Perspective on Risk - Dec. 19, 2023 (Systemic Risk)
Systemic Resolution Advisory Committee; Doomsday Book; The Fed and the Debt Ceiling Dilemma; Confessions of a Former SVB Executive; Margins, Debt Capacity & Systemic Risk; Bair on CRT; FOIA Request
FDIC’s Systemic Resolution Advisory Committee
Steve Kelly of the Yale Financial Stability Program dropped a short tweet that ended up sending me down a rabbit hole.
First of all, I didn’t know that there was a FDIC Systemic Resolution Advisory Committee, and that they webcast their meetings. Kudos for transparency.
There are quite a few heavy-hitters on this committee. I’m not sure I’ve seen a bigger set of heavy hitters (they need Geithner or Checki though) on systemic risk.
So I listened in to the 2023 Meeting (the whole thing is over 7 hours!).1 The agenda for the meeting included a review of the failure and resolution of SVB, Signature and First Republic. It also had a presentation by Eva Hubkes on the resolution of Credit Suisse. It concluded with discussions of the decisions that the FDIC made, the tools at their disposal, and policy changes they are pursuing. As with many of these meetings, some of the best insight came from the questions from participants, and I’d cite Dick Herring (Wharton) and Rodgin Cohen (Sullivan & Cromwell) in particular.
Business Model Failures
The consensus was that the failures of both the US Regionals and Credit Suisse were essentially business model failures.
Dick Herring noted that before failure each of the banks appeared well-capitalized with ample liquidity according to most observed metrics.
Gary Cohn (former Goldman; former Director of the National Economic Council) noted that the problem initially wasn’t the lack of liquidity, but that the banks were flooded with liquidity and had to buy something, so they bought zero RW assets (but didn’t hedge the IRR).
This, incidentally, seems to be a good reason to keep the Fed’s RRP facility in place, so that if banks do not want the excess liquidity it has some place to go.
Doug Peterson (S&P) stated that it did not look like SVB was a bank; he said it looked much more like a securities company, and he wondered whether SVB’s banking license should have been taken away, and whether there were other banks whose licenses should be revoked.
I would add that this was an interesting and provocative take that was not pursued. Throughout my career, the trend has been to allow banks to conduct more and more non-traditional activities, with the repeal of Glass-Steagall as the highlight. This does raise the important issue that we have discussed in the past of where the boundaries around regulated finance should be placed.
Tim Mayopolous (former CEO of Fannie Mae) noted that SVB was a State-chartered bank. And, he stated, there was a reason for this: the types of loans that SVB originated (loans to VC backed companies without profits and sometimes without even revenues) would per se be criticized assets if the firm had a national charter and was regulated by the OCC.
This is another area that was not pursued further. Our dual national/federalist banking system allows the states to grant authorities and powers that nationally-chartered banks may not possess, but FDIC deposit insurance gives them the same insurance backstop.
Resolving Banks
Much of the early part of the meeting discussed practically the resolution of the three US regionals.
Rodgin Cohen pointed to operational risks in the system, rather than just at the banks, that made assistance difficult. While the banks may not have prepositioned collateral at the Discount Window, he noted that the securities wire closes early and that there are large blocks of time in the day when you cannot move securities.
There was a discussion of the stigma of Discount Window borrowing, with everyone nodding in furious agreement that there shouldn’t be stigma, but Gary Cohn pointing out that as long as you had the Primary vs Secondary credit designation it would not go away.
Systemic Risk Exemption
As you are aware, in the US, Treasury approved the use of the systemic risk exemption in response to the US Regional bank failures. You also know that I am a bit critical of that decision.
Dick Herring asked the key question:
was any attempt made to avoid the systemic risk exemption?
The answer seems to be that, yes, initially they sought to resolve things in normal course, but then subsequently the fear of contagion changed their minds.
The FDIC showed the timeline of the SVB failure, and discussed how they and the other agencies initially planned for a normal bridge-bank resolution when SVB was closed on Friday morning March 10th. They first became concerned with reports of deposit flight from Signature on Friday evening. Signature, to them, was the indication that SVB would not be idiosyncratic, but rather part of a broader contagion. When they began to get reports of deposit flight from banks without the VC or crypto linkages they became increasingly worried that “anything less than 100% guarantee of uninsured deposits would continue the contagion.”
Rodgin Cohen later chimed in that without the systemic risk exemption, multiple banks would have failed within days.
