Felix Salmon

I missed this on Thursday:

Standard & Poor’s will stop rating new bonds composed of U.S. second mortgages, saying it’s too hard to assess the debt while the housing slump continues…

The unit of McGraw-Hill Cos. said today that it will continue to assess outstanding securities by looking at delinquency and default levels.

This raises more questions than it answers. Firstly, how many bonds does this affect? Remember it only applies to new bonds composed of closed-end second liens – S&P is continuing to rate bonds comprising Helocs. Are there any such new bonds to rate?

And secondly, isn’t this an admission that the mechanism for rating new bonds is broken, and has been for a while? Given that mechanism is based on a now-discredited model, and given that the wave of defaults on second-lien mortgages is very recent and largely unprecedented, how on earth can S&P consider its existing ratings on outstanding second-lien-backed bonds to be remotely reliable? If there’s no way of knowing how creditworthy a new second-lien-backed bond is, there’s no way of knowing how creditworthy an old one is, either.

Finally, if and when the market in these bonds ever reappears, what are the chances that S&P will start rating it again? I’d wager they’re pretty high. It’s all well and good proclaiming that you won’t eat cake so long as no one’s serving it to you; the real test is when you go to the birthday party and everybody else is stuffing their face.

(HT: MM via SAR)

If Steve Waldman could only dumb his prose down a bit, he’d be perfect for Slate’s new business site: he’s great for compellingly contrarian takes on the news. Today, he takes a shot at liquidity, saying that it’s not necessarily all it’s cracked up to be.

There is, in some sense, a "right" level of liquidity, defined by the uncertainty surrounding the present value of an asset’s future payoffs. We laud markets for "price discovery", their ability to distill complex economic facts into simple prices that put a value to unknowable future events. But we need markets to communicate the uncertainty surrounding those valuations as well. The depth-weighted spreads of assets whose values are nearly certain should be much narrower than those of assets whose payoffs cannot be accurately predicted. When that is not the case, it represents a market failure. The recently wide spreads on complex structured credits are not the crisis — those spreads accurately reflect the uncertainty surrounding what the instruments are actually worth. Nobody knows, so spreads should be wide. The real crisis was two years ago, when "oceans of liquidity" meant that whatever the underlying value of a thing, you could sell it quickly for near what you bought it, so spreads grew artificially narrow.

Clever. But wrong, on two pretty fundamental counts.

Firstly, there is such a thing as a pure liquidity premium, unrelated to the uncertainty surrounding future payoffs. Rajeev Misra of Deutsche Bank talked about it last week, at the Milken conference: he gave the example of credits where the cash bonds were trading 70-80bp wide of the credit default swaps. The uncertainty surrounding future payoffs is if anything greater with the CDS, since there’s extra counterparty risk there, but the liquidity of the derivative more than makes up for it.

Even more fundamentally, Steve seems to be under the misapprehension that there’s some kind of strong correlation between the degree of certainty surrounding a security’s future payoffs, on the one hand, and the bid-offer spead on that security, on the other. There isn’t. You want a security with high degree of certainty and wide bid-offer spreads? I give you no shortage of triple-A rated mortgage-backed securities, carefully structured so that there was no uncertainty in their payment streams. Even the Bank of England says they’re perfectly safe, but there’s simply no market in them.

Or you want a security with an enormous amount of uncertainty regarding future payoffs, but super-narrow bid-offer spreads? Let me point you to the equity market in general, and large-cap technology stocks in particular.

The mechanism of price discovery, in and of itself, does a very good job at judging uncertainty in future payoffs. A bond with high uncertainty regarding future payoffs will trade at a lower price than a bond with low uncertainty, because in the bond market risks are asymmetical. As the price falls, the yield rises to the point at which it becomes attractive enough to take on that extra risk. None of this has anything to do with bid-offer spreads.

To put it another way, I’m basically with Steve when the "depth-weighted spreads" he talks about are spreads over Treasury bonds. But he then subtly changes his tune and talks about spreads as though they’re a measure of whether or not you can sell a security quickly for more or less the same price you bought it at. That’s a bid-offer spread, not a spread over Treasuries. And while the bid-offer spread is indeed a measure of liquidity, it’s not something useful.