Monday, December 19, 2011

Godfried’s Low P/E with moderate dividend yield portfolio

Earlier on we discussed how to allocate funds in a portfolio based on multiple cash flow streams. Now we’re focusing on a strategy you may use to invest in specific stocks. Yes, you got it right, here we’re trying to outperform market index ETFs.

If you liked Ken Fisher’s “Market’s never forget…” then you want to read O’Shaughnessy's updated “What works on Wall Street”. Ken Fisher shows that ‘this time is not different’ based on selected headlines supplemented by some historical statistics. If you want to see lots of numbers and historical stats on various ‘value investment metrics’ then you have to read O’Shaughnessy.

Based on U.S. stock market database ‘Compustat’ dating back to 1963 and on CRSP (Center for Research of Security Prices) dating back to 1927, O’Shaughnessy shows that over the long term (10 years and longer) stocks selected based on value investing metrics such as low P/E; Price to Book Value; high(er)Dividend Yield; and high Interest Coverage do outperform the average stock market indexes and their emulating ETFs significantly.

He also shows that the risk with stocks selected based on some of those metrics is actually less than that of the market index stock basket and for sure they are less risky than high P/E and no-dividend paying market darlings. O’Shaughnessy is not alone in this; other authors often mentioned on this blog such as Dreman and Siegel show similar stats, although in less detail and with data that does not include the last 10 years.

All these authors show also that ‘the efficient market theory’ does not really work. There are many ‘Black Swan’ events that do not fall in the Gaussian probability distribution and that disrupt portfolio performance. O’Shaughnessy shows that bull markets and bubbles go through varies stages of irrationality and that the emotional investor is easy prey to those who stay cool and unemotional. I show below some of O’Shaughnessy’s stats for the lowest P/E decile of the U.S. stock market:

For relevant metrics, such as P/E ratio or dividend yield, O’Shaughnessy, subdivides the stocks that make up his database in 10 equal sized groups (deciles). When analyzing the impact of P/E on say stock market performance, he measures the P/E of each stock in the database and then groups those stocks in ten equal sized groups (deciles) ranging from low P/E to high P/E. From January 1, 1964 until December 31, 2009 a portfolio of all stocks listed on the New York Stock Exchange returned annually 13.26% while the stocks comprising the 10% of stocks with the lowest P/E (decile 1) returned 18.23%. 

The risk or standard deviation from this average historical return is 18.45% (above or below the average return) for the decile 1 stocks compared to 18.99% for the entire market. In other words: a return of 18.23% + or -18.45% for 'decile 1' versus 13.26% + or -18.99% for the overall market (All Stocks).

From: O'Shaughnessy: 'What Works on Wall Street' (published in 2011).
  O’Shaughnessy’s data also gives an idea about investment risk through various parameters such as the standard deviation, tracking error and Sharpe ratio. If you invested in the decile 1 stocks for 1 year (according to the data table above, your worst annual return if you hold the stocks for 1 year is minus 52.60% (that was in 2008-2009 recession); the best return, on a 1 year basis, is 81.42%. Based on a holding period of 10 years, the worst annual return was positive 6.07%, i.e. you would have made at least 6.07% ROI during any 10 year period since 1963 and as high as 28.2%.

The figure below shows the chance of beating the ‘All Stock Portfolio’ over time and by how much.

From: O'Shaughnessy: 'What Works on Wall Street' (published in 2011).
 Over any 10 year holding period since 1963 a decile 1 portfolio would have a 99% chance of outperforming the ‘All Stock’ benchmark by a whopping 4.63%. This shows the benefits of investing with a long term time horizon but also the fact that investing systematically in low P/E stocks should help you to outperform the overall stock market. If you want to learn more of these statistics and their merits, you’ll have to get O’Shaughnessy’s book. Here I just want to show you some statistics on low P/E stocks and how I used this to create my own portfolio.

I used the GlobeInvestor Gold Advanced Stock Filter. I am not interested in micro-cap companies so first I filtered out companies with less than $200 million market capitalization. The Globe’s database came up with 531 common shares, including ‘A’ and ‘B’ shares. ‘B’  shares are typically non-voting shares in family owned enterprises. Owning stocks without voting privileges is something I do not recommend because it means that management is not accountable to you the share holder but rather to the family (‘A’ share owners) that is in control. Examples of investors being fleeced by the families are recently the holders of Shaw B shares or those of Magna class B shares. Then I proceeded to identify the 10% of  TSX stocks that trade at the lowest P/E (price-earnings ratio)

I came up with 67 stocks that traded at a P/E of 8 or lower. I refined the search by demanding some dividends (but not excessively high dividends which may point to troubled stocks). The dividend yield had to fall between 2% and 6%. This resulted into a list of 26 stocks. I whittled the list further down by omitting ‘B’ shares and by ignoring companies I considered low quality. The end result was:

Toromont TIH
Sherritt International S
Rogers Sugar Inc RSI
Churchill Corp CUQ
Brookfield Office Properties BPO
Cannaccord CF
Canfor Pulp CFX
Bird Construction BDT

8 Stocks! I also searched for stocks with a P/E below 11% and a yield between 3% and 6% which counted 45 stocks, including many banks which I already owned in my other holdings as well as several of the stocks above. But 3 new stocks stood out:

Canadian Helicopters CHL.A
Cervus Equipment Corp CVL
Corus Entertainment CJR.B

Although Corus is a ‘B’ share which is controlled by the Shaw family, its other financial ratios (EBITDA, interest coverage, return on equity, etc) looked attractive enough to overlook the ‘B’ factor. So these are the 11 stocks that I selected for my ‘low P/E with moderate dividend yield’ stock portfolio. I will start this portfolio in real life for 2012 with equal proportions invested in each stock. At the end of 2013 I will revisit these stocks, rebalance them and/or replace them if they no-longer meet the above mentioned criteria.

We will also revisit this portfolio at various times during the year to track their performance based on a ‘$100,000’ portfolio starting January including commissions and dividends. At those times I will not make any changes to the portfolio itself. Note, I make no recommendation regarding these stocks; I am just following a strategy and testing how it works. I will own these stock in real life but I won’t disclose the real size of this portfolio – let the games begin.

1 comment:

  1. The article introduces the topic very well and in a layman language about the stock markets. Would like to know more about this in your next series.