How The Top 10 Stocks Ranked By Saevis Performed vs. The Benchmark

Last update on Sept. 24, 2018.

 

Jin's Note: This is a guest post by Pierre Guenette, who is a regular reader of this blog. Pierre works at the Alberta Health Services as a Senior Analyst for the Business Intelligence Team that's part of the Seniors palliative continuing care portfolio. Pierre has a master's degree in Economics from the University of Calgary, and has a general interest in data analysis.

 

On April 27, 2018, Jin announced in the MoneyGeek forum that the Saevis website will no longer be supported/updated. My first thought was “Poop on a Stick”. That sentiment changed to something stronger when I realized that my year and half experiment had generated a return of 38.55% and beaten my chosen benchmark (iShares Core S&P/TSX Capped Composite Index ETF - XIC) by 28.16%.

 

In the Beginning

When Saevis was created, I inquired about the methodology behind the “Most Highly Ranked Stocks”. Jin informed me that he drew inspiration from the book Quantitative Value: A Practitioner’s Guide to Automating Intelligent Investment and Eliminating Behavioral Errors by Wesley R. Gray and Tobias E. Carlisle. The authors wanted to find high performance value investment opportunities by applying the best aspects from quantitative investment and value investment towards stock selection and portfolio construction. They analyzed and ranked publicly traded companies using value indicators (measures of cheapness), quality indicators (franchise power and financial strength), as well as looking for signs of potential earnings manipulation and financial distress (hoping to avoid investing in an Enron, Nortel,  Bre-X, Valeant Pharmaceuticals, etc.).

Jin mentioned that he made two modifications to the book’s methodology. He added an additional value indicator EV/EBITDA (= Enterprise Value/Earnings Before Interest Taxes Depreciation Amortization) whereas the book relies solely on EV/EBIT (= Enterprise Value/Earnings Before Interest Taxes). Also, Jin included the company’s earnings over multiple years as opposed to only the past fiscal year. He believed these adjustments leveled the playing field for “resource companies (i.e., oil, gas, and mining) because they usually record large depreciation and their earnings fluctuate a lot from year to year” (MoneyGeek forum).

To emphasize the importance of sticking to a winning system, the authors of Quantitative Value used the example of Joel Greenblatt. Greenblatt devised a "Magic Formula" that was able to beat the market. However, on average, his self-managed clients failed to beat the market. Greenblatt found that this was because the self-managed clients tended to sell after periods of bad performance and buy after periods of good performance. Also, the self-managed clients tended to use their discretion to avoid purchasing many of the biggest winners because, at the time, they looked scary. Greenblatt concluded that these "are errors not made out of ignorance (the model presents us with the correct choice), but rather out of incompetence (we simply fail to follow the model)". The one exception among the self-managed accounts was an investor who faithfully bought and held the recommended stocks. This made me wonder if the same laissez-faire strategy (code word for lazy) could work using Saevis’ “Most Highly Ranked Stocks”.

 

Ground Rules

  • No Research (it wouldn’t be lazy if I actually had to do any research)

  • Buy $100 worth of stock from each of the Top 10 companies (regardless of size, reputation or my biased view as to their future prospects)

  • Start on January 3, 2017, and hold until the end of the calendar year

  • Compare against a benchmark that covers the TSX and requires no active management

  • Limit drag from Taxes and Transaction Costs

 

The reason that I used $1000 of my portfolio to run this experiment is to silence my “Fear Of Missing Out” cognitive bias. Also, this sum was small enough that, if this experiment proved to be disastrous, it would not crater my entire portfolio.

These invested dollars carry an opportunity cost, and, I felt it important to measure the Top 10 against a benchmark. This benchmark represents where I could have invested my money if I had wanted to passively track the TSX. I chose the iShares Core S&P/TSX Capped Composite Index ETF – XIC. BlackRock states that this ETF provides the investor with exposure to the entire Canadian stock market at a low cost (Management Expense Ratio = 0.06%). (Note: this is the vehicle that my wife uses for Canadian exposure in her portfolio.)

