Most investors assume that stocks, despite the occasional value trap, with low valuations always outperform more expensive issues as a matter of faith. Over the extreme long term, this faith is justified, particularly in terms of risk-adjusted returns.
In the past 13 years, however, the actual performance data are far more mixed – in most sectors, valuation levels had no relationship to future performance at all. In some cases, the results contradict the conventional wisdom that cheaper stocks are better investments.
Breaking the TSX into industry subsectors, this column will detail where a focus on valuations has worked for investors and where they haven't. There are numerous surprises.
The analysis uses the two most popular valuation measures – trailing 12-month price-to-earnings and price-to-book ratios – for each major subindex and compares them with index performance in the subsequent two years. The results can be seen in the top chart.
In each case, I was looking for the high negative correlation levels that would indicate that future performance rose as valuations declined or, conversely, that as valuations climbed and stocks became more expensive, performance weakened. In short, high negative correlations would indicate that buying cheap stocks added to positive market returns. (Correlation analysis provides a mathematical value between the minimum minus-1.0, which would indicate all monthly data points saw future performance go in one direction and valuation levels in another, and positive-1.0.)
Only one of the readings was positive: the relationship between price earnings ratio and future performance for materials stocks. This means that P/E ratios were no help to investors in the sector – materials stocks performed better the more expensive they got (in terms of P/E). Similarly, the very small negative correlation indicates that price-to-book-values ratios were very little help for investors in industrial stocks.
The most helpful results were, in order, price to book value for utilities stocks, P/B for banks, P/E for consumer staples stocks, P/E for energy stocks and P/B for consumer discretionary.
The second chart shows how remarkably well average price to book levels have worked for investors in Canadian utilities stocks. Note that the right hand scale, price to book value for the S&P/TSX utilities index, is plotted inversely to better show the trend (and also to account for the inverse relationship – future performance rises as valuation levels fall).
Time periods where price to book value has not effectively predicted future returns have been extremely rare since 2002. Outside of March, 2003, to April, 2004, there actually aren't any. In addition, current valuation levels project positive returns for the utility sector for the next few months.
The usual caveats apply – historical relationships between valuations and investment returns may not continue into the months and years ahead. The main point is that, in order to maximize returns, investors should pay careful attention to which valuation measures are used for each investment type.
Follow Scott Barlow on Twitter @SBarlow_ROB.