Was MF Global brought down by an accounting play?

By Felix Salmon
November 2, 2011
Bethany McLean has a theory: that accounting helped to sink MF Global, and that the $6.3 billion long position in European debt was made "for an accounting play".

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

Bethany McLean has a theory: that accounting helped to sink MF Global, and that the $6.3 billion long position in European debt was made “for an accounting play”.

The key part is that for accounting purposes, MF Global’s filings say the transaction was treated as a sale. That means the assets and liabilities were moved off MF Global’s balance sheet, even though MF Global still bore the risk that the issuer would default; that means the exposure to sovereign debt was not included in MF Global’s calculation of value-at-risk, according to its filings. And that also means MF Global recognized a gain (or loss) on the transaction at the time of the sale. The filings do not say how much of the gain was recognized upfront. But if it were a substantial portion, then these transactions would have frontloaded the firm’s earnings. That, in turn, may have helped cover the fact that MF Global’s core business was struggling.

Moving assets and liabilities off your balance sheet to make yourself look less risky? That’s a very Lehman move, redolent of the notorious “repo 105” trades.

But I’m not convinced by this story.

Firstly, the debt of Italy, Spain, and Belgium might be getting a lot of headlines right now, but in terms of price action it’s not actually as volatile as you might think. These assets weren’t actually significantly more volatile than the rest of MF Global’s balance sheet, and so including them on the balance sheet would not have increased MF Global’s value-at-risk very much. In fact, it might have reduced it.

Now that doesn’t mean, of course, that the assets weren’t riskier than most of the rest of the balance sheet. They were. They were sovereign bonds being held to maturity, in most cases at the end of 2012. And a lot can happen between now and then. That’s what the markets were worried about, rather than any immediate mark-to-market losses on the bonds. In fact, it’s not clear that there were mark-to-market losses on the bonds. But if you’re holding a bond to maturity, that’s an inherently much riskier position than if your prop desk is holding it as part of a bond position which it can sell at any time.

Secondly, it’s extremely unlikely that MF Global recognized a gain on this transaction at the time of sale. The bank bought this debt on the open market, and then immediately put the bonds up as collateral, getting cash in return. We don’t know who lent MF Global the money, taking the bonds as collateral. But whoever it was had no reason whatsoever to value the bonds at more than their market price. So there’s really no way that MF Global could have recognized a gain on the sale: much more likely, in fact, that it would have registered a modest loss. (In a repo transaction, it doesn’t matter if you sell the bonds for less than they’re worth, since you also contract to buy them back at a pre-set price later. So long as the repurchase price is also low, it’s fine if the initial sale is done at a low price too.)

And finally, there’s no indication that taking this trade off balance sheet helped to significantly hide the size of MF Global’s assets, or to understate its actual leverage. Here are the numbers from MF Global’s quarterly filings:

mfglobal2.jpg

What I’m charting here is the total size of MF Global’s assets, in green; its equity, in blue; and the ratio between the two, in red. Obviously, the leverage ratio is ludicrously high: there are more than $40 billion of assets being supported by just $1.2 billion of equity.

But can you see, in this chart, where Jon Corzine joined the company? (It was in March 2010.) And can you see where MF Global brought the European sovereign bonds back onto its balance sheet? (It was in the final numbers, for September 2011.)

The fact is that MF Global’s leverage ratio — and its total balance sheet — was higher before Corzine arrived than it was at any time afterwards. And that when the European sovereign-bond trade was brought back onto MF Global’s balance sheet, the total size of that balance sheet actually fell.

In other words, the trade which brought down MF Global, even if it had been on the balance sheet all along, would never really have been visible on the balance sheet, or in the leverage ratio, or in the value-at-risk figures. I’m not clear on why the trade was moved off balance sheet in the first place. But I think it’s a stretch to say that accounting brought down the bank.

More From Felix Salmon
Post Felix
The Piketty pessimist
The most expensive lottery ticket in the world
The problems of HFT, Joe Stiglitz edition
Private equity math, Nuveen edition
Five explanations for Greece’s bond yield
Comments
4 comments so far

Totally agree with your analysis. I was puzzled by Bethany McLean’s theory as I was reading it. It was nicely written but quite far out.

Posted by gauss | Report as abusive

This was a clear liquidity squeeze. Someone knew MF Global was under stress and demanded additional collateral. MF Global as far as we know went as far as posting client funds as collateral to buy more time.

It does not matter if this was Lehmann like repo 105 accounting. What matters is that the firm ran larger positions than it could afford. Moving the bonds off balance sheet just made the vulnerability less visible and ultimately obscured what the firm owed and owned at any one time. Maturity matching and liquidity management are banking basics, if you fail at them you deserve to be out of business. And if you fail at them while using client funds as collateral, then you belong into jail.

Posted by Finster | Report as abusive

Well very briefly, far far too much debt. Almost always the prelude to disaster.

Posted by Chris08 | Report as abusive

Am I the only person that thinks this is Refco Part 2. Moving debt around to fool investors. And doesn’t MF’s executive team have some former Refco executives? Did we really get scammed again by the same people at the same game?

Posted by JHilt | Report as abusive
Post Your Comment

We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see http://blogs.reuters.com/fulldisclosure/2010/09/27/toward-a-more-thoughtful-conversation-on-stories/