I wanted to minimize the tax and trading commission implications to ensure I was only scrutinizing the performance of the algorithm. I used my Tax Free Savings Account at Virtual Brokers so that I would not have to worry about tax implications (capital gains tax or taxes on dividends), and I would minimize my transaction costs. In a year and half, I have paid $9.20 in trading commissions and $0 in taxes. This equates to 0.899% of the original portfolio amount.

 

The Experiment

I bought the following top 10 “Most Highly Ranked Stocks” on January 3, 2017.

Throughout 2017, I must have checked on the Top 10’s performance on a daily basis. Finally, December 31, 2017 came and went.  For the 2017 calendar year, this portfolio returned $1,348.12 for a gain of $324.68 (31.72%). Once the feeling of euphoria subsided, I took a closer look and cursed the only two companies that generated negative returns: Data Communications (-47.14%) and Keg Royalties Income Fund (-0.05%). If only they could have been more like Epicore BioNetworks (111.57%), or Cogeco Inc. (61.44%), or Hammond Power Solutions (53.71%).

By comparison, had I used $1,023.44 to buy 41.94 XIC units (I know that in real world I would have had to buy 42 units, but I wanted to be able to compare exact dollar amounts) it would have been valued at $1,078.81 for a 5.41% return (includes dividends). Not too shabby. More or less what the TSX has returned on average over the last 30 years. However, this meant that my motley crew destroyed my chosen benchmark by 26.31%. Cue happy dance.

(Side Note: Always use a limit order when buying stock as the spread between what someone is offering to buy (bid price) and what someone else is willing to sell (ask price) can vary significantly. This spread can be especially devastating for small illiquid companies, of which there are several in this Top 10. By illiquid, I mean companies whose shares do not trade in high volume on a daily basis. It’s not uncommon for a limit buy/sell order to take days or even weeks to be filled.)

 

Captain Buzzkill Alert

Although this result is impressive, one year is a small sample. These monstrous returns could have been due to a perfect alignment of cosmic forces—a once in a lifetime flash in the pan that would only come crashing back to earth like Adrian Brody winning the Oscar for the Pianist, or Linsanity (NBA reference), or Tim Tebow (NFL reference).

In other words, as much as I wanted to unleash my inner ID (which demands immediate gratification) and turn over my entire portfolio to the Saevis Most Highly Ranked Stocks algorithm, more testing was needed.

 

Round 2 – 2018 Top 10

2018 Top 10 included two repeats from the previous year (Evergreen Gaming and Cogeco Inc.). In these instances, I recorded a sell price for 2017 which was the same as the 2018 buy price based on the market price at that time. However, no commission costs were incurred as I simply kept holding the stocks.

The 2018 Top 10 portfolio had a total Cost of $1,348.98--I spotted myself an extra $0.86 from the 2017 portfolio end value. There were permutations that would have gotten me closer to the real start amount, but I never envisioned that I was going to write an article and, for my own personal purposes, this was close enough. As we enter the fall, my portfolio currently has a value of $1,395.10 for a $46.12 (3.42%) gain. This year’s batch hasn’t been lighting the world on fire and there has been substantially more volatility. There have been times where the Top 10 has been above the benchmark by close to 10% and other times when it has trailed the benchmark by up to 2%. There’s still four months and I’ve got my fingers and toes crossed that last year wasn’t a fluke. My current top three are Evergreen Gaming (38.89%), Enerplus (29.37%), and Glacier Media (22.82%). The bottom three are currently Cogeco (-29.84%), Lucara Diamond (-20.57%), and GVIC Communications (-17.32%).

In comparison, the XIC has presently generated a 2.26% return to investors. The 2018 Top 10 is beating the benchmark ETF, albeit by 1.16 percentage points.

 

Recap

In one and a half years, the Saevis Top 10 Most Highly Ranked Stocks has transformed my initial investment of $1,023.44 into $1,395.10. This is a gain of $371.66 or 36.31%. Although, the majority of the gains were achieved by a phenomenal first year, it’s been able to beat my chosen benchmark in both years.

 

Large Cap Strategy

Mid way through 2017, it struck me that the amazing returns being generated by the Top 10 included small illiquid companies. These corporations would not be suitable for most investors. In the book, the authors manage an investment firm and aren’t able to invest in these itsy bitsy companies because they are too small. They focus, instead, on large companies. In this vein, I wondered if the Saevis Top 10 Large Cap could outperform the XIC. (Cap is short hand for Capitalization, which is calculated by multiplying the share price by the number of shares.)

On October 11, 2017, I decided to put the plan into action and bought the Saevis Top 10 Large Cap companies.

With only a month to go, this collection of stocks has generated a $146.70 gain (8.30%) on $1,766.79 invested. In comparison, the XIC has generated a gain of 5.58%. The Saevis Top 10 Large Cap Edition has outperformed the benchmark by 2.72%.

Undoubtedly, you’ve noticed that the Total Purchase Cost does not add to $1,000 and there are big disparities between investments (Canadian Tire $228.81 vs. CGI Group $131.68). I could not follow my initial rules because there are a number of corporations with a stock price greater than $100. I set up this portfolio to satisfy a curiosity, and, therefore, I was much more lax in terms of adhering to the ground rules. Interestingly, I was able to lower the drag associated with transaction costs. Higher individual purchase price meant I needed to buy fewer shares. This portfolio incurred $0.31 in commission fees ($0.01/share).

 

Large Cap Dividend Strategy

The demand for a re-occurring revenue stream has fueled the rise of dividend investment strategies. After building my Top 10 Large Cap portfolio, I noticed all but one company (CGI Group) paid out dividends. Curious as to whether or not a Saevis Top 10 Large Cap Dividend portfolio could outperform my chosen benchmark, I substituted CGI Group for Inter Pipeline to produce the following portfolio.

Currently, this portfolio is valued at $1,889.43 for a gain of $99.52 (5.84%). In comparison, the XIC has generated a gain of 5.56%. The Saevis Top 10 Large Cap Dividend Edition has outperformed the benchmark by 0.28%.

The difference in returns from the Top 10 Large Cap (8.30%) and Top 10 Large Cap Dividend (5.84%) is due to the superior performance of CGI Group (30.15%) vs. Inter Pipeline (-1.63%). However, if you’re looking for re-occurring income, Inter Pipeline has distributed the most in terms of dividends ($8.40). This is almost twice as much as the next closest company, Canadian Utilities ($4.57).

 

Concluding Thoughts

My personal bias towards evidence based value investing is screaming, pleading, and hoping that this methodology is sound. The initial results have been promising. The Saevis Top 10, Top 10 Large Cap and Top 10 Large Cap Dividend have generated greater returns than my chosen benchmark. However, it is still too small a sample to conclude definitively that this approach will consistently beat the benchmark over a longer period of time as well as over different business cycles. Small cap stocks are especially vulnerable to getting whacked during recessions.

The only reason that I’m writing this article is because Jin no longer has access to the data and I’ll no longer be able to track the performance of the Saevis Top 10 in all of its permutations. I’ll have to come to terms with this wonderful experiment coming to a premature end--by which I mean that I’ll have to day-dream of something other than $$$ dancing around my head when changing my triplets diapers (so many diapers). But, worst of all, I’ll actually have to research and analyze potential investing ideas or invest in the thoroughly defeated XIC.

I want to thank Jin for creating this tool. It’s often hard to find good Canadian stock investing content, and I really appreciate the time and effort that he has put into creating this Canadian centric website.

If you enjoyed this article, you might be interested in our free newsletter. Enter your email to get free updates.

Web Analytics