Wayne Thorp will speak at the 2015 AAII Investor Conference this fall; go to www.aaii.com/conference for more details.
Grant Henning is a former college professor who turned stock trader and has been developing and testing mathematical stock selection systems for over 20 years. In his book, “The Value and Momentum Trader” (John Wiley & Sons, 2010), Henning outlines three strategies for picking stocks—a technical-momentum model, a fundamental-value model, and a “hybrid” technical-fundamental model.
Henning’s research found that his technical-momentum model performed best during bull markets, but lagged his fundamental-value model during bear markets. Knowing that neither bull nor bear markets last forever, Henning devised a hybrid value-momentum strategy that caters to all market environments.
This article is first in a series in Computerized Investing discussing Henning’s three trading models. We start here with his technical-momentum approach and in the future will highlight his fundamental-value strategy and, lastly, his hybrid model. For a recap of Henning’s technical-fundamental stock selection strategy, you can also read the AAII Stock Screens column in the August 2010 issue of the AAII Journal.
A Pragmatic Approach to Stock Trading
Throughout his book, Henning advocates a pragmatic approach to trading. He is more concerned with the results—in terms of profitability—of a trading system than understanding exactly why the system works.
Henning considers a successful trading strategy one that will earn him an average monthly return of 10% on his invested capital. If a system fails to generate these kinds of results, Henning believes it is time to pause, reexamine and adjust or replace the trading system.
Impact of Market Timing
A key element of trading for Henning is timing—when to buy, when to sell, how long to hold, and when to be out of the market altogether. Depending on where the market is in its cycle, Henning also weights his stock selection more toward momentum and growth or toward value and earnings.
To help him react to market trends, Henning focuses on a “trading window” of a few days to a few months. As a “swing trader,” Henning can capitalize on the short-term trends that typically underlie a broader price trend. In his book, Henning offers a simplified example of a stock that moves from $10 to $15 over the course of a year. However, over that period, the stock makes a series of intermediate moves—from $10 to $8 to $12 to $11 to $17 to $14 to $15. If a trader were able to react to this movement—buying low and selling high—he would buy at $8, sell at $12, buy again at $11, sell at $17, and buy again at $14. While the likelihood of timing the market so successfully is doubtful, had the trader succeeded he would have earned $137.5%, versus a 50% gain from buying at $10 and selling at $15.
Beyond knowing when to buy and sell, Henning notes that that are times when it is advantageous for traders to be fully invested in stocks and times when it is better to be out of the market altogether. He mentions several “unreliable or quasi-reliable” market timing systems and two “reliable” market timing indicators—the CBOE Volatility Index (VIX) and the put/call ratio. The Technically Speaking column in the Third Quarter 2010 issue of Computerized Investing discusses the VIX (available at www.computerizedinvesting.com).
The put/call ratio is the ratio of put options to call options traded at the Chicago Board Options Exchange (CBOE) and is used by traders as a market timing tool. This data is available going back to October 2003 at the CBOE website (www.cboe.com).
In “The Value and Momentum Trader,” Henning illustrates how peaks in the five-day moving average of the put/call ratio and the VIX historically have usually marked near-term bottoms in the S&P 500 index. Henning suggests that traders can decide whether to be short, long, or out of the markets based on the movement of these two indicators.
A Technical-Momentum System for Stock Selection
For Henning, the biggest advantage to using mathematical trading models is that they lend consistency to the investing process and can help eliminate emotion from the equation (assuming the models are followed).
Most trading systems can be characterized as being technical, fundamental, or “hybrid” technical-fundamental. Early in his trading career, Henning preferred a value-oriented approach. However, as he discovered that different types of stocks performed better in different segments of the market cycle, he moved to a hybrid approach that utilized elements of technical investing and value investing. The focus of our discussion for this article is Henning’s technical stock selection model.
Technical traders focus their analysis on price, momentum, volume, patterns in price movement, etc. Technical “purists” believe that all relevant information regarding a company—earnings, prospects for the future, management, sector strength, etc.—is captured in the stock price.
