Quant Concepts: When Two Styles Become One

See more episodes of Quant Concepts here

Emily Halverson-Duncan: Welcome to Quant Concepts' virtual office edition. Investors typically look for stocks with similar attributes such as paying a long history of dividends, a low valuation or strong earnings reports to name a few. While it's good to know what you want, the one downside to screening for a lot of different metrics is that you can sometimes reduce your stock universe to a very small set of options. One way to help identify stocks with the characteristics you are looking for but without drastically reducing the number of possible companies is to look at these attributes separately. Today's strategy is constructed of two different sleeves, dividend growth and momentum. Each sleeve will focus on its respective style only and then the final portfolio will be a 50-50 combination of the two. So, let's take a look at how to build each of these sleeves.

First off, we are going to start with our dividend-focused sleeve. And again, we are using the CPMS Canadian universe here, which today for both sleeves, has a universe of 700 stocks. For the dividend growth portfolio, first off, we are going to rank stocks by five-year dividend growth. So, that's an annualized metric and we want higher values for that indicating the company has been growing their dividends over the last five years. Expected dividend growth is what the company is expected to be paying out in dividends compared to what they just paid out over the last four quarters. And again, we want higher values there. And then, lastly, quarterly earnings surprise looks at what a company is expected to post for earnings and whether or not they beat or missed that expectation.

On the screening side, once we've ranked our universe, we're going to look for stocks with a five-year dividend growth in the top third of peers or in other words, a value of 9.92% or higher. Expected dividend growth, we want that to be positive, so indicating they are maintaining or growing their dividends. Total return standard deviation across the last five years – this is a risk metric just to make sure that the return series isn't growing and declining drastically and is more consistent over the last five years, and we want that to be in the bottom half of peers. Payout ratio is looking at how much of their earnings they are paying out as dividends and we want that to be less than or equal to 75% to make sure that they are not paying out everything of their earnings in dividends and still retain money left over to invest back into the company. And then, lastly, five year sales growth, which just looks at similar to the dividend growth number across the last five years, how has the company been growing their sales and in this case, we just want it to be greater than or equal to zero, so again, flat or at least growing across the last five years.

On the sell side, once we've applied all those screens, we're going to sell stocks if that five-year dividend growth falls into the bottom half of peers which today has a value of 3.51% or below or if that payout ratio goes above 80%. So, that's the first sleeve of the two that we're looking at. Again, this is dividend growth.

The second one is a momentum sleeve. So, just pulling that up here. And similar here, again, that same 700-stock universe and we're going to go ahead and rank it. So, the factors we're looking at here – that same quarterly earnings surprise that we looked at before. Quarterly earnings momentum, which is looking at the growth in earnings quarter-over-quarter, we want to see a higher value there. And percent change from 12-month high. So, what that's looking at is a stock's price today and comparing that to their highest price over the last 12 months. So, in the last 12 months if today was their highest price, then the difference between those would actually be zero because it'd be equal to the highest price. And in general, you want a higher value for that particular metric.

On the screen side, we're going to look for a quarterly earnings surprise and quarterly earnings momentum both greater than or equal to zero or 0.01. So, in this case, we just want them to be positive, so either beating on their expected earnings and growing their earnings as well quarter-over-quarter. Price relative to 200-day moving average. This is a bit of a timing metric. So, what that looks at is a stock's price today relative to its 200-day moving average and we want that to be greater than or equal to 3%. And lastly, we want a market cap to be in the top two-thirds of peers. So, it's just to screen out some ultra-small cap names, although we are still allowing some small caps to filter into the universe.

On the sell side, we just have a simple one sell criterion here to look at which is the price relative to 200-day moving average. We're going to sell if it declines more than 15%. So, once we've applied all of the screens across the two strategies, we're going to combine the best 10 stocks from each to a 20-stock model and look at the back-tested returns.

So, just jumping into the backtest here, of that 20-stock model from January 2004 until end of May 2020, you can see the model performed 14.8% annualized which is very strong with both sleeves combined and results in outperformance over the benchmark of 8.2% and recall the benchmark here is the S&P/TSX Composite. Turnover is about 59%. So, it's a little higher than some of the ones we've seen in the past but not too bad considering there's a momentum component in there and those typically trade more.

And again, a couple of the metrics I always like to look at – downside deviation, which is the volatility of negative returns, for the strategy is 7.7% but for the benchmark is 9.6%. So, what we like to see it's doing better in down markets. And then, of course, my favourite green and blue chart down here – in up markets, the model outperformed 55% of the time and in down markets, outperformed 77% of the time. So, again, reiterating that strong down market performance and actually still doing better more often than not in up markets as well. So, again, if there's multiple styles that you tend to like rather than necessarily combining them into one portfolio, you can instead look at segmenting them into different sleeves and then combining that all together for one final portfolio.

For Morningstar, I'm Emily Halverson-Duncan.