Sunday, November 28, 2010

A Proposal for Money Market Fund Reform

There were two major areas of financial regulation that Dodd-Frank left rather conspicuously unaddressed: the GSEs, and money market funds (MMFs). The reason they omitted GSE reform is obvious: GSEs are a fiercely partisan issue, and including GSE reform could easily have — and, I think, almost certainly would have — killed the entire financial reform package. (Yes, the GSEs are that toxic.)

There were two reasons the administration omitted MMF reform: one, the SEC was already in the process of adopting substantial new regulations for MMFs, and two, there was nothing approaching a consensus on MMF reform. With other areas of financial reform, there was generally broad consensus on what needed to be done: an FDIC-like resolution authority, a systemic risk regulator, central clearing for standardized derivatives, a CFPB, etc. But with MMFs, no one really knew what to do yet (myself included). So the administration directed the President's Working Group on Financial Markets (PWG), which consists of the Treasury Secretary, and the Chairmen of the Fed, SEC, and CFTC, to prepare a report on MMF reform options. I give the administration credit for acknowledging that they simply didn't know what to do about MMFs yet.

The PWG released its report on MMF reform options last month. (The author of the report seems to share my title-writing abilities, as the report is creatively titled, "Money Market Fund Reform Options.") The report presents a menu of reform options, some better than others.

I want to endorse one reform option in particular: option (f), which proposes a two-tier system of MMFs, with stable NAV MMFs reserved for retail investors, and institutional investors limited to floating NAV MMFs. One of the main problems was that MMFs' ability to use "stable" NAVs — that is, to round their NAVs to $1 — fostered a perception among investors that MMF NAVs don't fluctuate. MMFs only have to reprice their NAVs (i.e., their share prices) if the mark-to-market per-share value of the fund's assets (known as its "shadow NAV") falls more than 0.5%, and doing so is known as "breaking the buck." (In MMF-land, 0.5% is a substantial loss.) Investors were blissfully unaware of any losses that were smaller than 0.5%.

As the PWG report notes: "By making gains and losses a regular occurrence, as they are in other mutual funds, a floating NAV could alter investor expectations and make clear that MMFs are not risk-free vehicles. Thus, investors might become more accustomed to and tolerant of NAV fluctuations and less prone to sudden, destabilizing reactions in the face of even modest losses."

Reinforcing the idea that MMFs are not risk-free is, I think, crucial. So why not move all MMFs to floating NAVs? Why only require institutional investors to use floating NAV funds? Well, for one thing, institutional investors were the problem in the financial crisis. I can't emphasize this strongly enough. The problem wasn't those supposedly-flighty, oh-so-irrational retail investors. It was the institutional investors. During the week of September 15, 2008, outflows from prime MMFs totaled roughly $300 billion, according to the ICI. Institutional investors accounted for over 90% of those redemptions. What's more, institutional MMFs have always had more volatile cash flows than retail MMFs. According to the ICI, between July 2007 and August 2008 — that is, from roughy the time the institutional investor-dominated ABCP market started blowing up to the time Lehman failed — MMFs received inflows of roughly $800 billion, over 80% of which, or $650 billion, came from institutional investors. This is, as they say, "hot money," and it comes from institutional investors.

In short, it was the institutional investors who panicked and fled the ABCP market in 2007, who panicked en masse again in September 2008, and who are most likely to panic again if one of their stable NAV MMFs "breaks the buck." They are the ones who need to be restricted, not the retail investors. Institutional investors are also the ones with the resources to closely monitor floating NAVs. They don't need to be protected by a stable NAV, nor should they.

One of the main arguments against limiting institutional investors to floating NAVs, which I find utterly unconvincing, is that internal investment policies may prohibit some institutional investors from investing in floating NAV funds. OK, then change your internal investment policies. It's not the SEC's job to make sure that its regulations don't conflict with investors' internal policies. If your internal investment policies conflict with the SEC's regulations, then that's your problem. Switching to a floating NAV doesn't make an MMF any more risky; the assets are still exactly the same. So if an investor's internal policies deem stable NAV MMFs to be appropriate investments, then there's absolutely no reason why the investor should be precluded from investing in floating NAV MMFs.

