Which performance measure is most relevant?
Let’s say a client asks you, “What was the best-performing fund last year?” Or perhaps they want to know, “What was the performance of the S&P 500 last year?”
Before answering, it would be best to ask some clarifying questions.
First, what are the assumptions the performance is based upon? Was it a lump-sum investment or periodic investments over time? Or is performance referring to a retirement portfolio in which money is being systematically withdrawn?
There can be a vast difference in the performance of a fund or portfolio when money is coming in versus when it’s being withdrawn.
The industry standard assumptions behind all published returns are the following:
- Lump-sum investment
- No additional money invested
- No money withdrawn
- No taxation
- No inflation.
As shown in the table “Performance in the new millennium,” the 20-year performance of five different investments was measured in three different ways:
- A lump-sum investment of $20,000
- An annual investment of $1,000
- Systematic withdrawal over 20 years from a portfolio with a starting balance of $250,000 (assuming a 5% initial withdrawal and a 3% COLA for the 19 remaining annual withdrawals).
Five different investments were considered. First, a 100% cash investment. Second, a 100% bond investment. Third, a 100% large-cap U.S. stock investment. Fourth, a 60% large-cap stock/40% bond portfolio. And fifth, a multi-asset portfolio consisting of 12 equally weighted indexes. Performance for each of the five investments was calculated using well-known indexes, listed in the footnotes of the table. The 60/40 and 12-asset portfolios were rebalanced annually.
As shown in the table of results, the best-performing investment over the past 20 years (Jan. 1, 2000 to Dec. 31, 2019) was the 12-asset portfolio, if we assume a lump-sum investment. The 20-year return for the multi-asset mix was 6.82%, followed by 6.12% for the 60/40 portfolio, 6.06% for the S&P 500, 5.03% for aggregate bonds and 1.64% for 90-day T-bills.
If we make a different assumption, such as an annual investment of $1,000 over the 20-year period, the best performer was large-cap U.S. stock. The second-best performer was the 60/40 portfolio, followed by the 12-asset portfolio, U.S. bonds and, finally, cash.
Under the assumption of a lump-sum investment, it makes no difference in which order the returns occur over the period in question. The ending outcome will be the same no matter how the annual returns are scrambled.
However, when assuming a scenario in which money is being invested periodically, the performance advantage will go to whichever investment had the best returns in the latter years of the time period.
For instance, the last year of this particular 20-year analysis period was 2019. In 2019, the S&P 500 had a return of 31.49%, well above the 22.38% return of the 60/40 portfolio and the 16.57% return of the 12-asset portfolio. In 2017, the S&P 500 had a return of 21.83%, compared to 14.52% for the 60/40 portfolio and 11.99% for the 12-asset portfolio.
The large margin of victory for the S&P 500 in those two years — at the end of the time period being measured — was the reason the index was the better performer when assuming a systematic annual investment over the 20-year period. As shown, under the assumption of a $1,000 investment at the beginning of each year, the S&P 500 ended with a final account balance of $59,408. That next closest was the 60/40 portfolio at $48,082, with the 12-asset portfolio at $41,481.
Let’s now turn to yet another way to measure performance, namely when money is being withdrawn from a portfolio, such as during retirement. This withdrawal scenario assumption is very sensitive to the performance of the investment in the early years of the time period. So, once again, the sequence of returns matters — in this case, a lot.
The differential in performance among the five investments is startling. The 12-asset portfolio had an ending balance of $327,685 after 20 years, assuming a starting balance of $250,000. The next-closest investment was U.S. bonds at $182,745. The 60/40 portfolio was third at $97,085. U.S. cash and U.S. large-cap stock both ran out of money before the end of the 20-year period.
How did this happen? In the case of an all-cash portfolio, the returns were simply too low to sustain a 5% initial withdrawal rate combined with a 3% COLA. In the case of large-cap U.S. stock, it was the performance during the first three years that doomed the retirement portfolio. In 2000, the S&P 500 had a return of -9.10%, followed by a return of -11.89% in 2001 and -22.10% in 2002.
By comparison, the 12-asset portfolio had returns of 5.43% in 2000, -2.09% in 2001 and -0.56% in 2002. The 60/40 portfolio had a return of –0.81% in 2000, -3.75% in 2001 and -9.16% in 2002.
If you’re not familiar with Excel, it’s never too late to start.
Which performance measure is most relevant?
Ironically, the standard industry assumption of a lump-sum investment is arguably the least common way to invest — and therefore the least relevant measure of performance. Very few people only invest a single amount of money once in their lifetime. Far more common is a monthly or annual investment, such as into a 401(k) or IRA.
Conversely, when a client begins retirement, they generally withdraw money monthly, quarterly or annually from their portfolio. Thus, those two performance assumptions are much more realistic than a lump-sum return.
Sadly, performance based on those two more common assumptions is not reported by any data provider I’m aware of. Rather, the performance reported by mutual fund companies and data providers such as Bloomberg and Morningstar is based on the assumption of a lump-sum investment. If you want to calculate performance under different assumptions, you need to do it yourself using Excel, or perhaps using web-based tools that allow you to specify performance assumptions.
My point here is not to be critical of using a lump-sum investment as the industry standard assumption. There has to be a common assumption across the industry and the various data providers, and lump sum is a defensible choice. It just happens to be a measure that doesn’t reflect how most clients actually invest money into — or withdraw money out of — their portfolio. Said differently, it’s a valid quantitative measure of performance, but not one that’s relevant to real people — meaning your clients.
As a result, it’s incumbent upon advisors to become familiar with calculations of performance that involve multiple investments over time, as well as when money is being withdrawn systematically from a portfolio. Spreadsheet skills will come in handy. If you’re not familiar with Excel, it’s never too late to start.