Dodd-Frank and Crapo, Meet SVB đ€
Regulators werenât looking because in 2018, we said âdonât look.â
The dust has settled. Startups met payroll. And SVB is channeling its way through Ch. 11 processes. Okay, theyâre stumbling. But just as the financial panic subsides, things in the regulatory world get interesting. Weâre starting to move beyond empty rhetoric to concrete proposals for reform. Sen. Elizabeth Warren (D-Mass.) suggests lifting the FDIC insurance cap from $250,000 to $1M per depositor. Robert Hockett, a law professor at Cornell, advocates for universal deposit insurance. Senate Banking Chair Sherrod Brown announced several hearings on the collapse of SVB and Signature Bank, the first of which happened on Tuesday.
In a prior post, I ask why regulators didnât catch this in time, and suggest that a passable answer is that they simply werenât looking. However, a more complete answer is that we told them to stop looking. To fully appreciate the irony here, we need to understand the historical wave of regulation (Dodd-Frank) and the subsequent wave of deregulation (Crapo Bill).
Meet Dodd-Frank.
The Dodd-Frank Wall Street Reforms and Consumer Protection Act was enacted in 2010 to promote financial stability in the aftermath of 2008. Calling it an act (singular) is somewhat misleading. It was an 848-page package of reforms that touched on nearly every aspect of the US financial system. Fun fact: though Dodd Frank is massive, itâs not the longest law on the books. At 5,593 pages, the Consolidated Appropriations Act of 2021 takes that title! (thatâs the one that authorized COVID stimmy checks)
Most relevant to todayâs happenings: Dodd-Frank established a framework for (1) identifying and (2) regulating âsystemically important financial institutionsâ (SIFIs).
Step 1: Identification
The first step - identification - is crucial. Dodd-Frank established the Financial Stability Oversight Council (FSOC) and vested it with the authority to label certain banks as SIFIs. Section 165 defined SIFIs as all bank holding companies with at least $50 billion in consolidated assets. Of which there were 38, including the usual suspects (e.g., JPMorgan, Bank of America, Wells Fargo, Goldman Sachs, and Morgan Stanley) and some smaller ones (e.g., Fifth Third Bank, American Express).
Step 2: Regulation
Once identified, SIFIs faced âenhanced prudential supervision.â Basically, more regulation. They were - to use the cliche - too big to fail. Regulators imposed extra burdens on them including routine stress tests, âliving willsâ, and enhanced capital and liquidity rules. Letâs dive in to each.
A) Stress tests: the Fed guesses the future
SIFIs are required to conduct an annual stress test called a Comprehensive Capital Analysis & Review (CCAR). CCAR tests whether a bank is sufficiently capitalized to weather a crisis. Itâs speculative but instructive. The Fed uses a variety of factors such as unemployment rates, GDP growth, and interest rates, to construct 3 hypothetical scenarios: baseline, adverse, and severely adverse. The baseline represents the Fedâs projection of what the economy will look like in the upcoming year, assuming no major shocks or disruptions, while the adverse and severely adverse scenarios simulate a severe economic downturn or financial crisis. SIFIs submit their data, and the Fed models how they will likely perform under the scenarios. If a SIFI fails, it must take remedial steps and resubmit a capital plan for approval. The goal is to reduce risk of failure, which reduces the need for a government bailout. But as with all models, the quality of output depends on the quality of inputs.
B) Living wills: planning for the worst
In addition to CCAR, the âliving willâ requirement ensures banks have plans in place for orderly resolution in the event of a collapse (without a bailout). Itâs a contingency plan. Per Section 165(d) of Dodd-Frank, banks must submit resolution plans to the Federal Reserve and the FDIC outlining how they would wind down in a way that minimizes the impact on the broader system. The largest banks are required to file a resolution every other year. Others, less frequently.
C) Liquidity coverage ratios: the bankâs CYA
Liquidity coverage ratios (LCR) requires banks to maintain a certain level of high-quality liquid assets (HQLA). These are assets that can quickly convert to cash in times of stress. For SIFIs, the LCR requires them to hold enough HQLA to cover 100% of their projected net cash outflows during a 30-day period of stress. For smaller banks, the LCR requirement is less strict. These requirements were designed to cushion against the risk of a sudden loss of funding or a run on the bank.
Dodd-Frank sent a clear message to the SIFIs of the world: we see you, and weâre watching. It identified and regulated. The issue was that we didnât see SVB. We failed at step 1. Why? Deregulation.
