Skip to main content

REVIEW OF LAST MONTH'S RECOMMENDATIONS Well, all streaks must come to an end. Yin Luo, executive director of quantitative strategy at CIBC World Markets, suggests a strategy for Number Cruncher each month. Last month, his strategy lagged the benchmark for the first time after beating it for nine months straight. For February, he recommended a value plus momentum portfolio. The portfolio rose 1.9 per cent, while the TSX equity index was up 3.2 per cent. His strategy therefore underperformed the index by 1.3 percentage points. THEME FOR THIS MONTH Mr. Luo is again exploring the relationship between value and price-return momentum this month. He has been doing some back testing and found that for low-momentum stocks (i.e., stocks that have not been doing so well in the past year), valuation helps pick good investments. How much better? Over the past 20 years, he found that low-momentum stocks with low P/Es performed on average 1.4 per cent per month better than low-momentum/high-P/E stocks. For the high-momentum stocks, however, difference in performance between low-P/E and high-P/E stocks was negligible. This means that for stocks with high intermediate-term returns, investors do not pay much attention to valuation. Looking at it a different way, expensive P/E stocks performed best when they had high price momentum. In fact, expensive stocks with high momentum beat expensive/low-momentum stocks by, on average, 2.5 per cent per month over the past 20 years. Among cheap P/E stocks, meanwhile, the high-momentum stocks beat low-momentum stocks by 1.3 per cent per month. MARCH RECOMMENDATIONS For this month, he recommends a value and momentum interaction portfolio. "We highlight that the value and momentum interaction strategy is essentially a non-linear model." Here's how it works. First, he sorted stocks by momentum and then sorted them by value. Those stocks with both attractive valuation and momentum fall into the long basket. On the other hand, companies with expensive valuation and low momentum are included in the short portfolio. As a comparison, a simple linear model (the strategy he recommended last month) calculated a value score and a momentum score for every stock in the universe, then sum up the two scores based on a certain weighting scheme; therefore, if a stock has a very high momentum score, it could still be included in the long basket, even if it is ranked poorly by valuation.

Interact with The Globe