S.74 and S. 75: Screens Using Analyst Price Targets and Earning Per Share

Spoiler alert: This is a quasi-public service post.

The reason is because there are times when financial journalists, maybe rushing to meet an approaching deadline or whatever, slap together some half-baked data analysis to support a sensationalist headline. These types of articles are the financial journalism equivalent of reckless driving, and have the potential to detonate some grievous financial injuries if taken seriously enough to propel action. Take for the example, the recent Barron’s Stock Screen article, “Amazon, Facebook, and 22 of the Most Undervalued Stocks, Right Now,” from February 16, 2021. The article’s author claims to found the most “undervalued” (in heavy quotes) companies using a stock screen, and yet provides no supporting data/evidence for a)why his claim should be believed or b)how the screen would have performed in the past. So, given this backdrop, what we will provide today is an outline of the Barron’s screen, and since the author didn’t provide any historical data, we’ll go about testing his screen for him, two different ways, and provide you with performance records so that you may draw your conclusions!!

The Barron’s Screen

The important point the author makes in the article are as follows:

  1. Stock analyst know their companies best.
  2. Stocks with heavy analyst coverage (10+) and with >50% buy ratings should be in the screen.
  3. Stocks with high analyst target price estimates relative to current represent value.

The screen basically requires us to take as a given that stock analysts know more than anyone else; that correctly forecast stocks that will outperform (especially those with a heavy coverage); and that they then give us the signal via buy ratings and/or price targets.  In short, a very big bunch of “trust me and the stock analysts.”   

So here are some counterpoints:

  1. I’d argue that even the mightiest stock analysts do not know companies as well as insiders (just ask those that covered Enron, etc.) and so statement 1. is easy to toss aside. 
  2. Let us say number two is a valid and we assume analysts only issue buy ratings because of objective analysis (and not from banking or other conflicts that often creep up).
  3. Then we come to number three, here we are asked to assume that the analyst’s predictions are accurate and that the analysts will tell the general public before their best clients who in some case pay their firm millions of dollars in brokerage fees. And if you believe this is how things work, I will ask you now to also start believing in tooth fairies! 

But let’s move on and get down to the screens and numbers. I tested the screen as presented in the Barron’s article, as best I could. I screened the Russell 3000 Index ex biotech looking for stocks with more than 10 analysts; greater than 50% buy ratings; and where the forecasted price was much greater than the current price. To get to a screen with roughly 20 holdings, like the article, I had to use a forecasted price increase of 28% above current price. I also made sure all companies had positive earnings, something not mentioned in the article, but which made sense.  I named this screen the Analyst Price Target Screen.  (The Barron’s articled mentioned weeding out commodity and biotech companies, but I kept the commodities in since because I am requiring positive earnings I will be avoiding the risky cyclical plays anyway.)

And I was not done.  The Barron’s screen only considered analyst price forecasting abilities, and yet analysts spend a lot of time fixated on earnings.  So, in S.75, I swap out price target for forward 12 month earning per share forecasts looking for stocks with the greatest forecasted earnings per share growth. I set this number at 400% earnings per share growth, which yielded a portfolio of just over 50 companies. This is the Analyst EPS Target Screen.  Let’s see which does better and how they compare versus the Russell 3000 Index.

S.74– Analyst Price Target Screen

Stock UniverseRussell 3000 Index, Ex Biotech
Rebalancing FrequencyAnnualIy
Screen Criteria12 Month Trailing EPS>0
Analyst Coverage >10
Greater than 50% Buy Ratings
Price Target>28% Above Current  

S.75– Analyst EPS Target Screen

Stock UniverseRussell 3000 Index, Ex Biotech
Rebalancing FrequencyAnnualIy
Screen Criteria12 Month Trailing EPS>0
Analyst Coverage >10
Greater than 50% Buy Ratings
Forward 12M EPS Target>400% TTM EPS  

Results

Play around with stock screens long enough and you get a sense for what works and what doesn’t even before you run a backtest, and I had a big hunch that using analyst price and earnings forecasts would fail as screens, and they did (in terms of beating the Russell 3000 benchmark index). To top it off, using analyst forecasted stock prices fared much worse than using analyst earning per share estimates as an indicator of “value.”  It was not close (5.8% per year return vs. 13.4%!).  I hope someone shares this information with Barron’s so that they don’t get a ticket for “reckless financial journalism,”  until then thanks for reading this public service announcement! 

