This is the third in a series of three articles outlining the stock trading models described in Grant Henning’s book “The Value and Momentum Trader” (John Wiley & Sons, 2010). The first article outlining Henning’s technical-momentum trading model was the August 2010 CI Online Exclusive, and the second covering his fundamental-value trading model was the September 2010 CI Online Exclusive.
Grant Henning, a college professor turned full-time trader, believes that following a quantitative, mathematical ranking system to select stocks is a good way to remove emotions from the investment process. Depending on where we are in the market cycle, either fundamental-value or technical-momentum strategies tend to outperform. To help alleviate the need to time the market, Henning developed a “hybrid” trading model that merged both fundamental-value and technical-momentum elements into a single system, which he discusses in his book “The Value and Momentum Trader” (John Wiley & Sons, 2010).
Just like with his other two models, Henning begins his search by creating a watchlist of roughly 200 stocks (depending on market conditions). To isolate the stocks in his watchlist, he applies three “qualifying variables”—share price, volume, and price gain (three-month and one-year):
For the Spreadsheet Corner article beginning on page 1, we used the SmartMoney Select website to screen for companies meeting the minimum share price and average trading volume requirements and that have also seen their stock price at least double over the last 52 weeks. We use this same watchlist of companies for this discussion, keeping in mind that it does not eliminate stocks whose prices have risen at least 30% off their low prices of the last three months. A portion of this watchlist is presented in Table 1.
Henning prefers to begin with technical variables, which he believes to be more restrictive, because they undercover stocks that are already under accumulation.
The 52-week multiple column in Table 1 represents the number of times the share price has multiplied from its 52-week low over the same period. It is calculated as follows:
Current Price ÷ 52-Week Low Price
Since one of the qualifying variables required the stock price to have doubled over the last 52 weeks, every stock in your watchlist should have a 52-week multiple of two or more.
The next column in Table 1 is percentage lag (% lag), which measures the degree to which the current share price lags the 52-week high price. It is calculated as follows:
(52-Week High Price – Current Price – $0.02) ÷ 52-Week High Price
To avoid the problem of dividing by zero if the current price and the 52-week high are both the same, Henning arbitrarily subtracts $0.02 from the current share price. Doing so, we discovered, also leads to the potential for a negative percentage lag value. Keep this in mind while performing your own analysis.
When evaluating the percentage lag, the smaller the value the better. This means that the current price is
|APP Pharmaceuticals Inc. (ABII)||2.977||0.027||167.539||38.077||4.675|
|American Oil & Gas Inc. (AEZ)||5.619||0.025||343.375||25.500||5.961|
|Atlas Pipeline Holdings LP (AHD)||2.910||0.082||53.242||138.786||1.628|
|Ashford Hospitality Trust Inc. (AHT)||3.186||0.042||112.722||42.373||4.531|
|Arlington Asset Investment Corp. (AI)||2.527||0.001||3791.071||19.808||5.250|
|Akamai Technologies Inc. (AKAM)||2.927||0.020||219.508||34.301||4.781|
|Allied Nevada Gold Corp. (ANV)||2.838||0.000||10635.683||58.917||1.731|
|Acme Packet Inc. (APKT)||4.512||0.005||1457.037||44.255||2.486|
|Atlas Pipeline Partners LP (APL)||3.163||0.041||114.371||96.507||1.969|
|ARM Holdings ADR Rept 3 Ord Shs (ARMH)||2.795||0.011||365.645||52.120||2.840|
|Aruba Networks Inc. (ARUN)||2.620||0.003||1537.195||48.892||2.209|
|Amtech Systems Inc. (ASYS)||3.249||0.033||146.115||92.503||1.968|
|AutoChina International Ltd (AUTC)||2.302||0.525||6.578||9.933||34.229|
|AXT Inc. (AXTI)||3.474||0.045||116.120||46.667||4.029|
|Brigham Exploration Co. (BEXP)||2.194||0.190||17.311||15.686||18.998|
|Baidu Inc. (BIDU)||2.582||0.045||85.500||25.074||8.375|
|Ballantyne Strong Inc. (BTN)||3.068||0.156||29.440||14.669||17.589|
|China Agritech Inc. (CAGC)||11.563||0.481||36.080||68.066||3.644|
|VocalTec Communications Ltd (CALL)||22.292||0.329||101.715||345.833||0.567|
|China Biologic Products Inc. (CBPO)||2.172||0.236||13.822||6.712||46.483|
close to the 52-week high. For Henning, anything above 3% presents an “overhang problem,” where investors who bought the stock at prices higher than the current price are waiting to sell their shares at a minimal loss. Therefore, while the price may increase, Henning believes it is reasonable to expect this rate of increase to be lower than shares without an overhang problem.
