Historical (Stock) Market Returns
What can you expect from the US stock market as an average performance? And Europe? Asia? Is there a difference between large, mid and small caps? What about bonds? How did other approaches do, like real estate funds, gold, or high dividend stocks?
Let’s try to find out what stocks returned over the long term. US, Europe, Asia, developed versus developing countries.
The S&P 500
I recently read this quote on the S&P 500:
The Standard & Poors 500 Index returned 7.4 % annually, before inflation and including gross dividends”
As if that was a universal truth, valid for every longer period of time. Apparently they took the 20-year 1998-2018 return as the truth. The truth is that this is a truth just as any other. If you take a different start and end year, you can get very different numbers, even averaged out over the long term. It depends very much if you start at a cheap market or at an expensive moment.
What is cheap or expensive? You could obviously look at the past and identify when the stock market was at its bottom or top, but that apprach would suffer from ‘hindsight bias’ (i.e. knowing the future for moments in the past). So I took another approach: I looked at the Cape – Shiller index and identified those moments when the stock market was at high or low levels as identified from the P/E ratio, not as identified from its future performance.
I will spare you the details, here are the results (average returns of multiple periods in the past)
|CAGR TR Inflation Adj.||5 Year||10 Year||20 Year||30 Year|
|Start in Expensive Market||-2.5 %||0.1 %||3.0 %||4.4 %|
|Start in Average Market||12.2 %||7.8 %||5.0 %||6.4 %|
|Start in Cheap Market||17.8 %||13.8 %||10.0 %||8.2 %|
That’s a very different thing from saying ‘the S&P 500 returns you 7% annually’, isn’t it? If you start in an expensive market (higher P/E), your expected return over the next 5 years is negative! To the contrary, in a cheap market, your expected return is a whopping 18 % after inflation (so 20 % in absolute terms, before inflation).
If you ignore your starting position, an average Total Return including gross dividends After Inflation would be 6.4 %.
US Inflation is from 2.3 % (most recent 10 years) to 3.0 % (most recent 3 decades). That makes a Total Return Before Inflation in absolute terms of 8.7 % (taking the most recent 2.3 % inflation).
That’s the very average number you can expect long- term as a Total Return from the S & P 500:
- 8.7 % before inflation adjustment
- 6.4 % after inflation adjustment
Dividends (on the S & P 500)
Dividend yields don’t jump up and down as wildly as the capital gains, but they still vary a bit between cheap and expensive markets. Which is logical: in an expensive market, prices are higher and dividend yields which are “dividend / price” are therefore lower.
|Gross Dividend Return||5 Year||10 Year||20 Year|
|Start in Expensive Market||1.9 %||2.1 %||2.6 %|
|Start in Average Market||2.9 %||2.7 %||2.9 %|
|Start in Cheap Market||3.8 %||3.3 %||2.7 %|
You get some 2.8 % on average. Half a percent more if you start in a cheap market, half a percent less in an expensive market. (2.0 % today, we come back on that).
S&P 500 returns
Very-long-term, you can expect this:
- 2.8 % Gross dividend returns (today much lower, 2.0 %)
- 8.7 % Total return (capital gains + dividend returns); after adjusting for inflation, you are left with 6.4 % real inflation-adjusted profits per year.
- 5.9 % Capital Gains (green arrow). This is what the SPX index (the number currently December 2018 at 2633) really does. Dividends are not included in that number.
All this is the index return, not your investment return; an index fund will have taxes and broker and fund expenses.
Side note: inflation has often been much higher in the past. During the seventies and eighties, it was from 5.5 % to 7 % annually. If you add that to the 6.4 % real inflation- adjusted returns, you have a before-inflation stock market return of 12% and higher. Which is why you sometime read about an historically average stock market return of 10 – 11 % and higher.
For instance, the Total Return (Dividends included) (before Inflation adjustment) of the S&P 500 from 1930 – 2013 is 9.7 %. From 1928 to 2013 the Total Return of stocks was 11.2 %. And so on. These are misleading numbers to predict the future. The 10 % gains contain 4 – 5 % average inflation in those days. Today we have 3 % inflation or less.
Boring numbers you will say but actually a very interesting very long-term statistic covering 1900 – 2013 (what you would call VERY long term averages 🙂 ):
What do we learn from this table?
