A Simple Screening Technique To Improve Returns

by: Kurtis Hemmerling


Stock screening can quickly isolate shares of companies you may want to purchase.

Ranking stocks according to your preferred screen factors can highlight the firms of highest potential.

The industry standard for screening and ranking stocks may not produce the best returns.

An alternative screening method is discussed with simulated performance charts.

As investors, we are always looking for new ways to find great stocks. Some start by looking for companies with the most intriguing story and then dig for the best value. My preference is the other way around. I filter out the best value stocks and then I look for any names which might have an interesting story. I screen first for a few reasons:

Time. With over 9,000 stocks to look at (major US exchanges and OTC), I simply do not have the time to read about all these firms first.

Emotions. As a human with a beating heart, I have a tendency to become overly attached to a company when I read an interesting story of what might happen in the future - and I tend to overlook or rationalize other aspects of a stock which are unappealing.

Value. I am a hard-core value investor. A company may become a world giant with over 1,000% year-after-year growth but if the share price is unfavorable today, I won't touch it with a 10-foot pole. Excess returns are made from the difference of what a company is expected to do - and what it really does do. If the share price reflects an expectation of 1,000% year-over-year growth expectations and future growth is a mere 900% year over year, this holding will underperform.

It is my personal view that screening for stocks is an important first step before performing due diligence on a handful of names.

The Simple Screening and Sorting Technique

But I digress, as mentioned in the title - what is the simple screening technique to enhance returns? The paper, Combining Value and Momentum (by Fisher, Shah, and Titman), discusses a certain method - although I wouldn't necessarily call it new. What is the process? First, we need to look at the standard screening/ranking process.

The common approach for screening/ranking is to rate a stock against other stocks based on numerous factors. Let me provide an example.

When building investment models, I use Portfolio123 which allows me to create a stock ranking system. In brief, you can select various factors or ratios (price to earnings, price to book, dividend growth, etc.) and bundle them into one comprehensive ranking system. You may believe the following:

  • Low price to earnings ratios are better
  • High 3 year dividend growth is better
  • High Piotroski F-Score ratings are better
  • High Altman Z-Score is better

This ranking system will look at these 4 factors in the stock universe of your choosing. It will then combine the scores and rate the stock between 1 and 100 compared to the other stocks in your universe. You generally consider the stocks with the highest overall rank. This process works well.

The second approach is to not choose stocks with the highest average score, but to choose stocks that are high in all the factor categories. In this second approach you would only purchase a stock if it had a high rank as regards PE, Div Growth, F-Score and Z-Score. Let me run though a quick example:

You have 100 stocks. You are screening for low price to book ratios and high 52-week returns thus having a value/momentum portfolio. You run your ranking system over the 100 stocks and have a relative score for each stock numbering from 1 to 100. Below is a random selection of 5 stocks.

· Stock ABC: PB rank 50 52W Return rank 10

· Stock DEF: PB rank 40 52W Return rank 20

· Stock GHI: PB rank 30 52W Return rank 30

· Stock JKL: PB rank 20 52W Return rank 40

· Stock MNO: PB rank 10 52W Return rank 50

If you simply take the highest average rank, then each of these stocks look identical with a combined score of 60. But these firms are very different even though they have the same score. Some have great value and low momentum while others have terrible value and high momentum.

The process discussed in the paper, Combining Value and Momentum, takes only the highest value and the highest momentum stocks. In this example, assume we have a value cut-off of 30 and a momentum cut-off of 30. Therefore, only stock GHI would be selected since it meets our minimum standards for both value and momentum instead of using a combined score.

Why this works is discussed in the paper and is a topic for a future article. Next, I want to construct a portfolio for dividend growth using this screening methodology. It is my belief that the relationship between value, dividends and future growth is strong - much like the relationship between value and momentum.

Creating the Dividend Growth Screen

The first step is to start with a dividend growth universe. Using Portfolio123's point-in-time engine, I have created a dynamic universe of stocks which have experienced 5 consecutive years of increasing dividends.

This will be my benchmark to determine the efficacy of the screening process. Overall, investing in dividend growth stocks has generated excess total returns with slightly less drawdown to the market. Even though annual returns are over twice as high as the S&P 500, the volatility of such a portfolio is less than the market.