Test Results and Chart

S.73: The Anti-GameStop Screen

Now is the time, in the middle of February, when in the northern latitudes, everyone seems to be on the ski slopes, and oh, how I wish I were one among those on the slopes of Zermatt, or maybe Gstaad or even better still among the young, power-elite couples of Crans-Montana, listening to apres-ski dance music, in the sundrenched outdoors at 2,000 meters above the sea! But alas, for me at least, this is just another (frequent) midwinter daydream, I am stuck here writing once again about meme stocks and GameStop!!!

This time around I take a different approach. Since I used my last two screening projects to demonstrate with a decade’s worth of annual return data how poorly GameStop-like short squeeze stocks had performed, I decided to flip the approach and screen for stocks that were anti-GameStops, that is, those stocks with very low levels of short interest! If stocks with high short interest relative-to-float performed poorly, logic would tell us stocks with low levels of short interest-to-float should perform better. So let us see how the anti-GameStop screen does.

S.73– Low Short Interest Screen (“Anti GameStop”)

Stock UniverseRussell 3000 Index
Rebalancing FrequencyAnnualIy
Screen CriteriaShort Interest/Float <10%  

Results

As I had suspected, our screen of stocks with low levels of short interest performed much better than our previous screens focusing on high short interest stocks. Anti-GameStop returned 11.6% per year over the 10 years ended 2020 while our GameStop screen of stocks with extreme short interest (60% plus of float) only returned 1.4% over the same period. Yet just because a stock has low short interest doesn’t mean it will outperform the broader market (there is more to stock performance than short interest, I’m sorry to say this, almost, given this will be heartbreaking to some) and indeed our screen lagged the Russell 3000 Index over the period. Well now that we’ve covered high short interest and low short interest stock screens, I look forward to moving on to other strategies, or perhaps maybe back to a little daydreaming!

Test Results and Chart

S.72: Short Squeeze Screen 2 (“Little GameStop”)

The 2021 GameStop short squeeze saga did more to bring the arcane art of short selling into the mainstream than anything before it. For a few days it even knocked the important stuff like the Super Bowl, the pandemic, and the situation in Burma/Myanmar off the front pages!!

It is too early to tell if retail and computer Bot traders that pile on to retail trades will drive a long-term change in market dynamics, and the situation is, as they say, fluid. All we can do here is to continue to inject data into the debate!

In our last screen, we looked for GameStop like stocks with short interest to float ratios of above 60%. At year-end 2020 there were six stocks that met the criteria, but as of February 2021 there are no stocks have short interest above 60% (short interest declined in these names after GameStop debacle). So, given this, we present our next screen, which I’ve dubbed “Little GameStop” that looks at the performance of stocks that have 30% to a max of just under 60% short interest to float ratios. There are 18 stocks that currently fall in this range, including GameStop, which saw a massive reduction in short interest (from over 100% of float to under 60%).

S.72– Short Squeeze Screen 2 (“Little GameStop”)

Stock UniverseRussell 3000 Index
Rebalancing FrequencyAnnualIy
Screen CriteriaShort Interest/Float <60% Short Interest/Float >=30%  

Results

The Little GameStop screen actually performed slightly worse than our original GameStop screen registering a trailing 10-year return of -1.2% per year versus 1.4% for our prior screen. Although returns were underwhelming the screen showed volatile swings with several positive years mixed in with the negative. Volatile and underperforming, this doesn’t seem like a winning combination to me!

Test Results and Chart

S.71: Short Squeeze Screen (AKA GameStop)

Emergency times call for emergency screens! The great Reddit GameStop short squeeze of January 2021 has caused the arcane practice of short selling to take over the public consciousness this week, generating mass confusion. As per usual, the situation playing out online and through media channels has devolved into “he said this” and “she said that” and stories of billionaires behaving, well, surprisingly like the TV show Billions! You can make this stuff up – really!! While we don’t partake in these cat fight debates here, we can inject some vastly more lucid stock screening data into the mixer for consideration, in, I might add,  almost real time. 

