See more episodes of Quant Concepts here
Emily Halverson-Duncan: Investors often use multiple sources of information when researching stocks, one common one being to check what equity analysts have to say about a particular company. While the reports maybe lengthy, analysts have ways of converting their highly skilled research process into output that can be easily understood by the average investor. One of the most commonly-recognized outputs from analysts are the stock ratings of buy, sell or hold. This information can be a useful way for investors to determine whether the companies they are considering should be added to their portfolio.
Today, I'm showcasing a strategy that looks across the CPMS Canadian universe for stocks that are viewed favorably by analysts. So, let's take a look at how to build that strategy.
First off, as always, we are going to rank our universe of stocks. So, again, in this step, we are ordering our stocks in order of our favorite to our least favorite stocks. Some of the factors that we are looking at here are the median recommendations, so that's looking at the median analyst recommendation, which is on a scale of 1 to 5, 5 being the most favorable, 1 being the least; the 3-month change in median analyst recommendation, so that looks at that median number three months ago and compares it to the number today; and then, the earnings per share estimate revision across three months as well. So, in this case, you are looking at what analysts are predicting the earnings per share number will be three months ago compared to the number today.
And then, we move on to our second step. Here we are going to screen out the stocks that we don't want to own. So, some of the screens we are applying here are that 3-month change in median analyst recommendation, we want it to be positive, meaning that the change is getting better, the recommendations are improving. We want a 5-year price beta versus the S&P/TSX to be less than or equal to 1. So, in this case, we are looking for stocks that are less sensitive to the market. And then, a median recommendation of at least 3. So, again, on that scale of 1 to 5, we are looking for a 3, 4 or 5 for stocks to purchase.
On the sell side, we are going to sell stocks when their median recommendation falls below that level of 3, so again, either a 1 or 2; if that 5-year price beta goes above 1.2, so the stocks are becoming too sensitive to the market; and then the 3-month change goes below zero, so the analyst recommendation is actually declining.
Now, let's see how that strategy looks over the long term. So, here, we are going to run our back test. In this back test, we are running from November 1996 up until June 2019. We're going to be holding a total of 15 stocks and let's give that a run.
Here we are looking at the performance of the model versus the benchmark. One of the important things to note about this back test is there is no survivorship bias. What that means is any stocks that were to bankrupt in this time period, the historical data still remains in the system and is still included in the numbers reported here.
All right. So, the model actually did very well. The annualized return is 19.1% across that full timeframe. So, that's an outperformance of about 12% over the benchmark which is a pretty heft outperformance. A couple of the measures that I always like to look at; the Sharpe Ratio, which is a measure of risk-adjusted performance, is at 1.2. Typically, we look at anything that's 1 or higher as being quite positive. So, that's good. Other metrics in terms of risk that I'll look at are standard deviation and downside deviation. Standard deviation for the model was 14.2% which is pretty much the same as the benchmark at 14.5%. But downside deviation which looks at just negative returns, the variability of negative returns, is 7.8% relative to the benchmark at 10.2%, so quite a bit better.
One other chart that I really like to look at is this green and blue chart here. In up markets, the model outperformed 67% of the time, but in down markets it actually outperformed 78% of the time, so very good downside protection. One caveat to this model; it does look really good, but the turnover is quite high. So, you can see the turnover here is 181%. What that means is, of those 15 stocks on average you're going to be trading them almost twice in a year. So, that could be up to 30 trades. So, it is maybe a little bit high turnover for someone managing their own money, but again, always something you can speak to your financial advisor about.
For Morningstar, I'm Emily Halverson-Duncan.
