I'm a big Paul Merriman fan, but noticed something on one of his recent podcasts that points to why Risk Parity may lead to better outcomes than traditional portfolio approaches. Basically, it's not just total risk that investors should focus on.
First off, I love Paul Merriman. I’ve been listening to his podcasts for years, and I’m a huge fan of his investing approach. He’s also a true benefactor in the personal finance space; through his foundation, he has educated thousands of young people about investing and financial literacy. He’s a role model for me, and I agree with his thoughts about 98% of the time.
This would fall into the other 2%. It connects to my post on what Risk Parity is not, namely the Mean-Variance Optimization approach which basically considers two asset classes and then slides up or down the stock/bond balance to find the comfortable level for each investor. This one short clip exemplifies the traditional approach and provides the jumping off point for understanding RP and its more nuanced understanding of risk.
Click here to listen to Sound Investing from January 26, 2022 ("Is the Market Going to Crash"). The key section is from 46:14 to 55:09
Around the 48:30 mark, Merriman is talking about how the 60/40 blend hasn’t done well lately, and Merriman suggests that it is because they have included Long-term Treasuries, which Merriman says are “too volatile.” But the issue really isn’t just how much risk there is, but rather what type of risk it is. In this case, Long-term Treasuries typically exhibit ups and downs that are dissimilar than that of equities, or in other words, are negatively correlated. This means their volatility works against that of equities, and we’d expect that 60/40 portfolios combining stocks with long-term Treasuries would generally have lower volatility than with short-term, or even intermediate-term Treasuries, as well as various types of corporate bonds.
True, there are time periods we can find in which this is not the case, and I use the word “typically” above intentionally. So, is the recent 14 month stretch typical? To check, I went to Portfolio Visualizer and ran three portfolios combining VTSAX (Total Stock Market Index) with: a) short-term Treasuries (VSBSX), b) Total Bond Fund (VBTLX), and c) long-term Treasuries (VUSTX). I then reiterated the backtest four tests, going back a little over one year to January 2021, then a little over three, six, and twelve years, so since January of 2019, 2016, and 2010 (see details and results under “Test One” below).
Merriman is indeed right about the very recent history, as the backtest confirms that you would have been better off in Short-term Treasuries for the past 14 months. But if you look beyond just one year returns and instead look at 3, 6 and 12 year returns, a different pattern comes into shape. At first glance, you’ll notice the returns are better with Long-Term treasuries than the other two, so that’s good, but maybe they had more volatile roads to get those returns. Take another look and you can see how the supposed risk of Long-term Treasuries actually works: you’ll see that maximum drawdown, Sharpe ratios, and standard deviation are better with Long-term Treasuries than in the other two portfolios. If it’s volatility that you are worried about, most of the time you’d be better off with Long-term Treasuries! Its volatility actually helps lower overall risk.
What about this year, since the performance since January 2021 shows the opposite? Yes, true, and that brings up the second point. For the past fourteen months, inflation has been the main risk factor for investments, and indeed, this (in the short-term) is a major headwind for both stocks and bonds. But in this case, the Risk Parity answer would be to have part of the portfolio devoted to an asset class that expects to do well in inflation: commodities. The conventional wisdom about commodities is that they too are volatile, and indeed they tend to be much riskier than any type of bond and are even riskier than equities, as measured by standard deviation. But the key is: that volatility is not correlated with either stocks or bonds, resulting in a smoother path when all three are combined.
Once again, I did a quick comparison. This time I compared the 60/40 with Short-Term Treasuries with two very simple Risk Parity style more portfolios: one a 55/35/10 split between equities, Long-Term Treasuries, and commodities (represented by PCRAX), and then a 50/30/10/10 split with equities, LT Treasuries, commodities, and gold (IAU).
Of the four periods backtested here, the 55/35/10 portfolio was the best in all four for return, and has better risk measures (standard deviation, maximum drawdown, and Sharpe ratio) than the 60/40 each time. The 50/30/10/10 beat the 60/40 three out of four periods for return, and four out of four for risk.
Returning to Merriman, the idea of using other types of assets to lower portfolio volatility never comes up – the framework seemingly only allows either more bonds or fewer. There are better ways!
Test One
Results of 60/40 using three different bond funds:
Portfolio One: Total Stock Market Fund (VTSAX) with Short-term Treasuries (VSBSX)
Portfolio Two: VTSAX with Total Bond Fund (VBTLX)
Portfolio Three: VTSAX with Long-term Treasuries (VUSTX)
Use this link to see these portfolios on Portfolio Visualizer
Since January 2021:
Since January 2019:
Since January 2016:
Since January 2010:
Test Two:
Comparing a 60/40 to two, very simple Risk Parity Portfolios
Portfolio One: 60% Total Stock Market Fund (VTSAX); 40% Short-term Treasuries (VSBSX)
Portfolio Two: 55% VTSAX; 35% Long-term Treasuries (VUSTX); 10% Commodities (PCRAX)
Portfolio Three: 50% VTSAX; 30% VUSTX; 10% PCRAX; 10% Gold (IAU)
Use this link to see these portfolios on Portfolio Visualizer
Since January 2021:
Since January 2019:
Since January 2016:
Since January 2010: