Title II of Dodd-Frank creates a new resolution authority, called the Orderly Liquidation Authority (OLA), for systemically important financial institutions — which, crucially, includes financial holding companies (all the major US banks are organized as FHCs now). The OLA is patterned on the Federal Deposit Insurance Act, which lays out the FDIC resolution authority for commercial banks. The OLA, like the FDIC resolution authority, gives the FDIC a range of tools to liquidate a large nonbank financial institution while also mitigating systemic risk. No one is arguing that commercial banks that are seized and resolved by the FDIC are being “bailed out,” because they’re not — that’s why we call them “failed banks.” The FDIC resolution authority is just an alternative insolvency regime; but, obviously, it’s still an insolvency regime. So presumably, the people who say that “Dodd-Frank did nothing to end TBTF!” are arguing that a large FHC would not be resolved through the OLA.

The real question, then, is:

If a large financial holding company were on the brink of failing, would regulators resolve the FHC through the OLA?
I think it’s almost certain that regulators would, in fact, resolve a large FHC through the OLA, because it would work.


As I noted in my previous post, one of the main reasons that Lehman’s failure was such a catastrophic event for the markets was Lehman’s complete and total lack of preparation for a bankruptcy filing. Any serious analysis of Dodd-Frank’s resolution authority therefore has to recognize that Dodd-Frank also requires any financial institution subject to the new resolution authority to regularly submit a “resolution plan” (a.k.a. “living will,” or “funeral plan”) to regulators. I’ve slowly become a big fan of the resolution plan. I was never against it, and it still won’t do what its original proponents intended it to do, but it will perform an absolutely vital function: ensuring that the banks keep all the information/data necessary to execute an orderly resolution readily available. Significantly, § 165(d) of Dodd-Frank requires that resolution plans include:
(B) full descriptions of the ownership structure, assets, liabilities, and contractual obligations of the company;

(C) identification of the cross-guarantees tied to different securities, identification of major counterparties, and a process for determining to whom the collateral of the company is pledged; and

(D) any other information that the Board of Governors and the Corporation jointly require by rule or order.
This is hugely important. The absence of this information made everything about Lehman’s bankruptcy significantly worse, as I discussed previously. In addition to the information in (B) and (C) above, the Fed and FDIC should require that the resolution plan describe the terms and conditions that a potential acquirer of the bank’s primary entities would have to be prepared to agree to (i.e., if you buy X, you’d also be legally obligated to assume the obligations of Y). They should also require the resolution plan to include a comprehensive index of customer accounts, and important documents such as clearing agreements and Global Master Repurchase Agreements with major counterparties. All of this is eminently doable, I promise you, and it’s something the banks should have on hand anyway, just as a matter of good business practices. (I know for a fact that at least some banks would be able to pull most of this information together in a single weekend.)

The process prescribed by Dodd-Frank for submitting resolution plans also gives the FDIC an opportunity to identify legal structures and processes that would cause serious problems in a resolution (like, for instance, Lehman’s now-infamous “RASCALS” process). Not only that, but the statute actually gives the Fed and FDIC the authority to do something about problem issues preemptively, which is huge.

In short, the presence of a resolution plan will make the decision to resolve a large FHC under the OLA significantly less daunting.


Now we get to the actual resolution. What would the resolution of a large financial holding company look like under the OLA? First, the FDIC would be appointed as receiver of the holding company, all its US primary entities (except for any insurance companies, which are still resolved by the states), and all its US booking entities. Technically, the FDIC would be appointed as receiver of the US commercial bank under the Federal Deposit Insurance Act rather than the OLA, but this is a distinction without a difference for our purposes, as the statutes are substantially similar. As receiver, the FDIC would succeed to all the rights, titles, powers, and privileges of these companies and their assets — in other words, the FDIC would be in complete control.

The FDIC resolves the vast majority of failed commercial banks through what’s known as “purchase and assumption” agreements (called “P&As”), in which a healthy bank purchases some or all of the assets of a failed bank and assumes some or all of the liabilities, including all insured deposits. For example, the FDIC resolved WaMu — which had over $300bn in assets — through a P&A with JPMorgan, in which JPM acquired all of the assets ($296bn), and almost all of the liabilities ($265bn) of WaMu’s depository instutition. Crucially, JPM left behind some $28bn of WaMu’s senior unsecured debt, subordinated debt, and preferred stock — and those creditors took huge haircuts. The particular structure of the P&A, and of the resolution, depends on the situation the FDIC is facing.

