Risk Parity Implementation vs Influence

On Friday I wrote a post asking why anyone would want to own risk parity in a mutual fund or ETF. This led to a discussion on Twitter with someone who I believe is more favorably disposed. He gave me another ticker to go learn about and a link to a podcast with Rodrigo Gordillo and Mike Philbrick from Rational/Resolve that was hosted by Jeff Malec from RCM Alternatives. Rodrigo and Mike manage the Rational/Resolve Adaptive Asset Allocation Fund (RDMIX). 

I've mentioned before that I am test driving RDMIX for possible use in client accounts. My concern about this fund is how complex it is. In the context of another recent post it is complex complexity or you could say multi-variable complexity versus single variable complexity. How complex? It took multiple attempts for Rodrigo and Mike to explain what RDMIX does before Jeff understood. The insights from them were still fantastic though, I'd recommend listening.

The fund is a mix of risk parity (there's more to it which we'll get to) as a base, the primary portion of the portfolio. Then on top of that, is a return stack (we've blogged a lot about that lately) where they seek alpha. They use many different strategies like global macro and trend (multi strategy) to try to add alpha, These are complex alpha seeking strategies that can themselves be multi-variable like global macro. Someone can correct me but global macro is like a catchall for a bunch of stuff. One thing I've heard twice from Rodrigo is what seems like an assumption that return stacking must add to returns, like it just does this no matter what. That is what I think I am hearing, whether my interpretation is right or wrong, return stacking does not always add basis points to return. I would not make that assumption beyond one example of putting 67% into WisdomTree Efficient Core (NTSX), aka the 90/60 ETF and then putting 33% in a T-bill that is so short term that the price doesn't move and doing that for as long as T-bills have a positive yield. Even then, not infallible.

I learned that I need to broaden how to define risk parity. It is not as simple as equal, risk-weighted exposure to equities and fixed income which is obtained through leveraged fixed income exposure. Commodities or inflation (seemed like those words were used interchangeably) should be part of the allocation and where RDMIX uses volatility in the fund, it seemed like they were saying that vol exposure is part of the risk parity base. Examples of long volatility, using volatility to protect against declines, would be products that go long VIX or tail risk which usually means long puts. 

The new (to me) fund is the American Beacon AHL Targeted Risk Fund. There's a bunch of symbols, I used AHTYX. Here's the allocation info.

 

Part of the description of the fund is to take a balanced risk approach to perform "positively in a range of market environments." Also "capture the risk premiums associated with being long each asset class."

The fact sheet compares it to 60/40 and that seems like an accurate comparison based on their chart and certainly that is easy to understand regardless of whether you agree/disagree.

 

And here's a comparison I built.

 

 And those same funds YTD.

 

AHTYX hasn't helped this year. An alternative to a 60/40 portfolio, in the context that we've been exploring would ideally spare holders from interest rate risk. It is certainly multi asset, not sure that it is true risk parity versus having been influenced by risk parity. Why buy a fund that is more expensive than 60/40 with no performance differentiation? If a fund isn't going to offer protection in the first half of 2022, when will it?

RDMIX clearly helped this year but in the podcast even Philbrick said it isn't always smooth sailing and the longer chart shows that to be the case. It did not bounce back with the broad market from the 2020 Pandemic Crash. MENYX which I am test driving is more of an equity proxy, did not avoid the 2020 Pandemic Crash but has done very well this year. Client and personal holding MERFX is immune to panics and bear markets apparently but is also immune to large stock market rallies but that type of protection has a place in a portfolio, the portfolios I manage anyway. 

Let's look at RDMIX with Portfoliovisualizer and compare to a portfolio we've looked at a few times and 100% VBAIX. Note that BTAL is a client and personal holding.

 

And the results.

 

Definitely a better CAGR with RDMIX but standard deviation is much higher with far less protection in 2022 in terms of max drawdowns. Note that the current managers of RDMIX took over in Feb 2018 so fair enough if you think the time frame is unfair. 

Back to the podcast. The idea that true risk parity takes in more than two asset classes is useful. They also talked about investors (pro or otherwise) using the wrong kind of diversifiers, using diversifiers that that correlate too closely to the components of plain vanilla 60/40, especially the fixed income portion.

If you've been reading my content for a while, this is something I've been talking about for ages. In life, I seek to be orthogonal to society in terms of how I live, what we do, how we spend money, how we take care of ourselves and so on and this has been directly applicable to interest rate risk and diversifying away from what most bond funds offer. We've talked frequently about seeking out the attributes that investors hope to get from bonds from alternative products or individual holdings while avoiding or greatly underweighting interest rate risk. This is what led me to so many of the tickers I've been using and writing about for many years.

I am still out on risk parity but I think influence can come from some of these ideas from the managers of these very complex funds like RDMIX. We've had quite a few blog posts in the last two or three months about how to construct portfolios using just a couple of much simpler alts to capture the effect, actually a little better, of plain vanilla 60/40 while avoiding interest rate risk. 

The discussion on the podcast was fantastic and I will listen to them on future pods but the conversation convinced me not to use RDMIX for clients.