Henning feels that a technical-momentum system is more suited for bull markets. While, as he says, bull markets do not last forever, he presents a technical-momentum trading system as a “preliminary step” in his overall stock selection methodology.
Building a Watchlist
Whether he is following a technical, fundamental, or hybrid technical-fundamental trading system, Henning employs a set of qualifying variables for stocks that allow him to narrow his focus. These variables include:
Sufficient Rate of Gain
As we discussed earlier, Henning’s goal is to generate an average monthly return of 10% on his invested capital. Therefore, he initially seeks out stocks with solid price momentum. To this end, he looks for shares that have risen at least 30% from their lowest price over the last three months and that have at least doubled in price over the last 52 weeks.
Sufficiently High Share Price
Henning also prefers to invest only in stocks that are priced above $5 per share. He feels that stocks below this price level are more susceptible to price manipulation. Furthermore, as a stock price crosses key threshold levels—$5, $10, $15, etc.—there is a likelihood of increased institutional interest and trading liquidity.
Sufficient Trading Volume
Lastly, Henning’s trading systems require stocks to have sufficient liquidity, which he defines as average daily trading volume over 10,000 shares. Investing in thinly traded stocks can make it difficult to close out a position quickly without forcing the price down. Also, stocks with low average trading volume tend to have higher bid-ask spreads, which increase the cost of trading.
Table 1 summarizes Henning’s qualifying criteria for generating his initial watchlist, as well as the variables he uses for this technical-momentum trading system. Using AAII’s Stock Investor Pro fundamental stock screening and research database program, we applied Henning’s set of qualifying variables to the stock universe as of July 16, 2010, and came up with a watchlist of 76 companies. A subset of these companies is listed in Table 2.
|Table 1. Grant Henning's Technical-Momentum Stock Trading Model in Brief|
Philosophy and Style
Grant Henning follows a pragmatic trading philosophy. He is more concerned with the overall results of a trading system than with fully understanding what is driving those results. For his own trading, Henning defines success as averaging a 10% gain per month on his invested capital. If a trading system fails to yield this type of result over an extended period of time, Henning will step back to re-evaluate the system and, if need be, either modify it or abandon it altogether. Henning also considers himself a “swing trader” with a trading window of a few days to a few months. By having such a short-term focus, Henning feels he is a better able to react to short-term trends that present themselves within a longer trend.
Henning uses a set of qualifying variables to arrive at a watchlist from which he selects the stocks he buys. In order to be part of Henning’s watchlist, a stock must meet these criteria:
After generating his watchlist, Henning examines a variety of variables tied to price action and awards points based on their values:
Once he derived his initial technical-momentum model, Henning admits he refined it by employing two additional indicators, which he also quantifies in order to arrive at a final technical scoring for the stocks on his watchlist:
Using the StockCharts.com website, Henning awards one point to stocks with a bullish point & figure signal; gives no points for stocks where no point & figure signal is generated; and deducts one point for a bearish point & figure signal.
Henning also looks for stocks with accelerating volume over the last 10 trading days relative to the average daily trading volume over the last three months. Since he does not outline a specific scoring method for analyzing volume acceleration, we chose to award one point when the ratio of 10-day average daily trading volume to average daily trading volume over the last three months is above 1.25; give no points when this ratio is between 1.0 and 1.25; and deduct one point for ratios below 1.0.
Portfolio Creation, Monitoring & When to Sell
After he scores the stocks on his watchlist, Henning assigns recommendations based on the overall scores. Stocks with a score of five or higher are considered a “buy,” while a value of four is considered a hold. Scores of three or less are considered a sell. Henning will only buy those stocks rated a buy. When trading technical-momentum stocks, Henning suggests holding at least 10 to 12 of them for adequate diversification. He also warns against investing more than 10% of your capital in a single stock.