GameStop is the most shorted stock as a percentage of float with over 122% of float based on the data that I collected. It stands out as a short squeeze target because no other stock approaches this level among the top 3000 stocks in the US market. In fact, only a half a dozen can claim short interest as a percent of float above 60% and they currently are:

 GameStop                       123%

Dillards                             83%

BigScreen Hldgs           74%

Bed Bath Beyond        66%

Ligand                               66%

National Beverage    63%

Our screen today involves testing an annual strategy that tracks the performance of stocks with short interest-to-float ratios above 60%! So grab a drink and see what happened had you followed the Redditor short squeeze strategy over the last 10 years!

S.71– Short Squeeze Screen

Stock UniverseRussell 3000 Index
Rebalancing FrequencyAnnualIy
Screen CriteriaShort Interest/Float >60%  

Results

Well the results are in and, the numbers show that buying stocks of high short interest companies hasn’t been a great investment strategy! I must admit that it is sporadic creature, some years it is up massively, some years it declines big, and some years there are no stocks that meet the screen! So far this year we are seeing the upside volatility, the numbers tell us though that these short squeezes don’t hold up for the long-term. Our emergency screen ends with this public service announcement! 

Test Results and Chart

S.69 and S.70: Multipurpose Value Sweden

As with many happy endings, the next two screens I’m about to discuss arose from the ashes of a failure. Faithful readers have seen me describe a screening approach to value equities that I call multipurpose value (“MPV”); a sort of Swiss Army Knife for value investing due to its flexibility. I’ve tested it with success in Canada, Germany, the UK and even the US (to a lesser extent). The screen’s flexible nature stems from its search for companies that have either decent free cash flow yields or dividend yields and thus allowing the screen to choose companies on either metric that passes.

A Scandinavian Backstory

For some time, I’d been increasingly in the mood to run some tests on the Scandinavian markets and with this mind I ran the multipurpose value screen on Norway. I chose Norway over Denmark, Finland and Sweden, for no reason other than maybe because I’ve met more Norwegians than Danes, Finns and Swedes!

Then it happened: The market-beating MPV screen that had worked so well in Germany simply failed up in Norway. Not only did few companies pass the screen, but the handful that did underperformed the Norwegian equity indices by wide margins. Lacking the time to figure out why Norway offered so little reward for my screen’s value stocks, I next directed my attention to its neighbor – Sweden.

I once again ran the same MPV screen (the one that worked in Germany and failed in Norway) and what came out the other end, after combing through 1,256 companies, was simply awesome and mind-bending! I’d just discovered that Sweden was the perfect market for multipurpose value stocks! Not only were the returns great, but several dozen companies passed the screen – a real festival of value companies!

But, digging into the data, I spotted a problem. I saw that there were many (shockingly) small nano cap companies coming through the screen, some even with market caps under 10 million kroner (which at an 8 to 1 exchange rate to the US Dollar meant values of a little more than $1 million, yikes!). This was bad news, very bad, because one of the lessons you learn with stock market data is that you can’t often generalize results from super-small companies. Knowing this maxim led me to run a second version of the screen, requiring the company’s market cap to be at least 400 million SEK, still small, yet threshing out those troubling nano caps. Did the MPV screen still work? Yes. Putting it in nonfinancial terms, if Bjorn Borg (the tennis player) were a stock screen, he would MPV Sweden large or small.  Take a look for yourself at the results.

S.69– MPV Sweden All Cap

Stock UniverseSwedish Stock Exchanges
Rebalancing FrequencyAnnual
Screen CriteriaCurrent Dividend or Free Cash Flow Yield >4% Liabilities to Assets<0.5
Quarterly YoY Revenue Growth>0  

S.70– MPV Sweden Large Cap (>400SEK)

Stock UniverseSwedish Stock Exchanges
Rebalancing FrequencyAnnual
Screen CriteriaCurrent Dividend or Free Cash Flow Yield >4% Liabilities to Assets<0.5
Quarterly YoY Revenue Growth>0
Market Cap> 400 Million SEK  

Results

I bet you haven’t heard of Litium AB and Betsson AB yet these are prime examples of companies that came through the MPV Sweden screen (and if you were wondering each had year-over-year revenue growth in excess of 30% and free cash flow yields north of 5%).  Hardly any global powers made the list of MPV Sweden names but it didn’t matter the little minnows were strong enough to rack up some huge returns over the 10 years ending 2020. Interestingly, what I thought to be a nano cap mirage was disproven as the MPV Sweden screen with large caps outperformed.  Sweden: What a country for value! 