The biggest unsolved problem with MMFs is still discretionary sponsor support. MMF sugar-daddies sponsors have repeatedly bailed out their MMFs, thus preventing them from breaking the buck. For example, banks like BofA and Wachovia bailed out their MMFs that had invested in SIVs and ABCP in 2007, and JPMorgan, Legg Mason, and Northwestern Mutual all bailed out their MMFs in September 2008. In fact, after Lehman failed, pretty much every major MMF sponsor had to bail out their MMFs. As the PWG report notes, "more than 100 MMFs received sponsor capital support in 2007 and 2008 because of investments in securities that lost value and because of the run on MMFs in September and October 2008." (Guess who didn't have a sugar-daddy sponsor? That's right, the Reserve Primary Fund!) For investors, the repeated bailouts from MMF sponsors have clearly created an expectation of sponsor support. This needs to be dealt with as well, though the solution is not straight-forward.

Probably the best way to deal with discretionary sponsor support is for sponsors' primary regulators to impose punitive (and automatic) penalties on any form of sponsor support. For example, punitive capital charges, or a 12-month moratorium on dividends. Something like that.

As I discussed in my previous post, the Dodd-Frank Act subjects "systemically important" nonbank financial firms to enhanced prudential standards and consolidated supervision by the Fed. All bank holding companies with over $50bn in assets — of which there are currently 36 — are also subject to the heightened Fed supervision. The enhanced prudential standards have to include capital requirements, leverage limits, liquidity requirements, and an ongoing resolution plan (or "living will"), among other requirements. So you can see why being designated systemically important would be a big deal for nonbank financial firms (which, for the sake of simplicity, I'll refer to in this post as simply "firms"). The Financial Stability Oversight Council (FSOC) is required to designate systemically important firms, and it has already gotten the ball rolling on this process.

 So what makes a firm systemically important? Section 113 of Dodd-Frank deems firms systemically important if either (1) "material financial distress" at the firm would pose a threat to financial stability, or (2) the firm's ongoing activities pose a threat to financial stability. The "material financial distress" option is clearly the more important of the two, as it's difficult to imagine a firm whose ongoing activities were systemically important, but whose failure wouldn't pose systemic risks.1

Section 113(a)(2) requires the FSOC to consider 10 specific factors in determining whether a firm is systemically important:

(A) the extent of the leverage of the company;
(B) the extent and nature of the off-balance-sheet exposures of the company;
(C) the extent and nature of the transactions and relationships of the company with other significant nonbank financial companies and significant bank holding companies;
(D) the importance of the company as a source of credit for households, businesses, and State and local governments and as a source of liquidity for the United States financial system;
(E) the importance of the company as a source of credit for low-income, minority, or underserved communities, and the impact that the failure of such company would have on the availability of credit in such communities;
(F) the extent to which assets are managed rather than owned by the company, and the extent to which ownership of assets under management is diffuse;
(G) the nature, scope, size, scale, concentration, interconnectedness, and mix of the activities of the company;
(H) the degree to which the company is already regulated by 1 or more primary financial regulatory agencies;
(I) the amount and nature of the financial assets of the company;
(J) the amount and types of the liabilities of the company, including the degree of reliance on short-term funding; and
(K) any other risk-related factors that the Council deems appropriate.


As I noted in my previous post, a lot of firms, including hedge funds, are arguing that leverage should be the most important factor. That argument has a superficial appeal to it — after all, everyone knows that leverage was one of the main causes of the financial crisis — but it's still misguided. A highly leveraged firm may be more likely to fail, but the point isn't to identify the firms that are most likely to fail. The point is to identify the firms whose failure would cause the most damage to the financial system. The FSOC should just assume in its analysis that the firm it's examining has already failed — then ask, "What would happen, and how bad would it be?"

How likely a firm is to fail should play almost no role in the FSOC's decision — we don't want the Fed to only regulate systemically important firms that the FSOC thinks have a decent chance of failing. In 2004, almost no one would have thought that an AIG failure was in the realm of possibility, and if the FSOC had been making this determination in 2004, I can easily imagine them giving AIG a pass based on its sterling reputation and perceived invincibility. We don't want to give firms like AIG-circa-2004 the chance to go to the FSOC and push the argument that, "Sure, our failure would be catastrophic, but we're such a strong company that the chances of us failing are practically nil. It's just inconceivable. [Do you know who I am? I'm Hank f*cking Greenberg!]" That argument should be per se illegitimate. (And you're crazy if you think that's not exactly what Hank Greenberg wouldve been saying back in 2004.) Moreover, it was precisely the inconceivability of an AIG failure that caused the markets to hit the Sheer Panic Button in September 2008. Allowing firms to avoid the "systemically important" designation because the FSOC thinks they have sufficiently safe levels of leverage just sets us up for another AIG-like shock.

 Of course, leverage can also magnify the damage caused by a firm's failure, so the FSOC should definitely still consider leverage in determining whether a firm is systemically important. My point is that the FSOC shouldn't be distracted by arguments about how firms with low leverage ratios shouldn't be deemed systemically important because they're less likely to fail.


I think it's hard to overstate the importance of interconnectedness. Size is important, yes, but it's interconnectedness that causes problems to turn systemic. Systemically important interconnectedness can take multiple forms as well. For example, LTCM, which by today's standards was a relatively small hedge fund, was systemically important not because of its size, or the number of counterparties it had, but because of its proximity to the dealer banks — LTCM was
deeply enmeshed with all the dealers, both in terms of raw counterparty exposure and ridiculously correlated trading books. What makes this kind of interconnectedness systemically important is, obviously, the identity of the counterparties, and the nature of the firm's relationships with the counterparties. (Even though this clearly falls under the third factor in Section 113(a)(2), I think this factor is best thought of as a subset of "interconnectedness.")

Lehman, on the other hand, was involved in every corner of the market. The number of counterparties and customers affected by Lehman's bankruptcy was staggering, and meant that practically every market in the world was negatively impacted in some way when Lehman filed. Lehman's interconnectedness transmitted the losses — and the panic — all around the world at a shocking speed. This kind of interconnectedness is marked by the number and variety of markets (and counterparties) that a firm touches.

Proximity to Core Commercial Paper Market

A separate factor I think the FSOC should take very seriously is a firm's proximity to the commercial paper (CP) market — or, more specifically, the effect that a firm's failure would have on the CP market. One of the key lessons of the financial crisis, which is constantly overlooked, is that our domestic economy is critically dependent on the CP market. Even a temporary disruption in the CP market can cause large-scale employment losses. A lot of companies spent months recovering from the temporary breakdown in the CP market in September 2008 — and by "recovering," I mean "laying people off."

Any firm that threatens a repeat of the CP market troubles is, without question, systemically important.

 Ultimately, I expect the FSOC to designate only a handful of nonbank financial firms (i.e., in the 4-8 range) systemically important and subject to consolidated Fed supervision. AIG, GE Capital, and Prudential are no-brainers. (No, I'm not forgetting MetLife — remember, it's a bank holding company now, so it's automatically included because it has over $50bn in assets.) It'll certainly be interesting to see who else is designated systemically important.

1 I suppose a firm would fit that description if the way it conducted its day-to-day business was important to the financial system, and it was also fairly easily replaceable. Like the DTCC, if it was easily replaceable. Which it is not.

Interesting story in Dealbook about how large nonbank financial firms are arguing to the Financial Stability Oversight Committee (FSOC) that leverage should be the key consideration in determining which nonbank firms should be deemed "systemically important," and thus subject to increased regulation by the Fed. (Dodd-Frank requires the FSOC to determine whether nonbank financial firms should be subject to Fed supervision based on a list of factors in § 113(a)(2), the first of which is "the extent of the leverage of the company.")

I guess this makes sense: banks (and broker-dealers) are naturally the most highly leveraged financial institutions, so if the FSOC determines that leverage is the most important consideration, then almost every large nonbank financial firm could point to how much less leveraged they are than the banks. Obviously asset managers and private equity funds would love it if the FSOC focused on nonbank firms' leverage.

This is a particularly interesting argument for hedge funds though. As Dealbook rightly points out, hedge funds, despite their reputation, are in fact generally not very highly leveraged — most hedge funds have leverage ratios between 2 and 3. (Most hedge funds are still equity-focused funds, where leverage is much lower.) But the issue isn't whether most hedge funds are systemically important; the issue is whether any hedge funds are systemically important, and if so, how many. And there certainly are hedge funds that run with high enough leverage, and are large enough, to be considered "systemically important." Regulators seem to be focused on funds that could suffer particularly sharp reversals ("panic-inducing" reversals, if you will), so I would expect those big global macro funds to get a very close look. My hunch is that the FSOC will also look quite closely at the big quant funds (e.g., RenTech, D.E. Shaw) — because let's be honest, they're a complete mystery. And they also have complex trading relationships with the various dealer banks, which at least has the potential to make the meltdown scenario look exponentially scarier.

