Online Investing, Second Edition

Author Jon D. Markman Pages 384 Disk N/A Level Beg/Int Published 01/17/2001 ISBN 9780735611238 ISBN-10 0-7356-1123-8 Price(USD) $24.99 To see this book's discounted price, select a reseller below.	More Information About the Book Table of Contents Sample Chapter Index Related Series Related Books About the Author Support: Book & CD Rate this book

Chapter 5: Counting

• Specialized Stats
• HiMARQ Analysis
• Identifying Patterns
• Sector HiMARQs
• Slugging Averages
• Buddy System
• The Nicoski Momentum Oscillator

Chapter 5 Counting

When she was little, the woman said, no one kept track of the date on which children were born. To determine whether a boy or a girl was old enough for school, the village schoolteacher would ask them to stretch and bend their right arm directly over their head and try to touch their left ear. If a child’s arm was long enough, she said, they were ready to start.

My daughter, Janie, immediately attempted the task and was disappointed to learn she didn’t measure up. She happily went on to kindergarten anyway, but the discussion got me thinking. The Togo technique for gauging the passage of years with the natural ratio of arm length to head size, not a calendar, luminously reflected my quest to discover alternative ways to measure stocks against the passage of days, months, and years.

Over the 18 months since the first edition of this book, I interviewed dozens of Wall Street strategists to gather fresh insights into hidden relationships between time and money, and I invented a few of my own techniques as well. In every case, the motivation was the same: to determine which quantitative factors characterize the best stocks to own at any given moment from the perspective of earnings growth, fundamental value, calendar or fiscal seasonality, group psychology, price momentum, and volatility. It’s a quest that quantitative analysts blandly call "the search for alpha"—that is, the search for price gain in excess of a benchmark, such as the Standard & Poor’s (S&P) 500 Index.

I shared most of the ideas as they emerged with readers of my weekly column at MSN MoneyCentral and will use this chapter to highlight and synthesize some of the better ones. They all require some proficiency in the use of spreadsheet software such as Microsoft Excel. I won’t explain how to exploit this software in detail; the point of this chapter is to encourage you to consider learning to combine the power of the Web and spreadsheets in ways that were not previously possible. In Appendix C of this book, found online at /mspress/products/4651/appendixC, I have provided homemade spreadsheets with macros that will enable anyone with modestly advanced Excel skills to begin exploring these ideas.

Specialized Stats

Visit any major financial news Web site, in contrast, and you’ll have to hunt for a list of even the current day’s top 10 performers, much less a quickly sorted list of the top 100 or 500 most heavily traded stocks. Moreover, the metrics are largely limited to the current day’s percentage gain or loss, plus the raw number of shares traded that day. Click a little deeper and you might find one-year, six-month, and three-month percentage returns for individual stocks. But it’s pretty hard to find and sort through year-to-date return, volatility measures, alpha, or any more-sophisticated measurements of performance that might help you get beneath the yadda-yadda-yadda of the market players and learn about the tectonic changes that are truly going on beneath the surface. Try to do the same in a mainstream metropolitan newspaper, and you’ll find a much shorter list of actively traded stocks, and nothing more exotic than a New Highs / New Lows list.

What accounts for this disparity between the quality of baseball stats and financial stats? Despite the fact that stocks appear on the surface to be all about numbers—growth rates, price/earnings ratios, returns on equity, and so forth—the high priests at brokerages and in the press typically profess wariness about statistics, preferring to talk mainly about ethereal factors like management, new products, and global economic conditions. Very few technical analysts or statisticians have managed to rise above the investment media’s aversion to math to offer just plain facts about the condition and direction of the only thing that ultimately matters about a stock—its price. The Web sites Investors.com (http://www.investors.com—published by the newspaper Investor’s Business Daily) and Barchart.com (http://www.barchart.com) are a couple of exceptions that prove the rule.

The fact that baseball has exalted stats to higher and higher levels is thanks in significant part to sportswriter Bill James, and his philosophy could also serve as a standard for statistics-loving investors. From 1977 to 1988, James published a series of books called "Baseball Abstracts" that extolled the advantages of understanding baseball by analyzing an aggregation of data about players and teams rather than just observing them on a daily basis and listening to their timeworn refrains. "Clichés are the soldiers of ignorance," he wrote, while "statistics look at games by the hundreds….Without them, it is impossible to have any concept of the game, save for meaningless details floating in space."