- the ‘Real Total Return’ on equities (last column) = inflation-adjusted cap gains + dividend return for the US over 113 years is 6.4%
- For ‘Europe‘ as a continent we need to take the average of a number of countries. The average is 3.9%. There is simply much less growth here compared to the USA.
- Japan: 4.1 %.
- Winners are Australia and South-Africa, both growth countries.
- Side note: dividend returns were much higher in the past (1900-1950). We don’t have 4 – 5% average dividend yields on an index today (except for Australia).
Check with stock market indexes
I took the S&P 500, Euro STOXX 600, Nikkei 225 (japan) and All Ordinaries (Australia) and calculated their long-term CAGR (annual returns)
|# years||€ STOXX 600||S&P 500||Nikkei 225||AORD|
|14 (2005 - 2018)||2.3 %||5.6 %||4.5 %||2.4 %|
|20 (1998 - 2018)||1.1 %||3.8 %||(bubble)||3.6 %|
|32 (1987 - 2018)||4.6 %||7.7 %||(bubble)||4.3 %|
|Average taken||3.4 %||6.4 %||4.5 %||4.3 %|
These are index returns, so before inflation adjustments, and no dividends included. Remarkable: the US is far ahead of the rest, Europe and Japan are lagging behind.
There is some relief in Europe with a higher dividend yield: 2.9 % (Europe) vs 2.0 % (S&P 500). The exceptionally low S&P 500 dividend yield comes from plenty of heavyweight stocks in the top of the S&P 500 with zero or low yield: Amazon, Google, Berkshire Hathaway and Facebook are examples of zero – dividend stocks in the top 10 of S&P 500. In the past, dividend yields were comparable to Europe (2.8 %).
Japan had the largest stock bubble in history at the end of the eightiest (after 1985), which lasted till 2000. This huge mispricing makes an objective return calculation difficult. Very long term we have 1965-2018 = 5.4 %. More recent years after the bubble give 4.5 % (2005- 2018). Let’s take that number because the years of super-growth-Japan (seventies, eighties) seem to be behind us.
So the Japanese returns are in the middle between Europe and the USA.
The Australian stock market index (All Ordinaries Index ) gives a 4.3 % return (1987 – 2018 = 32 years), but offers a historically very high dividend yield, thanks to plenty of high – dividend yielding (e.g. mining) stocks.
Conclusion on continents
|Country||Stock index return||Dividend yield (2018)||Inflation||Real (After Inflation) Total Return (calculated from Index - 32 years)||Real (After Inflation) Total Return (Dimson Table - 113 years)|
|USA||6.4 %||2.0 %||2.0 %||6.4 %||6.4 %|
|Europe||3.4 %||2.9 %||2.2 %||4.1 %||3.9 %|
|Japan||4.5 %||2.1 %||1.2%||5.4 %||4.1 %|
|Australia||4.3 %||4.0 %||1.8 %||6.5 %||7.4 %|
For an index investor, the USA and the world (-index) seem good choices. All continents have their bubbles and down-years, sometimes this goes in sync but sometimes not (like the Japanese bubble).
- the differences seem small, but a 2 to 3 % difference annually during 50 years make a huge difference
- these returns are before taxes, all expenses and inflation.
- this does not take the current state of the market into account. Some stock markets are overvalued and others undervalued; it is difficult to time the market but your starting position does matter enormously (see above with S&P 500 example).
Large Caps vs Small Caps
On the long term (last 80 years), small caps beat large caps with 2.3 %. That’s huge. Obviously that depends on the definition of small, mid and large caps. And that definition seems to depend on the country. A 300 million market cap is often considered a small cap in the US, while in Belgium it is at least a mid cap. But this higher return comes with a much higher volatility, as measured by the standard deviation.
They can also underperform the broad (larger cap) market during a long time (think decade(s)); investing is like fashion: one decade “value” is in full swing, other years “small caps”, etc. For their low volatility and more constant gains, large caps are considered ‘safer’ than small caps.
Also, there is a significant difference between ‘Value’ and ‘Blend’ (Value and growth blended). These are the numbers (Source: Dimensional Fund Advisors)
|Asset Class||Return 1930 - 2013|
|S&P 500 (= Blend)||9.7 %|
|Large Cap Value||11.2 %|
|Small Cap (Blend)||12.7 %|
|Small Cap Value||14.4 %|
Quick fact check: comparison of S&P 500 (largest 500 companies USA) with the Russell 2000 (2000 smallest companies of the Russell 3000) during 1987 – 2018:
|Period||S&P 500||Russell 2000|
|2005 - 2018 (14 years)||5.6 %||5.7 %|
|1998 - 2018 (20 years)||3.8 %||6.2 %|
|1987 - 2018 (32 years)||7.7 %||6.9 %|
On average, not much difference for small caps. For the total return, we need to add the dividend yield, but the yield on the Russell 2000 is on average a bit lower than the yield on S&P 500 (today: 1.35 % R2000 vs 2.0 % S&P 500). Which is also logical: the S&P 500 contains many established larger caps which offer less growth but pay out their profits to the shareholders. R2000 contains more small cap growers which rather re-invest their profits in company growth.
This fact check does not show the large advantage for small caps. They seem to be very much in line with large caps. Which still leaves an argument to diversify across all caps and not to focus only on small or large caps.
Governments pay you just a bit more than inflation. So looking at REAL returns (above inflation), there is not much left. The average over a century (1900-2000) for the US is 1.6 % above inflation; for Belgium it is even a negative number: -0.4 %. Many european countries have a negative return because of the wars during that period, which erased the value of most bonds. The USA 1.6 % number is more what you can expect in the future.
So bonds are safe to protect your money from erosion due to inflation, but not much more. A bond fund can also go up and down widely, as shown today (end of 2018: bond ETF’s have dropped this year). When interest rates rise, existing bonds (with low interest yields) drop in value compared to newer bonds (with higher interest rates).
Corporate bond returns are dependent on the bond quality as you might expect. “High-yield” bonds are also high-risk bonds, with a larger probability to default.
Investment – grade bonds (considered safe – high credit rating) had 4.3 % lower return than stocks from 1980 – 2013, or a yield almost the same as government bonds. High-yield bonds almost reach the returns of stocks (7 – 8 % range), but come with added risk and volatility.
A few basic simple facts on gold:
- they don’t produce profits like stocks
- the price depends on supply and demand
The long-term returns are this:
|Total Return Inflation Adjusted 1968- 2018||314 %||1424 %|
|TR Annual (CAGR)||2.9 %||5.6 %|
So gold performs some 2.7 % less than stocks, which equals the dividend return! Seems logical again: we are missing the dividend return of a profit-producing company. Somehow it is surprising that it returns even more than inflation and adds 3% price increase.
Berkshire Hathway as alternative to an index tracker
Why not? As a mega large holding it is kind of a fund itself. What returns did it give in the past and what can we expect in the future? The main difference between past and future is the ever growing size of this mammoth. You cannot take the results of the small BRK 1970-1980 and expect that BRK of today’s size will return the same or even comparable results.
|1980 - 2018 (38 years) Total return before Inflation correction||19.5 %||11.6 %|
|1980 - 2018 TR after infl correction||16.3 %||8.4 %|
|2004 - 2018 (15 years) TR before infl correction||8.7 %||8.4 %|
There you have it. Fantastic results for BRK 30 years ago, but hardly beating the average stock market today. What can you expect for the future? Very likely a result comparable to the most recent 15 years. Possibly 1 or 2 percent above the stock market return.
One more note on BRK: so far there have been no dividend payouts. All profits are used for new investments or for share repurchases. Whether this is good or bad depends on your personal preference: do you want dividend returns to live on or would you rather reinvest the dividends yourself?
Real Estate (funds) = REIT
reit.com is a very good site to get detailed information, so I will just try to give the bottom line.
There is an index since 1971, “FTSE Nareit U.S. Real Estate Index Series” which gives an excellent view over the last 46 years. The Total Return CAGR for “ALL REITS’ is 9.7 %, while the index return (excluding dividends) is 1.6 %. So 8.1 % comes from the dividends.Equity REITs (own property) have an even better return (11.8 % TR) while mortgage REITs (own mortgages) are a bit lower (5.6 %). Interest rates and inflation were much higher in the past, so a realistic expectation should subtract 2 percent or more, and a future 9.5 % annually would be achievable.
Returns for the rest of the world can be found here
Europe is much lower (almost half of US), Asia too; emerging equities are volatile and riskier.
As a conclusion, US developed markets equity REIT is the place to be, with 9- 10 % CAGR as a realistic total return.
High Dividend Yielding stocks
What if you invest in an index targeting the high dividend yielding equities, wouldn’t that be the best of both worlds’? Stable and high dividend income, together with growth of the equity prices? Google a bit and you find lots of contradictory data. From warnings such as The ‘dividend yield trap’: stocks having yields above 7 % often disappoint (they cancel their dividend), so instead of this nice high dividend you actually purchased lemons. To sites pimping high dividend yielding stocks.
A few facts:
- Dividend yields have decreased a lot over time; in the 1970’s, the average yield for US stocks was 5.50 %; today (2018) , the S&P has an averageyield of 2.0 %.
- Dividend yields differ per sector. The average for biotech is 0 %, the average for REITs is 6.6 %.
- Companies have to choose what to do with profits. In the past (1970), dividends was the obvious choice. Today, share buybacks are in fashion.
- Companies with high growth perspectives prefer to invest in their own growth, while the old-stable-low growth companies prefer to return profits by dividends. Which means that if you select high-dividend stocks, you also implicitly choose low-growth stocks (as a group). And your high dividend return may be compensated by a lower capital gain.
Just the numbers, please
The problem is that the numbers differ strongly depending on the historical period. Damodaran has done research in his book ‘investment fables’ and writes: “from 1952 to 1971, high dividend yielding stocks did 3% better than the rest. But from 1971 to 1990, this was reversed“. Over the very long term, there is hardly a difference, ‘higher dividend yield stocks generate a slightly higher annual return than did lower dividend yeld stocks ‘ (page 25). It should not be a surprise: whether a company chooses growth or dividends to return its profits, the shareholder benefits in both ways.
If you want factual numbers for the last 23 years (June 30, 1995 – november 30 2018), maybe this index serves as a good example
- Total return of S&P 500 over this 23 – year period: 9.1 %
- Net return of the high-yield index: 7.5 %
- Net return of the MSCI EAFE index (large & mid cap world wide dev. countries): 4.7 %
So our high-dividend yield index performs better than its broad index (MSCI EAFE) but still worse than the S&P 500.
Obviously, there could be tax implications and you should consider the applicable taxes on dividend and on capital gains.
The final table shows
- The future expected returns (given current economic conditions such as inflation and dividend policies),
- Excluding the current valuation (so assuming that all markets are perfectly correctly priced today and not too high / too low)
- Absolute total returns before inflation adjustments
- Gross returns excluding all expenses and taxes
|Capital Gains||Dividend||Total Return|
|S & P 500||5.9 %||2.8 %||8.7 %|
|Equities Europe||3.4 %||2.9 %||6.3 %|
|Equities Japan||4.5 %||2.1 %||6.6 %|
|Equities Australia||4.3 %||4.0 %||8.3 %|
|Small Caps||7.0 %||2.0 %||9.0 %|
|Gov Bonds US||0 %||4.3 %||4.3 %|
|Corp Investm Grade Bonds||0 %||4.4 %||4.4 %|
|Gold||6.0 %||0 %||6.0 %|
|BRK||10 %||0 %||10.0 %|
|US REIT (developed market)||2.5 %||7.5 %||10.0 %|
|High Dividend Yield||5.2 %||3.5 %||8.7 %|
Et voila, a long post but at the end usefull insights to help us choose where to make a bet when selecting investment vehicles for the next 20 years or so.
One more thing on risk. I ignore deliberately the concept of risk = volatility or Beta because I invest for the long term and I do not care how much my investment jumps up and down in between. If I build up a portfolio of, say, a S&P 500 ETF, and I want to use that money in 15 years from now, why would I bother about the fact that it goes down 30 % and then goes up 80 % versus climbing slow and steady to the end result? In that assumption that risk is not beta, risk is the possibility of permanent loss, which is not zero but small for all assets discussed above. There is a chance that the stock market (including all dividends of all these years) of the US, Europe or Australia is lower in 2033, or that a real estate ETF is cheaper by then even including the dividends, but it is a very very small probability and I wouldn’t have a clue how to estimate it.
In a next post we will combine the average returns with the current price level of each asset type.