Spotting a Financial Crisis Before It Happens
In the fall of 2008, as big-name financial institutions toppled, policymakers focused their efforts on saving the global economy from collapse. Now that the recovery is well under way, the nation is moving closer to establishing a new regime for monitoring systemic risk to make sure we don’t repeat past mistakes. It is high time, then, to discuss how we intend to do that. What could we have done differently to spot—and then stop—the impending financial crisis of 2008?
Researchers are making headway in answering that question. Last fall, the Federal Reserve Bank of Cleveland and the National Bureau of Economic Research sponsored a research conference on Quantifying Systemic Risk. Viral Acharya, a research associate with the Reserve Bank, presented a paper at the conference that helped inform the New York University Stern School of Business’s recently unveiled Systemic Risk Rankings service.
Cleveland Fed researchers are also studying a number of approaches to the systemic risk problem. Here are two of them. As always, we’d like to hear what you think. Contact us at email@example.com.
Let’s begin with the well-supported premise that financial crises happen when shocks to the financial system meet structural weaknesses within that system. If we have a strong shock but an equally strong financial system, the danger of a crisis is low. But if the system is fragile, even a moderate shock can wreak havoc.
To prevent a financial crisis, regulators must see both the big picture (the financial system as a whole) and the little picture (individual institutions). Only recently, however, have researchers looked at ways to combine “macroprudential” supervision of the entire financial system with “microprudential” supervision of individual institutions. By monitoring and analyzing both types of information, researchers can identify signs that potential shocks are building and compare them to potential structural weaknesses in the market.
Toward that end, Federal Reserve Bank of Cleveland researchers have recently developed a systemic-risk identification model called SAFE, for Systemic Assessment of the Financial Environment. SAFE can identify early signs of emerging shocks and structural weaknesses—a highly useful feature that enables policymakers to prevent those conditions from becoming reality. (If policymakers had only a few days’ notice of a financial system collapse, it would be far more difficult to develop an effective response.) This model’s key innovation is its use of confiÂdential supervisory information, gleaned from regular bank examinations, and data from supervisory tools to identify weaknesses in the institutions that make up the financial infrastructure.
Identifying the Shocks
Identifying a financial shock before it happens is difficult at best. Cleveland Reserve Bank researchers have approached this problem by thinking of a shock as a sudden change in investors’ expectations. In the SAFE model, these expectations are based on three factors:
- Return: how much an investor may expect to make on a particular asset
- Risk: the chance that an asset may lose some or all of its value
- Liquidity: the ease with which an investor may sell or trade an asset
The model’s central assumption is that investors are constantly making judgments about the return, risk, and liquidity of the assets they hold—the measures that determine the price of the assets. These measures are continuously compared to the historical norms for their assets. History shows that when significant, sustained gaps emerge between current measures and their norms, the likelihood of shocks increases.
The health of the financial market’s infrastructure strongly determines the potential for systemic risk. It directly affects financial firms’ ability to absorb shocks, which originate in gaps in investor expectations. To gauge the financial market’s condition, the SAFE model uses information on the nation’s largest financial institutions to assess three aspects of systemic structural fragility: connectivity, concentration, and contagion.
Connectivity indicators measure the volatility of each financial institution’s balance sheet compared to the volatility of the wider financial system. When the balance sheets of several large institutions move in concert with the entire system, institutions and the system are considered highly correlated. In this case, an emerging financial market shock will likely ripple through the country’s largest financial institutions as well as its financial markets.
Concentration indicators measure the intensity of asset holdings and market making—the ability to dictate prices—within the financial system. In general, the more concentrated the financial system’s asset holdings and the more narrow its market making, the more fragile the system. More specifically, when an institution or a small subset of institutions holds a large share of a market’s assets, its trades increasingly “make” the market, that is, move prices. Thus, if an asset price shock occurs, and these institutions sell concentrated assets, their disproportionately large holdings overwhelm buy orders, so that the market cannot function or does so only at very low prices. Likewise, if a single bank or a small group of institutions serves as the sole market maker, its failure would eliminate a liquid market for those assets.
Contagion indicators measure the relative ability of individual financial institutions to withstand a financial shock and remain solvent. If individual institutions can “internalize” the effects of a shock, it will not spill over into the larger financial system. On the other hand, if individual institutions cannot absorb the shock and remain solvent, the losses they sustain will probably affect other institutions’ health and spill over into the larger financial system.
The Cleveland Approach
To derive these three indicators, Bank researchers are using confidential supervisory information, including details about loans and liabilities that aren’t publicly available. Researchers are also tapping outputs from proprietary supervisory tools that are accessible to the Federal Reserve in its role of banking supervisor. It is this unique feature—the incorporation of supervisory information—that distinguishes SAFE from other models developed to identify systemic risk. Just as a weather forecaster uses radar tools to predict a coming storm, the SAFE model is being designed to help spot episodes of financial stress so as to head off a full-blown crisis. With ongoing test results, researchers will fine-tune the model with the overarching goal of making it more accurate.
Of course, policy actions don’t exist in a vacuum, and it would be useful to know how they might affect the financial climate. The short-lag variant of the SAFE model incorporates policy actions’ effects on emerging conditions to see if they are working as intended or if different policy actions are required. Taken together, the long- and short-lag versions of the SAFE model identify the advent of systemic risk and provide valuable feedback on policy actions that address those risks.
To validate the model’s effectiveness, researchers are building a financial stress index to chart previous episodes of stress in the U.S. financial system. Think of the index as a thermometer that tells regulators how hot or cold stress in the economy is running.
The work continues. Bank researchers are circulating the SAFE model among economists and bank supervision professionals in the U.S. and abroad for comment.