Using Analytics in Wealth Management: The Good and Bad (Part 2)
Advisor Perspectives welcomes guest contributions. The views presented here do not necessarily represent those of Advisor Perspectives.
Part 2 – Capital Markets Assumptions
This is the second part of a series on investment analytics in the wealth management industry and how advisors are using analytics to communicate with clients and prospects. You can read part one here. We will look at some of the more popular analytics and go over what they are, their good points, and their potential pitfalls.
Since the financial crisis, wealth advisory clients have shifted their relationships from one of blind trust to a desire to internally understand and validate that a particular advisor’s recommendation is the right thing to do. Advisors have obliged, taking their conversation from “trust me” to educating and partnering with their clients around their finances and future. Ironically, many advisors have reported that this education and partnership approach builds trust.
The challenge becomes how to educate clients about their finances when they may have little or no financial training. Turning to portfolio analytics is one part of the solution. But most financial analytics are confusing for people outside of the industry. Forward-thinking advisors have been searching for and employing analytics very carefully, choosing only those metrics that are simple, understandable, and meaningful for people with no financial training. This series will explore some of these metrics, along with their benefits and pitfalls. Today’s topic is capital markets assumptions.
Capital markets assumptions
Capital markets assumptions are an estimate of the going-forward long-term returns for various asset classes or investments. For example, a firm might say they expect the U.S. equity market to return an average of 7% per year over the next 20 years. Capital markets assumptions are a critical component of portfolio construction in wealth management and are taking on an even larger role with the rise of financial planning.
One of the key components of financial planning is the projection of a client’s retirement savings balance. While savings rate can be a critical input to this projection, the assumptions the advisor uses to project investment returns are equally critical, depending on the horizon.
Since the future is unknowable, coming up with a sound prediction of investment returns is a tricky business. It also causes compliance departments and legal disclaimer authors to narrow their eyes and stare intently.
In this article, we will look at some of the popular methods for developing capital markets assumptions and what makes each good and bad.
What are they?
One straight-forward way to forecast investment return is to look at what the returns have been in the past. An advisor will typically use the average return over some appropriately long historical time frame, preferably encompassing at least one business cycle. Things like volatility and correlation can be calculated the same way, if required.
Simple and straight-forward. Even investors with no financial training understand the notion of how an investment’s past performance could be a basis for predicting how it might perform in future despite the fine print that tells them exactly the opposite. This gives clients a point of validation that allows them to move forward with the rest of the analysis without too much in the way of anxiety.
Easy to calculate. The math behind calculating historical return and historical volatility is simple and quick. If the advisor is also calculating correlation, that can be a bit more computationally intense, depending on the number of investments being considered. But it is still a simple, closed-form calculation. While the rise in computing power has shaved away a bit of the appeal of simple calculations, scalability still has its proponents and rightfully so.
I occasionally still see people forgetting to include equity dividends when they calculate their historical returns. This is a big no-no and probably gives you some legal exposure since it is clearly not an academically accepted method.
Objective. Given the sensitive nature of predicting the future, it can be nice to say that your process is 100% objective. Of course, selecting the time horizon for the calculation is purely subjective, but once that is set your compliance department can rest easier knowing that you are using an objective and systematic approach. Just make sure that the client understands exactly what you are doing and agrees with it. Many compliance departments favor using broader asset classes as opposed to individual investments for regulatory reasons.
Things to be cautious of
Panic in a crash. Apart from getting new clients, one of the toughest parts of an advisor’s job is trying to calm clients down when the market is crashing. Unfortunately, using historical returns makes that job much more difficult.
Consider the case of a stock market crash of around 50%. Just prior to the crash, you provided your client with a financial projection showing they would have ample savings by retirement given their prior balance and your historical return assumption. Assuming that they were 100% in the equity market, that same analysis would show a significantly lower savings balance at retirement since they are starting from a much lower base and the return assumption is going lower as the recent market data points are incorporated into the average.
This can result in your client jumping to one of the following conclusions: you lied to them, you put them in the wrong portfolio, or the market is simply broken – probably all three. That’s a tough ledge to talk someone down from. Some of the other methods we will look at give you the ability to correct for this.
Sensitive to time period. While it is nice that the calculation is objective, the results can vary significantly depending on the historical time period selected. To help mitigate this, lengthy historical periods are typically selected, but an overly long period might not capture some of the nuances of modern markets.
Fails to account for current market conditions. Ignoring current market conditions has drawbacks. For example, there is a consensus around the notion that markets tend to have higher future returns when valuations are attractive. Bond returns are likely to be much lower than in the past given that yields are near historical lows and it is mathematically improbable that they will go much below zero. These conditions should figure into forward estimates.