Test Results and Chart

S.68: India Growth+Value Screen

I’ve never set foot in India and what I know of the country was acquired largely from reading vintage National Geographic magazines, V.S. Naipaul’s non-fiction, and (the excellent) Maximum City by Suketo Mehta; and yet, notwithstanding my lacking on the ground experience, I present, from halfway around our great blue marble of a planet, a stock screen for Indian equities!

Although it can be challenging (due to spotty data), I’ve slowly built out a small portfolio of country specific screens including screens for the US, Canada, China (mainland), Germany, and the UK markets. Since I created a nice screen for mainland Chinese equities it was only logical to try my luck at a screen for the second most populous country and third letter in the famous BRIC abbreviation. The easiest approach to the task was to transport what worked in China over the Himalayas to India. (As a side note, with stock screens, it may take several passes to come up with something that works in beating an index. Efficiency thus played into my logic of transporting a Chinese equity screening approach to the Indian market.)

The first problem I encountered was this: China is not India! And this holds true with respect to equity markets. The issue I faced was that the S.54 China Growth +Value screen involved extremely lofty earnings (100% year-over-year) and revenue growth (30% year-over-year) hurdles and a very low price-to-book ratio (1.5), hurdles that hardly any companies in India (or any market for that matter) would pass. Well, what next? When in doubt, divide by two (or multiply by two for the price-to-book ratio) and that is what precisely what I did: Earnings growth was tamed to 50%; revenue dropped to 15%; while price-to-book was increased to a cutoff of 3; to tailor the screen somewhat for the Indian market.

S.68– India Growth+Value Screen

Stock UniverseFTSE India All Cap
Rebalancing FrequencyQuarterly
Screen CriteriaYoY Quarterly Revenue Growth>15%
One Year Net Income Growth >50%
Price-to-Book Ratio<3

Results

India was a tough market and the screen, although decent, did not produce the numbers it had in China, struggling to add value to the standard indices while experiencing extreme (Himalayan is an apt word) volatility. Due to the data issues, I ran the screen over a eight year horizon rather than the usual 10 years. Even after dampening the screen’s revenue and income growth hurdles few Indian companies could pass the screen, and the resulting concentrated portfolios no doubt led to some wild swings. Part of me thinks we are on the right path, but I learned that India certainly isn’t China and to improve the screen I’d need some local insights regarding the nuances of this market!

Test Results and Chart

S.67: Negative Earnings Surprise and Negative Free Cash Flow Screen

This installment, borrowing from the 1966 rock album by The Who, could go by the title of “A Quick One,” since it will be a brief take on yet another earnings surprise screen. Simply put: Today we are looking at negative earnings surprises and searching for bad returns.

Last time out we tested a screen for companies with five consecutive quarters of negative earnings surprises, and so this time, to make the situation even more dire, we require the companies to also have negative trailing 12 month free cash flows. Negative surprises on top of a negative cash flowing business, can it get any worse? Well next we see how bad it can get for these companies annual stock returns.

S.67–RU3K Negative Earnings Surprise and Negative Free Cash Flow Screen (NegSurp2)

Stock UniverseRussell 3000 Index
Rebalancing FrequencyQuarterly
Screen CriteriaGAAP Earnings Negative Surprises in Last Five Quarters
Negative 12 month Free Cash Flow

Results

The long-term results were bad, yet slightly positive overall, but that is not the interesting part of this story. You see, while the screen only compounded at 2.7% per year it did produce some blockbuster returns more recently (2019 and 2020). Bad fundamentals haven’t led to bad results, recently speaking as investors have been unphased by the negatives! I suspect the current easy credit conditions might also have something to do with these returns (credit is available to all).  That is all for today! I hope you enjoyed this quick diversion.