If Henning holds a stock that becomes rated a sell, he considers how difficult it will be to sell off the position given its liquidity. Furthermore, he will consider holding onto a low-rated stock if its price remains close to its 52-week high. Lastly, the availability of better investment candidates plays a key role in Henning’s sell decisions. Since these technical variables are driven primarily by price, the stock ratings can change rapidly. Therefore, Henning suggests updating the data for the stocks in your watchlist after the close of each trading day.
After Henning arrives at his watchlist, he begins analyzing several different technical variables, which he ultimately uses to rate the stocks.
52-Week Price Multiple
The 52-week multiple column in Table 2 represents the number of times the stock price has “multiplied” from its 52-week low to its current level. Its formula is as follows:
Current Price ÷ 52-Week Low Price
Every stock in our watchlist has a 52-week price multiple above two, since one of the qualifying variables requires that the stock price has doubled over the last 52 weeks.
The percentage lag variable measures the degree to which the current price falls below, or “lags,” the 52-week high price. Henning calculates the percentage lag as follows:
[52-Week High Price – (Current Price – 0.02)] ÷ 52-Week High Price
Despite calling it percentage lag, Henning uses this field in decimal form. If you were to convert this to percentage form, you would get different results for other fields that make use of the percentage lag. To avoid the problem of dividing into zero when the current price and 52-week high are the same, Henning arbitrarily subtracts $0.02 from the current share price.
Another way to calculate the percentage lag is to subtract price as a percentage of 52-week high (in decimal format) from 1.0.
Henning feels that tracking a stock price’s proximity to its 52-week high is an important analysis tool. His rationale is that it is reasonable to expect good stocks to eventually reach new highs, otherwise there is no reason to buy them. Furthermore, Henning describes an “overhang problem” for stocks that are trading too far below their 52-week high. For stocks that have fallen significantly off their high there are many people waiting for any rebound in price to sell in order to minimize their losses. Henning feels that this creates a drag on the share price, meaning that even if the share price increases, it is reasonable to expect it to do so at a slower rate than other stocks without such an overhang problem.
Ultimately, for his technical-momentum trading model, Henning prefers stocks trading within 3% of their 52-week high (percentage lag no more than 3%) to potentially avoid the so-called overhang effect. However, he points out that especially volatile stocks can be good acquisition candidates even if their price is more than 3% away from the 52-week high.
The next variable Henning uses in his analysis is one he created—investment value. It is a weighted rate-of-ascent (price increase) value (52-week multiple) divided by a weighted percentage lag value (in decimal form). Specifically, the formula is as follows:
(3 × 52-Week Price Multiple) ÷ (2 × % Lag)
This investment value calculation rewards stocks with high 52-week multiples and small percentage lags. For his own trading, Henning prefers stocks with investment values over 100 to those with values under 100. However, he does not exclude stocks with values under 100 from his watchlist.
When he was first developing his technical-momentum trading model, Henning ended his analysis here and simply selected those stocks with the highest investment values. He discovered, however, that the 52-week price multiple can be unduly influenced by large stock price increases early in the year followed by flat price movement 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 level but then go no further. Therefore, Henning realized a need to “smooth” the rate-of-ascent curve and give priority to those stocks that continue to experience strong price gains throughout the year. This led him to add two additional elements to his analysis: three-month price gain and investment value rank.
Three-Month Price Gain
The three-month gain in price is the same variable used to arrive at the initial watchlist. However, it is not a simple three-month price change. Instead, Henning looks at the percentage change from a stock’s low price over the last three months to its current price.
The calculation, thus, is as follows:
[(Current Price – Three-Month Low Price) ÷ Three-Month Low Price] × 100
Investment Value Rank
The last variable Henning uses for this technical-momentum model involves ranking the watchlist from lowest to highest based on the investment value and then comparing this adjusted rank to the three-month price gain. He does this using the rank function in the Microsoft Excel spreadsheet program. Henning describes this investment value rank calculation as his primary technical stock selection variable; all of the variables we have covered thus far are building blocks for this rank calculation.
In Excel, Henning uses the rank function to first order and smooth the investment value for each stock in the watchlist. The results of this ranking are then divided by each stock’s three-month price gain to give priority to those stocks experiencing strong upward price momentum over the last three months.
The investment ranks shown in Table 2 are based on ranking the entire watchlist of 76 stocks. If you were to rank only the 20 companies in Table 2, the calculation would be as follows:
Inv Val Rank = (RANK (G6, G$6:G$25, 0) * 2 + 100) / H6
The companies in Table 2 are the 20 companies from the 76-stock watchlist with the lowest (best) investment value ranks as of July 16, 2010.
Additional Technical Analysis
In “The Value and Momentum Trader,” Henning admits refining his technical-momentum model over time, adding two additional “screens” to the analysis process—point & figure chart analysis and trading volume acceleration.
|Table 2. Companies Passing Henning’s Qualifying Variables With Lowest Investment Value Rank|
|Company (Exchange: Ticker)||Price||
|Wainwright Bank & Trust (M: WAIN)||18.65||18.9||5.7||3.27||0.01||490.5||140.03||0.74||bank|
|CTI Industries Corp. (M: CTIB)||7.5||8||1.8||4.17||0.06||104.25||170.76||0.77||flexible film prods|
|IDT Corp. (N: IDT)||15.33||16.65||1.88||8.15||0.08||152.81||136.21||0.91||telecom servs holding co|
|Power-One, Inc. (M: PWER)||8.5||9.3||1.16||7.33||0.08||137.44||111.44||1.13||power mgmt solutions|
|Abraxis Bioscience (M: ABII)||74.37||75.34||24.52||3.03||0.01||454.5||85.79||1.24||biotechnology|
|RADCOM Ltd. (M: RDCM)||5.36||6.8||0.42||12.76||0.21||91.14||94.2||1.44||cell network solutions|
|Cirrus Logic, Inc. (M: CRUS)||17.16||18.85||4.44||3.86||0.09||64.33||105.02||1.45||integrated circuits|
|Boston Beer Co. (N: SAM)||67.51||74.52||28.79||2.34||0.09||39||125.03||1.54||craft beer brewer|
|Impax Laboratories (M: IPXL)||17.3||22.39||7.2||2.4||0.23||15.65||140.28||1.71||specialty pharmaceuticals|
|Sybase, Inc. (N: SY)||64.83||66||31.75||2.04||0.02||153||67.35||1.81||enterpr software & servs|
|iGATE Corp. (M: IGTE)||15.97||16.45||6.27||2.55||0.03||127.5||66.35||1.93||information technology|
|Virage Logic Corp. (M: VIRL)||11.88||12.01||4.42||2.69||0.01||403.5||56.11||1.96||semiconductor intel prop|
|Arbor Realty Trust (N: ABR)||5.66||6.29||1.52||3.72||0.1||55.8||81.41||1.99||real estate finance|
|Nortel Inversora S.A. (N: NTL)||20.51||22.81||7.4||2.77||0.1||41.55||86.45||2.13||phone serv in Argentina|
|Gold Resource Corp. (O: GORO)||14.1||14.2||4.13||3.41||0.01||511.5||43.88||2.32||gold & silver production|
|Trimas Corp. (M: TRS)||10.8||12.89||3.37||3.2||0.16||30||92.51||2.31||pkg; engines; indus prods|
|Odyssey HealthCare (M: ODSY)||26.56||26.91||10.15||2.62||0.01||393||47.31||2.37||hospice care|
|Mindspeed Technologies (M: MSPD)||7.16||11.13||2.06||3.48||0.35||14.91||101.12||2.41||semiconductor solutions|
|Micrus Endovascular (M: MEND)||23.2||23.35||7.8||2.97||0.01||445.5||43.48||2.48||medical devices|
|Netflix, Inc. (M: NFLX)||118.39||127.96||39.27||3.01||0.07||64.5||60.81||2.47||subscription movie serv|
Exchange Key: M = NASDAQ; N = New York Stock Exchange.