Instead of using a one-size-fits-all approach like historical returns, some firms break the investible universe down into discrete asset classes and establish a separate estimation method for each. As a simple example, consider stocks and bonds. A firm could assign a valuation-based estimation method for stocks and a more mathematical method for bonds. Some asset classes even have long-term forward contracts that can be observed to determine what the market estimate for long-term returns will be.
Given the number of estimation methods available, I won’t go into them individually. Suffice it to say that they include valuation-based, market-based, mathematical, and even historical. I would like to rule out simply guessing, but I have seen that too.
Can incorporate current market conditions. The flexibility to choose more forward-looking estimation methods allows for the inclusion of current market conditions into the estimate. This overcomes one of the major disadvantages of the historical return method. Of course, it is still possible to use historical return as one of the estimation methods, if desired, but using current market conditions could help in a crash.
For example, if you adopt a valuation-based approach for the equity market, your equity return assumption will increase as the stock market decreases. If all goes well, you will have a much better time trying to hold your clients in the market during a crash because the increase in expected return will partially or fully offset the decline in value when looking out to retirement. It would be a great comfort to your clients to know that what you have been telling them all along is true: they will still be okay in the long run if the market crashes.
More robust than simple historical return. “Robust” is a term academics use when they are concerned about using the word “accurate.” In theory, by providing the ability to choose the most robust estimation method for each asset class, the overall result should be more robust as well. This assumes that different asset classes can benefit from different estimation methods, which seems like a reasonable assumption to us.
Can be updated automatically. Depending on the individual estimation methods used and the data feeds available, the assumptions can be updated automatically and rapidly as market conditions change. This would be particularly useful in our market crash example. Automation also reduces cost.
Things to be cautious of
More complicated. Depending on the types of methods chosen, this approach loses the simplicity and intuition of the historical-returns method. Even if the individual methods are simple and intuitive, the mere fact that the advisor must explain them for the client to understand what is going on makes it more complicated and prone to confusion/anxiety/lack of trust.
More complex. No need to google this; I already did. a I am using “complicated” and “complex” correctly. Unless you choose the exact same estimation method for each asset class, you are going to end up with several different estimation systems that need to be executed at roughly the same time. This introduces computational complexity that can increase costs. Still, depending on the methods chosen, it may still be possible for it all to be done in a spreadsheet, and as depressing as that is, it is likely the case.
What are they?
Macro approaches apply macroeconomic principals to the process of estimating future returns. Broadly, there are two main types.
The first is a building-block approach where you start with a base return, which is generally the risk-free real rate of return. Then, you add inflation. Finally, you estimate and add a series of appropriate risk/liquidity premiums for each asset class. This results in a set of returns that increases as risk increases and/or liquidity decreases in some sensible proportion. There is a good deal of arbitrage theory tied up in this, but most people accept the axiom of “more risk, more reward” in the long run.
The second is the full-blown macroeconomic model. This involves complex computer models of all the intricate relationships between various macroeconomic factors and markets. The result can be a massive global macroeconomic model with tens of thousands of inputs similar to a weather model, except people are more comfortable with the longer term forecasts than the shorter term ones.
Highly robust. From an academic standpoint, this is robust with a capital “R.” If you are impressed with PhDs and intensive study, and we can all admit this at least a little, then this is the method for you. There is some value in this level of pontification. And, while the wealth management industry is moving towards promoting greater client understanding and intuitiveness, there is a “shock and awe” associated with an effort that connotes value and expertise.
Things to be cautious of
Virtually unintelligible. While unintelligible isn’t the right word, it gets the point across. A typical client will not have any understanding or way to validate these methods. This means reverting to the standard “trust us” approach. Unless you have an excellent and trusting relationship with your client, no internal understanding can lead to anxiety, which can then lead to inaction.
Expensive. Depending on the deployment, you will need to hire at least one PhD and possibly several. This can make it prohibitively expensive unless you are a large bank or asset manager.
Laborious. Working up capital markets assumptions using these methods takes a lot of effort and time. As a result, many firms using this approach only update their capital markets assumptions once a year. This may be adequate if you are a large pension fund or small country, but the average wealth manager is having daily conversations and those conversations can suffer if the assumptions are not up to date. Consider again the case of an active market crash.
While capital markets assumptions are a uniform concept, the methods for calculating them are very different. As usual with analytics, there is no runaway best method, so advisors must understand the plusses and minuses of each method they have available.
Stay tuned for my next article, which will explore Monte Carlo analysis.
Kendrick Wakeman, CFA, is the founder and managing partner of qSpur, a consultancy operating at the intersection of finance, technology, and people. Prior to founding qSpur, he was the founder and CEO of FinMason, a financial technology company that built one of the world’s largest and most scaled investment analytics platforms. He has over 30 years of institutional investment and technology experience across a broad range of asset classes and has developed, built, and run institutional-level analytics and risk management programs at both large and small financial institutions.