I was playing around with a pretty cool tool called Portfolio Visualizer and got to thinking about the difference between how finance professionals and ordinary people talk about investing.
For finance types, an asset or a portfolio has an expected return, which might be calculated using historical returns or a forward-looking rule like Jack Bogle’s “reasonable expectations” formula, and then it experiences volatility, typically expressed as a standard deviation around that average return. So the stock market has higher expected returns than bonds, which are supposed to compensate for their higher volatility, which financiers also sometimes confusingly call “risk.”
But ordinary people aren’t concerned with volatility; they’re concerned with losses, and to a lesser extent gains. No civilian would say “I’m worried because I was expecting an 8% return but instead received a 20% return; this asset is too risky.” Nor would they say “I was expecting a 1% return but instead lost 0.5% thanks to the asset’s low risk.”
But it’s even worse than that! Even when financiers describe an asset or portfolio’s falling price, they express it in percentage terms, while ordinary people describe their gains and losses in dollar terms. Losing $20,000 in a $100,000 portfolio may hurt more than losing $10,000 in a $20,000 portfolio, despite the decline being 60% smaller in percentage terms.
While pondering this, I thought of a simple way of describing how risky a portfolio would be experienced by an actual investor.
The pain ratio
The pain ratio looks not at average annualized returns, but instead dollar returns, which means it takes into account the fact that smaller percentage declines in a larger portfolio can be more painful than larger percentage declines in a smaller portfolio. To calculate it, I plugged three simple portfolios into Portfolio Visualizer, a 40/60, 60/40, and a 90/10 stock/bond portfolio, and looked at dollar returns over a 30-year time horizon from 1987 to 2017.
Over this period, as you’d expect, the 90/10 portfolio outperformed the 60/40 portfolio, which outperformed the 40/60 portfolio, and was also much more volatile, with an average annualized return of 10.04% and standard deviation of 13.51%, compared to the 40/60 portfolio’s return of 8.19% and standard deviation of 6.56%, respectively.
Over 30 years, that means $10,000 invested in the 90/10 portfolio in 1987 returned a net of $184,117 compared to the $104,671 returned by the 40/60 portfolio. However, that net return is composed of two figures: the total gains experienced by the portfolio and the total losses. In the same 30 years, the 40/60 portfolio experienced just $10,273 in losses in just 3 down years, while the 90/10 portfolio lost $51,917 in 6 down years (meaning it also experienced $236,034 in gross gains).
You can find all the raw data I used and my calculations here.
Sequence of return risk and the pain ratio
When people refer to “sequence of return” risk they often mean the risk that a retiree will experience large losses early in retirement and that withdrawals from that lower base will permanently impair their ability to meet their needs in retirement.
But the pain ratio illustrates a different kind of sequence of return risk: that large dollar losses late in an investor’s accumulating years will permanently scare them into lower-returning assets, locking in both a lower asset value and lower future returns.
It seems to me that the answer is for investors and their advisors to look at expected returns and potential losses not just in percentage terms, but in dollar terms as well, and to consider moving into less risky assets before a relatively small percentage decline in a large enough portfolio scares them into locking in permanent underperformance.
Note that this is totally distinct from moving into less risky assets as you approach or enter retirement; it’s solely a function of the size of your portfolio and how you’ll subjectively experience the conversion of even small percentage declines into concrete dollar losses. This is just as risky for a young investor as an older one, since the pain ratio is based exclusively on the dollar size of the portfolio.
Why is it more painful for high value portfolio to lose $20k than low to lose $10k from &20k. Seems to me the latter hurts much worse. Somebody worked hard to scrap a bit of savings vs someone with lot of assets, often inherited or from something like rentals?
Mom,
Yeah, that’s the traditional finance way of looking at it, but what I’m saying is that in my experience talking to normal people about their investments, they’re much more focused on the dollar losses and gains than the percentage losses and gains. It’s much more common to hear “my investments made $40,000 last year” than “my investments made 4% last year,” even if it’s the same amount on a $1,000,000 portfolio.
And it’s not just for psychological reasons, although that’s what I focused on here. Taxes, for example, are levied on the dollar amount of dividends and capital gains, not the percentage, so it’s perfectly natural for someone who just filled out their tax return to know their dollar, but not their percentage, performance the previous year.
—Indy
Shouldn’t it matter whether the losses come after gains or after losses?
“Last year my portfolio went up $100,000; this year it’s down $20,000, so I’m averaging $40,000 per year. Cool.”
This is different from expectancy by only considering the short-term memory of the investor.