Source: AAII’s Stock Investor Pro/Thomson Reuters. Data as of 7/16/2010.
|Table 3. Technical-Momentum Scores & Recommendations|
|Company (Exchange: Ticker)||% Lag||Investment Value||3-Mo Gain||
|Power-One, Inc. (M: PWER)||0||1||1||1||1||1||5||Buy|
|Wainwright Bank & Trust (M: WAIN)||1||1||1||1||-1||1||4||Hold|
|Abraxis Bioscience (M: ABII)||1||1||0||1||1||0||4||Hold|
|iGATE Corp. (M: IGTE)||1||1||0||0||1||1||4||Hold|
|Odyssey HealthCare (M: ODSY)||1||1||0||0||1||1||4||Hold|
|Micrus Endovascular (M: MEND)||1||1||0||0||1||1||4||Hold|
|CTI Industries Corp. (M: CTIB)||0||1||1||1||1||-1||3||Sell|
|Virage Logic Corp. (M: VIRL)||1||1||0||0||1||0||3||Sell|
|Gold Resource Corp. (O: GORO)||1||1||0||0||1||0||3||Sell|
|Cirrus Logic, Inc. (M: CRUS)||0||0||1||1||0||0||2||Sell|
|RADCOM Ltd. (M: RDCM)||0||0||0||1||1||-1||1||Sell|
|Boston Beer Co. (N: SAM)||0||0||1||0||0||0||1||Sell|
|Impax Laboratories (M: IPXL)||0||0||1||0||0||-1||0||Sell|
|Sybase, Inc. (N: SY)||1||1||0||0||-1||-1||0||Sell|
|Arbor Realty Trust (N: ABR)||0||0||0||0||1||-1||0||Sell|
|Mindspeed Technologies (M: MSPD)||0||0||1||0||-1||-1||-1||Sell|
|Nortel Inversora S.A. (N: NTL)||0||0||0||0||0||-1||-1||Sell|
|Trimas Corp. (M: TRS)||0||0||0||0||0||-1||-1||Sell|
|Netflix, Inc. (M: NFLX)||0||0||0||0||0||-1||-1||Sell|
Exchange Key: M = NASDAQ; N = New York Stock Exchange.
Source: AAII’s Stock Investor Pro/Thomson Reuters. Data as of 7/16/2010.
Point & Figure Analysis
Using the StockCharts website (www.stockcharts.com), Henning generates point & figure charts for all the companies in his watchlist. Figure 1 is an example of a point & figure chart for IDT Corp. (IDT). Henning looks at the signals provided with each chart. In this case, we see a positive point & figure pattern, as indicated by the green “P&F Pattern” text at the top of the chart. On July 9, 2010, IDT’s point & figure chart had an ascending triple-top breakout. He also notes the price objective that is often provided with every signal. Here, the chart pattern indicated that the price would rise to $13.38 following the breakout. However, IDT shares closed at $15.33 on July 16, 2010.
To learn more about point & figure charting, see the Technically Speaking column in the Second Quarter 2010 issue of Computerized Investing (available at www.computerizedinvesting.com), or visit the Chart School at StockCharts.com.
Volume “Momentum” Analysis
The final screen Henning employs for his technical-momentum model considers growth in average daily trading volume. To this end, he compares the three-month average daily trading volume in a stock to its average trading volume over the last 10 days. Ideally he would like to see trading volume accelerate over the last 10 days relative to the average daily volume over the last three months. Dividing the 10-day volume by the three-month volume, Henning considers values above 1.25 to be positive indicators of volume growth.