The next column in Table 1 is investment value. It is a variable Henning created, which is a weighted rate-of-ascent (increase) value (52-week multiple) divided by a weighted percent lag value. The formula is as follows:
(3 × 52-Week Multiple) ÷ (2 × % Lag)
Based on this calculation, stocks with high 52-week multiples and smaller percentage lag values benefit relative to stocks with lower multiples and higher percentage lags. For his own trading, Henning prefers stocks with investment values over 100.
Once he started trading his models, Henning discovered a “flaw” with his investment value calculation: It could be unduly influenced by a large price, only to have the price stagnate for the remainder of the year. Likewise, a stock may decline during the middle of the year only to rebound quickly to its pre-decline value but go no further. Henning realized he needed a way to “smooth” the rate-of-ascent curve and give priority to those stocks that continued to experience strong price gains throughout the year. He does this by using two other variables—three-month gain and investment value rank.
The three-month price gain is the same variable we used when arriving at our watchlist. It is slightly different than a simple percentage change in price over the last three months. Instead, Henning uses the lowest intraday price over the last three months and calculates the percentage increase from that point to the current price. The calculation is as follows:
[(Current Price – Three-Month Low Price) ÷ Three-Month Low Price] × 100
For our calculations here, we do not use the intraday low price. Instead, we use the lowest closing price over the last 65 trading days, which approximates 90 calendar days. Again, since Henning is interested in averaging a 10% gain on his investments per month, he eliminates any stocks that have not increased by at least 30% from their low of the last three months. However, our analysis does not make these exclusions; but this should not impact the final results.
The last technical-momentum variable Henning uses involves the rank function in Excel. Henning describes this rank calculation as his primary technical stock selection variable; all of the variables we have covered thus far are needed to generate the rank data.
Using the rank function, Henning first puts the companies on his watchlist in ascending order by investment value. The rankings are then “smoothed” by the three-month price gain to give priority to those stocks experiencing strong upward price movement over the last three months.
Within Excel, the function is as follows:
Rank = (RANK (G2, G2:G50, 0) * 2 + 100) / H2
After calculating the variables in Table 1, Henning’s next step is to assign points based on the level of each variable for each stock.
Henning assigns points in the following fashion:
The Spreadsheet Corner article beginning on page 1 in this issue outlines how you can build a spreadsheet to automatically retrieve the necessary data for Henning’s technical-momentum analysis.
After performing the “quantitative” analysis of his watchlist using these technical-momentum variables, Henning uses two websites to perform additional technical-momentum analysis of the stocks on his watchlist.
First, he analyzes point & figure charts for the stocks on his watchlist using the StockCharts website (www.stockcharts.com). When you pull up a point & figure chart for a given stock, the site sometimes provides green buy signals or red sell signals based on common point & figure chart patterns. If a stock has a green bullish point & figure signal, Henning awards it one point. When there is a red bearish signal, one point is deducted.
Lastly, Henning uses the “Opinion” section of the BarChart website (www.barchart.com). Here you can get estimates of a stock’s short-, medium-, and long-term technical prospects based on over 13 different technical indicators. Henning awards a point to stocks with an overall average rating of 96% Buy to 100% Buy. For stocks with an average rating of 25% Buy to 95% Buy, zero points are awarded. Any stock with a rating below 25% Buy loses one point.
To learn more about how Henning uses these sites as part of his analysis, visit the Computerized Investing website, see the box at the end of this article.
The second half of Henning’s hybrid model involves using fundamental variables that—based on his statistical analysis—have performed well as indicators of future price growth.
Henning describes strong earnings performance as being the “hallmark of any value-based approach.” Henning prefers stocks with earnings per share that are 10% or more of price per share. This translates into a price-earnings ratio of 10 or less. However, he assigns points on this basis:
For Henning, strong earnings are only one piece of the puzzle. He also seeks out companies with records of strong, consistent earnings growth quarter to quarter.
To evaluate the earnings growth of a company, he compares the quarterly earnings history to forecasted or future earnings. To do this, he compares the current price-earnings ratio based on earnings for the trailing 12 months (last four fiscal quarters) to the forward price-earnings ratio based on estimated earnings. Ideally, the forward price-earnings ratio will be “substantially lower” than the historical price-earnings ratio. This indicates growth in earnings going forward.
Henning also compares the company’s price-earnings ratio (P/E) to the average price-earnings ratio for the companies in the S&P 500 large-cap index. Following the market analysis of Yale Professor Robert Shiller, we use his S&P 500 price-earnings ratio of 19.8 as of August 31, 2010.
Henning awards points based on the fundamental-value variables as follows:
Henning recognizes that there are numerous ways to estimate the value of a stock—price to earnings, price to sales, price to cash flow, etc. However, he chooses to use book value because he feels it most closely approximates the liquidation value of a company and, thus, offers a measure of safety.