In discussing the failure in the Credit Suisse case to ultimately use the resolution authorities that had been planned and instead merge the firm with UBS, the consensus was summed up that “the Swiss blinked” and that “this sets an unfortunate precedent.” No one seemed to feel it was a hypocrisy that the US took extraordinary measures to resolve regional banks.
At the end of the day, it comes down to a choice between contagion and market discipline. If history has taught us anything, it’s that regulators will be driven more by the immediate fear of contagion than the longer-term effects of market discipline.
Too Big To Fail
If Resolution Planning and new resolution authorities should have worked, it should have worked here (in the case of Credit Suisse). This was a very slow train wreck, it was idiosyncratic in a sense … there was not some wide spread market meltdown regulators were dealing with, … Some of the problems that should have been anticipated, and resolved far ago. And we owe it to the public, are we serious about this? Do we want to end Too Big To Fail? - Sheila Bair
Another area of missed opportunity.
Dick Herring did ask about how thinking about TBTF came into the resolution equation, noting that the sale of First Republic to JPMC, while putting the bank into a strong hand, only served to make the largest US bank even bigger.
Ryan Tetrick of the FDIC noted that they were bound by law to follow the least-cost test, and in the case of First Republic JPMC had by far the lowest cost bid.
We’ve discussed this in the past. If we are serious about ending TBTF, we need to have the law changed. Receiving a lower return to the FDIC Fund, which as a reminder is funded by deposit insurance premiums, seems a small price to prevent making a future problem potentially worse.
The NY Fed’s Doomsday Book
The WSJ has written Sun Shines on Fed ‘Doomsday Book’. It hasn’t yet gotten much commentary, but - like WOW.
I first remember the Doomsday Book being mentioned by ‘Ernie the Attorney’ - Ernie Patrikis the General Counsel of the NY Fed in the 1990s and subsequently its First Vice President (the #2 person at the Reserve Bank). I also distinctly remember Tom Baxter, the General Counsel during the Global Financial Crisis, carrying around this large binder and referencing and asking questions about this during the GFC.
Now, what they have disclosed here I don’t think is THE book, but rather the guide to the book. This is a massive set of references to other documents that detail the actual events, decisions, precedents and authorities.
From the WSJ article:
The so-called Doomsday Book, an internal document used to guide the Federal Reserve’s actions during emergencies, has long been the subject of intrigue and suspicion. Largely a compilation of legal opinions, the book has been a key resource for the Federal Reserve Bank of New York for decades, allowing it to play a unique and oversize role during financial crises. No other regional Federal Reserve bank has such a resource.
This was the bible: when in the past I have said that the Fed always considered it constrained by the law and sought to take actions within the boundary of the law, this was the interpretation of the law according to the NY Fed. These were the boundaries that the NY Fed was comfortable defending if necessary.
The document reveals a fascinating history of diverging perspectives on the Federal Reserve’s emergency powers. Instead of adhering strictly to clear legislative boundaries to justify its actions during financial crises, the central bank appears to ground many of its decisions in the New York Fed’s belief in the Fed’s discretionary authority. It relies on precedent for many of its actions, without explicit congressional authorization in some instances.
The book also exposes an apparent split in perspective between the New York Fed and the Fed Board of Governors. At the core of this disagreement are differing interpretations of the central bank’s legal powers, particularly Sections 13(3) and 14(b)(1) of the Federal Reserve Act, which allow the Fed to take extraordinary actions in financial crises. The New York Fed, wary of the complexities of financial markets and the unpredictable nature of crises, embraces a flexible interpretation of these laws. The board adopts a more cautious approach that underscores the importance of adhering closely to legal limits.
This is perhaps the most revealing statement in the book:
the powers of a Federal Reserve Bank are far greater than is commonly assumed.
I’m hoping that the Yale folks or some others go through this with a fine-toothed comb. There are so many tidbits linked to various historical crisis. The one that I found most intriguing, and that I have no historical knowledge, is that the NY Fed evidently looked into providing liquidity support to the BIS in the early 1980s. I would guess that this was so that the BIS could on-lend to help with the Latin American Debt Crisis.
The most recent additions are around the GFC lending facilities and the Maiden Lane loans to resolve AIG.
The Fed and the Debt Ceiling Dilemma
Nathan Tankus has received some notoriety recently for a series of FOIA requests and responses from the Fed. In his latest post, MORE FOIA FINDINGS: THE NEW NIXON ADMINISTRATION’S DEBT CEILING DILEMMA AND THE FEDERAL RESERVE’S SOLUTIONS, he covers his latest FOIA finding from the Federal Reserve regarding its response to Treasury debt ceiling struggles in 1968.
