What’s a Normal Stock Return?

Here’s are two trick questions: Are you normal? Am I normal? Common sense tells us it’s impossible to answer either without also asking: Compared to what?

For example, check out this image of the Passmore & Felix team on our newly updated website. If we’re talking about “normal” heights, let’s just say I stick out a little. I’m not normal. If you’re measuring our level of commitment to our clients’ best interests, I’d say we all stand equally tall.

What about your investments? As investors, we care about market returns. They are one of the biggest unknowns in planning for the future, and yet they loom large in determining how much you need to save along the way and how you’ll choose to structure your investments. Real returns – the kind you have after accounting for inflation – drive decisions like when you can retire and how much you get to spend once you do.

So, every investor wonders about two levels of normalcy. How is my portfolio doing? (Has it delivered at least “normal” returns?) And: How should I expect my portfolio to perform going forward? (What will “normal” look like in the future?)

Poking Into the Past

We know very little about the future, but at least we are blessed these days with easy access to historical data that allows us to understand the past.

The Credit Suisse Global Investment Returns Yearbook 2018 gives us data on stock returns going back 118 years, to 1900. It is updated every year. This study includes real returns for 23 countries, expressed in US dollars. For the full period, real equity returns for a global portfolio were 5.2% (in excess of inflation). This number includes total losses for Russia in 1917, and for China in 1949. After going to zero, both of those countries’ returns series re-entered the data set in the 1990s.

So going back as far as we have reliable data and including two stock markets going completely to zero, real returns have been 5.2% per year on average since 1900.

Referencing the same source and the same 118-year period, stocks have returned 4.3% per year compared to Treasury bills. We call this difference the “equity risk premium,” or the amount of additional return investors would have received for investing in stocks compared to investing in the risk-free asset.

These numbers might begin to bring past annual returns into focus, and potentially offer a better lens through which to gaze into the future. But as I’ll explain in a moment, it’s still only modestly useful in defining “normal.”

Normal Isn’t the Norm

The challenge is, while the data I’ve just shared may be accurate over very long periods of time, it tells us nearly nothing about what to expect in any given year, or even in any given decade. For example, while the average equity risk premium across the past 118 years may have been a little over 4%, the story changes dramatically if we look at rolling 10-year periods during that same timeframe.

In the US, stocks have underperformed Treasury bills 15% of the time over 10-year periods going back to 1926. The same is true for Canadian stocks going back to 1970. Ten years is the blink of an eye to the stock market, but investors tend to evaluate their returns at least annually. For them, a decade is an excruciatingly long time to wait for positive, risk-adjusted performance. Even if their retirement is decades away, many investors don’t have the patience to see it through.

Even considering one year at a time, the dispersion of annual stock returns is all over the map. I don’t have data going back to 1900, but I do have it going back to 1970 for Canadian, international, and US stock market indexes. In that 47-year period, compound average returns (after Canadian inflation and in Canadian dollars) were 5.22% for the S&P/TSX Composite, 5.26% for the MSCI EAFE, and 6.96% for the S&P 500.

That’s pretty close to the 5.2% long-term average for global returns cited above. But, in all three markets during those 47 years:

The exact average annual real return was not earned in any single year in any of these markets.
Annual real returns ranging between +3% and +10% occurred only 11 times in Canada, 6 times in the U.S. and 7 times internationally.
Annual real returns somewhere between –8% and +15% occurred about half the time. That means the other half of those 47 years saw either very high or very low annual returns.

Lessons from the Past

So again, I ask you: What’s “normal”? As an investor, here are your take-homes:

“Normal” isn’t the annual norm. Based on the evidence, it’s pretty clear that real returns are more likely to deviate from than reflect the norm each year, and even each decade. Often wildly so.
Hope for the best; plan for the worst. Choose an asset allocation you are pretty sure you can live with for a long, long time – through very thick and very thin. Remember, a worst-case scenario can be worse than anything we have seen in our lifetimes.
Stay the course. Most important, brace yourself. Stay invested regardless of how bad things look. Even if the market is crashing, there is no reason to believe that trying to avoid the downside will make you better off. On the contrary, there is plenty of evidence that it’s more likely to leave you worse off. Peter Lynch has claim to my favourite quote on this: “Far more money has been lost by investors preparing for corrections or trying to anticipate corrections than has been lost in corrections themselves.”

Preparing for the Future

You’ve now got a handle on the magnitude and dispersion of historical market returns, and why “normal” is such a squishy subject. Still, we’ve got to do what we can to prepare for the future. And while it’s a fool’s errand to try to predict outcomes for any given year, we can build some reasonable expectations for investors who can stay disciplined over the long-term. This remains an important part of the investment planning process.

We have talked about long-term risk premiums. If anything, it might be reasonable to believe that long-term risk premiums will persist in some form, at least on a relative basis, for assumptions such as stocks beating bonds over the long haul. We also know that there is information in prices. Research has shown a relationship between current valuations and future returns, where high valuations might mean lower future returns.

In a 2016 white paper entitled Great Expectations, my PWL Capital colleagues Raymond Kerzérho and Dan Bortolotti explained the process we use to estimate future returns for financial planning purposes. To summarize here:

First we estimate inflation, which currently comes in at 1.7%.
Next we look at the 50-year historical real return of each asset class. The fancy term for this is the “equilibrium cost of capital,” which helps us estimate an historical equity risk premium.
As we might expect, relying on the historical return is not an ideal way to know what the future has in store. So we also use a second estimate based on current yields according to something known as the Shiller CAPE method. This method tells us the current price over 10-year trailing earnings, adjusted for inflation.

Both methods of estimating expected returns have shortcomings, but they complement each other well. Their average results in a figure that we feel reflects current conditions, but is also more stable due to the inclusion of the 50-year risk premium. Again, for a more detailed walk-through of how to estimate future returns, check out Raymond and Dan’s paper.

Have we figured out what “normal” is yet? The truth is, we’ll never know for certain, except perhaps in retrospect. Still, having forward-looking return expectations is a very important part of investing. Taking care to make a reasonable estimate for future returns should be considered essential for building a sensible financial plan. It’s equally essential to ignore those same estimates when evaluating returns over relatively short periods of time.