When examining the price-to-book value of a stock, Henning awards points as follows:
Henning views a positive divergence between rolling 12-month earnings per shareand the price-earnings ratio as a strong indicator of future price growth. Henning defines this type of divergence as an increase in earnings per share that does not lead to an increase in the price-earnings ratio or an increase in share price. Holding all else equal, an increase in earnings per share would lead to a decline in the price-earnings ratio. However, watching the price action of many stocks around earnings announcements illustrates that many investors tend to buy stocks with recent positive earnings announcements, which could offset the increase in earnings and cause the price-earnings ratio to increase.
Henning awards points based on EPS-P/E divergence in the following manner:
Henning believes that cash flow is perhaps the best indicator of a company’s future earnings and, thus, price growth.
He defines “cash flow” as the net cash flowing into or out of a company. He also defines “free cash flow” as cash that is “unencumbered by operational expenses.” This seemingly corresponds to the traditional definition of free cash flow, which is often defined as cash from operations less capital expenditures and dividends paid. While Henning views free cash flow as the more important measure, he considers both because the two often differ greatly from one another.
Henning uses a unique metric to consider cash flow and free cash flow—he takes the average of cash flow per share and free cash flow per share. He awards points in the following manner:
One point is awarded when the average of the current price-to-cash-flow and price-to-free-cash-flow ratios is less than 12.5;
Zero points are awarded when the average of the current price-to-cash-flow and price-to-free-cash-flow ratios is greater than or equal to 12.5 but less than 25; and
One point is deducted when the average of the current price-to-cash-flow and price-to-free-cash-flow ratios is greater than 25 or if either current cash flow or free cash flow is negative.
Once he has calculated the variables used in his technical-fundamental hybrid approach, Henning tallies the results and assigns a recommendation to each of the stocks in his watchlist.
There are six technical indicators and five fundamental indicators in Henning’s hybrid model. In theory, a stock can have a rating that ranges from –11 to +11. However, since he eliminates stocks with a percentage lag greater than 50% and three-month price gain of less than 30%, he rarely assigns negative values for those indicators.
Henning assigns his final recommendations as follows:
Table 2 shows the 17 highest-rated stocks based on our fundamental and technical analysis for Henning’s hybrid stock trading model. Out of over 120 stocks, none of them received any type of a buy recommendation. The stock receiving the highest rating—weak hold—was ExpressJet Holdings. As we can see from this listing, several stocks were rated very highly based on Henning’s technical-momentum criteria. Radcom Ltd. even received a perfect score—six out of six—but had a –3 fundamental-value rating. It is probably not surprising that even if a stock is performing well based on price momentum that its fundamentals would be weak, given the difficult operating environment over the last 18 months.
|ExpressJet Holdings Inc. (XJT)||5||0||5||Weak Hold|
|Grupo Financiero Galicia ADR Rept 10 Ord Shs Cl B (GGAL)||2||2||4||Weak Sell|
|Arlington Asset Investment Corp. (AI)||3||1||4||Weak Sell|
|Impax Laboratories Inc. (IPXL)||1||2||3||Sell|
|Radcom Ltd (RDCM)||6||-3||3||Sell|
|AXT Inc. (AXTI)||2||1||3||Sell|
|TriMas Corp. (TRS)||4||-1||3||Sell|
|Petrobras Energia Participaciones SA (N: PZE)||0||2||2||Strong Sell|
|Amtech Systems Inc. (ASYS)||3||-1||2||Strong Sell|
|IDT Corp. (IDT)||0||2||2||Strong Sell|
|KKR Financial Holdings LLC (KFN)||0||2||2||Strong Sell|
|Diamond Mgmt & Technology Consultants Inc. (DTPI)||3||-1||2||Strong Sell|
|Prospect Medical Holdings Inc. (PZZ)||3||-1||2||Strong Sell|
|Sauer Danfoss Inc. (SHS)||3||-1||2||Strong Sell|
|Polypore International Inc. (PPO)||3||-1||2||Strong Sell|
|CTI Industries Corp. (CTIB)||-2||4||2||Strong Sell|
|Energy Partners Ltd (EPL)||-2||4||2||Strong Sell|
Just because a company does not receive a favorable rating now, however, does not mean you should cast it aside. Henning warns of the transitory nature of his ratings, especially considering that more than half of the variables of his hybrid model relate to price activity. For this reason, he suggests never investing more than 10% of your available capital in any one stock.
While stock-picking strategies are often rooted in either value and fundamentals or momentum and technicals, Grant Henning attempts to blend the two.
His reasoning is that growth-momentum approaches typically do better during bull markets, while value-fundamental strategies tend to outperform during bear markets. Instead of trying to time the market to switch between the differing strategies, Henning combines fundamental and technical criteria into a singular trading model. His goal is to identify stocks that are likely to rise in price based on both historical performance and expected future performance.
