Friday, March 27, 2009

What Should Be Happening to Toxic Bond Prices?

On by Felix Salmon:

Consider three assets. Asset A is a basket of subprime mortgage-backed bonds, sitting on the balance sheet of JP Morgan Chase. Asset B is an identical basket of subprime mortgage-backed bonds, being traded in the secondary market. And Asset C is a credit default swap written on that basket of subprime mortgage-backed bonds.

The Geithner bank bailout plan is released. What would you expect to happen to the prices of Asset B and Asset C?

If you're Krishna Guha of the Financial Times, you'd reason something along these lines: Thanks to the availability of cheap government funding to buy it, Asset A will rise in price. Since Asset B is identical to Asset A, then Asset B will rise in price too. And since the spreads on Assets A and B have both tightened, then the spreads on a CDS written on that asset must be tightening too, thanks to the CDS-cash arbitrage. So expect the CDS spread to narrow, just as the yield on Assets A and B has fallen.

Well, guess what. If you look at the market, Asset B has not risen in price, and the spread on Asset C has not tightened. And Krishna Guha is worried:

The plan's modest impact on toxic asset prices raises questions as to the sustainability of the rally in bank stocks. It is a reminder that even this plan, which most experts believe is well crafted, may not work.

I, on the other hand, would not expect the price of Asset B to rise, nor the spread on Asset C to fall.

After all, Asset B is likely not eligible to be purchased as part of the Geithner bank bailout plan. So why should its price rise? If anything, one would expect its price to fall: investors holding Asset B and who are eligible to get funding to buy Asset A are likely to dump Asset B today, in anticipation of buying Asset A tomorrow. Now there might be a bid for Asset B from banks hopeful that they will then be able to turn around and flip it in the government auction, but somehow I doubt that's a trade many banks would get particularly excited about these days -- although there are always exceptions.

As for Asset C, the availability of cheap funding to buy Asset A has no effect whatsoever on the underlying default probabilities for that particular basket of subprime mortgage-backed bonds -- so you wouldn't expect CDS prices or spreads to move at all. If anything, you'd expect the CDS spraeds to widen, since investors who are about to buy lots of subprime assets might well want to start hedging that position in the CDS market, giving themselves some downside protection.

Yet Guha is perfectly happy quoting ABX and LCDX prices -- which are credit default swap indices -- as indicia of what's happening to toxic-asset prices.

The real problem here is that financial journalists find it pretty much impossible to get out of the no-arbitrage mindset: we're used to living in a world where a thousand hedge funds descend on any possible arbitrage and close it in milliseconds. But that's not today's world, not by a long shot. We live instead in a world of massive CDS-bond basis spreads and many other arbitrages too.

And of course, there's the narrower problem that journalists -- including Guha's colleague Gillian Tett, whose book I'm reading now -- have a habit of looking at the level of the ABX and using that as the actual trading level of subprime mortgage-backed bonds, in cents on the dollar. It isn't.

So never mind all of Guha's elaborate theories about how the lack of price action in the ABX, in the wake of Geithner's announcement, might mean that there isn't as much of an illiquidity premium as policymakers seem to think, or that the Geithner plan simply isn't big enough. Both of those things might be true -- but there's no way you can possibly deduce them from looking at bond and CDS prices.

No comments: