Go to page Previous 1, 3, Next      

Changes taking place industry-wide get competed away.  For example, one might argue industrials have benefited from decreased power among labor unions.  But these benefits are transitory and do not allow companies across an industry to sustain higher margins.  Price competition ultimately destroys above average margins; it is just a matter of time.

I analyzed historical margins in each of the ten sectors comprising the S&P 500.  Six of these sectors are severely inflated.  Three (energy, materials, and industrials) are 50% higher than their 1988-2006 historical average margins, even though industries in these sectors haven’t changed much structurally.   A slowdown in the global economy will be a great deflator in those sectors.  Financials are 33% above historical averages – we are already seeing profit margins contract here as banks are taking write offs and defaults are going up.  Information Technology and Telecommunications Services are 135% and 25% above historical levels, respectively, although I don’t attach as much significance to this because of the enormous structural changes which have affected these sectors.  It will probably take the longest for profit margins to adjust in these sectors. 

A lot less margin expansion has occurred in Consumer Discretionary, Consumer Staples, Health Care, and Utilities.  It comes down to competitive advantage and industry structure.  You want to own companies that create a large moat around their business and will be able to defend their position and maintain their margins. 

The S&P 500 P/E was at 18.92 as of 1/31/08, versus a historical 25-year average of 20.72 and a historical 50-year average of 17.60.  Shouldn’t we feel pretty comfortable with today’s P/E ratio, compared to historical levels?

If you look at P/E levels over last 100 years, the average is about 15.  The higher number over last 25 years encapsulates a very strong 1982-2000 secular bull market.

You need to look at a period including full secular bull and range-bound markets, which together last 30 years or longer.  That is what I did in the book. P/Es over the last 25 years include a full blown bull market and only a small portion (seven years) of a range-bound market.

P/E ratios have actually spent very little time at the average level of 15.  Only 27% of time have P/E ratios been between 13 and 17.  P/Es are at average levels only when they go from one extreme to another.

Shiller argues that P/E ratios should reflect historical (e.g., 10-year) trailing earnings, rather than just the trailing 12 months.  His data also suggests that P/E ratios are inflated.  How much emphasis do you place on historical data, such as Shiller suggests, versus contemporary ratio values?

I agree with Shiller.  Historical data helps to understand the environment and helps to make better decisions.  In the book I took a fairly detailed look at P/Es based on 10, 5, 3, and 1 year trailing earnings. 

I’d like to ask about non-US markets.  The MSCI World Index P/E ratio is currently below its 20 year average.  Have overseas profit margins expanded to the same degree as in the US?  Should overseas investors be worried about P/E contraction?  Earnings contraction?  Which regions do you see as currently over- or undervalued?

Unfortunately, I don’t have a good answer.  It is difficult to get good data on these markets for the type of analysis I require.  The math works the same way.  Returns come from price appreciation (due to earnings growth and/or P/E expansion) and dividends.

We stay away from hot markets.  China has been bid up, and the returns that follow will be low.  We have 25% of our portfolio in ADRs, all in companies we determined are undervalued and meet our stringent Quality, Valuation and Growth criteria.  Mostly these are European, as this is where we find most opportunities (i.e. cheap high quality stocks).   We also own plenty of global companies with a significant portion of their revenue from overseas markets.  They should benefit from faster growth in international markets.

Display article as PDF for printing.

Would you like to send this article to a friend?