A (little bit) further evidence on the dollar carry trade

Heiko Hesse, using an econometric tool known as Dynamic Conditional Correlation, finds that the co-movements between the dollar and several asset classes has increased lately. I don't know much about autoregressive conditional heteroskedasticity (the framework from which stems this tool), but in my understanding, this just barely confirms that 80%-90%+ correlations between the dollar and these asset classes are, as suspected, both significant and much higher than they have in the past. (Simpler analysis as in my previous post showed correlations were higher but only very likely to be statistically significant...)

What it does not, however, is state that these high (and higher than unusual) correlations are a definite proof that the dollar carry trade is a major culprit for the froth in risky asset prices. We are still looking for definitive evidence on that, and I'm afraid this is not going to come before regulators start asking banks to report the numbers.