Why does the market discipline that financial intermediaries face seem too weak during booms and too strong during crises? This paper shows in a general equilibrium setting that rollover risk as a disciplining device is effective only if all intermediaries face purely idiosyncratic risk. However, if assets are correlated, a two-sided inefficiency arises: Good aggregate states have intermediaries taking excessive risks, while bad aggregate states suffer from costly fire sales. The driving force behind this inefficiency is an amplifying feedback loop between asset values and market discipline. In equilibrium, financial intermediaries inefficiently amplify both positive and negative aggregate shocks.

We build a general equilibrium model with financial frictions that impede monetary policy transmission. Agents with heterogeneous productivity can increase investment by levering up, which increases liquidity risk due to maturity transformation. In equilibrium, more productive agents choose higher leverage than less productive agents, which exposes the more productive agents to greater liquidity risk and makes their investment less responsive to interest rate changes. When monetary policy reduces interest rates, aggregate investment quality deteriorates, which blunts the monetary stimulus and decreases asset liquidation values. This, in turn, reduces loan demand, decreasing the interest rate further and generating a negative spiral. Overall, the allocation of credit is distorted and monetary stimulus can become ineffective even with significant interest rate drops.

We identify and track over time the factors that make the financial system vulnerable to fire sales by constructing an index of aggregate vulnerability. The index starts increasing quickly in 2004, before most other major systemic risk measures, and triples by 2008. The fire‐sale‐specific factors of delevering speed and concentration of illiquid assets account for the majority of this increase. Individual banks' contributions to aggregate vulnerability predict other firm‐specific measures of systemic risk, including SRISK and CoVaR. The balance‐sheet‐based measures we propose are therefore useful early indicators of when and where vulnerabilities are building up.

Inspired by experimental evidence, we amend the recursive utility model to let risk aversion decrease with the temporal horizon. Our pseudo-recursive preferences remain tractable and retain appealing features of the long-run risk framework, notably its success at explaining asset pricing moments. In addition, our model addresses two challenges to the standard model. Calibrating the agents’ preferences to explain the equity premium no longer implies an extreme preference for early resolutions of uncertainty. Horizon-dependent risk aversion helps resolve key puzzles in finance on the valuation of assets across maturities and captures the term structure of equity risk premiums and its dynamics.

We model how a cyber attack may be amplified through the U.S. financial system, focusing on the wholesale payments network. We estimate that spillovers of an attack on one of the five most active banks would impair 31% of the network, on average, with some days significantly worse. When other banks respond by liquidity hoarding, forgone payment activity can reach more than 2.5 times daily GDP. A reverse stress test shows attacks on groups of small banks can also impair a significant amount of the network. Interconnectedness through third-party providers and financial market utilities poses additional risks.