How to Evaluate Investment Performance
When evaluating an investment portfolio, it's critical to review returns as well as the path to those returns. The optimal portfolio is not the highest-returning portfolio, but the highest-returning portfolio that you can stick with over the long run.
Ground Rules for Evaluating Investments
Before getting into the investment statistics I like to focus on, let's start with some ground rules. Two prerequisites for any performance analysis are:
- Theoretical Backing
- Data across multiple economic environments (and the more the better)
Without both of these items, evaluating a strategy could be interesting, but not all that valuable.
Theoretical Backing
Firstly, before evaluating any portfolio, you should understand and be able to articulate why the strategy should continue to work in the future. Otherwise, you may be simply extrapolating past performance or cherry-picking a winner.
"It's done well in the last five years" is not sufficient theoretical backing. And if you hear or are tempted to say "It's different this time," then you should question the investment's long-term viability.
Changing Economic Environments
Although it's tempting to think time (for example, 10 years of returns) is ample data to evaluate a strategy, it's far more important to test the strategy in different economic environments. Even a decade or more may not have a lot of variety in the economic factors at play.
For example, the decade from 2010 through 2019 was dominated by economic growth, stable prices, and below-average inflation. Not a lot of economic variability during this period and also an environment where stocks and bonds tend to do exceptionally well.
Economic environments can broadly be classified across two axes: economic growth (growth or recession) and price changes (inflation or deflation). Below is a visual of this broad characterization, with corresponding historical periods for reference.
Ideally, you want to evaluate a strategy across all four of these economic environments.
How to Compare Investment Strategies
Once you've met these prerequisites, you can evaluate and compare an investment strategy to other strategies using the following data points. Although the overall return is important, the path to realizing those returns and the consistency of those returns is just important. And that is just what these statistics aim to quantify.
The following statistics are not exhaustive but represent some of my favorite data points in evaluating a strategy.
Annualized Return
Also called Cumulative Annualized Growth Rate (or CAGR), this is the average rate of growth of an investment over a specific period of time, usually multiple years. The CAGR takes into account the compounding effect of the returns on the investment, which means that it assumes that the investment earns interest on the interest earned in previous years.
While it is tempting to only focus on returns, the CAGR can mask significant differences in the experience of an investor.
For example, two strategies over identical time horizons could have yielded 9% CAGRs. However, the path to those returns could be massively different.
By focusing only on CAGR, you might be tempted by strategy 1 with its slightly higher return. However, a quick glance beyond CAGR to some additional statistics we will review in this post will likely have you shun that same strategy.
Standard Deviation
Standard deviation is a measure of the volatility, or ups and downs, of an investment's ride. More specifically, standard deviation measures how much an investment's return tends to deviate from its long-term average.
All else equal, a lower standard deviation is preferred.
For example, if an investment strategy has a 9% CAGR and standard deviation of 16% (similar to Strategy 2 above), it is not unusual to see annual returns significantly different than 9%. In fact, returns significantly different than the CAGR should be expected.
Here are seven consecutive years of annual returns for Strategy 2. Note how different these returns are from 9%.
Ultimately, an investment with a high standard deviation is expected to generate returns far different than its long-term average. While investments with lower standard deviations are expected to generate returns closer to the long-term average.
For this reason, all else equal a lower standard deviation is preferred.
Sharpe Ratio
The Sharpe Ratio, developed by Nobel laureate William Sharpe in 1966, is a nice statistic that ties return and risk together.
The Sharpe Ratio measures the investment's excess return per unit of risk. Excess return is calculated by subtracting the risk-free rate (the rate an investor could earn by holding cash or short-term Treasury bills) from the investment's return.
All else equal, the same return with less volatility is going to yield a higher Sharpe Ratio.
The Sharpe Ratio is most helpful in comparing similar investments of like return profiles. Used on its own (without considering overall return), the Sharpe Ratio can yield suboptimal results.
For example, consider a comparison of global stocks vs. U.S. investment-grade bonds for the 20 years from 2002 through 2021. Bonds had a far superior Sharpe Ratio during this timeframe, 0.86 vs. 0.54. However, long-term investors with the ability to withstand volatility may still prefer stocks over this time frame.
Despite a greater Sharpe Ratio, bonds returned less than half the returns generated by stocks in this timeframe, 4.2% vs. 8.5% CAGR. An investment of $10,000 in global stocks would have grown to over $50,000, while the same investment in bonds would have grown to less than half that, $22,686.