Test Results and Chart

S.65 and S.66: Positive and Negative Earnings Surprise Screens

We continue our analysis of earnings surprise screens and today we offer up, for your viewing pleasure, S.65 ESurp (a screen for positive earnings surprises) and S.66 NegSurp (a screen for negative earnings surprises). Since these screens are near mirror opposites, we can calculate and isolate the impact that positive earnings surprises have on stock returns relative to negative surprises (CEO’s should take note of this if they care about share returns). Our findings (results are shared below) have ramifications for long only as well as long short investors.  

The screens are near mirror opposites with the only difference being what types of earnings we use to determine the surprise. For positive surprises we look for adjusted earnings, which management can influence, while for negative surprises we look at GAAP earnings (e.g., the raw data stripping out the management influence).

S.65–RU3K Positive Earnings Surprise Screen (ESurp)

Stock UniverseRussell 3000 Index
Rebalancing FrequencyQuarterly
Screen CriteriaAdjusted Earnings Surprises in Last Five Quarters

S.66–RU3K Negative Earnings Surprise Screen (NegSurp)

Stock UniverseRussell 3000 Index
Rebalancing FrequencyQuarterly
Screen CriteriaGAAP Earnings Negative Surprises in Last Five Quarters

Results

As you’ll see, there is no surprise: Positive earnings surprises outperformed negative earnings surprises over the 10-years ending 2020. The ESurp screen did well by all metrics returning 15.2% per year versus NegSurp’s 8.3% annualized return returned. Take a minute though to think about the preceding statement: Over the last 10 years negative earnings surprisers generated positive equity returns, in the high single digits no less! (And this is why short selling is a thankless occupation for true masochists only.) 

Then there is the “spread,” which is the difference between the two series and a metric widely followed, and ultra important for long/short investors. Think about this: If you went long the positive earnings surprisers and shorted the negative earnings surprisers your return was 6.9% per year*, which was less than you would have earned from just buying the negative earnings surprisers! This indicates that long/short investing, at least using an equal-weighted long/short scheme, is also a pretty tough way to earn returns!   

*Simply speaking, because we did not subtract financing costs.

Test Results and Chart

S.64: Good Company + Earnings Surprise Screen

It is always good to revisit old friends and today I revisit one of the more interesting screens we covered last year and then add an interesting twist to it.

I define “good” companies as those with consistently strong financials and balance sheets and, on this theme, last year we profiled S.34, a Good Company screen, that produced nice pro forma results. But financial markets are funny, even if a company looks good on quality metrics, the investors with the most clout might be fixated and place more value on something else besides its quality (or goodness). Based on observations over the years I’ve come to accept, for example, that the horde of active analysts and traders at buy side investment funds value, perhaps more than all other factors, earnings surprises. There must be a reason fixation. Extra returns perhaps? I’ll test this hypothesis here by adding earnings surprise screening variables to our original Good Company screen to see whether the market does indeed offer evidence of an earnings surprise premium.

But first why are earning surprises the rage? Well, it might just be because they are powerful signals. For example, it could signal the CFO and CEOs at the company have their act together or know how to arrange the earnings to eek out the surprises sought by investors. Beyond signaling talented management teams (and/or earnings managers) surprises could further signal that the company is maybe tapping into an earnings or revenue stream that is stronger than anyone, including management expected (or was willing to acknowledge, e.g., low balling).  The secret sauce of the next screen lies in variables that look for serial “earnings surprisers,” i.e., those companies that surprise quarter after quarter after quarter. Additionally, because we want to screen in those companies that manage earnings well, we will focus on adjusted earnings, which can be influenced by management more so than GAAP earnings.

S.64–Good Company + Earnings Surprise Screen

Stock UniverseRussell 3000 Index  
Rebalancing FrequencyQuarterly
Screen CriteriaLiabilities to Assets<0.5
Six Years of Positive Free Cash Flow
12 Month CFO/Divided By Assets > 10% Decreasing YoY Share Count
Adjusted Earnings Surprises for Five Consecutive Quarters

Results

The results were as expected. The Good Company + Earnings Surprise screen outperformed our original Good Company Screen by 1.8% per year over the last 10 years. Both Good Company screens outperformed the Russell 3000 Index over the trailing 10 years with mid-to-high teens returns. The earnings surprise strategy was especially powerful in 2020 as the market wrestled with the uncertainty of the COVID pandemic. It seems from the numbers that investors were (logically) willing to pay up for positive surprises in a year of high uncertainty!  Well that concludes, for now, our test drive of earnings surprises, there are many more tests we can perform with this variable, so stay tuned!