To my mind all of these different stories fit into a much longer history of accounting gimmicks to avoid the debt ceiling.
The two memos I’ve secured relate to a more expansive list of accounting gimmicks, some of which require the Federal Reserve’s cooperation.
Continuing the tradition of anodyne memo names, the first memo dated January 30th 1969 was simply entitled "Treasury Cash and Debt Ceiling Dilemma." by Alan Holmes. Holmes was the New York Federal Reserve’s Manager of the Fed’s “System Open Market Account” (SOMA). In other words, he decided what the Fed bought and sold. The second memo “Legal aspects of proposals for assisting Treasury in connection with cash and debt ceiling problems” was authored by head legal counsel Howard Hackley.
I’m not going to cover the substance of his analysis, feel free to click over if you’re interested. I cite this to further note how the NY Fed will carefully look at the law to determine what they can and cannot do.
In this case, it’s important to remember that the analysis is being conducted to assist a client: the Federal Reserve, and in particular the Fed’s Market Desk, conducts Treasury operations on behalf of the Treasury. This is not an independent decision.
Confessions of a Former SVB Executive
William Cohen, the author of numerous financial tell-alls, writes in Puck News ($) Confessions of a Former SVB Executive
I spoke with a former bank executive for SVB, who gave me a sense of what it was like to work at Silicon Valley Bank in the years leading up to the bitter end.
Among its myriad problems, this person said, SVB had lost sight of the fact that it was… a bank. “It was a bank that thought it was a technology company,” the executive told me. “It didn’t run itself like a bank.”
Essentially, the banker explained, it was like one big party, until two weeks before the bank failed on March 10.
The problems at the bank were compounded by the fact that SVB, despite its very provenance, was behind the curve on banking technology. And this weakness was exacerbated during the pandemic when it was flooded with deposits. In effect, SVB didn’t have the technological infrastructure to handle the influx of deposits that came into it during the pandemic, nor the investment skill to manage all that new money.
We thought, like, Why would you want to invest in risk management or technology infrastructure? Because, as we all were told all the time, well, that’s going to dampen down on profits, and that’s gonna dampen down on the formula we use to get our bonuses.
Margins, Debt Capacity & Systemic Risk
Here is another paper that aligns with Minsky’s Financial Instability Hypothesis. Minsky asserts that financial crisis’ are endogenous to the system; this is often stated as “stability breeds instability.” For that reason, this paper caught my eye.
Margins, debt capacity, and systemic risk (CEPR) seems important as our financial system increasingly has moved from unsecured to fully secured borrowing.
In particular, the paper’s discussion of the recursive effects of leverage in the paper provide an almost mechanistic representation of the dynamics Minsky described more qualitatively in his FIH framework. Similarly, the signaling effect of margin increases leading to deleveraging and cash hoarding provides empirical support for the regime shifts in financing conditions emphasized by Minsky.
Abstract
Debt capacity depends on margins. When set in a financial system context with collateralized borrowing, two additional features emerge. The first is the recursive property of leverage whereby higher leverage by one player begets higher leverage overall, reflecting the nature of debt as collateral for others. The second feature is that the “dash for cash” is the mirror image of deleveraging. In any setting where market participants engage in margin budgeting, a generalized increase in margins entails a shift of the overall portfolio away from riskier to safer assets. These findings have important implications for the design of non-bank financial intermediary (NBFI) regulations and of central bank backstops.
Body comments
Traditionally, systemic risk narratives have relied on the “domino” model of cascading defaults. According to the domino model, if Bank A has borrowed from Bank B, while Bank B has borrowed from Bank C, and so on, then a shock to Bank A’s assets that leads to its default will hit Bank B as well. If the hit is big enough, Bank B’s solvency will be impaired, in which case Bank C would be hit, and so on further down the line. Insolvency is seen as the driver of systemic risk in the domino model.
Fluctuations in leverage working through shifts in risk-taking capacity can also be a potent channel of propagation of stress, especially in settings with market-based intermediation. Margin is posted using own funds (equity) so that the ratio of total exposure to margin corresponds to overall leverage.
Conclusion
Systemic risk is born out of externalities that originate from the distress faced by a market participant and that adversely affect others, typically in the form of non-fundamental price movements. The propagation of systemic risk does not require insolvency, and can be routed through disruptions to liquidity/maturity transformation or through fluctuations in leverage that magnify changes in prices and intermediation activity.
