Risk Parity and Diversification
Ask A CIO #13
My parents have unbelievably green thumbs. Multiple plants in their house reach the ceiling (they’re beautiful), and their organic garden overflows with vegetables each year. This past summer, we collected buckets of cherries from the 20 year old cherry tree in the yard. And soon, the apples will be ripening on the two apple trees. Mind you, my parents live in a typical suburban cul-de-sac, so you wouldn’t even notice these are fruit trees outside of the fruit bearing season.
Having grown up among their talents, it feels like I’ve always known how to do things like propagate a branch into a new plant. Or germinate a seed. I’ve always known that for most plants, it’s better to water them heavily and less frequently rather than lightly and often (it forces stronger and deeper roots).
But it wasn’t until I was an adult that I learned the terms propagate or germinate. And it wasn’t until I learned the terms that I started to think about why my parents do the things they do. It just was. I still appreciate that experience of re-exploring something I already knew.
I think it’s helpful to periodically pause and revisit the why behind investment decisions. This week’s question gives me an excuse to revisit the purpose behind one of the most well-known practices in investing – diversification.
I've been listening a lot recently to an interesting investment podcast [Wendy’s note: I’m omitting the name for compliance reasons]. By analyzing historical data, the host concludes that the main asset class that has negative correlation to the stock market is long-term treasuries, and there isn't much sense in holding something like a total bond ETF or corporate bonds generally, unless your goal is something other than diversifying your portfolio (e.g., goal is to generate income).
However, it seems that most people still think of the 60/40 stock/bond portfolio as mainly a total stock market ETF + a total bond ETF. What do you think of using long-term treasuries as the main (or only) bond component in a portfolio?
I hadn’t heard of that podcast previously, so thanks for sharing! I listened to a couple of episodes and read through some of the site. If my understanding is correct, the podcast uses the term ‘risk parity’ loosely and isn’t referring to the technical approach. And so before we dive into your specific question, let’s take a detour into the traditional meaning of risk parity. I promise it’s relevant and will help answer your question (which by the way, is a great question that gets at the purpose of diversification).
First, a caveat – while I find risk parity to be a highly intellectually interesting approach, I’ve never implemented it as a practitioner.1
Second, a warning – this next part is going to get very dry, very quickly. The TL;DR version: risk parity is one way to manage a portfolio. The approach involves picking a portfolio risk target (with risk defined as volatility), picking some asset classes to invest in, leveraging up the “less risky” asset classes to add more risk to the portfolio, and diligently rebalancing across the asset classes to stay at the risk target. If done well, a risk parity portfolio theoretically should withstand various market drawdowns better and outperform a 60/40 stock/bond portfolio over multiple cycles.2 Ok, you can jump to the end if you like.
Now for the fuller version. In a traditional portfolio, say 60/40 stock/bond, the portfolio is allocated according to asset class weights. And when the stock and bond markets fluctuate, the volatility of the portfolio fluctuates too. In other words, the asset class weights are the inputs, and the portfolio volatility is one of the outputs.
A risk parity portfolio inverts this portfolio construction. The portfolio has a defined volatility target as the primary input, and the required asset class weights are the outputs. As its name suggests, risk parity is all about balancing risk across the portfolio. If a portfolio includes three asset classes – let’s say equities, treasuries, and commodities – risk parity dictates that each should account for 1/3 of the total volatility of the portfolio. Since some investments (e.g., treasuries) have significantly lower volatility profiles than others (e.g., equities), leverage is added to increase exposure to lower volatility/lower risk investments and raise their risk contributions.
To be clear, when I talk about asset classes in risk parity, I’m referring to investments in the cheap betas of the asset classes (e.g., passive indices). Since investment weights are an output to this process and need to be adjusted regularly to maintain the portfolio’s targeted volatility, it’s important for each investment to be 1) cheap to access and 2) highly liquid. That’s why the most common asset classes considered for risk parity portfolios include equities, treasuries, and commodities. Alternative and private strategies are necessarily precluded.
A core tenet of risk parity is that the asset classes selected for the portfolio need to be uncorrelated to each other. The mix of equities and asset classes uncorrelated to equities then produces a portfolio with lower equity risk versus a traditional 60/40 portfolio.
