Clearing up Miller-Moore

By Felix Salmon
November 24, 2009
Andrew Ross Sorkin's column today is not far from that of Yves Smith: he seems to have been captured by the Fed (and/or Wall Street more generally), and is decidedly uncharitable when it comes to the motives of legislators trying to fix a broken system. Yes, there are unintended consequences to any legislation. But as we've seen, the unintended consequences of the lack of decent regulatory oversight are much, much worse. Virtually anything would be an improvement on what we've got right now -- and that key fact is heavily obscured in the column.

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

Many thanks to Brad Miller, one of the co-sponsors of the Miller-Moore amendment, who has been doing the rounds of the blogs trying to explain what it does and doesn’t do. He left a comment on my blog on Sunday, and another on Yves Smith’s today; I then spent a good chunk of this morning on the phone to him, nailing down the thinking behind the amendment.

Conceptually, what’s happening here is that the FDIC is getting resolution authority not just for the community banks and thrifts which it’s been overseeing up until now, but also for systemically-important financial institutions like Goldman Sachs or AIG. It’s important for the FDIC to have a certain degree of freedom to maneuver — but a lot of these firms, especially when things start falling apart at the very end, find themselves pledging substantially all of their assets in a desperate search for liquidity at any price. “What precipitated the final collapse of AIG was a collateral call,” says Miller. “It complicates life for the FDIC to walk in to a firm which has all of its assets pledged as collateral.”

When a firm like AIG fails, its list of creditors on the date of failure can look very, very different to its list of creditors a few weeks earlier — because all of its assets have been pledged to last-minute emergency funders. Unsecured creditors who took solace in the large number of assets on the firm’s books can wind up with little or nothing as a result; and the people who get paid out in full are people who were most aggressive in seizing collateral in the final days and weeks.

The Miller-Moore amendment seeks to make it more dangerous for those last-minute secured creditors to jump the queue and get first dibs on a failing firm’s assets. In doing so, it gives the FDIC the ability to seize some of those assets itself, to minimize the hit to taxpayers. As Yves Smith says,

Creditors ARE risk capital, if they made bad decisions, they should take their lumps before innocent bystanders like taxpayers. That is a perfectly sensible idea.

But neither Yves nor Andrew Ross Sorkin nor just about anybody else seems to be able to get a grip on what exactly the Miller-Moore amendment does, despite the fact that it’s all of 160 words long. Here’s Smith:

One can argue that the bill is too liberal. It allows loans to secured creditors to be cut only by 20%. What if the collateral is worth only 50% of its face value? Even a 20% reduction would be too low.

This isn’t true — in fact secured debt is only secured to the value of the collateral. That’s perfectly standard, and it’s in the original bill, and therefore doesn’t need to be amended by Miller-Moore.

Meanwhile, Andrew Ross Sorkin today recapitulates the Barclays fallacy — indeed, he approvingly quotes the Barclays research saying that the amendment could turn secured loans into “an unsecured loan at 80 percent of the original amount”. But that’s not what the amendment states.

My reading of the amendment was that a 100% secured loan will get converted into a loan which is a combination of 80% secured and 20% unsecured, with the unsecured debt being treated as being pari passu with all the bank’s other unsecured debt.

But Brad Miller, in his comment on my blog, said that

Secured creditors would lose nothing until shareholders and unsecured creditors had lost everything.

Is the 20% of the formerly-secured debt which is now unsecured pari passu with old unsecured debt? Or is it actually senior? (Either way, the Barclays/Sorkin reading is massively off-base.) Miller says that in practice the old senior creditors will remain senior, since the FDIC won’t impose the haircut unless and until unsecured bondholders are being wiped out. “The way the bill as amended now works, that 20% is treated as unsecured but as a practical matter it would still be senior to wholly unsecured debt,” he says.

He adds, in his comment:

The haircut provision wouldn’t apply when a systemically significant financial firm teeters over into insolvency, it will only apply in spectacular, catastrophic collapses. Think of the Hindenberg.

It wouldn’t take a lot of underwriting to see something like that coming. The most likely candidates for the haircut would be existing creditors who demanded more and more collateral as the firm collapsed, jumping queue and exacerbating the liquidity crisis at the collapsing firm.

Also, the haircut would be discretionary. The FDIC could allow secured creditors who had taken collateral when the firm was solvent their full security interest, and just impose the haircut on creditors who had grabbed collateral from a firm that should already have been in receivorship.

What Miller’s talking about, here, is a system whereby haircuts are imposed on secured creditors on a case-by-case basis at the discretion of the FDIC. Some secured creditors (he cites as an example “existing creditors who demanded more and more collateral as the firm collapsed, jumping queue and exacerbating the liquidity crisis at the collapsing firm”) get given haircuts; others (such as “secured creditors who had taken collateral when the firm was solvent”) get no haircut at all.

There’s an echo, here, of the silly attempt by the likes of Maxine Waters to carve out an exception for vulture funds, saying that they’re worse than other creditors and therefore should have fewer rights. But the difference in this case is that we’re not talking about tradeable, fungible secured bonds: instead we’re talking about bilateral contracts between a failing bank and a group of creditors scrambling to secure as much collateral as they can.

Clearly the FDIC, if it were so inclined, could allow all repo transactions to remain unscathed — thereby addressing at least part of the objection of my colleague Agnes Crane. The key question is whether the markets would believe that the FDIC would keep repos untouched — and that’s a question which would probably have to be handled by the head of the FDIC at the time. Since the FDIC has every interest in preventing runs on banks, my guess is that the FDIC would make it clear that it would not seek to impose haircuts on repos.

But at the same time Miller does want to reduce banks’ reliance on the repo market. “What we are talking about is short-term secured debt,” he says. “We don’t want systemically significant firms to be relying on that kind of financing.” And in that sense Agnes is right to be worried about the effects of this amendment on the repo market: it probably would become smaller and less liquid.

Is that a bad thing? I’m not sure that it is. A large and liquid repo market means a large and liquid market in Treasury securities — making those Treasuries even more of a safe haven in times of crisis. One of the big problems in crises is the flight-to-quality trade, where people sell any kind of risk assets that they own, and buy instead the safest and most liquid securities they can find, invariably Treasuries. The flight to quality is inherently destabilizing, and sends the correlations on all risk assets soaring unhelpfully towards 1. If the Treasury market were slightly less liquid and attractive, that might be no bad thing: we want anybody with a large sum of money to be OK with taking a small amount of risk, rather than panicking and looking for something entirely risk-free.

What’s more, a move from secured lending and repo-market funding to good old-fashioned interbank lending is something to be encouraged, since it forces people lending money to banks to actually do some underwriting first. The Federal Home Loan Banks opposed the Miller-Moore amendment, since they blithely advance billions of dollars in secured financing to banks around the country. That’s not a smart use of taxpayer funds, and it’s a great idea to force them to do some very basic underwriting first.

More generally, my feeling about the Miller-Moore amendment, and the other pieces of legislation being discussed in Sorkin’s column today, is that it’s all well and good to worry a little about unintended consequences, as he does at length. But against that it’s also imperative to bear in mind that the unintended consequences of the lack of decent regulatory oversight are much, much worse. Virtually anything would be an improvement on what we’ve got right now — and that key fact is heavily obscured in the column. The Miller-Moore amendment makes the world a better place — even (perhaps especially) if it does end up damaging the repo market. So let’s be a bit more constructive about it.

7 comments

Comments are closed.