Risk Parity. Alex Shahidi
Читать онлайн книгу.
I have devoted a single chapter to each major asset class to provide greater insight into the relationships between the environment and returns, so I won't get into too much detail here. Most important, all the cause‐effect linkages are logical and reliable and can be supported by long‐term data. With an appreciation of how each asset class is biased to perform in various economic environments, we can thoughtfully select the appropriate asset classes to include in our balanced portfolio.
Four Diverse Asset Classes
There are numerous asset classes to pick from. The key question is how we can take advantage of diversification across various asset classes to reduce risk. Our goal isn't to select or weight them based on return or risk but based on how they diversify each other (i.e., what drives their performance). The core observation is that asset classes have two key drivers in common – growth and inflation – and what distinguishes them from each other is how they respond to those drivers.
Based on this logical sequence and our desire to build a well‐diversified portfolio, the next natural step is to select a group of assets that perform differently from each other because they respond differently to changes in growth and inflation. By temporarily ignoring the returns and risks of various asset classes (in contrast to the way most people think about building a portfolio), the steps should seem reasonable and, in many ways, very obvious.
There are four general economic environments that should be considered: rising growth, falling growth, rising inflation, and falling inflation. There are four asset classes that cover these economic outcomes:
1 Global equities – rising growth, falling inflation
2 Commodities – rising inflationIndustrial commodities – rising growthGold – falling growth
3 Treasuries – falling growth, falling inflation
4 TIPS – falling growth, rising inflation
You may notice that some commonly used asset classes like intermediate‐term core fixed income, high‐yield bonds, and real estate investment trusts (REITs) were excluded from our list. I believe Treasuries are a better environmental fit for falling growth than core fixed income because their higher credit quality means they will typically outperform during extreme adverse environments like 2008 and first quarter 2020. High‐yield bonds have the same bias as equities (rising growth, falling inflation) since their returns are largely driven by changes in creditworthiness, so they don't add much incremental benefit from a diversification standpoint. You'll see later when we discuss returns that they also have a lower return profile and are less tax efficient than equities, which is another reason to exclude them. REITs also are similar to equities in terms of how they respond to shifting economic environments. This one is a closer call because real estate is a real asset and has some inflation‐hedging components. However, real estate is often purchased with debt, which makes it sensitive to rising interest rates, and rents can often lag inflation – particularly with fixed, longer‐term leases. Therefore, it has historically not provided as reliable and consistent an inflation hedge as TIPS and commodities, which are included in our mix of assets.
STEP 2: HOW TO STRUCTURE EACH ASSET CLASS TO HAVE EQUITY‐LIKE RETURNS
The first step was to select a diverse set of asset classes that are biased to perform differently in varying economic environments. This creates the potential to reduce risk, since the positive performance of assets that do well in a given environment can offset the negative performance of assets that do poorly. Thus far, we have not discussed returns, because we separated out risk reduction from return maximization at the outset. By isolating the key drivers of asset‐class returns, we can follow a relatively straightforward logical sequence to arrive at the asset classes selected for our portfolio.
We can now turn to the second step that will enable us to design the selected asset classes to deliver an expected return competitive with stocks over the long run. One problem that investors face is that some of the diversifying assets are low returning. This apparent obstacle commonly turns investors away from investing in many of the most diverse asset classes. However, this is a problem that can be easily solved.
Importantly, investors are not forced to take the chosen asset classes in the packages in which they are typically offered. For instance, Treasuries and TIPS are normally considered low‐risk, low‐return asset classes. Commodities are also often thought to be lower returning than equities. Fortunately, we can take simple steps to structure each asset class to offer a long‐term expected return competitive with equities. I appreciate that the preceding italicized statement may sound unreasonable, counterintuitive, and possibly too good to be true (which is precisely why I emphasized it). Imagine for a moment if Treasuries, TIPS, and commodities could be held in a way that would offer equity‐like returns over the long run. You could see the power of a portfolio that is balanced across a reliably diverse set of asset classes, each of which would be expected to earn returns comparable to equities. With these asset classes, we should be able to achieve attractive returns while minimizing risk. Furthermore, all of this is possible without any reliance on active management. A simple passive portfolio can outperform even a 100% equity allocation over the long run.
This is where the “parity” component of risk parity is relevant. Parity refers to the state of being equal. Within the context of building portfolios using a risk parity framework, we take the major step of increasing the risk in each asset class so that it has a roughly equal long‐term expected return as equities. Since most asset classes have a comparable return‐to‐risk ratio, this step basically involves equalizing the risk of various asset classes to raise the expected return to the level of stocks. Using long‐term volatility as a measure of asset‐class risk, we can now consider a wide range of investment options to create a diversified portfolio. Equalizing the risk to synthetically create long‐term equity‐like expected returns across multiple and diverse asset classes brings us closer to the smoother path described at the beginning of this book. By limiting our universe of investment options to those that are not correlated yet high returning, we eliminate the need to have to choose between low‐ and high‐returning assets like those in a 60/40 framework.
Years ago, I had a conversation with a highly successful investor. He had earned a PhD in Economics from the Booth School of Business at the University of Chicago, one of the most prestigious and well‐respected business schools in the world. He was well trained in portfolio management and had over 30 years of experience with investing. He would easily qualify as a sophisticated investor by any measure. I asked him a simple question: if you were investing for 50 years and had to construct a portfolio that consists of stocks and bonds that you couldn't change, how would you allocate between the two? His quick and confident response of 100% stocks because everyone knows that stocks beat bonds over the long run was not a surprise. If I posed the same query to 100 people with his background and expertise, I would likely receive the same answer the majority of the time, if not every time.
His response was expected, but my rebuttal completely caught him off guard. “Why would you say 100% stocks when the two asset classes have about the same return over the long run?” I added, “why wouldn't you split them 50/50 since you can get a similar return with much less risk?” He was puzzled. I clarified my point by explaining that bonds could easily be structured to have similar return and risk as equities with some leverage and did not have to be held in their conventional form. I spent a few minutes describing that the only reason bonds are considered to have lower returns than stocks is that they are less risky, and that characteristic can be easily adjusted. I shared some long‐term data to back my point. He said that in his 30 years in the business no one had ever brought up this idea to him because stocks beat bonds over the long run was such a basic assumption that it was beyond reproach. In short order, he restated his answer as 50/50 and was amazed at how quickly and easily a core tenet that he had assumed to be a fundamental investment truth for decades was disaffirmed.
We can do the same thing for other asset classes. That is, we can structure multiple asset classes to have equity‐like returns by adjusting their risk to match equities. Hence,