Test Results and Chart

S.63: A BDC Growth Screen

To BDC or not BDC? Is our Shakespeare inspired question. Public BDC (business development companies) stocks have now garnered enough of a following among big yield hunters to merit their own S&P index. If you are not familiar with these securities, BDCs hold portfolios of small-to-middle market loans (aka not ready for prime-time loans to just about anything), leverage them around 2 to 1, and then distribute enticing dividends (e.g., in the range of 7% to 10% is typical) back to shareholders after the BDC’s manager deducts some hefty management and incentive fees. It is beyond doubt a yield play and somewhat gimmicky, yet a bit similar to how REITS work in passing through pre-tax income back to investors.

Before I get into the BDC screen, I wish to share, so you can get a feel for some of the actors in this space, a curious encounter I had with a person directly responsible for making BDC type small-to-mid-market loans. The setting was a coffee shop in the vicinity of Grand Central Station in Midtown, Manhattan, and this was where I’d agreed to meet an ex-colleague who was now involved in small-to-mid market loans. I’ve changed his name to Chalanoglu (Chah-lah-no-glue) in the near verbatim transcript of the informative part of our conversation.

ME:

So, what sort of due diligence do you perform when you make a middle market loan?

CHALANOGLU:

I get the documents from the investment bank working on the deal and make a few suggestions for changes, that is it. I don’t work on the weekends so I have them make the changes over the weekend — the bankers, they work on the weekends all the time so they will make any changes I suggest.  In addition, they know the changes I will likely make anyway since they are usually a few minor things that I make on all the documents. I’m glad in my job I don’t have to work weekends like the bankers, I just work during the week.

ME:

So, what happens when there is a problem with the loan? Like for instance the borrower can’t pay?

CHALANOGLU:

If there is a problem and the borrower can’t pay, all the lenders get together. Since these are small loans all the investors (lenders) know each other; we just have a call and agree to change the terms of the loan, usually we give the borrower more time. Since all the lenders in these loans know each other we always agree.

This brief exchange shares a window into the clubby little world of mid-to-small market lending and the wafer-thin due diligence performed by (some) of its operators. Since I like due diligence, I definitely would not trust Chalanoglu with any of my own money. I could go on about BDC dumpster fires (e.g., look up Fifth Street Finance) and pitfalls but that would get wordy and my point is simply the underwriting standards can be a bit loose and, well crazy things can happen! (For a nice summary of the BDC pitfalls the https://www.simplysafedividends.com/ website.)

The BDC Screen

Now back to the BDC screen. The S&P BDC Index was hard to beat and I learned this after I’d tried several screening approaches. Searches for high yielding, the most obvious reason to invest in a BDC, and BDCs priced below book value didn’t lead to index beating returns.  So then, what came to me was the idea that maybe for BDCs, like REITs, asset growth could do the trick. That was my lightbulb moment!  I set the screen to search for BDCs that had good asset growth year-over-year (10%+) and to make sure my friend Chalanoglu and his ilk didn’t stuff the BDC with bad loans, I’d also require positive trailing 12 month earnings per share.

S.63–BDC Asset Growth Screen

Stock UniverseS&P BDC Index
Rebalancing FrequencyAnnual
Screen CriteriaYoY Asset Growth> 10%
Trailing 12 Month EPS>0

The Results

Since the S&P BDC Index doesn’t have a long history (i.e., 2013) the test results only cover a five-year period (2015 to 2019). The screen yielded some excess value over the period due mostly from the strong 2016-2018 numbers. Overall, the returns from the BDC screen and the S&P BDC Index where in the range of the stated yields on these securities, indicating that there wasn’t much capital appreciation in the loan portfolios (even though most risk assets saw nice appreciation).This is a quirky part of the capital markets to be sure! Now that we’ve checked off our first screen of the year, we can happily move on to other interesting diversions!

Test Results and Chart

Create your website with WordPress.com
Get started
%d bloggers like this: