Office of Financial Research

The Office of Financial Research (OFR) is a data-gathering and research bureau within the US Department of the Treasury, created by the Dodd-Frank Act to feed information to the Financial Stability Oversight Council. Rather than write rules or punish firms, the OFR assembles granular data on interconnections between banks, asset managers, and counterparties, then publishes research on emerging vulnerabilities that might threaten the financial system as a whole.

Origins and mandate

The financial crisis of 2008 exposed a critical blind spot: no single institution had a complete picture of who owed what to whom, which major firms were interconnected, or where leverage was concentrated. Bank regulators kept silo’d views of their own charges; the Federal Reserve knew some things; the SEC and CFTC watched equities and derivatives respectively. But there was no bird’s-eye view of the system. When Lehman Brothers collapsed and AIG nearly followed, the interconnection shocks propagated in directions regulators hadn’t anticipated.

Dodd-Frank created the Financial Stability Oversight Council to coordinate across regulators and, crucially, gave it the OFR as a research and data arm. The OFR’s charge was to collect information on systemic risk, analyse it, and report back to the FSOC and Treasury Secretary. It was not a regulator itself—it has no enforcement power, cannot declare firms systemically important, cannot order firms to reduce leverage—but it was meant to be the system’s data hub.

Data collection and authority

The OFR has a subpoena-like power, called “information requests” or “data calls,” to compel banks, asset managers, insurance companies, and other major financial firms to submit data on their balance sheets, counterparty risk, and interconnectedness. Unlike a regulator’s supervisory exam, the OFR’s data calls are broad-brush: they ask the same questions of everyone, not targeted supervisory deep-dives.

In practice, the OFR has struggled to standardize these data requests. Each large bank has its own accounting system and terminology. The OFR wants to know, for example, how much exposure Bank A has to Bank B and other counterparties, broken down by asset class and credit rating. But one bank’s internal data may not map cleanly to another’s, especially if they use different derivatives pricing models or valuation methodologies. The OFR has pushed for a common data standard and has gradually made headway, but the process is slow.

The OFR also buys market data from exchanges, data vendors, and self-regulatory organizations—equity and bond prices, derivatives positions, CDS spreads, and so on. It maintains a Financial Research Database and publishes a good deal of this data in de-identified form for academic researchers. This is valuable public good: outside economists can use real market and system data to study financial fragility.

Research and systemic monitoring

Once data is collected, the OFR publishes analysis. The flagship product is the annual Financial Stability Report, a dense, readable document that flags emerging risks: concentration in repo markets, rapid growth in leveraged buyouts, surging commercial real estate valuations, or unexplained correlations between seemingly uncorrelated assets. The OFR also publishes working papers on specific topics—stress testing methodologies, duration risk in bond funds, the stability implications of passive index funds.

This research is meant to inform the FSOC’s decisions and alert regulators to problems before they metastasise. The OFR might conclude, for instance, that counterparty risk concentration in the derivatives market is rising unsustainably, and the FSOC (which includes the Fed, OCC, FDIC, and SEC) can then press banks to reduce exposures or raise capital buffers. The OFR itself doesn’t order anyone to do anything; it just shines a light.

Strengths and limitations

The OFR has genuine value. Its work on systemic interconnectedness, shadow banking growth, and emerging market volatility has influenced regulatory thinking. Its data infrastructure, though imperfect, is far better than the blind guessing of the pre-2008 era.

But it has clear limits. First, it is severely understaffed for the scope of the task. A bureau of ~100 people is expected to monitor a financial system worth many trillions of dollars, with thousands of complex firms and opaque markets. The OFR’s data lags by months—it cannot spot a crisis in real time, only after the fact. Second, the OFR has no enforcement tool and depends on the FSOC for action. The FSOC is a coordinating body of regulators with competing mandates and constituencies; getting agreement on a systemic risk response is glacially slow. Third, the OFR’s research, while solid, is not predictive. It can identify risks but cannot easily forecast which ones will blow up and when.

Finally, the OFR’s reliance on Congress for annual appropriations makes it politically vulnerable. In lean budget years, it has been starved of funding. The Trump administration considered shuttering it entirely, viewing it as a redundant bureaucracy.

Data gaps and the cryptocurrency question

The OFR’s historical focus has been on traditional banking and derivatives. As cryptocurrency and decentralised finance have grown, the OFR has struggled to collect comparable data. A crypto exchange or decentralised protocol may have no traditional regulatory submission and be registered in a jurisdiction with light oversight. The OFR has gradually added crypto-related data requests to its calls, but true real-time monitoring of systemic risk in crypto—which moves fast and is highly leveraged—is beyond its current capacity.

International and peer institutions

The OFR’s model of a unified data hub and systemic-risk researcher is paralleled internationally. The Bank for International Settlements plays a similar role for the global system, though with less enforcement power even than the OFR. The European Systemic Risk Board (ESRB) has a comparable mandate in the EU, with similar trade-offs between broad visibility and narrow enforcement reach.

See also