In the case of a large financial holding company, there are essentially four types of situations that the OLA needs to be able to handle:
(1) A buyer for the entire FHC can be identified prior to triggering the OLA;
(2) A buyer for only some of the FHC’s assets can be identified prior to triggering the OLA (most likely);
(3) Potential buyers have been identified, but no agreement has been reached by the time the OLA is triggered;
(4) No buyer can be or will ever be identified (least likely).
Let’s take these situations in turn.

(1) A buyer for the entire FHC can be identified prior to triggering the OLA

This is an easy one. Dodd-Frank gives the FDIC broad authority to arrange for the sale of entities in receivership, or for selected assets of the entities. The FDIC would just do a straight P&A for all the assets and liabilities of the FHC and its subsidiaries in receivership. Done and done.

(2) A buyer for only some of the FHC’s assets can be identified prior to triggering the OLA

It’s likely that a prospective buyer of one or more of the primary entities (i.e., the broker-dealer and the commercial bank) will insist on leaving behind a portfolio of problem assets — for Bear Stearns, it was a $30bn pool of assets that JPMorgan left behind; for Lehman, it was a $62bn pool of assets that Barclays was originally going to leave behind (before the FSA torpedoed the deal, of course). To do this, the FDIC would likely do one of two things. The first option would be to enter into a “loss sharing P&A,” in which the buyer would acquire all of the FHC’s assets, but the FDIC would agree to absorb most of the losses (>80% usually, I think) on a fixed pool of assets.

The second option would be to “bridge” the FHC and the relevant entities being acquired, then sell the pool of problem assets back to the receivership, and then sell the good assets to the buyer. Section 210(h) of Dodd-Frank authorizes the FDIC to create a “bridge financial company” — a temporary, pop-up financial institution that automatically has all the necessary charters and licenses, and is run by the FDIC — to purchase selected assets and liabilities of the entities in receivership. Crucially, before a broker-dealer is handed over to the Securities Investor Protection Corporation to resolve, the OLA allows the FDIC to transfer assets and customer accounts from the broker-dealer to a bridge financial company. This will allow the FDIC to ensure that the full suite of key services can continue uninterrupted: depositors can have uninterrupted access to their bank accounts, prime brokerage clients can have uninterrupted access to their cash and securities, etc.

Moreover, if a buyer can be identified ahead of time, the FDIC can prevent the bank’s derivatives counterparties from seizing the bank’s posted collateral and liquidating it at fire-sale prices (which they’d be allowed to do immediately under the Bankruptcy Code). Section 210(c) prevents derivatives counterparties from terminating the contracts and seizing the collateral for essentially a full business day after the FDIC is appointed as receiver, during which time the FDIC can transfer the derivatives to a bridge financial company. If the FDIC can sell the bank’s derivatives to another financial institution or a bridge financial company within one day, then there will be no interruption in the bank’s derivatives book.

The key, in my opinion, to making this work is that the OLA allows the FDIC to bridge not just the holdco and the primary entities being acquired, but also any booking entities that engage in significant intercompany transactions with those primary entities. This allows the FDIC to avoid the kind of endless intercompany disputes that have absolutely plagued the Lehman bankruptcy. If assets need to be shifted among the booking entities and the primary entities in order to minimize the disruption to clients, or to pave the way for a sale of the entire broker-dealer (and with it the entire “trading book”), then the FDIC has the discretion to get that done.

Now, what would happen to unsecured creditors? That depends entirely on what liabilities the buyers are willing to assume. In normal commercial bank resolutions, the P&A usually includes the senior unsecured and subordinated debt. This was famously not the case with WaMu, where the P&A with JPMorgan left the senior unsecured and subordinated debt holders out in the cold. This came as a huge shock to the market, and we can debate whether or not it was the right move by the FDIC at the time, but one indisputable benefit is that it created a precedent for haircutting senior unsecured and sub debt. Market participants won’t be nearly as shocked if the FDIC does that again.

In any event, I think the FDIC could certainly allow third-party buyers to leave the senior unsecured and sub debt behind without causing a systemic meltdown. The FDIC has the tools to ensure that the bank’s key services continue uninterrupted, which will prevent most of the contagion from spreading. Past that, I don’t really care if unsecured creditors get a haircut.

(3) Potential buyers have been identified, but no agreement has been reached by the time the OLA is triggered

This type of situation is the reason the “bridge bank” option was created back in the S&L crisis, and the bridge financial company option can be used here. The FDIC would likely bridge the FHC, the primary entities that potential buyers are interested in, and any booking entities that engage in significant intercompany transactions with those primary entities. It would be necessary to bridge the key booking entities here in order to prevent any derivatives counterparties from terminating the trades and seizing and liquidating the collateral, which could easily kill any potential sale of the primary entities or holdco.

