Quant Concepts: A Portfolio of Horrors

A cautionary tale of picking Canadian companies poised to lose money, with CPMS's Emily Halverson-Duncan

Emily Halverson-Duncan 30 October, 2020 | 4:28AM

 

 

See more episodes of Quant Concepts here

Emily Halverson-Duncan: Welcome to Quant Concepts' virtual office edition. With Halloween just around the corner, this is arguably Canada's eeriest time of the year between scary movies and haunted houses all the way to witches, ghosts and goblins. In the investing world, there's nothing scarier than the idea of purchasing a stock that experiences significant declines. What's even worse to imagine, however, is picking a whole portfolio of stocks with dramatic losses.

In today's video, the model we're going to explore is a cautionary tale, looking for Canadian companies that are poised to lose money. We'll do this by picking stocks with characteristics that go against the typical fundamental perspective we usually take and instead search for companies with negative attributes. So, let's take a look at how to build that.

Jumping into CPMS here, we're using our usual CPMS Canadian database, which has 703 stocks as of today. And first off, we're going to go ahead and rank that universe of stocks. So, one thing you'll notice here is the factors that we're using to rank, they're all highlighted in red, or they've got the text here in red, which means we're looking at them in the opposite way we usually would. So, for example, for long-term debt to equity, typically, we would look for stocks that have a lower value there, meaning they have lower debt to equity on their books. But for this particular stock, because we're looking for stocks we don't want to own, in this case, we want that value to be higher.

Other examples of that cash flow to debt – we want a lower value for this model than that, meaning we want less cash flow on hand. That's not usually the case that you want to do. Quarterly earnings surprise, so whether a company has beat or missed on their earnings. We want them to essentially have missed or have lower reported earnings compared to what analysts expected. And another factor here we're looking at is their industry relative price to book, which is looking at on an industry relative basis, how they're valued compared to peers, and we want them to be highly valued, meaning overvalued for this particular model.

On the screening side, once we've applied all those ranks, one of the screens that we're looking at here is quarterly earnings momentum, which is looking at their earnings quarter-over-quarter and seeing how they've grown them. In this case, we're putting a maximum on it of negative 1%, meaning that we want them to be declining their earnings and have negative earnings on a go-forward basis.

On the sell side, we're going to get rid of companies that fall below the top 150 stocks, and that order, that top 150, is based on those ranking criteria that we looked at earlier.

So, now that we've gone through how to build the model, let's take a look at how it performed. Just jumping into our back test here, our back test is going to be holding 20 Canadian stocks and the benchmark here is going to be the S&P/TSX. Looking on this page, we ran the back test from December 1991 until the end of September 2020. We can see across that timeframe pretty opposite to what we usually look at. The returns are negative 7.2% annualized or an underperformance of the benchmark by negative 15.2% on an annualized basis. Turnover is pretty high as well within this at 157%, meaning on a 20-stock portfolio, you're probably placing a little over 30 trades per year, or turning over the portfolio about 1.5 times.

But a few of the key metrics to look at that we usually do – and again, they're going to have a bit of an opposite connotation with this – downside deviation, which is the volatility of negative returns for the strategy, is 19.4% and for the benchmark is 10%. So, you can see it's significantly more volatile in this model. And then, my favourite green and blue chart, it looks exactly how I would expect it to, which is in up markets, the model has outperformed about 42% of the time, so a little less than the benchmark, and in down markets, it's outperformed only 16% of the time, so being beat about 84% of the time when the market is declining and we can see how that's compounded, of course, across the returns here.

So, again, as a cautionary tale, it just goes to show if you're picking stocks with those negative attributes, it's not going to work out over the long term. You might get a short period of outperformance, but really, across a long-term strategy, we can see here that it doesn't work out.

For Morningstar, I'm Emily Halverson-Duncan.

 

Get the Latest Insights in Your Inbox

Subscribe Here

About Author

Emily Halverson-Duncan  Emily is Director, CPMS Sales at Morningstar

© Copyright 2020 Morningstar, Inc. All rights reserved.

Terms of Use        Privacy Policy        Cookies