Saturday, 28 June 2014

Long Term Stock and Bond Return Expectations for Canada - the Current Outlook

A view of probable future returns is essential for individuals to develop realistic planning assumptions for retirement savings growth. Two years ago we wrote about the prospects for long term returns (10 years or more) of the Canadian stock market, the S&P/TSX Composite Index. Our estimate then was an average compound real after-inflation return of 4.2% per year.

The last two years have seen undeniably outstanding stock market growth - almost 32% in price gains alone (leaving aside another yearly 2.5% or so of dividends) as the Yahoo Finance chart below shows. It's time for an update look forward, which we'll also expand by looking at probable bond market returns and inflation.


Stock Returns

Method #1 - First, we'll apply the same method as we did two years ago:

Future Return = Current Dividend Yield + Estimated Growth Rate of Dividends

Current Dividend Yield - The current S&P/TSX Composite dividend yield (the current annualized sum of dividends being paid by the 245 stocks in the index divided by the current stock prices in proportion to the market cap size of the companies) is not easily found. It is not unfortunately not published by the index providers so the individual investor must use an approximation, as Globe columnist John Heinzl described in How to find index yields. The answer is to use the yield on an ETF that tracks the index, such as the iShares S&P/TSX Capped Composite Index ETF (TSX symbol: XIC). Both the Trailing Yield and the Distribution Yield round out to the same 2.4% and the recently lowered MER is now down to a paltry 0.05% so we'll use a Current Dividend Yield of 2.4%.

Estimated Growth Rate of Dividends - A declining working age population and slower population growth in Canada cause forecasters to predict markedly lower economic long term growth rates, as in the chart below from BMO Capital Markets recent publication Long Term Outlook: Destiny Dictated by Demography? (and similarly elsewhere e.g. this report from the OECD where table 4.1 estimates a likely 1.2% per capita GDP annual growth rate)

Taking the OECD figure of 1.2% times the 0.6 proportion that actually flows through to the shareholder (see our previous post and its links as to why all the growth isn't gained by the investor) gives us 1.2% x 0.6 = 0.7% in real future growth rate of dividends.

Thus our total is 2.4% + 0.7% = 3.1% future estimated real long term Canadian stock returns

It is perhaps not surprising that the estimate would be much lower today than two years ago, given the large gains since then, far above the projected average at that time and the long run historic averages.

Method #2 - For comparison, we'll add another method, described in Jay Ritters' paper Economic growth and equity returns, namely the earnings yield, or Earnings / Price.

According to the data for XIC on the iShares website, its P/E is currently 17.6. Turing that upside down gives us E/P, or 1/17.6 = 5.7%. But the ratio should use the past 10-year average of earnings according to Ritter, to smooth out economic booms and busts, not just the current earnings number embodied in the 17.6. Again the data for Canada is hard to to obtain for the individual investor. A rough estimation can be obtained using the TSX earnings figures taken at various points over the last twelve years by InvestorsFriend.com in his evaluation of whether the TSX is fairly valued. The average Earnings since 2005 over the current TSX Price (Index Value) of 15,000 gives an E/P of 4.7%. That number is still too high since the Earnings for each year have not been adjusted downwards for inflation but it gives an upper bound on the probable value.

Another comparison is the projection in the BMO Capital Markets paper, which is 7.0 (before inflation). Taking away 2% for inflation, for example using the figure derived below and which BMO also estimates, that would give a net return of 5.0%. Finally, using yet another method (and higher growth assumptions), InvestorsFriend estimated at the beginning of June based on a lower TSX index value of 14,375, that the TSX Composite would return about 7% annually, which gives the same 5% after inflation.

Our conclusion: over the next ten years, it is reasonable to assume that Canadian stocks are poised to deliver from 3% up to 5% real total return.

Bond Returns
The future return for bonds is simpler to estimate and much more certain. As we have written about previously here and here, bond duration tells us how long we need to stay invested in order to attain, within fairly narrow variation, the yield to maturity of the bond or a bond ETF at the time of purchase. Thus, if we take a bond ETF with long duration of 10 years or more, we know almost exactly what the total return, before inflation, will  be.

Take an average with a third each of the BMO Long Federal Bond Index ETF (TSX: ZFL) with a 14.84 years duration and a yield to maturity of 2.79%, the BMO Long Corporate Bond Index ETF (TSX: ZLC) of 12.85 years duration and 4.34% yield to maturity and BMO Long Provincial Bond Index ETF (TSX: ZPL) of 14.21 duration and 3.69% yield to maturity. That gives a future bond return of 3.6% before inflation, or 1.6% after inflation of 2%.

BMO's estimate for bond returns over the next ten years is 4%, or 2% net of inflation.

Inflation
The best forecast of inflation is the difference between the yield on long term Government of Canada ordinary bonds (2.82% as of June 25) and long term real return bonds (0.78%), which automatically adjust for inflation. These figures from the Bank of Canada webpage with daily updated values for both types show almost exactly a 2.0% (2.82 - 0.78 = 2.04) difference between the two. 2% also happens to be the policy target rate for inflation set by the Bank of Canada. The market rates set by supply and demand obviously believe the Bank of Canada can and will achieve 2% on average. Thus our estimate of 2.0% inflation for the long term.

Wishes, and estimates, don't always come true
It is necessary to keep in mind, as we noted two years ago, that for stock returns, and for inflation, the relationships driving forecasts are not deterministic and there can be considerable variability in actual outcome. It might be a lot more, or a lot less.

A future below historical averages
A 50-50 portfolio of Canadian stocks and bonds would thus produce a blended annual nominal total return of 4.3% to 5.3% and a real return of 2.3% to 3.3%.  That's less than the averages from 1900 to 2013 reported in the Credit Suisse Global Investment Returns Yearbook 2014 for Canada of 5.7% real stock return and 2.1% bond return, or 3.9% in a 50-50 mix. Over a long period that 0.6 to 1.6% gap would make a significant difference. The implication is that investors need to save more to build the same size retirement fund, or take longer to do it and then, in retirement, can withdraw at a lower rate than the 4% rule which worked in the past.

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: