Allocation Change
The stock bond timing model changed allocations at todays close. The relative performance component switched over to stocks. The new allocation is 53% SPY, 47% TLT.
Here is how each component is allocated:
Relative performance: SPY
Correlation: TLT
Moving average: SPY/TLT mix
Posted on September 24, 2012, in Stock Bond Timing Strategy, Transactions. Bookmark the permalink. 6 Comments.
How do you determine when to rebalance. Is it relative performance of who is leading in the last 13 weeks or something like that. Why the 53/47 ratio in favour of stocks. Why not 66/33 or 90/10 etc
Thanks
Hi Bert, the strategy gets rebalanced whenever an allocation change happens (on average about 5 times a year during the test period). You can read more details about the system here: http://dontfearthebear.com/stock-bond-timing-model/
The system is a composite of three different timing strategies. Each of the three strategies is given equal weight and contributes to the final stock bond mix. Let me know if you have more questions not answered there.
Best, Mike
Mike, question – why do you make these strategies public? Specifically your alpha strategy? If you truly found something epic, wouldn’t you keep it for yourself to make sure others didn’t get in on it, exploit it, and ruin what advantage you had?
I built an XIV/VXX strategy last year as well, and it has similar performance to yours. However I am not telling others about it. If institutional desks start playing in VIX Futures it could drastically change their behavior during VIX moves, which would screw up all the progress we’ve made.
Thanks.
Hi Pete,
I understand what you’re saying to an extent. The market does have a way of eliminating the effectiveness of strategies that used to work in the past. I believe this modest blog, despite its transparency, will have little impact on that.
Institutional investors have been and will continue to trade VIX futures. I’ve spoken with people from several companies who’ve worked on similar projects to my Alpha strategy. The market will change over time and we will have to adapt.
There are a variety of reasons for why I post stuff here. These reasons include, among other things, getting feedback and ideas from people with similar interests, keeping a lot of my research organized and in one place, and helping other system designers generate fresh ideas. If it weren’t for feedback from a reader, my Alpha strategy likely wouldn’t have included a term structure component. In this case, sharing my research led to a benefit for me. Same thing for the recent VIX follow through series. That idea also originated from reader feedback.
I’m not saying everything should be shared, but there is some benefit to it in general.
I agree. It is hard enough to be a retail trader/investor, but what makes it worse is that the community is awful and is very reluctant to share ideas. If a retail investor has thought of it, some institution or HF has probably come across years ago (e.g. pair trading), so I don’t think the impact is as big as they think it will be. Compare this to open-source programming community and it is night and day between the two.
That being said I suppose there could be some merit that inefficiencies can get weeded out. For example, in 2009 everyone got really excited by the DV2/DVO indicator and it immediately stopped working (although it could be due to QE influences).
Also, MarketSci seems to be selling a similar strategy as the alpha strategy, so it isn’t a huge secret. But I think the strategy has legs somewhat because the alpha comes from selling vol, which isn’t really an inefficiency and is well known. You are just being paid for taking tail risk – XIV could drop 30-50% in one day on a 1987 event.
Well said, Mike. It’s interesting that, sometimes, being generous serves one’s self-interest more than being selfish.