In a van, down by the river. | Back in late May I was thinking an inflection point was nigh, and was starting to see the correlation with '09, but it appears I was off by a week or two, and the other analog years hadn't started to show any correlation at that point: http://talk.newagtalk.com/forums/thread-view.asp?tid=635478&posts=1...
I should have taken another look at the correlations in mid-June, but too many other irons in the fire. Shame on me I guess. You are also correct that analog comparisons are only useful up to a certain point, and they are really only useful when you have a strong chart "feature" in the rearview mirror to create correlation. I have found these types of comparison to be useful in the past in setting forward expectations, and I believe they are also useful if you can identify the point when a strong correlation stops working (i.e. "heads up, something just changed...").
I have this idea I like to call "temporarily-persistent correlation" in seasonal markets, where certain groups of years tend to be highly correlated over long stretches of time (but nothing lasts forever). If you can catch the correlation early enough, it can be useful for awhile. In many cases, once a strong analog pattern appears it will persist for more than a year. Basically I just take the idea of using a seasonal average and juice it up a bit by choosing only years that are displaying some similarity over a trailing period (or similarity in some other factor such as basis or spreads). I believe that, in markets, history doesn't always repeat, but it often rhymes.
Also, thanks for sharing your suggestions on better methods we might use for guiding our marketing moves over the next few months...
Edited by Rob.Cogdill 7/23/2016 11:49
|
When lows will be made...?