Within AAII’s Stock Investor Pro program, there is both 10-day average daily volume and average monthly volume over the last three months. Assuming, on average, there are 21 trading days in a month, dividing the monthly average trading volume by 21 will approximate the average daily trading volume over the last three months. We can then divide the 10-day average trading volume that is in the program by the adjusted three-month average daily trading volume to arrive at an approximation of the ratio Henning uses. Henning also mentions the Yahoo! Finance website as a source for this data for individual companies. Note, however, that the data in Stock Investor is updated once a week, while sites such as Yahoo! Finance may be updating their volume statistics on a daily basis. This could lead to a significant difference in your results, depending on the data source.
Scoring the Technical-Momentum Model
Once Henning compiles all of his technical-momentum variables, he scores the stocks by assigning points based on the level of each variable for each stock. Henning assigns points to the indicators in his technical-momentum trading model in the following manner:
While Henning mentions in the book that he refined the technical-momentum model to include point & figure analysis and volume acceleration analysis, he does not provide any guidance on how to score this analysis in the context of the technical-momentum model. However, when discussing his hybrid technical-fundamental model, Henning outlines his scoring based on StockChart.com’s point & figure alerts: One point is awarded for green/bullish point & figure signals, no points are awarded when no signal is provided, and one point is deducted for red/bearish point and figure signals.
For scoring the volume acceleration analysis, Henning is even vaguer and does not outline a point structure for the analysis. Instead of ignoring volume acceleration altogether, we instead chose to apply our own scoring schema for this variable: One point is awarded for ratios of 10-day volume to three-month volume that are above 1.25; no points are awarded for ratios between 1.00 and 1.25; and one point is deducted for ratios of less than 1.00.
Table 3 shows the scoring for each of the variables in Henning’s technical-momentum model, as well as the overall tally and his “recommendation” based on each stock’s overall score. He assigns recommendations for his technical-momentum model as follows:
Strong Buy: +6
Sell: +3 or lower
Of the 20 companies listed in Table 3, only IDT Corporation (IDT) and Power-One, Inc. (PWER) received “buy” recommendations based on Henning’s scale.
Henning only considers stocks rated “buy” or higher when looking to add positions to his portfolio. He does not recommend buying stocks rated as “hold”; instead he encourages readers to wait to see if the rating moves higher. Any stocks that are currently in his portfolio and not rated as a strong buy, buy or hold would be prospective candidates for sale. However, even then, there is a level of subjectivity. Henning points out the difficulty of quickly closing out a position in a thinly traded stock, so it may take some time to unwind the entire position. Furthermore, he suggests holding on to stocks that remain close to their 52-week highs (have low percentage lags), even if they do not rate highly based on other criteria. Finally, he tempers his sell decisions based on whether there are clearly better investment candidates available.
Monitoring Technical-Momentum Model Stocks
Once an initial watchlist is created and the necessary data elements are collected, Henning encourages readers to update the data at the end of each trading day. He may even monitor stocks of special interest on an intraday basis. Since all of the variables of the technical-momentum model are price- or volume-driven, their values and the overall scoring of stocks in the watchlist can change very quickly. I hope to discuss in a future article in CI how to use Excel plug-ins to automatically update price data on a list of stocks, which would automate the entire process and speed up the time it takes to perform your daily analysis of these technical-momentum stocks.
If you are looking to build a portfolio of stocks based on Henning’s technical-momentum model, he suggests choosing at least 10 to 12 stocks in order to maintain adequate diversification. Ultimately, however, this decision is based on the amount of money you have to invest, overall market conditions, and the number of available investment candidates. Henning does warn against investing more than 10% of your capital in a single stock.
Grant Henning believes that using mathematical models to select stocks can eliminate the guesswork and emotion from the investment process. Henning’s technical-momentum approach tries to identify stocks that are experiencing strong price and volume momentum.
While, ultimately, Henning believes in following a hybrid technical-fundamental approach for his own stock trading, this model is an interesting approach for those with a shorter-term investment timeframe, especially given the rapidity with which these variables and the overall stock ranking can change.
In future articles, we will explore Henning’s fundamental-value and hybrid technical-fundamental models.