Friday 22 June 2012

What Long Term Return Can We Expect from the TSX?

In our last post we noted that the level of stock prices, or stock valuation, versus dividends, earnings and Q (replacement value of corporate assets) give an indication of how risky it is to invest in stocks as well as how much future returns are likely to be. As of Thursday June 21st market close, the TSX Composite Index per the TMX Money website is paying a dividend yield (dividends/price) of 3.22% with a price/earnings ratio of 14.3 at an index price level of 11,408. Let's work through an estimation of future returns using dividends, which though simple is well-grounded in research and finance theory.

The Magic Dividend Formula

It may surprise some people, but it is a fact that historical returns from equities come mainly from dividends, reinvested and compounded. Though stock market performance is not identical, Canada has been quite similar to the USA over the long term, so we use (it's hard to find nicely graphed Canadian data) the chart of the US market below taken from What Risk Premium is Normal? by Robert Arnott and Peter Bernstein to show the dominance of dividends over capital gains in cumulative total returns.

The same paper lays out a simple formula (called the Gordon model) for estimating future stock returns:
Return = Dividend Yield + Growth Rate in Dividends

The numbers to plug in:
  • Dividend Yield is the current TSX value of 3.2%. The higher this value is the better since a high dividend yield is the higher the future total stock return will be according to John Cochrane's Presidential Address: Discount Rates.
  • Growth Rate in Dividends depends on growth of the economy, or more precisely, as established by research, on real GDP per capita, which is less than total GDP growth (for several fascinating reasons per Arnott and Bernstein, such as, 1) the fact that a good part of economic growth comes from new businesses that we investors cannot buy in the index of existing companies and 2) stock dilution/issuance to managers). Based on the past 50 to 60 years, real GDP per capita has risen about 1.8% to 2% on average (see Jay Ritter's Economic Growth and Equity Returns William Bersntein's The Two-Percent Dilution and the Google chart of nominal non-inflation adjusted GDP below), of which around 60% actually flows through to the equity investor i.e. about 1%.
 
The result is, as of today,  we expect a long-run real net-of-inflation return of 3.2 + 1 = 4.2%.

How Much Chance of More Downside Before the Upside Comes?
To cross-check the chance that the market equity price might decline further, we can compare the current dividend yield and the P/E, or its inverse E/P aka earnings yield, against historical values to see if they are high or low. This chart from Are Equities Cheap? by Derek Holt of Scotia Capital going back to 1956 puts the current 3.2% dividend yield just above its long term average and the earnings yield of 7.0% about 1% above its long term average. Those numbers are reassuring, indicating reasonable valuation. However, we also see that the red and blue lines have ranged considerably above the average in past decades, which indicates times of low prices. Reasonable current valuation does not guarantee the TSX cannot drop further, possibly a lot, before it rises.
Another measure of over- or under-valuation of the market that is often-cited for the USA but hard to find for Canada is the Cyclically-Adjusted P/E (CAPE), which averages inflation-adjusted earnings over the past ten years to smooth out the effects of recessions. The same Scotia Capital presentation shows the CAPE value for the TSX as of January 2012 when the TSX traded around 12,200 to be just over the 20.2 average for the time since the mid 1960s i.e. slightly over-valued at that level.

Actual Return Might be Well Above or Below Our Expected Value
Our calculation of the expected return is subject to fairly wide uncertainty about the actual future return. What we are promised is not necessarily what we will get.  The chart below from an early 2001 study of deriving expected returns from valuation by John Campbell and Robert Shiller indicates how widely spread around the expected value of the line the subsequent 10-year returns (the vertical axis) ended up depending on an initial dividend yield (the D/P horizontal axis). What we obtain as a prediction from higher dividend yield is better odds, not an assurance, of a more favourable return.

Disclaimer: this post is my opinion only and should not be construed as investment advice. Readers should be aware that the above comparisons are not an investment recommendation. They rest on other sources, whose accuracy is not guaranteed and the article may not interpret such results correctly. Do your homework before making any decisions and consider consulting a professional advisor.

No comments: