After investing for ten long years and ending up with less capital than they invested, many are questioning the wisdom of long-term ownership of stocks. While I think bonds have a place in even the most aggressive portfolio, it would be a mistake to extrapolate the recent poor run in stocks far into the future and give up on stocks altogether. This post on Falkenblog (which I found via Larry MacDonald’s Investment Ideas blog) captures the prevailing mood by arguing that the equity premium does not exist by making the following points:

  1. Geometric versus Arithmetic averages: I’m not entirely clear who the target of this criticism is. Almost every book I have uses geometric averages in talking about long-term market returns. Jeremy Siegel, for instance, clearly breaks out both geometric and arithmetic averages and his data indicates that stocks have exceeded bond returns by 3.3% during the 1802-2006 time period.
  2. Survivorship bias: I find it interesting that Falkenblog chose not to cite the Dimson, Marsh and Staunton paper, which found that equity premiums existed in every major market database for the 1900-2005 time period: “The annualized U.S. equity premium relative to bonds was 4.5% compared with 4.1% for the world ex-U.S. Across all 17 countries, the equity premium relative to bonds averaged 4.0%, and for the world index it was also 4.0%”. In other words, equity premiums are not unique to US stock markets.
  3. Taxes: Applying a tax rate on the equity premium is simplistic. A more careful analysis would apply past tax rates on returns from stocks and bonds and arrive at an after-tax risk premium.
  4. Adverse market timing and transaction costs: Many studies have established that the returns obtained by the average investor falls far short of market returns. However, intelligent investors who pay careful attention to expenses and emotions have high odds of outperforming all other asset classes by sticking with stocks over most long holding periods.

While the debate about past risk premiums may be interesting, a more pertinent question for investors is: where do we go from here? Today, 10-year bonds are yielding 3.5%. The dividend yield of the S&P 500 is about 2.5% and the TSX Composite roughly 3%. Stocks seem to be priced to deliver a healthy premium in the future even with very modest returns.

This article has 5 comments

  1. All you need to say is “If you don’t want those stocks anymore I’ll take them – but since they’re so bad now I won’t pay full price” 🙂

    I think the point about adverse timing is the best one. From what I remember of other studies I think adverse timing creates more than a 3% difference between the average investor and simple buy-and-hold investing. In that case it explains everything; if you hang on to the index you get the premium but if you buy high and sell low you lose it (when you’re affected a little too much by the risk in the risk premium). That merely means that many people are bad at investing in stocks, not that investing in stocks is a bad idea.

  2. I think the key thing is to look at Gordon equation.

    Expected Return = Dividend Yield + Earning Growth + Change in P/E ratio

    Dividend yield is 2.5%. Earning growth should be above average for the next few years due to the economic recovery (maybe 7%). Change in P/E ratio is the trickiest and might be negative due to the fact that S&P 500 trailing P/E ratio is at 31 times according to Wall Street Journal (maybe -2%).

    According to Gordon Equation, expected return right now for S&P 500 should be positive at 7.5% using my estimation and higher than 3.5% for 10 year Treasuries. However, this is not always the case. On the top of the market in 2007, Gordon Equation suggested negative returns due 1% dividend yield, 0% earning growth (that is what John Bogle said), and possible decrease in P/E ratio (S&P 500 had a P/E ratio of 25 at that time, very high in comparison with a historic average of 15.) 10 US Treasury was yielding from 4.5% to 5.0% in 2007.

    I am wondering if Gordon Equation is the best way to calculate expected returns, which can be used as a good estimation for risk premium of investing in a broad stock market index.

    I think the stock market has become less stable in the last two decade due to the fact that a significant portion of market return is depended on earning growth and change in P/E ratio. I believe dividend yield is most stable and reliable part of Gordon Equation and today’s dividend yield is much lower than the historic norm. I expect a lot of volatility in equity markets in next few years due to the fluctuation of earning growth and P/E ratio.

    Now, I use Gordon Equation as my fundamental analysis tool for tactical asset allocation investing and SMA 200 as my technical analysis tool for market timing and risk management.

    Tell me what you think of Gordon Equation.

  3. Canadian Capitalist

    @Silicon Prairie: That’s a very nice way of putting it. It’s not that stock returns are bad; it’s just that the average investor is bad at investing in stocks. And yes, I’ll be happy to accumulate stocks at low prices. I don’t want another bull market for another 20 years.

    @Henry: I think p/e ratios are fair now. If you use Robert Schiller’s method of averaging the earnings of the previous ten years, S&P 500 trades at about 16 times earnings. Earnings growth over 10 year periods is fairly stable around 6%, so we get:
    Expected returns = 2.5% + 6.0% = 8.5%

    That leaves us with changes in valuation. I’d go with 0, as p/e is about average now.

    Compared to this, bonds are trading at 3.5%. Stocks provide a comfortable margin of safety today.

  4. @CC: I agree with you that Robert Schiller’s averaging earnings of the previous ten years is another way to estimate the Gordon’s equation.

    The reason for the long post is that I believe Gordon equation is extremely useful and can calculate the expected returns of the next few years. I don’t see a lot of people using Gordon equation, but I feel it is a must use tool in investing. Buying stocks when expected equity returns is higher the risk free return is logical, because there is a risk premium. Selling stocks when expected equity returns is lower the risk free return is also logical, because there is no risk premium or in fact a negative risk premium.

    My conclusion is that using Gordon Equation satisfy the first principle of investing: investing requires adequate return and the safety of principal even in a volatile asset class.

  5. Pingback: Intelligent Speculator | Financial Ramblings