It's unlikely that the FSOC will adopt the kind of narrow focus on leverage that buyside seems to have in mind, but it's a good effort nonetheless. I'll have more to say about the criteria I think the FSOC should use soon.

Continuing my Volcker rule theme, the current issue of Risk has an article titled, “Dealers Confident on Volcker Exemptions,” which quotes various banking lawyers talking about how easy it will be to get around the final Volcker rule language in Dodd-Frank. From the article: (emphasis mine)

“Last month, we sat down with our counsel and after an hour of question-and-answer, the lead counsel turned to me and said ‘it is not going to be absolutely clear how the provision will work until the rules have been written, but given so much of proprietary trading has a client nexus to it, I’ll be embarrassed if I don’t manage to exempt all your activities from the rule’,” says one head of equity derivatives structuring in New York.
Early reports on the Volcker rule caused astonishment in the banking industry, with predictions of massive changes to the structure and business of certain dealers. Now lawyers have had some time to absorb the clause within Dodd-Frank, many feel the prop trading requirement will be relatively easy to negotiate.
This is exactly what I said Wall Street’s real reaction to the Merkley-Levin version of the Volcker rule would be. (“Merkley-Levin” was the amendment that the conference committee based the final Volcker rule language on.) As I wrote way back in May, when the Senate bill was still being debated: “I spent the majority of my career as a lawyer for one of the big investment banks, and my first thought after reading Merkley-Levin was: ‘Wow, this would be cake to get around.’ Wall Street is scared of the Volcker Rule, but believe me, they’re not scared of Merkley-Levin.”

To be fair, Merkley’s office fixed some of the most glaring problems that I talked about in that post, so the final language is at least an improvement over the original version of Merkley-Levin. (You’re welcome, by the way.) Unfortunately, the final language still has several major flaws, some of which I’ve discussed previously.

The Risk article talks about other problems with Merkley-Levin as well:
[The lawyers’] confidence stems from the language used in section 619. Prop trading is defined as a sale or purchase conducted as a principal on behalf of the trading account of the firm. The trading account is described as any account used for acquiring or taking positions in securities with the main purpose of selling them in the near term or the intent to resell them to profit from short-term price movements.

Lawyers say these few sentences are riddled with ambiguities. “There are many possible unintended exemptions in this definition - it is just not clear at all,” says Scott Cammarn, special counsel at Cadwalader, Wickersham & Taft in Charlotte, North Carolina. “For instance, what happens if the trade occurs outside of the trading account? The definition explicitly states the transaction must be booked in the trading account, so I would argue that if any trade is transacted in a different account, it is not prop trading.”
I don’t think the “trading account” issue is as simple as Cammarn makes it out to be, but it definitely is another potential loophole. I have no idea why Merkley and his staff decided to define prop trading by reference to a trading account, and then include the real definition of prop trading in the “trading account” definition. The real definition of prop trading under Merkley-Levin is “taking positions ... principally for the purpose of selling in the near term (or otherwise with the intent to resell in order to profit from short-term price movements).” So why not just use that as the definition of “proprietary trading”? Why force regulators to regulate accounts rather than trades? It’s bizarre, frankly, and ultimately detrimental to the overall rule.

For instance, seeing as the language in Merkley-Levin’s definition of “trading account” was lifted directly from the Form Y-9C definition of “trading account,” and the Form Y-9C definition is explicitly based on accounting treatment, it’s possible that banks could put on prop trades outside of a “trading account” as long as they agree to use a different accounting treatment for the prop trades. And what if accounting standards change? What if FAS 115 is overhauled in the future? What happens to the definition of “trading account” then?

Also, are we talking about the account where the trade originated? What if a trade is originated in the “banking book,” and then subsequently transferred to the trading book? Since the banking book is (obviously) not a “trading account,” it's not clear if the trade still prohibited. What is clear is that using this incredibly roundabout definition of prop trading was entirely unnecessary, and just created more loopholes for banks to exploit.

