"The Cost Of Being Different"

Morningstar has an article up about the importance of diversification but warns about over diversifying. Here's the money quote from author Amy Arnott. "In my opinion, most investors need exposure to three core asset classes: US stocks, international stocks, and investment-grade bonds." She adds that some people might want TIPS exposure too. 

Over the long term, both portfolios (note I did not see suggested weightings beyond 60/40) compounded just fine at 7.17% and 7.59% respectively. Those numbers, combined with an adequate savings rate will get it done.

But neither version offers any real diversification. They track the market, they are the market, they don't differentiate from the market at all. Put differently, what is it the Morningstar thinks is being diversified away? If someone just went 100% SPY at the start of this back test, they'd have had larger drawdowns every time on the way to a higher growth rate; no differentiation, just bigger swings. 

The notion of overdiversification is worth raising though. I think we explore that here by try to keep things simple, relatively simple or in trying to allocate more heavily to simplicity versus complexity. A lot of the portfolio construction ideas we pull in to blog about flirt with overdiversification. The Cockroach Portfolio might be a tad busy, so too is some of the work the ReturnStacked guys do with their model portfolios and there are others. But we can learn from all of them. 

That brings us to a paper from ReturnStacked titled "What Is The Optimal Stack?" The focus seems to be trying to create a portfolio that has the same volatility as plain vanilla 60/40 but improve on every other metric. The real answer from them was a very levered up split with 24% to equities, 71% to bonds, 7% gold, 55% merger arb and 39% managed futures. They had a funny bit about how unworkable that is in real life. I spent some time trying to recreate it such that it had a volatility the same as 60/40 but I couldn't get there. 

Corey Hoffstein Tweeted out this image that was not in the paper. I think it was an output from their new optimizer tool.

I tried several different things with this blend.

Note that you have to be comfortable with AGG-like exposure to actually implement any of these. I am not but I am pretty sure the ReturnStacked guys are. 

All three are better in terms of CAGR and the volatility of Portfolios 1 and 3 are almost identical to VBAIX but I think the way in which they all went down much less in 2022 adds a favorable skew. In the other drawdowns, they don't look much different. 

If you really want to diversify, they say there will be an "expected cost of being different" which is a great line. Just about every backtest you can run that has a large weighting to managed futures looks fantastic but the "cost of being different" is that there can be long periods where it underperforms. Ditto gold. Do you like merger arb? I certainly do but a very strong year for merger arb might be up 7% which looks paltry compared to a strong year for equities. 

We could create more differentiation by including exposure to negatively correlated assets that are more immediately reactive to market declines as we've done in dozens of posts. Using those types of strategies pretty reliably creates a much smoother long term result but can be difficult in shorter periods, that is the "cost of being different," frustration in the short term. 

When you really invest the time to understand how something like managed futures actually works, it becomes much easier to hold on during that frustration. 

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.