Correlations - breaking down or not?
Posted 06-12-2008 at 02:36 PM by GammaJammer
Updated 06-12-2008 at 02:41 PM by GammaJammer (add quick explanation)
Updated 06-12-2008 at 02:41 PM by GammaJammer (add quick explanation)
I have seen a few bits and pieces of research lately suggesting that key correlations are no longer functioning as well as they might. Certainly if the macro community is anything to go by this seems to be causing people pain.
With this in mind, I started doing a bit of digging tonight and usd/jpy piqued my interest. In short, one historically reasonable correlation (with the US 2s/10s spread) remains at very acceptable levels, while another (vs the VIX - front month futures in this case, but could have used cash, futs was just what I had up at the time) has almost totally broken down. Both showed their troughs (for the time I have taken) at around the same time - when subprime exploded last year. But since then, and particularly since the start of the current quarter, their fortunes have markedly diverged. Assuming the blog attracts any interest, I will possibly extend this entry and discuss my thoughts on what to do about this in a subsequent post.
edit - p.s. usdjpy is inverted in my correlation
With this in mind, I started doing a bit of digging tonight and usd/jpy piqued my interest. In short, one historically reasonable correlation (with the US 2s/10s spread) remains at very acceptable levels, while another (vs the VIX - front month futures in this case, but could have used cash, futs was just what I had up at the time) has almost totally broken down. Both showed their troughs (for the time I have taken) at around the same time - when subprime exploded last year. But since then, and particularly since the start of the current quarter, their fortunes have markedly diverged. Assuming the blog attracts any interest, I will possibly extend this entry and discuss my thoughts on what to do about this in a subsequent post.
edit - p.s. usdjpy is inverted in my correlation
Total Comments 3
Comments
|
very interesting stuff GJ, look forward to anymore you add.
|
 
|
|
|
Talking about correlations, lately a whole lot of people seem to note an inversely correlated reaction in the stock market as oil fluctuates.
But I don't think it's that simple. Oil goes up, markets go down. Yes, on some days. But these correlations have not presented them in the past (think of last year when markets were going steadily up, and oil as well). Imho, it's important to be aware that any such correlations of unrelated markets may be nothing other than a temporary situation with other fundamental meaning. This is not the case comparing the ES and the YM for example, which are and continue to be, correlated to a certain extent. I don't remember where I read the stats, but I'm pretty sure it was 80% or more. |
|
|
|
|
Fair points FW, but some of these intermarket correlations are important for two reasons; 1) They are hostorically strong under a wide variety of market conditions. and 2) As a result of this they are very widely watched. Thus for a whole bunch of reasons, material changes in them really do mean something.
|
|
|
Recent Blog Entries by GammaJammer
- Sterling - Buy the Number, sell the fact - part deux (07-02-2008)
- Sterling - buy the number, sell the fact? (06-19-2008)
- Some thoughts Pre G8 - "Roll up for the mystery tour" (06-13-2008)
- Correlations - breaking down or not? (06-12-2008)
- JPY - something's gotta give (06-09-2008)