Sortino Ratio
The Sortino Ratio, developed by Frank Sortino in the early 1990s, is a close cousin to the Sharpe Ratio. The Sortino Ratio tweaks the Sharpe Ratio to only include downside deviations from the average return, instead of "penalizing" upside deviations.
This makes a lot of sense. As investors, we are happy when we get upside surprises. The volatility we are really concerned about is the volatility to the downside.
Oftentimes, comparing the Sharpe and Sortino Ratios of two strategies will yield the same conclusion. The portfolio with a superior Sharpe Ratio tends to have a greater Sortino Ratio. However, the Sortino Ratio can add additional color to the risk-return profile of an investment strategy and can act as a tiebreaker in the event of similar Sharpe Ratios.
Maximum Drawdown
The max drawdown is the worst peak-to-trough decline of an investment over a specified time period. Put another way, it's the maximum decline an investor would have had to withstand on the way to the return (CAGR) over that time period.
Let's go back to our original example of two investments returning over 9% per year over the same time period. The maximum drawdown for each of these strategies is drastically different. And any investor should question whether they would be able to stomach and ride out a 72% loss in their portfolio.
Although this statistic is quoted in percentage terms, it's helpful to put this percentage decline in perspective by applying it to your investment portfolio. What would this max drawdown statistic equate to in dollar terms?
When looking at historical data, perhaps a 20% or 30% drawdown may not seem all that significant. But as you build your portfolio and approach retirement, these dollar values grow exponentially. A stock market crash in your 50s likely feels a lot different than a market crash in your 20s.
Length of Drawdowns
In addition to understanding the worst drawdown in an investment's history, the time it took the investment to recover is also important.
Did the investment bounce back within a couple of years and set new all-time highs? Or did it take several years to recapture lost ground (as was the case for stocks after the global financial crisis in 2008)?
For example, you may come across three strategies with the following rates of return over 20+ years: Portfolio 1 has the best returns, Portfolio 2 has lagged Portfolio 1 by 1.0% per year, and Portfolio 3 has lagged 0.5% behind Portfolio 2 annually. While 1% to 1.5% per year is meaningful, reviewing the severity and length of drawdowns helps to explain the ride to those returns.
- In order to achieve Portfolio 1's returns, an investor would have had to suffer through four separate 20% declines, including a more than 50% decline that took over 5 years to recover.
- With Portfolio 2, you would have missed out on 1% returns per year, but would only have had to stomach about half the downside as Portfolio 1.
- Portfolio 3 is about 1.5% behind Portfolio 1 annually but provided its investors with a much smoother ride. This strategy is by far the easiest to stick with over time.
The length of drawdowns is particularly important if you are withdrawing from your portfolio. A down market combined with withdrawals can negatively compound to make it very difficult for your portfolio to recover.
Rolling Returns
Reviewing rolling returns helps understand the consistency of returns. By reviewing the average, minimum, and maximum return of an investment over 1-, 3-, 5-, and 10-year timeframes, you can understand the variability of returns, even over longer time horizons.
Continuing with the three example portfolios from the previous section, a similar story is told from a rolling return perspective. Portfolio 1's returns came at the risk of significantly varying 3-year returns (ranging from over 24% per year to -15% per year). Portfolio 3 never experienced a 3-year decline and its 3-year annualized returns had a much tighter range of 2.5% per year to 6.4% per year.
While helpful in assessing any investment, these statistics are particularly relevant if you are withdrawing from your portfolio and want to mitigate the "Sequence of Returns" risk.
Sequence of Returns risk is a concept underscoring the importance of the order of returns, not just the overall return. Two investors with the same long-term return can have very different outcomes based on the order of returns and their cash flows.
A 30-year annualized return of 6% may seem perfect for an investor only withdrawing 4% per year. But if the investment declines for the first decade of the investor's retirement, before rebounding and producing outstanding returns in the next 20 years, this could lead to issues for the sustainability of the plan.
Understanding rolling returns, and specifically the worst returns over 3- and 5-year periods is particularly relevant to investors nearing and in retirement.
Conclusion
Analyzing investments goes far beyond returns. Ultimately, we want to maximize our returns for the risk, volatility, and stomachaches we have to endure in our chosen strategy.
Before selecting an investment strategy based purely on performance, use some of these statistics to put past performance in context and differentiate between seemingly comparable strategies.