Why the Smarter Rule Can Lose
This series examines a specific question raised by the Federal Reserve paper A Static Capital Buffer is Hard To Beat: when should a bank regulator vary capital requirements, and when is a simple fixed buffer safer? The paper studies a dynamic economy in which banks choose between safe and risky lending while limited liability and government deposit insurance protect them from much of the downside. The practical value of the analysis is not a universal instruction to freeze capital requirements. It is a way to see why the right response depends on the shock changing banks' incentives—and why a policy that looks more sophisticated because it follows the credit cycle can still miss that mechanism.
That distinction matters because “cyclical” sounds like common sense. If rapid credit growth often precedes trouble, raising capital during a boom seems prudent. If a downturn threatens to make capital requirements constrain lending, lowering them seems supportive. The Basel III countercyclical capital buffer is built around that intuition. Yet a credit boom, a recession, or a quiet financial-market shift can each arise from different forces. They need not alter the relative profitability of safe and risky loans in the same way.
The Federal Reserve paper turns this tension into a model-based comparison. It asks what happens when a regulator can set capital exactly high enough to prevent a bank from crossing into excessive risk-taking, but cannot realistically observe all the shocks that determine that threshold. It then compares that benchmark with simple rules based on aggregate indicators and with a modest static buffer. The answer is surprising only if capital regulation is treated as a smoother for the business cycle rather than as a guardrail around a bank's payoff.
A policy question hiding inside a model
The subject of the paper is a dynamic macroeconomic model of banks, not the Basel framework by itself. Households supply labor, hold deposits, and own bank equity. Firms borrow to acquire capital. Some firms are safe: they face aggregate economic shocks. Others are risky: they face the same aggregate environment plus an idiosyncratic shock with a negative average effect and positive variance. Banks choose how much of their loan portfolio to direct toward each type.
Under ordinary social accounting, the risky project is unattractive on average. Its additional shock lowers the expected return on a loan compared with a safe project. But the bank does not experience the two outcomes symmetrically. If a risky loan pays off, the bank's equity holders receive the upside. If the bank's net worth becomes negative, limited liability means those equity holders receive nothing, rather than having to cover the remaining loss. Depositors are protected by government insurance. The bank can therefore find risky exposure privately attractive even when the project is socially inefficient.
That is the first important shift in perspective. The problem is not simply that banks lend too much during a credit boom. It is that the composition of lending can change when the expected return on safer assets falls, when risky returns become more volatile, or when capital requirements leave too little of the bank's own equity at stake. A regulator watching only the volume of credit sees an indirect symptom. The bank is responding to a relative-payoff comparison that may be hidden inside the same aggregate numbers.
The paper evaluates policies by their consequences in this environment. A rule can prevent failures, but it may also displace deposits that households value for their liquidity services. A rule can preserve more deposits, but if it allows the bank to cross the risky threshold, the resulting failures reduce output and income and require insurance-financed taxes. “Better” therefore means more than a higher capital ratio, fewer fluctuations in credit, or a smoother time series. It means balancing protection against incentive-driven failure with the continuing value of bank deposits.
The qualification is important. The paper is a Finance and Economics Discussion Series working paper, and the Federal Reserve landing page says that its analysis and conclusions are preliminary and represent the authors' views. Its conclusion is a disciplined result within a stylized model, not a claim that every jurisdiction should abandon a countercyclical buffer or that every static rule dominates every dynamic rule.
Why the countercyclical idea is persuasive
The Basel III countercyclical capital buffer provides a serious institutional benchmark for the paper's comparison. The Basel Committee describes it as a macroprudential tool that makes banking-sector capital requirements take account of the macro-financial environment. Its primary purpose is to build a cushion during periods of excess aggregate credit growth associated with system-wide risk. In a downturn, releasing the buffer is intended to reduce the chance that regulatory capital constraints will further restrict credit and damage the real economy.
The framework is more concrete than a slogan. For an internationally active bank, the buffer is a weighted average of the buffers in the jurisdictions where the bank has credit exposure. It is implemented as an extension of the capital conservation buffer, uses Common Equity Tier 1 capital, and constrains capital distributions when breached rather than automatically shutting down ordinary banking operations. Increases can be announced in advance—by up to 12 months under the Basel Committee's guidance—while decreases take effect immediately. These design features reflect a real administrative problem: banks need time to adjust upward, but a release is most useful when stress is already present.
The paper does not deny the rationale. Its calibrated model reproduces a mild empirical relationship in U.S. data: a credit-to-GDP ratio above trend is associated with somewhat lower real GDP two years later. That relationship helps explain why the indicator became attractive. But prediction is not the same as identification. A measure can say something about what may happen later without telling a regulator which underlying shock is responsible or how capital should respond today.
This is where the paper's core theme enters. The relevant question is not “Is credit high?” but “Why has the bank's relative incentive changed, and can the regulator observe that reason?” A persistent productivity shock may lower the expected return on safe capital during a downturn. An investment-specific technology shock may increase the future supply of capital during an expansion and also lower the safe return. A volatility shock may make risky lending more appealing while barely moving GDP. The same credit indicator cannot be assumed to encode all three mechanisms.
The diagram captures why the choice among rules cannot be made by looking at one arrow in isolation. Shocks change the relative attractiveness of projects. Capital changes the bank's exposure and its funding mix. Risk-taking can end in failures, while tighter capital can reduce deposit liquidity even when the economy remains in a safe state. The policy instrument sits in the middle of a system with competing consequences; it is not merely a dial marked “more” or “less” credit.
Three tests for a capital rule
The first test is safety. Does the rule keep the bank from taking the protected downside of risky lending, including when the shock is not a conventional recession? In the paper's terminology, the strongest benchmark is a “no-failure” rule: set capital just high enough, given the current shocks, to prevent the switch into the risky regime. This benchmark is useful because it identifies the minimum capital needed for the safety objective. It is not automatically usable, because calculating it requires information about the full state of the economy.
The second test is liquidity. Deposits are not treated as interchangeable with equity in the model. Households value deposits as a liquid, safe asset, and that value lowers the deposit rate relative to the return households require from bank equity. Banks consequently prefer deposit funding, subject to the capital requirement. Raising the requirement forces more equity into the balance sheet and reduces the deposit supply. A rule that avoids every failure by demanding capital far above the safe threshold may impose a continuing cost even in periods when the risk it fears never materializes.
The third test is information. A policy can be theoretically precise and practically fragile if it needs variables that a regulator cannot observe in time or cannot distinguish reliably. The paper's most successful dynamic rule responds to the complete constellation of shocks and states. Its performance tells us what an ideal policy would need to know. The simpler rules are then judged not by whether they imitate that ideal in every period, but by whether their limited information still produces acceptable welfare and failure outcomes.
Those tests explain the title's challenge. A static buffer does not claim to know whether a change in credit reflects productivity, investment technology, or financial volatility. It places a modest amount of extra equity between the banking system and the risky threshold. In the paper's experiments, that restraint performs nearly as well as the information-rich benchmark, while optimized cyclical rules largely turn themselves into static buffers. The result is not that variation is always bad. It is that unsupported precision can be worse than a deliberately limited rule when the signal does not identify the mechanism.
The rest of this series follows that logic from the bank's funding deal to the regime switch, then through the paper's three shock experiments and its quantitative comparison with the Basel-style rule. The central judgment will remain conditional: simplicity is attractive here because it is robust to missing information, not because capital regulation has no cycles, no trade-offs, or no need for better evidence.