Think sad thoughts.

Photographer: Mark Wilson/Getty Images

Living Wills Make Banks Think About Death

Matt Levine is a Bloomberg View columnist. He was an editor of Dealbreaker, an investment banker at Goldman Sachs, a mergers and acquisitions lawyer at Wachtell, Lipton, Rosen & Katz and a clerk for the U.S. Court of Appeals for the Third Circuit.
Read More.
a | A

One important thing to realize about U.S. banking regulators' responses to the big banks' "living wills" is that they are, on their face, absurd. From JPMorgan's:

JPMC's 2015 Plan did not sufficiently disclose or provide comprehensive support for estimating liquidity needs in resolution beyond the assumptions used for intraday reserves, buffer for postfailure sever stress outflows, and operating expenses. Specifically, the 2015 Plan did not sufficiently provide daily cash flow forecasts for the period following the parent bankruptcy filing required to stabilize the material entities and did not provide a breakout of all interaffiliate transactions and arrangements that could impact JPMC's liquidity forecast estimates. For example, the 2015 Plan only provided daily cash flow for the first seven days and last four days of the runway period, and for the first three days after JPMC's bankruptcy filing.

JPMorgan's bankruptcy filing is in the far distant and hypothetical future. Yesterday, JPMorgan announced that it earned $5.5 billion last quarter, and has $250 billion of shareholders' equity; its stock rallied. JPMorgan is doing great. But meanwhile, somewhere at JPMorgan, someone is in trouble for only projecting daily cash flows for the first seven days after that bankruptcy filing, whenever it is, and whatever causes it. What happens on the eighth day? That's the sort of hard factual information that JPMorgan's regulators at the Federal Reserve and the Federal Deposit Insurance Corp. need to know. And JPMorgan will have to go back and figure it out.
QuickTake Too Big to Fail

The living wills -- they are technically called "resolution plans" -- are the most straightforward manifestation of the regulators' post-crisis efforts to make sure that no bank is too big to fail any more. Each bank has to file a resolution plan explaining how it could be resolved in bankruptcy with a minimum of disruption to the financial system. On Wednesday, the Fed and FDIC found that seven of the eight biggest banks had submitted resolution plans that were "not credible or would not facilitate an orderly resolution under the U.S. Bankruptcy Code." (Citigroup passed.) FDIC Vice Chairman Thomas Hoenig said:

Most importantly, no firm yet shows itself capable of being resolved in an orderly fashion through bankruptcy.  Thus, the goal to end too big to fail and protect the American taxpayer by ending bailouts remains just that: only a goal.

But remember that the evidence for this is things like JPMorgan's lack of a detailed daily cash flow statement for its eighth day after bankruptcy. Or that Bank of America's plan to wind down its derivatives portfolios "lacked detailed portfolio information and specificity regarding implementation of the wind-down." Or the insufficiency of Goldman Sachs's "triggers designed to escalate information to senior management and its board through multiple phases as the condition of the firm worsens." Or Wells Fargo's lack of "legal entity rationalization criteria" -- that is, rules about not creating too many subsidiaries -- that "not only provide for the rationalization of current entities, but also provide for adequate controls for future strategic actions." 

Of course, much of this stuff really is relevant to whether these banks could be wound up tidily in bankruptcy.  But it is difficult to make predictions, especially about the future, and most especially about the conditional counterfactual future in which some unknown event has caused the imminent bankruptcy of a bank that is currently huge and profitable. Who knows what JPMorgan's cash flow will be like eight days after it files for bankruptcy? Who knows what Bank of America's derivatives portfolio will look like in a world where it has gone bankrupt? Who knows how Goldman's board will hear about Goldman's bankruptcy filing? I'm pretty sure they'll hear about it.

It seems implausible that any big bank could lay out a plan, in the level of detail that regulators seem to want, that could reliably guide it through any actual bankruptcy. The actual context of a big bank's failure will determine what the bank does; a cash-flow budget concocted years earlier to fulfill a regulatory requirement seems unlikely to be a useful guide. This means that the banks with inadequate living wills may be able to fail gracefully, and it means that the one bank -- Citigroup -- with an adequate living will may nonetheless be ugly if it fails.  

I mentioned earlier a relatively simple, relatively dumb, relatively reliable way to let banks fail gracefully. You capitalize the parent company (with equity and also long-term unsecured debt), and if a systemically important subsidiary (a bank with deposits, or a broker-dealer with derivatives) becomes insolvent, you write down the parent company's securities to recapitalize the subsidiaries. This "single point of entry" mechanism is a part of Dodd-Frank, and an (as yet somewhat untested) alternative to the bankruptcy process. If you like this approach -- and it is after all the approach crafted by regulators specifically to respond to the too-big-to-fail problem -- then there is less reason to worry about how credible the banks' plans are to resolve themselves in bankruptcy. 

Or you might prefer other solutions to too-big-to-fail, like vastly boosting capital requirements so banks are safer, or breaking up the banks so they're smaller, or "narrow banking" or various other proposals to make banking immune to disorderly collapse. There are lots of ideas to address too-big-to-fail, and they all share the salient advantage of not requiring JPMorgan to project its daily cash flows for weeks after a hypothetical bankruptcy filing. Those ideas work by trying to find structural solutions, rather than by taking a microscope to the banks' complex and ever-changing positions and cash flows.