Crapo changed the game.
đĄ Regulators werenât looking because in 2018, we said âdonât look.â
The Economic Growth, Regulatory Relief, and Consumer Protection Act raised the asset threshold from $50 billion to $250 billion. The "Crapo Bill", affectionately nicknamed after its primary sponsor Sen. Mike Crapo (R-Idaho), was a bipartisan cry for deregulation of mid-sized banks. Crapo changed the SIFI identification game. If a bank had at least $250 billion in assets, the full arsenal of regulations would apply. Below the threshold, regulations became discretionary. The Fed could choose to apply key regulations like CCAR, living wills, or stronger liquidity requirements, but it didnât have to.
Several banks that were previously designated as SIFIs under Dodd-Frank gladly shed the designation. The higher asset threshold exempted 25 of the largest banks, including Deutsche Bank, UBS, and Credit Suisse. Their smaller size and lower risk profile allowed them to fly under the $250 billion radar. Crapo was designed to give mid-sized banks like SVB more flexibility in their regulatory obligations. The consequence: you canât effectively regulate what you donât see. Sometimes, you canât even effectively regulate what you do see.
Flying under the radar.
Banks with between $100 and $250 billion in assets have routinely flown under the radar. By design, they are subject to some of the least stringent capital, stress testing, and liquidity requirements among large banks. Under Dodd-Frank, SVB would have been big enough to warrant strict federal supervision when itâs assets surpassed the $50 billion mark. But post-Crapo, SVB was able to fly under the radar.
Because SVB was under the Crapo threshold, it didnât have to comply with burdensome legislation like CCAR, living wills, or LCR. One way to conceptualize what happened is to think about bikes. When cycling, riders can change gears to adjust their cadence and effort in 2 ways: big lever (chainring) and small lever (cassette) adjustments. Raising the asset threshold for federal scrutiny is a big lever (chain ring) move. It delays risk identification. Modifying regulations that apply post-identification is fine tuning (cassette) work. Still important, but only relevant once weâre addressing the right group of SIFIs.
So the question becomes: if stricter regulatory compliance was required (i.e., in a pre-Crapo world), would it have made a difference?
SVB did submit a living will in 2022. That didnât help much. And itâs likely they would have passed LCR since they held a lot of âlow-riskâ assets (e.g., treasuries, Ginnie Maes). Even in the unlikely scenario that they didnât breeze past LCR, the solution would have been to add more treasuries (i.e., more HQLA). But recall: having too many treasuries was precisely SVBâs problem. In todayâs high interest rate environment, adding treasuries is a trojan horse. SVBâs core problem was mismanaging interest rate risk. That leaves us with CCAR.
CCAR: hero or false promise?
Interest rate is a significant input for CCAR. The stress test scenarios include 6 distinct interest rate variables: the rate on 3-month Treasury securities, the yield on 5-year Treasury securities, the yield on 10-year Treasury securities, the yield on 10-year BBB-rated corporate securities, the interest rate for 30-year fixed-rate mortgages, and the prime rate. Interest rate risk was theoretically quite well covered. Some believe that CCAR stress test requirements would have caught SVBâs high risk holdings. But Iâm not so sure.
The 2022 tests asked banks to assume a baseline of 2.2% for the 10-year Treasury in 2023-Q1 and then an adverse scenario of 0.9%. They never considered rates as high as they actually were this year (~3.9%). Itâs a classic garbage in = garbage out problem. The Fedâs 2023 scenarios recast rates so the new 2023-Q1 baseline is 3.9% and the new adverse case is 1.1%. Perhaps, SVB would have failed CCAR under the revised scenarios.
In 2023, for the first time, the Fed has published an additional, âexploratory market shockâ scenario. This wonât feed into a bankâs capital requirements, but will be used to capture a wider array of risks in future stress test exercises. Itâs designed to make future models better. Maybe CCAR based on 2022 data offers false promise, but CCARs based on 2023 or future scenarios have hero potential.
What complicates this is that bank risk is managed both by company leadership and regulators. Co-parenting makes supervision both more important and more difficult. SVB didnât have a Chief Risk Officer for months. Greg Becker, CEO of SVB, was on the board of the Federal Reserve of San Francisco until March 10. This is more than just plain bad optics. It speaks to a fundamental supervisory failure. One that is likely to have complicated even the most theoretically sound regulatory regime.