A Systematic Earnings Revision Model To Help In Your Stock Picking Process

A Systematic Earnings Revision Model To Help In Your Stock Picking Process

On Validea.com, we run 22 public facing models. Most of the models are a combination of fundamental factors, combining measures like profitability, valuation, cash flows, levels of debt, relative valuation, price strength and other factors that indicate the health of a company and attractiveness of a stock. However, there is a small group of models we run that are laser-focused on one key investment attribute and screen for the stocks that exhibit the very best of this characteristic.

One of those models is the Earnings Revision Investor model we deploy based on Wayne Thorp’s (an analyst at The American Association of Individual Investors, AAII) paper, How to Profit From Revisions in Analysts’ Earnings EstimatesBefore we get into the details on the model, it’s important to understand the drivers behind earnings estimate revisions and how in the current environment these profit revisions could prove very helpful for the active stock investor.

Earnings Expectations Matter, A Lot

In late July, Barron’s published a piece, “Why This Rally Still Has Room to Run”, and in it there was a very important earnings revision stat that is important to today’s market:

“More than 300 S&P 500 index companies have reported second-quarter numbers so far. About 85% are beating Wall Street earnings estimates by an average of 22%. Both numbers are huge. Companies always beat numbers, but not like this. In the first quarter, about 70% of S&P 500 companies beat analyst estimates by an average of 3%.”

What typically happens during downturns is that companies guide earnings estimates down as demand for goods and services fall. As a result, Wall Street analysts follow suite by bringing down their own profit estimates. In many cases as earnings are troughing, Wall Street estimates become too pessimistic, thus giving companies the opportunity to beat actual results once reported as the downturn subsides for many companies. That is exactly what the Barron’s stat is telling us. If in aggregate, the 300 firms that beat earnings were expected to deliver $1.00 in earnings in Q2, they exceeded that and instead generated $1.22 in earnings. In fairness, this downturn wasn’t your typical, run-of-the-mill recession and trying to project earnings, especially for some businesses most impacted by the shutdown, was impossible, but the trend of overestimating to the downside and companies beating estimates, and then analysts having to play catch up with revisions upward, is a very persistent pattern that takes place during most economic and market contractions.

Earnings revisions don’t just take place in bear markets though. Growth companies can deliver higher growth, thus resulting in higher earnings estimates and upward revisions, and firms with slow or even negative growth can see upward revisions if expectations are too low. One of the theories as to why investing in stocks of companies with upward earnings revision works is centered around investor reaction to the news, with the idea that better expected earnings takes time to reflect itself in the market and there tends to be a underreaction to good news.

So, while there are many examples and situations that can drive upward earnings revisions, the key really comes down to an improved outlook for earnings over market expectations, which increases the intrinsic value of a stock. Having a systematic approach to identifying those companies with improving earnings can be accretive in one’s active stock selection process over time.

Five Simple Earnings Revision Criteria In One Model

The model we run on Validea based on Thorp’s screen and paper is relatively straightforward.

First, stocks need to have sufficient analyst coverage on Wall Street so only those stocks with 4 or more analysts are considered for eligibility. This acts as the way to define the investable universe and screens out very small caps and microcap companies with very little analyst coverage to begin with.

Next, the screen favors companies where the current year EPS today is greater than one month ago, which indicates recent revisions upward for the current year earnings.

It then applies the same measure to the next year and wants to see that next year’s earnings estimates have also been revised upward compared to one month ago.

It then looks for more than one upward revision over the last 30 days for both the current and next year and that there are no downward revisions. This helps ensure that one estimate is not driving the passing of this criteria and that there is more herding of revised upward estimates by Wall Street.

A Concrete Example

Let’s Look at an example based on real-time data as of September 4th, 2020. We’ll use Dick’s Sporting Good (DKS). All of this data can be accessed in real-time using our Guru Analysis tool. The example of Dick’s is a good one when you think about it. During the shutdown, people have flocked to outdoor activities – biking, camping, home fitness and kayaking – and the migration to online buying has fueled sales for the retailer causing analysts to play catch-up with their upward revisions as Dick’s produces impressive financial results compared to expectations a few months ago.