Over the past decades, non-bank financial intermedaries have become increasingly central in the supply of credit and liquidity. Collateralized borrowing, such as repurchase agreements, and partially funded exposures, such as centralized or over-the-counter derivatives, are instrumental to the activity of NBFIs. In both instances, margins determine the debt capacity of intermediaries. For a given amount of capital available, margins are directly linked to the risk that market participants can take and the liquidity they can provide.
… We find that the recursive nature of debt capacity – namely, that higher leverage by one investor allows others to increase their leverage – magnifies fluctuations in financial activity and generates procyclicality. We also find that sudden reallocation from risky assets to safe havens (dash for cash) is a manifestation of deleveraging.
From a regulatory perspective, the presence of externalities highlights the importance of macroprudential policies. Leaning against procyclicalty generated by leverage fluctuations can potentially take the form of minimum initial margins, which would constrain leverage increases and dampen the ensuing declines. Such provisions would complement appropriate self-insurance against adverse shocks, so to minimize the need for leverage adjustments in response to shocks. In the presence of very large shocks, the public sector – in particular central banks given their flexible balance sheet – may need to provide backstops (Markets Committee (2018)), thereby supporting debt capacity. Such prospect can alter behaviors and ex-ante incentives to self insure. Hence it is important to make sure that any public backstops be accompanied by an appropriate regulatory framework.
I Disagree With Bair On Credit Risk Transfer
Sheila Bair writes in the FT Regulators must resist banks’ magical thinking. I think she makes a number of unsubstantiated claims.
The late, great Charlie Munger once said that derivatives trading desks made witch doctors look good. That was certainly true of credit default swaps, used by banks to circumvent capital requirements in the run-up to the 2008 financial crisis.
That is a very strong statement. I know of no case where US banks were inappropriately applying the capital rules. Yes, banks were able to take on increased risk based on the use of the Market Risk Amendment rules (1.6%), rather than the traditional credit risk (8%) treatment had the funded positions been on the balance sheet, but this is hardly circumvention. Positions were margined and marked-to-market, a different risk management paradigm.
CDS were supposed to make banks safer by magically transferring the risk of loan defaults to counterparties outside the system. Since regulators had assumed the risk transfer was real, they allowed banks to lower their capital levels. But when the 2008 crisis hit, banks found the risk transfer was not always legally binding and even when it was, counterparties could not perform.
The large losses experienced by Citigroup and others were more a function of the mischaracterization of backup lines as liquidity facilities than as instruments involving credit risk. In essence, for reputational and legal reasons, in many cases the banks were forced to take the exposures onto their balance sheet.
The other generator of the loss was that the super-senior positions were much more volatile than expected. In this case, she does have something of a case once more of the tranches were sold than were in existence.
She goes on to admit that credit-linked notes have structural advantages over credit default swaps.
With CLNs, banks get the money upfront by issuing bonds with repayment obligations that depend on borrower performance. If borrowers default, the bank’s repayment obligations are reduced, passing along the losses to bond holders. But while this eliminates counterparty risks, CLNs still suffer from fundamental problems which should make regulators wary.
Her subsequent concerns come down to the following unpersuasive arguments:
CLNs do not provide the same level of resiliency as good old tangible common equity.
Synthetic risk transfers may also make the financial system less stable. Nonbanks buying them are mostly lightly regulated or unregulated entities such as private funds and insurance companies. They are not subject to the same level of capital regulation and disclosure requirements as banks. They are not supported by deposit insurance and central bank lending for liquidity needs. They are not as expert in understanding and pricing credit risk.
Her first point would take us back to the 1970s, where the financial sector was dominated by low rate of return banking. Bringing all of these exposures onto bank balance sheets would either exacerbate the too-big-to-fail problem as existing banks grow (unlikely due to leverage ratio constraints) or force the business entirely out of the regulated financial sector (thereby shrinking regulated banking relative to the unregulated sector). Her second point perhaps points to this being her preferred path.
And to her second point, insurance companies are lightly regulated? Private investors and funds are insufficiently sophisticated to understand and price the risk? Go talk with Blue Mountain Sheila.
I’d also add that the European experience has insight to bear on these questions. For years their banks have managed to successfully sell principal protected structured notes, even to their retail bases.
The fundamental business of banking is liquidity intermediation. A related business, arguably more investment banking than banking but the line has so blurred over the years, is risk intermediation, and finding ways to transfer risk from those less optimal to bear the risk to those more able. We shouldn’t inhibit this function.
FOIA Request
Just a note that I have not yet had a response to my FOIA requests. Interestingly, Jason Leopold of Bloomberg is on the same beat (and a step ahead of me):
You can watch the whole thing in about 3 hours by skipping the breaks and running it at 1.25x