Knowledge Problem

Michael Giberson

The list of people who agreed, after the fact, that yes there was a {internet company | real estate | … | tulip bulb} price bubble is frequently longer than the list of people who publicly announce a bubble in unequivocal terms in advance of a crash.  But here you have someone willing to stick their neck out:

“It seems quite clear to us that the (Nymex) futures market is currently part of an dollar-led asset bubble,” said Olivier Jakob of Petromatrix in Switzerland.

FT Energy Source provides some context from Jakob:

Remember the days when hurricanes and geo-political events made oil fly?

Well, according to Olivier Jakob at Petromatrix, those days — for the time being at least — should be forgotten. The correlations between the Dow, the dollar and oil are now so well set, traders simply can’t afford to ignore them.

FT Energy Source quotes from an unnamed source document (but Petromatrix produces subscriber-only reports and I assume it is from one of these; emphasis added by FT):

WTI is still not able to break away from its pure correlation to the exogenous markets of Dollar and Equities. For the last two days WTI and the Dow Futures have run an R square of 0.9 on the intraday 10 minutes and at such ratio it is just possible to beat the theme of purely trading the Dow on oil futures.

I’m not sure I’d stake too much on a two-day regression correlation.  More:

It does not make sense per se but that is the way it is and not trading that theme would only be a proposition to provide liquidity to those who are. The problem remains that the real economy works on different principles than computer games and the current asset correlation would not allow an economy recovery to materialize. At current correlations the Dow at 11′000 would translate in WTI at 100 $/bbl which will hurt consumer confidence and demand and cap the recovery.

Here we extrapolate out from our two-day correlation up to a 11,000 point Dow – a level we haven’t seen for a year and may not see for a while longer.  I’m no statistical genius, but we seem to be forecasting pretty far out of sample.  Analytically, it makes me nervous.

The bubble statement comes next:

It seems quite clear to us that the WTI futures market is currently part of a dollar-led asset bubble and irrespective of the oil fundamentals the next input that will be decisive in the direction of oil prices will be the Fed meeting of next week (November 3rd and 4th ). No action is currently expected from the Fed, but it must be also realizing the across asset bubble in formation and at one stage it will have to decide if it wants to start deflating it or letting it run at the risk of having a burst that it can not handle later on.

It isn’t quite clear to me whether the the dollar-led asset bubble conclusion hangs on any evidence more substantial that the two-day price correlation.  Color me unpersuaded (unless it turns out to be true, of course, in which case I will claim to be among the cognoscenti from the beginning on this whole new dollar-led asset bubble thing).

N.B.: I’m not disputing the value of the larger analysis from which these quotes were ripped by FT Energy Source, which I haven’t seen, just gently poking fun at the idea of trading oil futures based on two-day correlations in prices.  Since I am not a trader and not familiar with real-world trading analytics, it may be that I’m entirely off base and two-days worth of 10-minute data is great empirical stuff.