Fun With Portfolio Theory

Long time readers might know my fascination with Nassim Taleb's idea about barbelling portfolios to concentrate risk into a small slice while having the vast majority in safe assets. He used to frame it a 90% in T-bills from around the world and 10% in very high risk or as he might have said later, asymmetric risk and what's more asymmetric than Bitcoin so let's have a little fun.

It turns out Portfoliovisualizer can model actual Bitcoin which is more accurate than using Grayscale Bitcoin Trust. Now that there are spot ETFs, maybe they will work as far as tracking the underlying and make this sort of post a little truer in terms of backtesting. What I am curious to see is if we can combine this barbell idea with the 75/50 portfolio to get a market equaling (or beating) returns over longer periods. As a reminder, 75/50 seeks to capture 75% of the upside of the equity market but only 50% of the downside. 

LFMIX in Portfolio 1 is the LoCorr Macro Strategy Fund and in Portfolio 2, QGMIX is AQR Macro Opportunities. VBAIX in Portfolio 3 is the Vanguard Balanced Index which is a proxy for a 60/40 Portfolio. LFMIX and QGMIX both have very low volatility but are complex funds and complexity has its risks and drawbacks too. I could have gone back further than 2018 but Bitcoin skyrocketed in 2017 so I wanted to take that out in case that sort of year never repeats again.

Apparently LFMIX had a bad year in 2018, otherwise it's longer term result would be a little better. With a 5% weighting to Bitcoin, if it really does go to zero, it would have just been a bad idea but not a ruinous idea. The backtest clearly meets the 75/50 goal. Portfolio 1 lagged VBAIX but is far above 75% of the VBAIX return. So this is interesting.  

Putting 95% into one alternative strategy fund is off the table for me entirely. I can't really see putting 20% into just one fund either. To try to get a little closer to reality, the next back test takes in eight funds each weighted to 10%, 5% in Bitcoin and the remaining 15% in cash to deal with sequence of return risk. Most of the funds come from a report that Standpoint puts out every month that tracks the "top 20" alternative mutual funds. 

Portfolio 1 is exactly as I described above, I just picked funds at random with no thought and included Bitcoin. Portfolio 2 is the same eight alt funds, no Bitcoin and 20% in cash and Portfolio 3 is VBAIX. Just using the alts and cash comes up short on the 75 but does much better than 50% downside.

Putting 10% into one alt is not my preference but I might be overly cautious on that front. I think this is constructive for the idea of trying to budget volatility and understanding that small allocations, in this case to Bitcoin, can have a big impact. Take Bitcoin as just an example. The more useful context of small allocations going a long way usually relates integrating a negatively correlated strategy into a much simpler, plain vanilla portfolio. 

80% into alts flies in the face of a big priority for me which is building mostly simple portfolios and hedging them with just a little complexity. Today's post though challenges the idea that a portfolio full of alts hedged with a little bit of equity is likely to lag a simple portfolio. VBAIX is pretty simple but the alts plus BTC are very competitive. 

Last one. This looks at cash plus Bitcoin. So very simple, we avoid the alts but to make it work in terms of 75/50 we went 6% Bitcoin, not 5%.

Taleb's theory stands up even though the last I heard, he hates Bitcoin, even though it is possibly the most asymmetric thing ever. 

The gaping flaw in all of this, lets assume for just a minute that Bitcoin does not turn out to be complete bullshit, there is no way to model forward what Bitcoin will do. There's no fundamental work to do, there's very little history to draw on, so who knows? The various alt funds that are now available, the strategies by and large have been around for ages, just recently available in mutual funds, and are somewhat predictable in terms of how they will react to different events and how they might correlate with other assets. 

Actually, one more idea. Along the lines of capital efficiency, maybe a little Bitcoin can be swapped in for quite a bit of equity exposure to equal a VBAIX portfolio. 

NTSX is leveraged to offer 90% equity exposure and 60% fixed income such that putting 67% of a portfolio into it equals a fully invested 60/40 portfolio and it has tracked pretty reliably. With only 67% invested to replicate 60/40, investors could then leverage up with the remaining 33% or as would be my preference, leverage down which we've talked about in many past blog posts.  

Portfolio 1 in the above image works out to 27% equities, 18% fixed income (pretty much Aggregate Bond Index type of exposure), 5% Bitcoin and the rest in cash. If the portfolio could somehow be close to 60/40 but with that much cash earning 5% for now, that would be very resilient against any sort of adverse return sequence and be a pretty good example of leveraging down.

The back test only goes back to 2019 because that is how far back NTSX goes. Portfolio 1 outperforms Portfolio 2 and almost hits the 50 in 75/50 with decently lower volatility. 

If Bitcoin never goes up again, then all of these would fall far short of 75/50. Even though 5% is more than I would want to start with in Bitcoin, that sized speculation ending badly is not enough to be ruinous. While I do find value in this sort of exercise, it needs to be repeated that I have no expectation about forward modeling of what Bitcoin might do. 

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.