Understanding Market Returns
Having a background in financial analysis, I’ve learned you should always question rates of return when you hear them. A “10% return” can mean almost anything if you don’t have the backup for how it was calculated and the assumptions which went into it. Never take a return statement at face value! This post is about stock market returns – we’ll look at understanding historical returns and then discuss what I feel is a reasonable long term stock market return projection.
One of the most important things to remember when creating a forecast (and the truest) is it will not be accurate! Therefore, I tend to err a little bit on the conservative side in this area because I’d rather save too much than run out of money later in life. No one has a crystal ball which can show us what the stock market is going to do, aside from a very long history demonstrating it tends to move upward when you look at it with a long enough time horizon (with the occasional nasty drop!).
But the fact that most of us can’t predict the future doesn’t mean there’s no value in making forecasts. In fact, I think its incredibly important to use a reasonable rate of return forecast so we can back into things like what our savings rate should be.
The Average Return Fallacy
One common way of stating returns is the ‘average return’. For example, if you have $100 and it goes up during the first year by 100% and then loses 50% the next year, your ‘average return’ was 25%. Let’s look at the math…. (100 – 50) / 2 years = 25%. But what actually happened year by year? We started with $100, then it went up by 100% and we now have $200. Then next year we lose 50% and are left with $100. So, we ended up exactly where we started. Now is it still fair to call this a 25% return? No way! This would be incredibly misleading.
And yet this is what so many people out there do – even Dave Ramsey, whom I greatly respect for all the people he’s helped, uses this very flawed approach when he says we can expect 11% to 12% stock market returns. His logic on this seems to be that, since we can’t predict the future, we may as well use the highest number mathematically provable (the S&P has technically produced this as an ‘average return’ over the last 70+ years) as a way to entice people to invest more, which is to their overall benefit. I strongly disagree, and feel this may even lure people into a false sense of safety. I’ve even heard some people reference this figure as a safe withdrawal rate (yikes!). In addition to ignoring something we call ‘volatility drag’ which results in real returns being lower than the simple average, it also ignores inflation.
Because of these factors, I call this madness the average return fallacy.
Fallacy: a mistaken belief, especially one based on unsound argument
The correct way of calculating stock market returns is to use the compound annual growth rate (CAGR), also sometimes called the Geometric Mean Return. The formula for this is a bit more complicated than a simple average, but still fairly easy and it properly accounts for volatility drag. It looks like this:
[ (1 + r1) x (1 + r2) x (1 + r#…) ] ^ 1/n – 1
In our example, we just take (1 + 1) x (1 + .5), find the square root, and subtract one to arrive at 0% – which would be accurate, since we started and ended with the same amount of $100. Once we obtain the CAGR, we can then back out inflation.
So, if 11% to 12% isn’t the right number to use, what is? To get that, let’s take a look at the historical returns for the S&P 500 which, for purposes of illustrating the concept in this post, we’ll say is reasonably representative of the overall market. Below is a chart which breaks down historical returns in 10 year increments through 2017 with and without dividend reinvestment, and with and without inflation. For fun, I also included the ‘simple average’ so we can see just how misleading this is. Note that obtaining accurate historical return data is more difficult and nuanced than you might think – when going back to the pre-computer age you have to take returns with a bit of a grain of salt and accept ‘close enough’.
Historical Market Performance
I’ve calculated all this with data complements of Yale (http://www.econ.yale.edu/~shiller/data.htm). You may see slightly different results on other calculators floating around the interweb – it appears the Yale data is using monthly averages for their historical prices rather than end of month prices (to each their own!). The difference isn’t large, and it was nice having a single free data source for S&P returns and inflation to create this post.
When looking at the difference between the simple average and CAGR, it’s typically anywhere from one to two percentage points. While this may not sound like a lot, it is! I compared the difference in growth from a $1 investment in 1947 through 2017 in the chart above using the two methods. As you can see, the difference is substantial – with the ending result from the simple average being two and half times higher than the real number (CAGR). If someone were using the simple average return in a financial calculator or spreadsheet for whatever purpose, it could really overstate the ending result. Think if I used the average return to help me figure out how much I needed to save for a retirement that was decades away – that could really understate how much I need to be saving!
Additionally, the above rates of return are more theoretical than practical. It would be impossible for an investor in the S&P 500 to have actually obtained those returns for the following reasons:
- Trading Fees
- Management Fees (at an advisory and fund level, if applicable)
- Bid/Ask Spread
- Imperfect timing
All the factors listed above, plus some others, result in returns lower than the stated historical figures to actual investors. Some of these things can be reduced and managed effectively, but there’s still an impact.
And don’t forget – when looking at return projections, it’s best to adjust your numbers for inflation. This way, we keep things more or less apples-to-apples. $1,000,000 might seem like a lot of money today, but it probably won’t be in 100 years. If we take 2.5% inflation out for 100 years, then that same $1,000,000 is only worth about $85,000 today. So, even if I was using 12% for the market return projection (wrong), and thought I could actually get that (you can’t – at least not by holding S&P 500 index funds), you’d still be off quite a bit because you hadn’t accounted for the impact of inflation.
So, a lot of words to say – look at the column on the far right, highlighted in blue. That is the truest measure of apples-to-apples historical stock market performance – inflation adjusted CAGR with dividends included. And its a far cry from 11% to 12%! We also glean from this chart that reinvesting dividends is important!
I think a reasonable long-term forecast, based solely on historical data, for stock market returns is around 7% (inflation-adjusted) which, after adjusting for implicit and explicit costs of managing a portfolio, is likely closer to 6% (could be more or less depending on one’s level of discipline, behavior, and ability to keep fees low).
The 6% figure (or 7%, before I adjusted it down!) seems to be relatively consistent with the historical data. Nothing fancy here – simply taking a realistic look at the relevant data. 6% is quite a bit different than the 11% to 12% figure we often hear, and the impact of this is difficult to understate – it is very important for people to understand this.
Let’s say I want to accumulate $1,000,000 in 25 years. Using the the ‘simple average’ return of 12%, to accomplish this I would need to save $7,500 per year for 25 years. Of course what we really mean is “I want to save today’s equivalent of $1,000,000 25 years from now”. One might assume using 6% (which, as we discussed, includes an inflation estimate) would require saving twice as much, but its actually more than twice as much because of the nature of compounding. To accumulate $1,000,000 at a 6% rate of return assumption, it would require saving $18,227 per year.
Even if you think I’m too conservative at 6% or even 7%, it gets difficult to argue for a much better number. Meanwhile, Jack Bogle (founder of Vanguard / legendary investor) is talking about 4% return expectations for the next decade. I have to point out again, this discussion is only based on a review of historical data. And as we’ve all heard, past performance does not indicate future results.
What do I think the market will do? I have no idea – and I tend not to trust people who say they do! I see reasons in spade for both optimism and pessimism (or skepticism, at least). But long term I’m positive about the human capacity for ingenuity which will, over time, drive us forward.
I’ll end by encouraging you to double check the rates of return you are using in your own forecasts. I think there are a lot of folks out there who may need a reality check on their savings rate if they’ve been using much higher numbers.