In times of market uncertainty, using models can help investors – whether they rely on fundamentals or adopt quantitative strategies – to improve their security selection. A recent study demonstrates that combining two key StarMine models can assist investors find stocks with strong momentum and value as they try to steer clear of that all-too common peril: the value trap.

One of the biggest challenges confronting investors today is a perennial one: how to identify those securities that are undervalued or mispriced, without falling into the trap of buying a cheap stock that is simply going to get cheaper still or, nearly as bad, mis-timing a purchase. The greater the uncertainty, the more limited the visibility, the more difficult it becomes to steer clear of these value traps, and the more important earnings quality becomes.On a stand-alone basis, two StarMine models have proven to be helpful in preventing investors from tumbling into these traps and in generating above-average returns over the last 12 months. The StarMine Valuation-Momentum (Val-Mo) model has generated, on a stand-alone basis, a top/bottom decile spread (the difference between the returns of the top decile and the bottom decile) of 27.5% in the United States over the last 12 months -,while the StarMine Earnings Quality (EQ) model, also has performed well on a stand-alone basis, generating a decile spread of 26.7%.

Accordingly, to help institutional investors who use screening techniques to search for new portfolio ideas or quantitative investors who are looking for alpha-generating models that enhance their investment process, we decided to evaluate the benefits of combining these two models to produce a new screen – one that will help both fundamental and quant investors from accidentally finding themselves trying to “catch a falling knife”, as the old market adage puts it. The aim is to assist both groups, as the gulf that once separated fundamental investors from their quant counterparts is narrowing. Today’s quantitative investors increasingly likely to use fundamental metrics to confirm their analytic work, and fundamental investors are turning more and more to quantitative signals as a way to screen the vast universe of potential investment ideas and identify inexpensive stocks with strong fundamentals.

In our study, we screened all stocks trading on U.S. exchanges with a market capitalization of more than $1 billion (all 2,804 of them) using the StarMine Val-Mo model, and selected those that score 80 or higher on that model. Then we went on to screen the 594 companies that passed that first hurdle, using the StarMine EQ model in order to winnow out potential value traps. Only 101 companies with a Val-Mo score of better than 80 could also boast an EQ score of 70 or higher. To see how employing that approach to building a portfolio works over time, we back-tested the strategy for the last decade; each stock was given an equal weight, dividends were reinvested and the portfolio was rebalanced at the end of each month, removing those stocks that no longer passed muster on both screens and adding those that did. -

The results were compelling. As you can see in the chart below, combining the Val-Mo + EQ models and building a model portfolio composed of those stocks with high scores on both created a portfolio (represented by the blue line) that handily outperformed both the S&P500 (represented by the green line) and the total investable universe of stocks (represented by the red line). Over the last decade, the Val-Mo + EQ portfolio soared by 451%, while the S&P500 advanced only 54%. (The universe of investable stocks – those listed on a U.S. exchange with market capitalizations of $1 billion or more) did better than the S&P 500, but not nearly as well as the sub-universe created by employing the two StarMine models.

During the 120 months of the study period (the number of months in which the portfolio was rebalanced), combining the StarMine Val-Mo and EQ models outperformed about two-thirds of the time, or during 77 of those months. For instance, during August 2012, this Val-Mo/EQ portfolio generated a return of 3.8%, compared to a 1.9% gain for the S&P 500.

Aetna Inc. (AET.N) is one company that contributed to that August outperformance. Its scores on both StarMine models rose to the point where it was added to the portfolio as of December 1, 2011; it was removed on May 1, 2012 when it failed the Val-Mo screen. By then, the stock’s price was already declining, signaling that investors feared it may have been overvalued. It re-entered the portfolio July 1 when the Val-Mo score reached the top quintileonce more, and so its gain of more than 5% during August was captured in the strategy’s returns for that month. From the period between December 1, 2011 and May 1, 2012, Aetna’s share priced gained 6.4% -, after which the strategy’s screen correctly predicted that its momentum was breaking down after the swoon in April. The chart below captures the performance of Aetna’s stock during periods when it was part of the model portfolio (represented by the sections of the chart shaded in grey) and when it wasn’t. After being dropped from the portfolio as of May 1, 2012, the stock extended its decline, falling from $48.52 a share to $38.77 for a total loss of 11% during May and June, the two months in which it wasn’t in the portfolio. The stock’s decline put its Val-Mo score back into the top quintile, prompting its return to the portfolio on July 1; while that was, perhaps, a month early, the insurer’s share price bounced back in August. (It remains in the portfolio today.)

We ran the Val-Mo + EQ screen to see what stocks are in the modelportfolio today andidentified five companies that may be interesting investment candidates to scrutinize at this point in time. We also ran the inverse of these two screens, trying to identify companies with both poor Val-Mo scores and low EQ rankings, in search of companies that may prove attractive candidates for investors looking for stocks to short. Here are the results:

Positive Screen : Val-Mo >80; EQ >70; Market-cap > $1 billion

1) Aetna Incorporated (AET) – Coventry acquisition diversifies earnings by giving Aetna access to the government segment.

2) Foot Locker Inc (FL) – Same Store Sales growth is strong; and demand for athletic footwear is robust, driving store traffic.

3) Papa John’s International Inc. (PZZA) – Strong international growthand positive guidance are driving profits.

4) Microsoft Corp (MSFT) – A strong product cycle is approaching, led by Windows 8 and Surface RT Tablet.

5) Wells Fargo Company (WFC) – Strength in mortgage revenues continues to drive earnings.

Negative Screen : Val-Mo <20; EQ <30 ; Market-cap > $1 billion

1) Applied Materials Inc (AMAT) – The company released negative guidance on weak solar energy demand.

2) Hyatt Hotels Corp(H) – The acquisition of new properties and investments may dilute earnings.

3) Tesla Motors Inc (TSLA) – Orders are back-end loaded for the year, and look weak.

4) Hillshire Brands Co (HSN) – Flat sales growth expected in FY13; accounting irregularities in Brazilian operations cloud outlook.

5) JC Penney Company Inc (JCP) – Suspended guidance and its dividend payments amid sales dip and management restructuring.

These represent a few of the specific investment ideas captured in this model-based approach to building a portfolio. Combining these two StarMine models helps us identify stocks that are cheap (relative to others in the region) but at the same time offer momentum and the probability of sustainable earnings. Investors can overlay their own screens on top of this, and potentially create a customized portfolio that performs still better. Of course, no strategy is ever foolproof, but as the results of our study show, this approach does tilt the odds in your favor. In future AlphaNow articles, we’ll look at how these models have contributed to investor returns in other regions.

 
Learn more about how QA Studio and StarMine Quantitative Models can help you discover alpha.
Request a demo today.