If it’s starting to look less and less likely that a deal will be reached, the FDIC can start taking steps to minimize the disruption to the markets , like allowing prime brokerage customers to transfer their accounts to other brokers, and allowing derivatives counterparties to novate trades to other dealers. If a deal is struck, however, the bridge companies would transfer the assets being acquired to the buyer, and sell whatever is left back to the receivership.

(4) No buyer can be or will ever be identified

Next stop: Liquidation Station! Honestly, this scenario strikes me as extremely unlikely. You need to have a business that’s important to the functioning of the global economy just to be eligible for resolution under the OLA in the first place, so any company being resolved under the OLA is almost by definition going to have value as a going-concern. Anyway, in the unlikely event that this scenario did occur, the FDIC could keep the assets off the market for a good 2 years, and possibly even longer if necessary, as bridge financial companies have a 2-year life under the OLA, with an option for extensions. Again, it would bridge the FHC and its subsidiaries, and sell the assets off over time, avoiding a messy fire-sale that could depress asset prices and transport the bank’s problems to other financial intermediaries.

What about all those thorny international problems? Well, the truth is that in terms of systemic risk, there’s only one other jurisdiction that really matters: the UK. New York and London are still the two dominant financial centers, and the vast majority of transactions at the major US banks flow through either New York or London. It’s important to understand that it was the UK’s ridiculously backward somewhat dated insolvency regime that forced the liquidators of Lehman’s European broker-dealer to seize so many client assets and assets of affilates. Fortunately, the UK now has their own version of the OLA, which they call the “Special Resolution Regime,” and was enacted as part of the Banking Act of 2009. The Special Resolution Regime is, like the OLA, modeled explicitly on the FDIC resolution authority, and gives the Bank of England the same wide-ranging tools to wind down a London broker-dealer in an orderly fashion — including, significantly, the power to create “bridge banks” to ensure that key functions can continue uninterrupted. Cross-border problems that aren’t identified and dealt with in the resolution plan can, if necessary, be dealt with by bridging the relevant entities until a solution can be fashioned.


I’m not claiming that the Dodd-Frank resolution authority is perfect by any stretch of the imagination. But I think it absolutely provides the FDIC with the tools necessary to resolve a major US bank in an orderly fashion, and that, understanding this, regulators would indeed use the OLA to resolve a major US bank.

This post started out as a defense of Dodd-Frank’s resolution authority, and a description of what the liquidation of one of the major US banks would likely look like under the new law. But I quickly realized that in order to understand how a resolution of one of the major banks would work in practice, you really have to have an understanding of how the major banks/investment banks are structured, legally, and why that structure causes so many problems in bankruptcy. I don’t think this is something that’s ever been explained in the blogosphere (I’d be extremely surprised), but it’s crucial to understanding the real issues surrounding financial reform and the major banks, so I think this post can be useful. Anyway, this is part one of a two-part post; part two will describe what an orderly liquidation of a major US bank under the Dodd-Frank resolution authority would actually look like. That post will come sometime tomorrow (I’m too tired to finish it right now).


Major international banks are organized as holding companies — in the US, the major banks are all financial holding companies (FHCs). The holding company (known as the “holdco”) sits at the top of the legal structure. Right below the holdco is the layer of primary regulated entities, which conduct the vast majority of the bank’s day-to-day business, and employ most of the bank’s employees. The most important primary entities at the major US banks generally include registered broker-dealers in New York, London, and Asia, a chartered US commercial bank, a smaller continental European bank, and an asset manager.

Below the primary entities is a layer of product-specific “booking entities.” (See my very rough org chart below.) These entities are structured to hold specific types of instruments that the primary entities deal in (e.g., FX spot, interest rate swaps, crude oil futures), and are set up for regulatory, capital, tax, or other such noble reasons. Booking entities are generally fully guaranteed by the holdco as well. Here’s how this works: when a trader in New York enters into, say, an foreign exchange swap with a hedge fund, the primary broker-dealer or commercial bank is usually the initial counterparty; however, the bank can immediately assign the swap (which is sometimes done automatically) to the designated booking entity for foreign exchange swaps, as there will likely be a clause in the bank’s ISDA Master Agreement with the hedge fund that allows the bank to unilaterally assign the swap to any subsidiary that’s guaranteed by the holdco. (Sometimes traders execute the trade directly on behalf of the booking entity, although for reasons having a lot to do with excessive laziness and/or ignorance at the trader level, this used to happen a lot less than it should have.)

(Yes, I’m ignoring things like intermediate holding companies and captive issuers in this org chart, but they’re not important for our purposes.)