Grant Henning uses a variety of technical and fundamental data when tracking his hybrid technical-fundamental stock trading system. Much of his analysis revolves around the collection and manipulation of raw price and volume data along with various income statement, cash flow and balance sheet items. Beyond this data analysis, however, Henning also uses two websites to supplement his technical-oriented analysis and one site for fundamental analysis. These sites are BarCharts.com, StockCharts.com and BigChart.com.
Barchart.com provides a wide variety of free and fee-based equities and commodities data, charts and analysis. Henning uses the site’s “Opinion” section, which analyzes a stock (or commodity) using 13 short-, medium-, and long-term analytics. Each of these analytics generates a buy, hold, or sell signal based on their current behavior and an overall percentage buy or sell summary figure is also generated.
The Barchart Opinion page shown here for homebuilder PulteGroup Inc. PHM is as of September 7, 2010. Here we see the indicators broken down into four-groups—composite, short term, medium term and long term. For each of the 13 indicators there are ratings for the signal, strength, and direction.
There are also average short-, medium-, and long-term indicator values, along with an overall average, which is what Henning uses for his analysis. PulteGroup’s 40% Buy short-term average, 25% Buy medium-term average and 67% Sell long-term average translate into an 8% Buy overall average for Pultegroup.
The Snapshot opinion section is a graphical representation of the overall ratings for today, yesterday, last week and last month, thereby allowing you to see the trend in the overall rating. The Snapshot page also contains a ratings box for the current and previous ratings, which ranks the overall strength and direction of the respective ratings.
Henning awards a point to stocks with an overall average rating of 96% Buy to 100% Buy. For stocks with an average rating of 25% Buy to 95% Buy, zero points are awarded. Finally, any stock with a rating below 25% Buy loses one point, which is what would have happened to PulteGroup based on its 8% buy rating.
The StockCharts website is a CI favorite for stock charting and technical analysis education. The site offers a range of free and subscription-based products, with prices ranging from $14.95 a month to $45.90 a month.
StockCharts.com is unique relative to most other charting sites in that it offers point & figure charts, which study pure price movements while ignoring the passage of time in the plotting of price action. To learn more about point & figure charts see the “Technically Speaking” column from the Second Quarter 2010 issue of Computerized Investing or visit the ChartSchool at the StockCharts.com website. Point & figure charting is free at the site.
The point & figure chart for PulteGroup, Inc. is shown here. When Henning creates a point & figure chart at StockCharts.com, he is looking to see if the site generates a signal for the stock, which is displayed just below the company name above the chart. Bullish signals are in green text and bearish in red. If a stock has a green bullish point & figure signal, Henning awards it one point. When there is a red bearish signal, one point is deducted.
In the case of PulteGroup, the company’s stock price had a double-bottom breakdown on May 25, 2010, which is a bearish signal. This means that, under Henning’s weighting system, PulteGroup would have lost a point. In addition to the bearish or bullish signal, the site also offers bearish or bullish price objectives, depending on the signal. The site is predicting a possible $3 price target once the double-bottom pattern runs its course.
The final website Henning uses when analyzing stocks for this hybrid technical-fundamental trading model is BigCharts.com. It is here that Henning attempts to identify companies experiencing a positive divergence between their trailing 12-month earningsand their price-earnings (P/E) ratio. In other words, the company has seen its trailing 12-month earnings increase, but it has not attracted enough buying pressure to drive the price-earnings ratio upward. While, holding price constant, an increase in earnings per share would lower the price-earnings ratio, Henning notes that investors are drawn to companies with increasing earnings, which would ultimately cause the price-earnings ratio to increase.
Shown here is an interactive chart from the BigCharts website for clothier J. Crew Group JCG. It consists of a six-month daily price chart, the rolling 12-month earnings per share of the company and its price-earnings ratio.
On August 26, 2010, the company reported earnings of $0.53 per share for the quarter ending July 31, 2010. This boosted the company’s trailing 12-month earnings from $2.25 to $2.49, a 10.7% increase. Looking at the chart, we see the increase in the rolling EPS line at the end of June, which corresponds to the end of the fiscal quarter (to which the latest earnings announcement belonged). Looking at the trend in the P/E line since the end of June, which reflects the increased rolling EPS, the price-earnings ratio has been relatively flat. In addition, we see that the price of JCG shares is still slightly lower than it was at the end of June when rolling EPS increased.
Henning awards one point when the rolling 12-month earnings per share is increasing and the price-earnings ratio is flat or declining, where the price is below the level seen prior to the increase in earnings. If the rolling earnings per share is rising and the price is increasing or rolling 12-month earnings per share is negative, he awards zero points. One point is deducted when rolling 12-month earnings per share is declining.
Using Henning’s weighting criteria, JCG would have been awarded a point.