Is there a resource that quantifies correlations during market stress events? I want to hedge a vanilla asset allocation strategy against systemic market upsets. If risk asset correlations move towards 1 during systemic crisis, then an optimal asset allocation hedge could be developed by buying the relatively cheapest (quarterly/monthly/weekly) OTM spreads on systemically important securities (SPY, HYG, TLT, GLD, VXX). To study this I would like to look at historical periods of market stress to determine the correlations of stocks/treasuries/junk/gold. Any tips for an under capitalized working stiff to do this?
Pull in free Yahoo or Google data, analyze with appropriate R (or Py) packages. It's at most a dozen lines of code. There are several threads with relevant examples.
In 2008 dispersion became non-existant. Trust me, when shit hits the fan you will know it, no sophisticated model needed (heck, I say that as practicing quant trader) . And when you start hedging when it's too late you are just one additional person trying to squeeze through the bottle neck. Hedges will be very expensive. To some problems there simply are no optimal solutions, only consequences to choices, made beforehand. May I suggest you concentrate your efforts on identifying market behavior at a much earlier stage and implement optimal solutions upstream.
@Zzzz1 Thanks for the input. I'm working on my breadth and intermarket studies too. Any other aspects I should monitor for system stress? Do you think we'll get as much advanced warning for the next debacle as we had for the 2008 train wreck? With all the intervention it seems to me that a pending bear will be masked until it just snaps free.
There are always warning signs. It's just a matter of degree. The money markets seized up a year before the last big crash. I would make the case that because of that crash it's harder for the next one to happen, as some will always have one foot out the door. Nevertheless it's a really old bull.