Learning, Not Emulating

Sophisticated influencers in portfolio construction.

We've spent a lot of time in the last few days looking at some strategies/funds that blew up during this decline, why they blew up and what some of the warning signs might have been in terms of the risk taken by those strategies. Predicting a blow up is pretty difficult to do but recognizing that something might have risks that you don't want to take on isn't so difficult.

This was how I looked at risk parity the other day, doubting that, as the WSJ asserted, it contributed to the decline in equities but that I don't want to take on an exposure that levers a bond position. Since that post I found a couple of other things about risk parity that don't change my mind about taking on the exposure but as I mentioned it is sophisticated and there is plenty to learn from how a risk parity portfolio is constructed.

First is this article from the FT written by someone who used to work at Bridgewater, whom I believe to be the biggest risk parity shop. The first point to consider is "since stocks tend to be more volatile than bonds, a typical portfolio of 60 per cent stocks and 40 percent bonds is going to have almost all of its upside and downside determined by the performance of the equity component over time." This is an easy point to understand and is a key building block for risk parity. It is trying to mitigate the potential consequence of having so much riding on one asset class. But if rates are indeed going to normalize then the exposure to the wrong part of the bond market will tilt the equation expressed in the quote. The PIMCO 25 Year Zero Coupon ETF (ZROZ) is the most vulnerable fund to rising rates I can think of, it would get decimated if rates normalized. It is down 12% since mid-December and the ten year is yield is only up 35 basis points. Although an extreme and maybe obvious example, levering up to own ZROZ would devastate a risk parity portfolio.

There are plenty of other exchange traded fixed income proxies that take very little interest rate risk or even hedge out interest rate risk to help. It is looking like a move higher in interest rates, if it happens, will be at least partially driven by inflation (breakevens among a couple of other indicators pointing to this possibility) which makes TIPS exposure interesting and there are of course plenty TIPS ETPs. Inflation protection comes up in the FT article. In several different pie charts the unquantified TIPS exposure appears to be about 1/3 (not sure whether leverage is expressed in the pie charts) of the portfolio.

Heavy TIPS exposure is also favored by Zvi Bodie whom I have written about a few times. He has a similar approach as Nassim Taleb, putting the vast majority in low risk assets, Bodie likes TIPS for this, and then a small slice in very risky assets. A large TIPS weighting, clients own PIMCO 1-5 Year TIPS ETF (STPZ), combined with an interest rate hedged fund could be an interesting combo in creating a risk parity-lite portfolio, avoiding the leverage.

In the FT article was a link to a blog post from AQR, they run the AQR Risk Parity Fund (AQRIX). It is a long post but includes the idea that there are different ways to assess the volatility of the different asset classes. Correlations are also discussed. Commodities by themselves are viewed as risky they say, but when combined with other risky but uncorrelated assets like equities reduce the overall volatility which of course is well known even if not universally implemented.

Bloomberg had an article titled Brokerage App Robinhood Thinks Bitcoin Belongs In Your Retirement Plan. The title was either misleading or the article was edited such that the part that actually said that was cut. I've written about Bitcoin plenty enough, don't take this as recommendation either way but what role could something that volatile play in some version of risk parity. Would the risk of a 1% weighting in Bitcoin equal the risk of 99% in TIPS? If not that ratio, maybe some other ratio. As AQR says there are different ways to asses the volatility of different assets so as a thought experiment, use any ratio you want. If Bitcoin goes up ten-fold in some given period of time and STPZ or any other TIPS proxy goes up 5% then the overall return would be 14.95% which is interesting in the context of getting a market-like return with very little exposure to volatile assets and/or a very low correlation to equities. If Bitcoin goes to zero in this example, that's too bad but it was 1%. Again, not a suggestion but an interesting way to look at how to build risk in a portfolio and how to use non-correlating assets.

In constructing a portfolio, these are some concepts that can be influencers even if you choose not to emulate. I want no part of levering up a portfolio but the idea of concentrating risk in a smaller amount of assets is interesting, I was long ago influenced to believe that commodities, notably gold, and other alternatives in small doses have a role in managing risk and volatility and have seen that play out satisfactorily over more than a full stock market cycle.