So what is the point of the living wills? I can't imagine that the answer is that they are the best way to make sure that, if a big bank becomes insolvent, it will be able to go bankrupt quietly. But I also can't imagine that the point of the stress tests is that they are the best way to make sure that, if there is an economic crisis, all the big banks won't become insolvent. The living wills, like the stress tests, are complicated stylized exercises that make assumptions that are unlikely to be true in any real crisis. So why do them?

Because they are exercises. The point of the living wills, like the stress tests, is to sit banks down and make them comb through their businesses in excruciating detail, with a focus on grim aspects like liquidity crunches and operational risks in bankruptcy. A useful result of the living wills is that, if they're done correctly, they give regulators a good overall picture of how a bank works, how money flows between its parts, what its pressure points are, and how it responds to crisis. But a much more important result is that, if they're done correctly, they give bankers themselves that same overall picture: They force a bank's executives and directors to understand the workings of the bank in a detailed and comprehensive way. And if they're done incorrectly, that's useful too: They let the regulators and bankers know what they don't know. 

So here's a passage from the Wells Fargo letter:

The 2015 Plan contained material errors that required resubmission of the 2015 Plan's financial information. For example, substantial changes were required for projections supporting the least-cost test analysis and the volume of available liquid assets at the time of the transfer of Wells Fargo Bank, National Association (WFBNA) to the FDIC as receiver. The latter was especially important as the original volume of liquid assets and its trajectory of decline did not support WFBNA's failure state. These errors call into question the executability of the 2015 Plan, as the lack of effective resolution planning governance raises concerns regarding quality control, senior management oversight, and recovery and resolution planning staffing.

The 2015 Plan represented that the firm's leadership steering committee, recovery and resolution planning office, and lines of business/control and support functions all had input into the 2015 Plan prior to its submission. The material errors noted above call into question the extent to which there was appropriate internal review and coordination with respect to the 2015 Plan prior to its submission.

The message here isn't as simple and catchy as "Wells Fargo is still too big to fail." The message here is: Wells Fargo's senior management needs to spend a lot more time thinking about failure, and liquidity, and resolution planning, than it now does.

The great purpose of the living wills, it seems to me, is to re-focus banks' attention. It's to make sure that banks, at their most senior levels, are thinking deeply and carefully and critically about the things that regulators are worried about. It's to change how bankers think. The natural state of a chief executive officer is one of optimism, growth, aggressiveness.  In the current regulatory environment, they are supposed to think a bit more about pessimism, decay, defensiveness. The living wills are a way to make them think sad, nervous thoughts -- and to punish them if they don't think those thoughts as rigorously as they should.

There are certainly simpler ways to regulate! Addressing too-big-to-fail structurally, by changing the rules around bank resolution or by requiring banks to be smaller or better capitalized, might be a more direct approach. Some people who like the direct approach balance it with a deregulatory aspect: Once banks are small or well-capitalized or whatever enough that we don't worry about their failure, we can let them do whatever they want, without worrying too much about how they think.

But there are also more complicated ways to regulate. Risk-based capital requirements and liquidity coverage rules and prudential standards and much of the rest of the banking regulatory apparatus involve substituting the judgment of regulators for the judgment of bankers, at least to some extent. The bankers think that something is a good idea, the regulators think it's risky, and the regulators make banks cut it out, or at least make it more expensive for them to keep doing it. But this is a difficult game for the regulators to win, since the bankers will always be better paid, and better staffed and more motivated. It's hard for regulators to get into bankers' heads; the bankers can always stay a bit ahead.

The new approach isn't (just) to have regulators second-guess bankers, though obviously there's a lot of second-guessing going on when seven out of eight banks get failing grades on their living wills. The new approach is to make the bankers get into the regulators' heads, to fill banks with people who spend so much time worrying about bankruptcy that those worries bleed into the banks' regular operations. If you're forced to revise your bankruptcy plan over and over again, and to fill in more and more details about where your cash will go in a crisis, that crisis will become much more vivid and immediate, and much less attractive. If you think constantly about how unpleasant failure will be, maybe you'll care more about preventing it.

  1. Strictly, they "jointly determined that each of the 2015 resolution plans of Bank of America, Bank of New York Mellon, JP Morgan Chase, State Street, and Wells Fargo was not credible or would not facilitate an orderly resolution under the U.S. Bankruptcy Code." The other two banks, Goldman Sachs and Morgan Stanley, got mixed reviews: Goldman passed at the Fed but failed at the FDIC, while Morgan Stanley passed at the FDIC and failed at the Fed.

  2. Other stuff is even more clearly relevant, a matter of bankruptcy regimes rather than specific predictions about cash flows. For instance:

    Part of the problem is that the banks relied on getting liquidity from other pockets of their far-flung empires. JPMorgan and Morgan Stanley, whose living will was found not credible by only the Fed, are among the banks that count on help from foreign units. It’s a problem because overseas governments could ring-fence the banks in their jurisdictions in a pinch. The banks were ordered to ensure they had enough liquidity without relying on that assistance.

  3. From the regulators' response to Citi:

    Nevertheless, the Agencies identified a shortcoming because Citigroup also made optimistic assumptions about continued access to bilateral OTC derivative markets to hedge its portfolio risk and about the ability to novate bilateral OTC derivatives without sufficient specificity on the nature, concentration and illiquidity of the bilateral OTC derivatives.

    I don't know why that shortcoming isn't as bad as the shortcomings of the banks that failed the living-wills test.

  4. This is generically true, but probably less true of bank CEOs than of those in other industries. Still, it's safe to say that bank CEOs are more optimistic about their banks than some of their critics are.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

To contact the author of this story:
Matt Levine at

To contact the editor responsible for this story:
James Greiff at