Archive for September 15, 2010

Mounting Tension

September 15, 2010 Leave a comment

There was more odd market behavior today, but at least it was odd in a different way. Stocks rallied 0.4% while Treasury yields backed up to 2.72% (10y note), so the normal relationship resumed for at least a day. Moreover, the dollar rallied – sort of. It was unchanged against the Euro, but 2 big figures weaker against the Yen as the Bank of Japan intervened to buy as much as $20bln versus yen (Reuters story here). Japan has become concerned at the increasing strength of its currency against the dollar (near all-time highs), partly since a strengthening domestic currency helps to lower domestic inflation…which is just about the last thing that Japan needs. Word was that the sale of yen wasn’t sterilized, implying that the BOJ also was adding to its money supply.

I don’t know a lot about currency intervention, but here is what I do know and I think it’s pretty close to all that I need to know: intervention tends to work only when the market was overextended anyway relative to the fundamentals, not merely because it is inconvenient for the central bankers. If Japan’s currency is strengthening relative to the dollar, it is partly because the US appears to be moving towards quantitative easing with much more determination than Japan ever showed. If the BOJ wants to weaken the yen, it should print a lot more of them. If the U.S. increases the supply of dollars, relative to other currencies, appreciably, then its relative price will decrease (this is one reason that we ought to be wary of concerted QE, since not many countries likely want to see their currencies strengthen a whole lot against the buck…especially when they own trillions of dollars).

The unusual move today was in the VIX index, which despite recent low volumes and fairly uninspiring ranges (and higher equity prices, which tends to move us down the skew) rose abruptly in the morning and ended 0.5 higher at 22.10. What caused the bid for vol today? Why is the tension mounting? Was it the currency intervention in Japan? Was it the fact that the EU is moving to ban all naked short-selling, allow national regulators to ban all short-selling for temporary periods, and require mandatory reporting of “short” positions to a central database (why not longs)? (Link to story.) Can you imagine the carnage if levered long-short portfolios are forced to unwind? Covering the shorts will require covering the matched longs as well!

Changes like this, no doubt, create additional risk in the marketplace, just as the long-standing crusade against market-makers (see yesterday’s comment) drives away liquidity to the same effect. I really have no idea what these politicians are thinking, but perhaps the problem is that I am trying to use “politicians” and “thinking” in the same sentence.

If implied volatility continues to rise, especially if it rises in the context of another deflationary financial shudder, one player who will be unhappy is Pimco. An article in Bloomberg today revealed that in a regulatory filing Pimco’s mutual funds sold $8.1bln “deflation floors.” This is an OTC structure that is designed to mimic the embedded principal floor in TIPS. A 10-year deflation floor would have a one-time payment of Notional * max(0, endCPIindex / beginCPIindex – 1). That is, if the price level is lower in 10 years’ time, Pimco will owe money to the buyer of this protection; in exchange, Pimco receives an up-front premium.

It is usually a bad idea to sell options on highly-unlikely events, because it is extremely hard to evaluate the true probability of those events. Pimco is convinced, as am I, that it is very unlikely that prices will fall on balance for a whole decade. Pimco is also convinced, as am I, that the price of the option is too high relative to the expected value of the protection they are selling. However, unlike Pimco I’d be extremely reticent to sell this option, for two reasons. One is that when I was an OTC options trader I learned a very important lesson from a more-experienced options trader. The rule was “Never be a weenie and sell a teeny.” (A teeny is 1/64th of 1%, the lowest price at which you used to be able to sell an exchange-traded option. They regularly trade when there is essentially no chance of the outcome in question.) The point was that you need to be unlucky only once for your losses to amount to huge multiples of your all-time cumulative gains. Moreover, we really don’t have a good feel for what the tails of the distribution…where you’re playing, if you sell this option…look like because they are so unlikely we can’t ever have enough history to evaluate that probability. Unless, of course, you assume lognormality; if you are tempted to do so, I recommend reading either of Nassim Taleb’s books, Fooled by Randomness: The Hidden Role of Chance in Life and in the Markets or The Black Swan: Second Edition: The Impact of the Highly Improbable.

The other reason I wouldn’t do it is that the seller of these options can’t just “fire and forget.” If implied volatility doubles tomorrow, Pimco will show to its investors a large mark-to-market loss, and if the investors all flee there is no way for the firm to buy back this exposure – the market isn’t that well-developed. If another financial calamity hits, which is hardly a scenario we can discount completely, then a sharp move lower in inflation expectations, combined with a sharp move higher in volatility, would result in a mark-to-market loss far in excess of any likely eventual loss Pimco may have. I think that if I was a betting man I might make the same bet Pimco made, but I’m an investing man and I think this risk has too many downsides. But perhaps they have something on the other side…

Today’s economic data were ho-hum, and tomorrow’s PPI (Consensus: +0.3%, +0.1% ex-food-and-energy) doesn’t really matter. But Initial Claims (Consensus: 459k) ought to be fun. Last week’s number, remember, plunged to 451k; this happened coincidentally at the same time that eight states were unable to get their actual claims into the BLS and those numbers had to be estimated. For some reason, the estimate showed a sharp improvement in the labor market. It is really odd to me that economists, in forecasting a bounce only to 459k, are essentially giving that number full credibility. Before last week, the 4-week moving average was 487k and the 8-week average was 473k. Where is the evidence of a marked improvement? I may be eating my words, but I would expect something more like 475k with a sharp upward revision to the prior week (unless, that is, those states are still behind on their paperwork). The highest estimate recorded on Bloomberg of the 42 economists who forecast the number is 476k, and quite a number of economists are projecting in the 435k-445karea. That would be in the neighborhood of the best prints since 2008. I just don’t see the evidence of such a sharp improvement in the labor market.

Categories: Economy, Liquidity