Understanding Risk Parity

My test of how well I understand a topic is to try explaining it to my friends. They’ll listen patiently and ask questions as I’m explaining the concept. As I talk, I often begin to realize that my narrative is unraveling in the face of their questions, and I am repeating the same lines over and over (at a louder and louder volume).

Ugh. Clearly, I don’t know enough to explain the topic simply and therefore, I don’t really understand it all. My thoughts aren’t translating to my mouth, and in the case of this blog, won’t translate onto the page. This happened recently with risk parity.

I know. I know. Some of you are yelling at your screen. “It’s intuitive! Balance risk not dollars! Use some leverage! Ray Dalio!” Others may not have heard of risk parity and you’ve already forgiven me. Thank you.

But, for retail investors the practical application of a theoretical risk parity portfolio is a bit more complicated. And, I think a strong understanding of risk parity and the challenges facing retail investors is essential before diving in. 

I would also say many of us, myself included, have a long relationship with adding equities to increase expected returns. While we may “get” risk parity, it could take some time to get comfortable building portfolios in a different way. 

‍Back to school: Finance 101 

Bear with me a minute while I get finance-y. As simply as I can say it, to create a risk parity portfolio, investors weigh assets by risk (volatility) instead of dollar amounts. Traditional portfolio construction looks at expected returns and volatility (risk) and creates a portfolio that minimizes risks for a desired return. An investor may end up with a portfolio that is 60% equities and 40% bonds (i.e., weighted by dollars). 

However, if we assume volatility of 15% for stocks and 5% for bonds and a correlation of 0.2, then stocks will account for over 90% of the risk in a 60 40 portfolio. SeekingAlpha crunched the numbers from 2000 - 2017 and found the risk contribution was actually greater than 100%! This concentration of risk is what risk parity seeks to address.

In a risk parity portfolio, an investor looks at the risk contribution of each asset and builds a portfolio that balances the assets without considering expected returns. Once the portfolio is risk-balanced, leverage–borrowing money to amplify returns–is applied to increase volatility and achieve the desired return. The underlying logic here is rooted in financial theory. The portfolio on the edge of the Efficient Frontier is theoretically the optimal portfolio because it delivers the most return for the risk taken (i.e., the highest Sharpe ratio). In practice, the expected return of the optimal portfolio is usually below the required return of investors. That’s because this portfolio has a much higher allocation to low-risk, low-return assets like bonds.

We have talked about how higher risk = potentially higher return before. To achieve a higher required return investors have two options to take on more risk. In traditional portfolio construction, investors allocate more to the riskier asset (i.e., equities) and increase concentration risk. In risk parity, investors take on the risk of leverage. I’m not here to tell you one risk is better than other, simply that investors have options in portfolio construction.

So why does no one talk about risk parity? 

Unfortunately, there are a number of challenges for retail investors to implement risk parity strategies. First is the issue of leverage. For many investors, leverage is a foreign concept. And, while I can babble on about how leverage and concentration risks are both viable options, investors have clearly voted with their dollars. A recent Yahoo Finance survey pegged the number of investors who have traded on margin in 2020 at 20%. However, this includes investors who sold stocks short and traders who used margin to amplify gains on speculative stocks. My gut says the number of investors using margin to create a risk parity portfolio is much lower. 

Further, to successfully implement risk parity, the cost of borrowing needs to be low enough that borrowing costs do not eat away all the benefits of using the optimal portfolio. And, I think it's fair to say, you and I probably don’t have access to the same lending rates as hedge funds through their prime brokers and banks. Well, I don’t know your situation, but I got student loans.  

So, lack of familiarity and cost are two reasons why investors may avoid leverage. Indeed, research from AQR points to leverage aversion as a theoretical underpinning for why risk parity works. AQR shows that this reluctance to take on leverage leads riskier assets to have lower risk-adjusted returns and less risky assets to have higher risk-adjusted returns. Investors willing (and able) to undertake leverage can take advantage of this fact by reducing their allocation to equities and increasing their exposure to bonds. 

Another issue for retail investors is gathering data, calculating the appropriate risk metrics, and executing the trades to balance the portfolio. Whether using forward or backwards looking risk measures, investors need to continually update the data and adjust the portfolio accordingly. 

And with leverage, portfolio weights are not the only thing investors need to monitor and adjust. Leverage increases the volatility of a portfolio and amplifies losses as well as gains. Investors need to have a plan for how they will manage this risk. Again, from AQR: 

“Having a plan to systematically reduce leverage in the face of severe losses, and systemically add it back when things start to get better (which we think is actually the harder part!) is a much better plan than reducing leverage only when your back is against the wall and the decision is forced and likely expensive.”

Risk Parity with Composer 

So, how can investors implement a risk parity strategy using Composer? Let’s start with the issue of leverage.

Off the top of my head, there are a few ways to access leverage (none of this is financial advice!). You can open a margin account with your broker, take out a personal loan, or use leveraged ETFs. One of the easiest ways, taking a page out of Hedgefundie’s book, is to use leveraged ETFs. My reasoning is that the costs are relatively low and the administrative burden is much lower than taking out a loan. It is worth noting though that investing in leveraged ETFs is not the same as adding leverage to the entire portfolio, which is what theoretical risk parity calls for [1]. Through the Composer platform, you can backtest and evaluate leveraged ETFs and then easily add them to a symphony.

Per the AQR quote above, having a plan to reduce leverage is also important. Fortunately, systematic, rules-based trading is Composer’s bread and butter. In our blog profiling Hedgefundie’s Excellent Adventure, we offered Composer’s take on the classic strategy, Hedgefundie’s Excellent Adventure Refined. Our refined symphony incorporated trading logic to exit the leveraged ETFs and shift into safer assets if the 10-day drawdown of the S&P 500 reached a certain level. The trade, moving to safer assets, is executed automatically through Composer’s automated trading platform. 

Further, Composer makes the execution of risk parity easy with inverse volatility weighting for assets. The platform looks at the volatility of assets, over a specified time period, and weights the portfolio so that more volatile assets receive a lower weighting. 

Looking at a simple portfolio of stocks (SPY), bonds (AGG), and commodities (DBC), we use the inverse volatility weighting to balance risks and create a risk parity portfolio. As you can see from the table below, the allocations are dynamic, based on the previous 252-day volatility. As expected, there is a much higher weight in bonds (~70% as of September 24th, 2021) than equities (17%). Commodities make up the difference.

And, as discussed above, the unlevered risk parity portfolio has much lower annualized returns than SPY (20.4%) and the 60 40 portfolio (14.6%). However, the risk parity portfolio does have a higher Sharpe ratio.

We can also implement inverse volatility weighting for Hedgefundie’s Excellent Adventure, so the allocation of UPRO and TMF dynamically updates with changes in recent volatility, keeping the portfolio risk balanced. Here are the allocations over the past three years based on 252-day volatility. While the split hovers around 50/50, the allocation to UPRO went as low as ~13% in March 2020.

Next week in the blog, we continue down the risk parity rabbit hole. Why? Because the ability to build symphonies that mimic some of the most sophisticated investors in the world (e.g., Bridgewater, AQR, endowments) is what gets me so excited about Composer. I will offer some more examples of risk parity strategies built using the power of Composer and investigate strategies to manage risk. Cheers!