In the first part of " Hunting For Quality At A Reasonable Price (QARP) In Canada" we discussed the rationale and details of the "Quality Minus Junk (QMJ)" factor before applying the analysis to the TSX. The results showed a mix and expected and unexpected results. For more details, please see that article.
Now that we've done some sifting through the TSX, potentially weeding out some of the lesser-quality companies, it is time to turn our attention to the "At A Reasonable Price" portion of QARP. Although I think that having a portfolio with a lot of exposure to the QMJ factor is probably a good idea, finding those gems of companies that offer quality without a large premium is really what we are after.
In part 1 we ended up with a quality score for each eligible firm; these scores were the averages of the z-scores across a number of metrics within four broader categories (profitability, growth, safety and payout). Because we used z-scores for quality, if we utilize a similar approach for the price component we can simply look for the largest difference.
QARP = max (Quality score - Price score)
To put it simpler, we will isolate the firms with the highest quality for the lowest price. The firms with the lowest prices will have a negative score (because the scores are normally distributed with a mean of zero) and will actually be additive with the quality score.
There are a number of ways to measure price, each with its own advantages and disadvantages. Below is the list of selected metrics:
- Price to book (P/BV): The is the traditional approach to value from an academic view and is what the authors used in the " Quality Minus Junk" paper that inspired this analysis. In order to utilize this metric a company needs to have a positive book equity value which will further condense our list of eligible firms. Additionally, book value holds less significance for some industries where a lot of investment doesn't show up on the balance sheet, such as internet companies.
- Price to earnings (P/E LTM): Investors look at the potential earning power of a company to better understand its worth. Obviously a company that is better able to produce earnings should be worth more than one that isn't. However, historical results are often not indicative of the future and the data may be skewed by firms that experienced restructuring in the last 12 months (temporarily inflating their P/E ratio). Again firms need to have a positive ratio to be included in our analysis.
- EV/EBITDA ((NYSE:LTM)): Similar to P/E, enterprise value/EBITDA (earnings before interest, taxes, depreciation and amortization) is a measure of value over earning power. Similar to P/E we should note that historic results don't necessarily indicate the future. By looking at an earning metric that is higher up on the income statement, EV/EBITDA has the advantage that it should exclude a smaller number of firms (when compared to P/E). However on the flip, for some industries, EBITDA doesn't reflect the true cost of operating, this is particularly true for industries with significant capital expenditure and where D&A is very reflective of asset deterioration.
Although next twelve month (NTM) forecasts are available for a number of firms and would help correct the data for firms who are expected to have significant year-over-year growth, the dispersion of analyst coverage and analyst quality across the list of eligible companies could lead to biases and data quality issues. It may be fair to utilize these forecasts just for the group of larger firms, this may be revisited at a future date.
Industry value results
Table 1: Small Cap P/BV - Lowest 10% (left), highest 10% (middle), eligible firms (right). Source: Author's analysis.
After a quick glance at these charts and the charts in the first article, they certainty indicate a relationship between quality and price. The lowest priced 10% contains a significant amount of energy firms, and when combined with materials it makes up 88% of the cross section. Energy firms don't make a single appearance in the highest price category, but those industries that previously showed up being higher quality such as IT do.
Table 2: Small Cap P/E - Lowest 10% (left), highest 10% (middle), eligible firms (right). Source: Author's analysis.
Each metric brings with it different information and different biases, in this case P/E shows a significant negative bias against energy firms. Much of that industry continues to struggle and adds restructuring costs that further push earnings down, resulting in a lot of firms having a negative net income. Outside of energy, the lowest price bucket shows more diversification (compared to P/BV). Materials appear to be higher priced relative to other firms, with the remaining highest 10% split fairly evening amongst the other industries. It should be noted that P/E had the most impact on the sample size, reducing the number of eligible firms by nearly a third.
Table 3: Small Cap EV/EBITDA - Lowest (10%), highest (middle), eligible firms (right). Source: Author's analysis.