Here is are the criteria for the model along with the earnings revision data for Dick’s.

ANALYST COVERAGE This strategy requires that the stock has at least 4 analysts covering it. The number of analysts covering DKS is 18, which passes this criterion.
CURRENT YEAR ESTIMATE REVISIONS The first requirement is that the current year EPS estimate be greater than it was one month ago. DKS’s current year EPS estimate is $2.09 as compared to $0.58 one month ago, which passes this criterion.
NEXT YEAR ESTIMATE REVISIONS The second requirement is that the EPS estimate for the next fiscal year be greater than it was one month ago. DKS’s next fiscal year EPS estimate is $4.03 as compared to $3.46 one month ago, which passes this criterion.
CURRENT YEAR UP AND DOWN REVISIONS The company should have more than one current year positive estimate revision in the previous thirty days and no negative revisions. DKS has 19 up revisions and 0 down revisions in the past month, which passes this criterion.
NEXT YEAR UP AND DOWN REVISIONS The company should have more than one positive estimate revision for the next fiscal year in the previous thirty days and no negative revisions. DKS has 19 up revisions and 0 down revisions in the past month, which passes this criterion.
SUMMATION The model ranks stocks on a score of 0-100% and those stocks that pass all five criteria are the top rated stocks according to the model.
Revisions Can Augment Other Strategies

Because this model focuses on stock’s exhibiting only one general characteristic – positive and upward earnings revisions – some investors may feel more comfortable combing the strategy with another approach that takes into consideration value, quality or other fundamental factors. For example, using the Guru Stock Screener we can search on stocks that get 100% based on the Earnings Revision approach and augment that with a value model like the Value Composite approach based on the work and research of Jim O’Shaughnessy in What Works on Wall Street. Combining these two models culls the list down from 250 names to 10 stocks that look cheap, at least based on traditional value metrics like price-to-sales, price-to-cash flow and other measures but are also seeing upward revisions in their earnings estimates. There are 20+ other models on Validea that can be combined with the Earnings Revision and this is just one simple example.

When constructing our focused list of 10 and 20 stocks for the model portfolios we track based on the strategy, we narrow the list of stocks down using our composite ranking, which looks at stocks through the lens of all of our quantitative strategies. Each stock is assigned a composite ranking, which creates a systematic way to get the list of stocks to a manageable set of 10 or 20 names and also incorporates all of our other strategies, which consider a wide array of fundamental factors.

The Results Are In

Using the optimal model portfolio tool on Validea, the Earnings Revision Investor model performs best under the 10-stock version with quarterly rebalancing as of this writing. Since 2009, that version has beaten the market in 9 out of 12 years, which is a good percentage of outperforming years. The best year was 2013, where the model was up 54.7% on a back-tested basis. In 2018, the model fell on tough times and was down 32.5% vs. a loss of 6.2% for the market. So far this year, the model is up slightly compared to a 6.1% gain for the S&P 500. Overall, the 10-stock quarterly rebalanced portfolio has bested the S&P 500 since inception. (note: none of these returns are actual returns and please see our disclaimer at the bottom of this page for more on the performance metrics relating to our model portfolios).

Over the past two quarters we’ve gone from shutdown to re-opening to signs of a recovery. Some businesses and industries will be changed for years, while others have seen increases in demand for their goods and services as consumer preferences change and as certain trends have been accelerated forward. There has been a bifurcation in the markets with companies that are seeing increases in business and those being hit the hardest. Paying attention to where the earnings revisions are taking place is one-way investors can find those companies that are seeing upticks in their business and underlying profitability. Finding those stocks using strategies like the Earnings Revision model presents an opportunity for today’s active stock investor.




Leave a Comment

Your email address will not be published. Required fields are marked *

Related article

United States Investing Championship 2023 Final Results

https://www.businesswire.com/news/home/20240126217806/en/United-States-Investing-Championship-2023-Final-Results LOS ANGELES–(BUSINESS WIRE)–The United States Investing Championship just announced the final standings for the 2023 Investing Championship. There were three hundred fifty-two competitors. Prior

Read More →