In the following steps, I will apply various value and growth filters to the dividend growth universe - one at a time. Each filter will be applied to the entire universe of stocks and not to the reduced pool of screened securities. This is to say that each filter works independently of each other and we are not combining factors for an average score, each factor will remove the bottom scoring half of the original universe. Thus, it matters not which factor we start with.

First, I filter out the lowest dividend yielding 50% from our dividend growth universe.

This one filter has improved returns, lowered drawdown and decreased overall volatility. The succeeding steps will be modeled after the process in the Combining Value and Momentum paper.

This next filter still removes the lowest 50% dividend yielding names from the original universe, as well as removing the bottom 50% ranked stocks according to earnings yield from the original universe.

The combination of high earnings yield and high dividend yield is a good one. The earnings yield is what provides the base support for a sustainable high dividend yield.

The next factor we will screen for is a variation of the PE to projected growth ratio, or better known as the PEG ratio. This factor compares the projected PE ratio to the projected long-term growth ratio which includes dividend yield in this instance. This concept made popular by Peter Lynch attempts to put perspective on current valuation by comparing it to future growth. A higher PE ratio may be rationalized by higher growth perspectives and a low PE value may not be good value if future growth is low or negative.

In the following screen we also exclude the bottom 50% ranked stocks according to the PEGLTY ratio.

So far, we have taken a handful of sensible factors that relate to our dividend growth universe and have removed the bottom 50% ranked names. Remember, each factor is run across the entire dividend growth universe. Each succeeding factor is not used to reduce the preceding sort by an additional 50%. There can be overlap between factors as to which stocks are removed.

Running 10 Stock Sorts

I will stop the screening process as outlined in the paper as our current list of names is only 50 positions long, although it has been as high as 150 stocks around the turn of the century.

The next step will be to take the final screening process above and apply a sorting process to it. What will be the effect of holding only the 10 highest ranked names from our screen above? For instance, what if we prefer small market capitalization? Or the big yields? Or high dividend growth? I will outline below the outcome if we kept the 'best 10' stocks from the screening process above.

Annualized Return

Max Drawdown

Sharpe Ratio

Sortino Ratio

Standard Deviation



Highest Yield








Highest Yield vs. 5-Yr Avg








Highest annual div growth








Highest Earning Yield








Lowest Price to Book








High % Return on Assets








Worst trailing 4 week Performance








Lowest volatility
















One word of caution: don't focus on the highest annual returns only as turnover is also very high in certain sorts (low volatility and low 4-week returns). Once you factor in transaction costs, the additional returns would be greatly reduced. This chart is merely to demonstrate the sorting process according to investor preference.

Let's have a quick look at one of these sorts - the highest % return on assets:

Ticker Name Rank SectorCode MktCap
(NYSE:LO) Lorillard Inc 99.44 STAPLE 21769.2
(NYSE:BR) Broadridge Financial Solutions Inc 98.87 TECH 5033.34
(NASDAQ:ARLP) Alliance Resource Partners LP 98.31 ENERGY 3580.11
(NASDAQ:AHGP) Alliance Holdings GP LP 89.83 ENERGY 4223.93
(OTCQX:CSVI) Computer Services Inc 89.27 TECH 510.22
(NASDAQ:XLNX) Xilinx Inc. 88.42 TECH 11322.37
(NASDAQ:INTC) Intel Corp 84.75 TECH 172883.1
(NYSE:EMR) Emerson Electric Co. 83.62 INDUSTRIAL 45165.29
(NYSE:HP) Helmerich & Payne Inc. 83.33 ENERGY 10807.02
(NASDAQ:MSFT) Microsoft Corp 82.2 TECH 371990.9

The purpose of this article is not to discuss the merits of each of these companies - that is the secondary process where the investor performs his due diligence before buying. This article is focused on how to screen for stocks of high potential using the aforementioned process.

Take Home Message

It is an industry standard practice to build a ranking system which gives stocks an average score by analyzing various factors. However, this may not be the best approach. Instead of using a high average score, there may be a significant benefit to selecting stocks which have a high score with each factor.

This portfolio creation technique was used with a handful of value and growth factors in a dividend growth universe. Future articles will explore this technique in conjunction with other common strategies.

Disclosure: The author has no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.

The author wrote this article themselves, and it expresses their own opinions. The author is not receiving compensation for it (other than from Seeking Alpha). The author has no business relationship with any company whose stock is mentioned in this article.

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.