StumpGrinder’s performance similar to other allocation funds.
related article: Why diversified portfolios lost value in May/June 2013
Overview: Our “sanity checks” are failing, so it is time to move to cash
As I said in a previous post, you should never blindly follow an investment strategy. I’ve heard so many times “strategies work until they don’t work.” You should be prepared to exit when things seem to be headed south. But how can you tell the difference between a short term down trend and a systemic failure?
An important component of any strategy is a set of “sanity checks.” Sanity checks are assumptions you’ve made about your portfolio’s performance, that you can check at any time to see if they are true. When they become not-true, that’s when to go to cash.
I mentioned one of the assumptions we make about StumpGrinder above, namely that we assume BOND is anti-correlated with the low-vol equity ETFs USMV and SPLV. Those three ETFs are, together, our largest holdings, so this relationship is important.
As towards the end of May we saw many days when BOND and SPLV were both down together — becoming correlated, and not performing as they should. In fact on the last day of May the two were exactly — not-correlated (more detail to come in another blog). I was poised to move to cash on the June rebalance, but for the next few days saw anti-correlation creeping back.
Now by mid June, BOND and SPLV are correlated, and both moving down.
Time to step to the side and reassess
Stay tuned for a deeper analysis, and for details on how we’ll know when to step back in.
Travis Deyel
June 21, 2013
You’re measuring correlation every day? Perhaps this is just a “repeated significance testing error”?
http://www.evanmiller.org/how-not-to-run-an-ab-test.html
Tucker Balch
June 24, 2013
I’m using various lookback periods. I’m going to take this time in cash to assess what that lookback should be. Glad we went to cash by the way, the portfolio would have been down 1.5% today (so far) otherwise.
Ben
July 10, 2013
So did you liquidate the entire BOND and SPLV holdings? (What about USMV? It’s kind of the same thing.) Any other changes? I kept looking at the “StumpGrinder: Allocations” post to find any changes but didn’t realize you wrote this one. Thanks!
Tucker Balch
July 10, 2013
Yes, I should have also logged this on the “StumpGrinder: Allocations” article. StumpGrinder and allocation funds in general had a tough May/June. See this article on Bridgewater: http://www.reuters.com/article/2013/06/24/us-hedgefund-performance-bridgewater-idUSBRE95N0YF20130624
Several of our sanity checks failed towards the end of May and mid-June: Namely anti-correlation of BOND and USMV. So on June 21 we exited the portfolio to cash.
I will follow up with two blogs: 1) Details about the sanity checks and why they are important, and 2) Plans for re-entry.
Fortunately for you, if you didn’t exit on 21 June, you’re in decent shape because the portfolio has rebounded somewhat. In terms of what you *should* do: If you want to continue to track StumpGrinder, go to cash. If you want to stay in the market, and wait for my next blogs, the main thing you want to avoid now are longer term bonds. So I would exit BOND. I think SPLV and USMV have a good prognosis though.
Tucker Balch
July 24, 2013
See this new blog about May/June. I’m working now on a plan for getting back in. http://wp.me/p11WgN-nj
Johan
December 3, 2013
What has happened to Stumpgrinder? No longer an active project?