Based on the diversification EV/EBITDA appears to keep things a bit more industry agnostic than P/BV and P/E, although I caution before jumping to conclusions. EV/EBITDA had a smaller impact on the number of firms than P/E but more than P/BV. It is intriguing to see energy make up such a significant component in both the top and bottom 10%. One hypothesis that comes to mind to explain that divide is a flight to quality both in terms of assets and capital structure, where investors are paying more for quality and less for poor quality (relative to historic averages).
Table 4: Large Cap P/BV - Lowest 10% (left), highest 10% (middle), eligible firms (right). Source: Author's analysis.
Energy is once again the topic of interest here as it represents c.25% of the eligible firms, but has no place on either the top or bottom 10%. Also interesting to note is how largely Utilities and Healthcare are represented on the low relative valuation, with only a small contribution to the overall sample. It is possible that some of the biases of using P/BV may be coming out in this visual, as industries associated with higher asset intensity are coming up as lower priced.
I elected not to show the industry breakdown for P/E, simply because the number of eligible firms was too low to really take value from it.
Table 5: Large Cap EV/EBITDA - Lowest 10% (left), highest 10% (middle), eligible firms (right). Source: Author's analysis.
The results here are particularly interesting, particularly for energy and IT. Investors may think that energy prices are going to rise in the future and thus pushing valuations higher relative to the last twelve months EBITDA. Information Technology also represents a large portion of the higher priced firms, this could be growth similar to energy or perhaps lower risk.
Small Cap QARP Results
Table 6: Small Cap QARP top ten. Source: Author's analysis.
Nobilis Health (NYSEMKT:HLTH) came across with the largest differences between Quality and Price for all three statistics, Granite Oil (OTCQX:GXOCF) was also in the top 10 for all three measures. While Chalice Gold Mines (OTCPK:CGMLF), Serabi Gold (OTC:SRBIF), Amaya (NASDAQ:AYA), North American Energy Partners (NYSE:NOA), Corus Entertainment (OTCPK:CJREF) and Leons Furniture (OTCPK:LEFUF) were in the top ten for two statistics. I separated the average results into two columns because personally as an investor negative book value, negative earnings or negative EBITDA can be a bit concerning. The final column on the table above shows the top ten firms of those that had meaningful statistics for all three price metrics.
It is certainly interesting to see the amount of overlap between the measures, which highlights a likely correlation between the variables. I think if I were to utilize this method of analysis as a screening tool, I would start my due diligence with the firms in the right most column before looking at the overall averages and then individual measures.
Large Cap QARP Results
Table 7: Large Cap QARP top ten. Source: Author's analysis.
Well, Valeant (NYSE:VRX) reared its head once again, somewhat expectedly after its high quality score, but this time it only showed up in the top ten when Price/Book Value was the measure. Valeant had negative earnings for the last twelve months so it was excluded from that measure and simply did not rank high enough from an EV/EBITDA point of view. In the large caps there was only one firm that consistently appeared across all three metrics, Stantec (NYSE:STN). Five firms made two appearances: Turquoise Hill Resources (NYSE:TRQ), CGI Group (NYSE:GIB), Canadian Tire (OTC:CDNTF), Canadian National Railway (NYSE:CNI) and TELUS (NYSE:TU). Lundin took the top spot when the average results are considered, this is largely due to a very low score in the EV/EBITDA price category.
The true meaning of these results will become clearer as time progresses and we see the relative performance of the quality, value and QARP firms. Prior to that, I think the analysis is a useful tool in conducting investment due diligence and portfolio management. Should the results from the paper continue to hold, it could prove prudent to hold a portfolio that maximizes exposure to these factors and reduces exposure to the firms exhibiting the opposite traits.
Editor's Note: This article discusses one or more securities that do not trade on a major U.S. exchange. Please be aware of the risks associated with these stocks.