Between the Lines

If you were looking for a proof that the IMF needs to revamp its communication tactics, and/or that (some) journalists need to learn how to read, you need to go no further than the Fund’s $4.1 trillion estimate of the global losses in the financial sector.

The number sounds staggering, even compared to the IMF’s own estimate back in October 2008, which had losses on US-originated loans and securities at roughly half the size of the current estimate. Which is why I find it all the more puzzling that the Fund chose to flash it on the front page of its WEO/GFSR introduction, just like that, without any qualifiers.

Starting with what the $4.1 trillion is really about… or what it’s not. So, it is not the net cost of the collapse of the global financial sector. In estimating writedowns on securities, the Fund used current market prices and spreads of relevant credit market indices. But according to the going wisdom (including the IMF’s), these market prices overstate the actual default rates that we are likely to see in the underlying loans, due to “abnormal” funding constraints faced by market participants.

The IMF’s own numbers reinforce this point. For example, market-implied loss rates on US-originated commercial mortgage securities are shown to be 35% on average. This compares to the Fund’s own estimate of 9.8% for commercial mortgage loans, which was derived based on a model driven by fundamental factors such as lending standards and real private consumption.

In other words, the writedowns for the securities (almost $2 trillion out of the $4.1 trillion) do not reflect the “fair” (fundamentals-driven) value of the securities, which would be realized once the securities mature. In fact, in what is admittedly a very crude analysis, the application of the Fund’s own estimated loss rates on the securities portfolios would reduce the projected writedowns by $830 billion.

Journalists’ headlines regardless, the Fund never actually said that the $4.1 trillion is the estimated net loss to the financial system. The purpose of the exercise was a whole different one: To use the results as an input for deriving the capital hole in the banks of the major global financial centers.

So what do we get? Well, for starts, the $4.1 trillion reflects the potential writedowns for the entire financial system, including hedge funds, pension funds, etc, i.e. not just banks. The relevant number for banks is $2.47 trillion globally.

So far so good. Then we get the “Bank Equity Requirement Analysis” (here), which breaks down the capital hole by region. The hole, of course, depends on (a) the losses each country or region will likely experience; and (b) the existing capital cushion (if any) that banks have to absorb these losses.

Here things become a bit unclear. First, the numbers don’t exactly add up. I mean, where is Japan?? Somehow, the Fund suggests that the entire $2.47 trillion of writedowns will be borne by US and European banks—i.e. that Japanese banks had the foresight to avoid every bit of toxic asset originated elsewhere, and even to pass on the entire stock of their own toxic loan and securities to those American and European fools. Impressive, if one believed it…!

Importantly, we have the striking difference between the capital ratios of Eurozone and US banks—with the former looking much more unfavorable than the latter, even before the bulk of the writedowns begin. But it’s a well-known fact that different accounting standards in Europe and the US (IFRS vs. GAAP) tend to overstate the leverage ratios of European banks, compared to what would have been under GAAP, because IFRS treats derivatives on a gross basis rather than net. I am not saying that European banks are hale and hearty—only that the numbers used by the Fund are not comparable and, as such, are likely the wrong starting point for calculating the capital hole.

I don’t want to downplay the cost of the financial crisis which, however one looks at it, is astounding and grossly unfair to those who played by the rules. But we are in the middle of the stress test season, and catchy numbers can become counter-productive if interpreted by a hasty pair of eyes.

My own pair of eyes see a simpler message.. or two. One is that we have a long way to go in restoring capital adequacy, though perhaps not as long as the Fund suggests, given the caveats above. Another is that estimating the capital hole involves as much art as science. So let’s hope markets don’t find the art too abstract for their taste…


Originally published at the Models & Agents blog and reproduced here with the author’s permission.