On Sunday and Monday, I went down a bit of a replication rabbit hole. Replication is pretty much what it sounds like. Usually a replication strategy will build a portfolio based on reported hedge fund holdings filed on a 13f or in the case of managed futures will sample maybe the ten biggest futures markets believing they can get 90% (or some high number) of the full effect, do it for cheaper such that the cost advantage ends up being the difference in performance.
Some of the managed futures funds we've mentioned in previous posts are replicators and some are the full strategy. Alpha Simplex has one of each. There used to be an ETF that replicated hedge fund portfolios based on 13f filings, the Alpha Clone ETF (ALFA). It closed due to low AUM but the performance was fine.
There's been plenty of research on replication. It definitely is valid. It may or may not be optimal, but it is valid, there is research that concludes replication is superior to the thing being replicated but that's not the point of my post, I can't say that it is or isn't but I can say it is valid, repeated for emphasis.
I looked at three different portfolios to take a stab at replicating. Some of the holdings used are pretty good fits and one or two of them really are not but that's ok, it's just a blog post. First up, the Harvard Endowment which posted the following asset allocation.
The private equity piece isn't great. I used PSP. Some of the names have done phenomenally well and others not. Modeling with PSP evens that issue out and I also compared using all public equity.
If you pick the right hedge fund proxy, then the results are compelling. I just considered multi-strategy not absolute return or managed futures. I would want to divide that sleeve up more if this was any sort of real portfolio I was going to implement. The standard deviation of the version with PSP is quite a bit higher. Here's an article at theStreet.com from 2007 where I bagged on PSP. Average returns with high volatility. The version with VOO was down a lot less in 2022 and only trailed VBAIX in 2023 by 89 basis points.
Then I looked at Dartmouth.
Like all of them, a lot of private equity but it appears to be a little simpler.
Again, the replication with VOO appears to be superior.
In 2022, the VOO version was down less than half of VBAIX and it also outperformed in 2023.
Arguably neither one is very close in terms of how it replicates but borrowing the asset allocation from the top down yields what I would call a valid result. The weighting to QDSIX in each is way more than I would ever put into just one alternative which I am also repeating for emphasis.
One more to look at is the Destra Multi-Alternative Fund (DMA) which is a closed end fund. I saw a Tweet about it and I was intrigued by the name. I'd never heard of it before Monday. Here's their allocation.
And my attempt to replicate it.
JAAA and PPFIX are both in my ownership universe.
And again using VOO instead of PSP delivered noticeably better results here but not dramatically so. In 2022, the VOO version outperformed VBAIX dramatically but lagged by about 8% in 2023.
The actual closed end fund backtests much shorter and has been far more volatile than our attempt to replicate it due probably to the fund eroding to a huge discount to NAV.
One obvious flaw to remember to this whole process is that the strategies we're replicating are not static but the backtests are. The asset allocation ideas are no less interesting to learn from it's just that what we've done is replicate a recent snapshot, we haven't tracked the portfolios.
The information, analysis and opinions expressed herein reflect our judgment and opinions as of the date of writing and are subject to change at any time without notice. They are not intended to constitute legal, tax, securities or investment advice or a recommended course of action in any given situation.
No comments:
Post a Comment