Now we get to the real flashpoint, which is how you distinguish “market-making-related activities” from trades done “principally for the purpose of selling in the near term (or otherwise with the intent to resell in order to profit from short-term price movements).” As the Risk article notes:
By far the biggest ambiguity centres on the use of the word ‘intent’. The rule only seems to apply if banks had the intention from the start to sell a position to profit from short-term price movements. “How on earth do you police motive? It is almost impossible. A bank could claim to enter into a five-year put option on the S&P 500 and intend to hold the contract until maturity. If the index then drops 20% in two days and the bank monetises the gain by selling the option, it is arguably not prop trading. The intent was never there to profit from a short-term price move,” says Cammarn.
This is where I disagree. Obviously traders can lie to you about their intent in putting on a trade. But there are signals which are indicative of proprietary trades, and market-making trades can be distinguished from proprietary trades by looking at those signals. For instance, was the position opened on the bid side or the offer side of the market? If the position was opened on the offer side and was legitimately related to market-making, then the trader should also be able to identify the specific risk that he was hedging.

Also, how long has the security been held in inventory? Ace Greenberg, the legendary Bear Stearns trader and former CEO, famously used to require traders on his desk to cut any position that they’d been holding for longer than, say, 30 days. I’m not saying that this should be the hard-and-fast rule for every market — every market really does differ in terms of how long the securities need to be held in inventory. I’m just saying that this is absolutely a legitimate tool, and the inevitable claims from the Street that this would be the end of the world, and would prevent them from effectively managing their risk, etc., should be ignored. This was a risk-management tool for Greenberg, who knows a little something about trading.

Unfortunately, Merkley-Levin exempts not only “market-making-related activities,” but also “risk-mitigating hedging activities,” so even if the Fed writes a perfect definition of “market-making-related activities,” most prop trades will probably still be able to fit under the “risk-mitigating hedging activities” exemption. And this is before we get to the problem of banks setting up Section 2(a)(13) “employees’ securities companies” to do their prop trading. Merkley-Levin inexplicably misses those, which creates a loophole big enough to drive a truck through.

Don’t look for these issues to be mentioned in the notice-and-comment period though; the Street doesn’t generally tip their hand there. Better to make policy arguments during the rulemaking process, and save the legal arguments for the no-action letters. But just be aware that these issues are out there, and if they’re not dealt with during the rulemaking process, then they will be exploited by the Street.

There’s been a lot of discussion recently about the Volcker rule in the popular press, and how it’s going to be implemented. Most prominently, Michael Lewis wrote a widely-read column last week about how the banks are simply shifting their prop trading to market-making desks. While I’ve said before that the poorly-drafted language of the Volcker rule will make it very difficult to enforce effectively for precisely this reason, Lewis’ column is an absolute disaster. Lewis badly misstates the law, identifies the wrong loophole (which is amazing, since there are so many actual loopholes), and generally demonstrates no understanding of the issue whatsoever.

Lewis claims that the loophole is the bill’s definition of proprietary trading as investing “as a principal.” He then goes on to claim that the government agency who must “determine precisely what the phrase means” is....the Government Accountability Office!

First of all, the GAO has absolutely nothing to do with it. The GAO doesn’t write banking regulations. It’s the Fed’s job to write the Volcker-rule regulations for the big banks. The GAO just has to do a study on proprietary trading sometime in the next 12 months, which is completely separate from the rulemaking process for the Volcker rule. Hence why the GAO spokesman clearly had no idea what Lewis was talking about when he called. (The FSOC also has to do a study on prop trading, which actually is part of the rulemaking process for the Volcker rule. Don’t ask.)

Second, the phrase “as a principal” isn’t a even loophole. Everyone understands that both proprietary and market-making trades are entered into “as a principal,” and no one has a problem with that. (That is, after all, what a market-maker does.) There’s no need for banks to try to pretend like they’re not entering into trades “as a principal,” because trading as a principal isn’t prohibited! What’s prohibied is trading as a principal “for the purpose of selling in the near term (or otherwise with the intent to resell in order to profit from short-term price movements).” If Lewis doesn’t understand this much, then he really has no business writing multiple columns about the Volcker rule.