Booking entities are generally funded by a combination of intercompany repos and securities loans with the main commercial bank and broker-dealer, and unsecured loans from the holdco. So, for example, if a client of the US broker-dealer wants to borrow a security, the broker-dealer might actually enter into back-to-back repos with the booking entity and the client — in effect, the broker-dealer would borrow the security from the booking entity and then lend it to the client. The bank’s primary entities engage in significant intercompany transactions with each other as well. For example, one standard practice at large banks is “local settlement,” the main result of which is that the US commercial bank or broker-dealer clears and settles US securities trades for the European broker-dealer, and the European broker-dealer clears and settles European trades for the US entities.

The reason all of this is important is that the major banks, and particularly the major broker-dealers, are all responsible for customer assets in the hundreds of billions (out of proportion to the banks’ total assets, which is a truly silly metric for systemic risk). The fact that the banks are constantly shifting assets — some of which are the bank’s, some of which are client assets that have been rehypothecated — among their primary entities, booking entities, and holding company makes it difficult, when a major bank fails, to quickly sort out which entity controls which assets, and what rights each party has in those assets. And sorting this out is a precondition to a successful resolution of the bank.


Now, it’s easy to overstate the complexity of all this. There are a lot of moving parts, but 99% of the cash flows follow a standard protocol, and are repeated hundreds of times a day. It’s even easier to overstate the difficulty of dealing with the failure of one of the major banks, given these complicated interrelationships. Will resolving a major bank be a difficult task? Of course. Can it be done? Absolutely.

What’s important to understand — and I don’t think enough people do understand this — is that one of the biggest reasons that Lehman’s failure was such a catastrophic event for the markets was Lehman’s complete and total lack of preparation for a bankruptcy filing. I know this isn’t what most people want to hear or believe, since it doesn’t fit neatly into any of the popular narratives, and fixing it wouldn’t be gut-wrenchingly painful for the banks. But it’s the truth — and if you ask people who worked through the Lehman failure, they’ll all tell you the same thing: the complete lack of preparation was devastating.

It was honestly as if no one in the entire Lehman organization had thought about what they’d need to do in the event of a bankruptcy filing until the moment they filed. As a result, it was the most chaotic, uncontrolled failure imaginable.

To take one example: Lehman’s holding company (LBHI) filed for bankruptcy, but at the last minute its US broker-dealer (LBI) was kept out of bankruptcy by the NY Fed. The problem was that no one knew about this — most people thought LBI had filed too. Lehman had all sorts of problems getting employees to even show up for work; JPMorgan, which was LBI’s clearing bank, unilaterally shut off LBI’s access to its accounts for several days, and actually started seizing assets of LBI’s prime brokerage clients (a huge no-no); clearinghouses improperly limited LBI’s trading activity; the NSCC mistakenly seized a large amount of LBI’s customer securities; Lehman’s European broker-dealer (LBIE) stopped payments to LBI’s omnibus account even though LBI continued to make payments to LBIE; incoming customer securities to LBI weren’t getting properly segregated; counterparties simply stopped posting collateral they owed on OTC derivatives with LBI; and so on. That first week, the biggest challenge was simply getting someone at Lehman on the phone. (I saw a 63-year-old senior partner do a fist-pump you’d have to see to believe when he finally got an account executive at Lehman on the phone. Unquestionably the highlight of my week.)

You get the picture: it was utter chaos, in no small part due to sheer confusion about what was going on. Another crucial consequence of Lehman’s lack of preparation was that there was no sustained effort on Lehman’s part prior to bankruptcy to make sure that client assets were in the right place to effect a smooth transition at the time of the bankruptcy filing. This led to one of the biggest problems of the entire episode, when a large number of US hedge funds discovered, to their surprise, that their assets had been transferred to LBIE, and had been frozen when LBIE filed for insolvency. The LBIE debacle effectively caused about $40bn to unexpectedly disappear from the market at the worst possible time.

What all this shows is that the failure of a major bank is not some deep-seated, inherently insurmountable problem, as some people would have you believe. By contrast, it shows that a well-planned and well-executed resolution would have avoided many of the most significant problems that arose in the Lehman bankruptcy.

And tomorrow, I’ll explain how this can be done under the Dodd-Frank resolution authority.

I wish I could take credit for this, but I can't. A colleague came into my office today with a proposed enforcement mechanism for the Volcker rule, and when I opened my mouth to disagree with him, nothing came out. I just laughed.

The proposal: If a trader is found by regulators to have violated the Volcker rule more than once in a calendar year, his pay cannot be more than $100K for that year. No exceptions. No million dollar bonuses. If the trader has already received more than $100K in base salary for the year, then he has to pay back any amount over $100K. Believe me, once a trader has violated the Volcker rule once, he'll be falling all over himself to document why and how every trade is related to market-making.

I like it: short, sweet, and devastatingly effective. It uses deterrence to compensate for the inevitable inadequacies — that is, over-permissiveness — of the final rule.