”

This study explores how emissions are financed by financial institutions could represent an indirect source of risk. Transition risk, albeit not being directly borne by banks, may indirectly affect financial intermediaries via their financing activities. As firms are under pressure to comply with transition, their relative credit risk increases thus exposing banks and, ultimately, the credit system. By using a sample of 8753 syndicated loans from DealScan, this study builds an empirical network model to relate lenders and borrowers. Such a structure allows us to measure the amount of lenders' financed emissions while considering also their centrality in the lending market network. The results show that the most central financial intermediaries are positively associated with higher shares of emissions with a joint probability around 60\%. Several robustness checks as well as different ways to account for emissions are considered, confirming the main results.

Climate change is increasingly turning water from a local environmental issue into a broader source of economic and financial risk. In this context, this paper investigates whether the mechanisms of shock transmission between commodity markets and the water sector differ across market regimes, especially during periods of energy and climate stress. The analysis spans from November 16, 2001, to December 31, 2025, and examines the interconnectedness among three water-sector equity indexes: the S&P Global Water Index, the Dow Jones U.S. Water Index, and the MSCI Europe Water Utilities Index, and selected commodity market segments, including Brent crude oil, industrial metals, grains, softs, and livestock. These variables represent three possible channels through which commodities may influence the water sector: an energy-cost channel, a climate-agricultural channel, and an industrial-input and infrastructure channel. To address nonlinearities and asymmetries in shock transmission, the study uses a quantile connectedness approach.

The results show that the water–commodity transmission network is strongly state-dependent. Total connectedness is markedly higher in the tails of the distribution than around the median, indicating that spillovers intensify under extreme market conditions. Dynamic estimates and structural break tests further reveal substantial regime shifts over time. The findings do not support a uniformly dominant role for crude oil. Instead, agricultural and soft commodity markets exhibit more asymmetric spillovers, especially in downside states, consistent with a climate-related transmission channel.

This paper develops an early warning system for predicting distress

for large European banks. Using a novel definition of distress derived

from banks’ headroom above regulatory requirements, we

investigate the performance of three machine learning techniques

against the traditional logistic model. We find that the random forest

model shows superior performance both out-of-sample and out-oftime. Unlike previous studies, we also employ a series of sampling

techniques showing that they significantly improve the ability to

identify distress events irrespective of the model used. Moreover, we

show that ensemble techniques can help improve performance

relative to the single best performing model. Finally, using the latest

machine learning interpretability tools, we show that the variables

closely tied to bank profitability and solvency are important drivers

for predicting bank distress. Overall, our paper has important

practical implications for bank supervisors and macroprudential

authorities who can utilise our findings to identify bank weaknesses

ahead of time and adopt pre-emptive measures to safeguard

financial stability.