Let’s walk through a very quick scenario of how this portfolio hypothetically works. The equity market experiences a drawdown, and relatedly, equity volatility goes up. Remember, the goal is to maintain a total portfolio volatility target, so as equity volatility goes up, equities start adding too much to the portfolio’s overall volatility. The risk parity portfolio then begins trimming equity exposure. Because a larger portion of the portfolio is exposed to non-equity correlated investments, the overall portfolio should experience a lower loss versus a 60/40 portfolio through a major drawdown.3
Now let’s circle back to your original question – based on data shared on the podcast, should long-term treasuries be the main or only bond component in a traditional 60/40 stock/bond portfolio? Should all other categories of bonds, particularly corporate bonds, be excluded from balanced portfolios?
From what I can gather, the podcast borrows three primary qualities from traditional risk parity: 1) an emphasis on asset classes that are non-correlated to equities; 2) a requirement that all investments be cheap and liquid; and 3) a rebalancing discipline (in this case, at periodic calendar intervals versus market driven). The podcast also focuses on creating portfolios that can support 3% to 5% annual withdrawals through market cycles. When annual spend is a driving factor, it makes sense to prioritize reducing overall portfolio drawdown during a market drawdown. Which means it makes sense to prioritize non-equity correlated investments. And so in this context, I would say the answer to your question is definitely yes.
But wait, there’s more. What traditional risk parity and the podcast both capture implicitly, and what a trusted mentor often reminds me explicitly – diversification is most important in moments of market stress. I think this is one of those statements that is both obvious when said but also easy to overlook.
The average correlation between an asset class and equities matters far less than the correlation during market dislocations.4 In those moments when markets are stressed, the investments that are non-correlated to equities are the best diversifiers. And those are the investments to hold in the name of diversification. Ultimately, investments either go in the equity bucket, or they don’t.
Bringing it back to your question – treasuries are generally uncorrelated to equities in periods of market stress (with some notable exceptions, including the 2022 rate hike cycle). Credit risk, on the other hand, is somewhat correlated to equity risk, and that correlation tends to increase in periods of market stress. Treasuries have no credit risk, whereas corporate bonds do have credit risk. The greater the credit risk of those bonds, the less useful those bonds are for diversifying the portfolio.
In a general sense, I think you’re correct to say that a 60/40 portfolio of stocks/treasuries is better diversified than a 60/40 portfolio of stocks/bonds. But I wouldn’t go so far as to say that all 60/40 portfolios should abandon corporate bonds. Any individual investor might value higher everyday returns (corporate bonds should offer higher yield and better total returns than treasuries) over smaller losses in a market drawdown.
I’ll wrap up on a note that reflects on the work of endowment and foundation investment offices (and the underpinnings of what we do here at Ivy Invest). E&Fs have long capitalized on their advantages – flexible investment approach, in-house expertise, access to a wide range of managers and strategies, long-duration capital – to build truly diversified portfolios.5 It’s hard for me to affirmatively say that any one way of constructing a traditional 60/40 portfolio is better, because I believe the best long-term portfolios hold more than just traditional assets.
I know this week’s topic went a little (or a lot) into the weeds, but I love this stuff. As always, if you have a question, please reach out: askacio@ivyinvest.co!
Until next week,
Wendy
I know just enough to be dangerous on this one and welcome additional perspectives. My views on risk parity are also shaded by a regrettable experience investing in an alternative risk premia fund.
Reality is always messier. The track record on risk parity portfolios and strategies is decidedly mixed.
There are two primary ways that the risk parity portfolio breaks (from what I’ve seen historically). The first is a forced deleveraging process. Here, the leveraged assets experience a large drawdown and increase in volatility, and the portfolio is forced to sell assets while deleveraging because the value of the collateral is also declining. The second is a large shift in the underlying correlations across the portfolio’s asset classes. Think back to 2022 when equities and treasuries declined in tandem during the rate hiking cycle (and collateral management was also a challenge during that period).
This is largely why it’s a mistake for investors to think that they are creating a diversified portfolio by combining large cap equities with small cap equities, or international equities, or any other mix of equities. When the S&P 500 experiences a major dislocation, these other markets are experiencing as large, or in many cases, even larger dislocations. In other words, the diversification does not exist when you need it most. There can be compelling reasons to invest in equities outside the S&P 500, but diversification is rarely one.
E&Fs also support annual spending for their organizations. In my experience, typical spend is around 5% of a portfolio’s multi-year average balance. These portfolios are built to be resilient, as they are often called upon to spend more in years where markets are down (market crises tend to also correspond to economic crises).