My favorite Jamesian conceit is the invention of new statistics from time to time to uncover trends that aren’t well explained by conventional means. So in that vein, in this chapter I will present HiMARQ analysis, which reveals stocks’ historical monthly and quarterly performance; slugging percentages, which reveal stocks’ daily and weekly volatility; the buddy system, which attempts to combine stocks with negatively correlated performance and volatility; and the Nicoski M26/M52 oscillators, which reveal whether stocks are overbought or oversold compared with the average historic ratio of their current price to their price 26 to 52 weeks ago. An important caveat before starting: unlike the screens presented in the previous three chapters, most of the ideas presented here cannot easily be historically tested to determine their usefulness as predictors of future stock prices. Yet that should not prevent you from examining each and thinking about how they might be installed in your arsenal of investment weapons. (No matter whether they all work all the time—and I guarantee that they won’t—you will still be allowed to progress to kindergarten.)

HiMARQ Analysis

That means 3Com scored an average annual gain of 93% for half a decade, thanks to enthusiasm for its network-interface card, a high-margin industry standard that was practically an emblem for the rapid, worldwide growth of business productivity in the mid-1990s. Yet as Christmas lights were coming down that month, 3Com’s stunning run was about to end. An earnings shortfall in February 1997, precipitated by Intel Corporation’s price-slashing entry into its market, led to one of the most prolonged slides in high-tech stock history up to that time. Over the next three years, 3Com’s shares fell 65%. It didn’t happen all at once. After an initial sharp drop, the rest came in heartbreaking dribs and drabs. Many longtime shareholders cursed a troublesome and ill-timed merger with modem maker U.S. Robotics and waited patiently for the turnaround.

But by August 2000 they were still waiting—despite the firm’s successful spin-off of a division that designs and manufactures the wildly popular Palm family of personal digital assistant devices (acquired, incidentally, in the U.S. Robotics deal). So if they were smart they were wondering two things: would there ever be a decent bounce again that would let them get out whole, and what they should do the next time that "buy and hold" turned into "buy and die"?

One way to solve this question is to determine whether—all fundamental factors aside—there is one time of year that tends to be more favorable to an individual stock than other times. And second, is that time of year so favorable that it’s worth waiting for?

The answer is hard to come by with pen and paper, but it turns out to be fairly simple if you’re handy at downloading data from the Web and manipulating it in a spreadsheet program like Microsoft Excel. I’ve provided detailed instructions in Appendix C (found at /mspress/products/4651/appendixc), but essentially you just need to use MSN MoneyCentral or Yahoo! Finance features that permit the export of historical monthly stock price data from charts, then create a column for monthly percentage change, and finally use Excel’s PivotTable and Group functions to combine and group that data into months and quarters. The result is something I call "Historical Monthly Average Return Quotients," or HiMARQ analysis for short.

HiMARQ analysis helps you figure out the best month to sell or buy both individual stocks like 3Com and whole industrial sectors, such as networking equipment makers, oil drillers, or airlines. It works best for growth companies as well as for industries characterized by seasonal sales patterns. First I’ll offer a broad overview of the concept, and then I’ll suggest a theory as to why it works—and along the way I’ll propose ways to take advantage. If it sounds like too much work to do yourself, in Appendix C you will find the HiMARQ stats for the 100 largest companies that have traded at least five years through November 2000. Also in Appendix C, you’ll find the 25 large-cap stocks with the best HiMARQ stats for each month.

Identifying Patterns

Table 5-1. 3Com HiMARQ Analysis, 1989–2000

Throughout its 12-year history, 3Com stock has consistently rallied sharply from September to December and declined from January through April.

1st Quarter
January	-0.8%
February	7.2%
March	-5.4%
2nd Quarter
April	-4.3%
May	5.0%
June	1.5%
3rd Quarter
July	0.9%
August	1.4%
September	12.8%
4th Quarter
October	5.3%
November	8.6%
December	12.0%
Average	3.0%

Skeptics of this sort of analysis might immediately imagine that a few stray months’ returns are skewing the data. But a closer look at the returns of individual months suggests that’s not the case. 3Com has experienced only two negative Septembers in its history, as shown in Table 5-2, and they were way back in 1989 and 1990. Even during the recent downtrend, the stock has managed nice advances in the month. In 1999, for example, the stock was up 16.2% in September even as the Nasdaq Composite Index and Dow Jones Industrial Average were down 0.7% and 6.6%, respectively, and industry stalwarts Cisco Systems and Intel were down 3.1% and 9.9%, respectively. In 2000, the stock was up 13.3% while the Nasdaq Composite Index plunged 13.3%.

Table 5-2. 3Com Septembers, 1989–2000

Since 1991, 3Com has never experienced a negative September.

Septembers	Price Chg.
2000	13.3%
1999	16.2%
1998	17.9%
1997	2.0%
1996	32.5%
1995	18.6%
1994	15.4%
1993	16.3%
1992	45.2%
1991	4.5%
1990	-25.7%
1989	-10.2%

3Com likewise has experienced only two negative Decembers in its history (-6.5% in 1997 and -4.0% in 1996) and three negative Novembers (1997, 1995, and 1991). But its Januaries are a different matter: 8 out of 11 have been negative. And it has suffered two of its greatest one-month declines in Februaries: -46.6% in 1997 and -28.8% in 1999. You also seldom want to own this stock in April, it appears: 8 out of 11 have been negative. So the message seems clear: buy 3Com just before Labor Day, sell it after Christmas, and avoid it the rest of the year.

These seasonal price-change patterns play out for most growth and cyclical companies—particularly ones in industries, like technology and retail, that are characterized by strong seasonal buying patterns for their products. Semiconductor and semiconductor-equipment companies such as Intel, Micron Technology, and Applied Materials, for instance, regularly experience fabulous rallies in January that have little to do with the so-called "January effect" (which is a market theory that suggests small-cap stocks tend to outperform the market in January). And consistently the best months for America Online on average since 1992 are September (+11.5%), November (+15.6%), December (+19.9%), February (+11.3%), and March (+15.1%). The April–August period is regularly just barely better than neutral for the stock.

Indeed, for some stocks there have been can’t-miss months in the past—months in which they have recorded a gain every year that they’ve been public. In Table 5-3, I’ve listed 50 stocks with market caps greater than $1 billion that have perfect records of 6-0 or greater and a 6% return in a single month.

Table 5-3. Perfect Records

Some stocks have never recorded a negative result in certain months. These 50 have market caps greater than $1 billion, and they have perfect records of at least 6-0 and a minimum average gain of 6% in at least one month through October 2000.

	Month	Average Gain	Positive	Negative
Dana Corp.	Dec	6.1%	13	0
Harley-Davidson	Dec	6.6%	12	0
Kinder Morgan	Sep	6.4%	12	0
Dollar General	Mar	15.2%	11	0
Paychex	Sep	14.1%	11	0
St. Jude Medical	May	8.7%	11	0
Cisco Systems	Nov	14.0%	10	0
Qualcomm	Mar	19.5%	9	0
SunGard Data Systems	Feb	10.9%	9	0
Foundation Health Systems	May	10.6%	9	0
C&D Technologies	May	10.3%	9	0
Kimco Realty	Dec	7.5%	9	0
ImClone Systems	Nov	32.1%	8	0
Alkermes	Jan	20.9%	8	0
Valassis Communications	Dec	16.0%	8	0
Express Scripts	Dec	15.1%	8	0
Boston Scientific	Dec	7.3%	8	0
Marine Drilling	Mar	19.7%	7	0
Coflexip	Mar	16.6%	7	0
Dura Pharmaceuticals	Dec	13.6%	7	0
Barnes & Noble	Mar	12.8%	7	0
Plantronics	Jul	12.3%	7	0
Panamerica Beverages	Nov	11.7%	7	0
Wind River Systems	Aug	10.7%	7	0
Dynegy	Apr	9.3%	7	0
North Fork Bancorp	Sep	9.0%	7	0
Martin Marietta Materials	Apr	6.9%	7	0
PartnerRe	Dec	6.2%	7	0
CMGI	Dec	38.8%	6	0
Incyte Genomics	Dec	34.9%	6	0
American Eagle Outfitters	Mar	23.4%	6	0
Flextronics International	Nov	20.8%	6	0
Inhale Therapeutic Systems	Jan	20.6%	6	0
JDS Uniphase	Nov	19.8%	6	0
Legato Systems	Sep	19.3%	6	0
VERITAS Software	Dec	15.7%	6	0
Affiliated Computer Services	Dec	13.7%	6	0
Darden Restaurants	Sep	13.5%	6	0
Northwest Airlines	Mar	12.6%	6	0
Inhale Therapeutic Systems	Nov	11.5%	6	0
Plantronics	Dec	10.9%	6	0
Orthodontic Centers of America	Sep	10.9%	6	0
Adecco SA	Feb	10.4%	6	0
QLogic	Nov	10.0%	6	0
PMI Group	Oct	9.5%	6	0
Nippon Telegraph & Telephone	Jun	9.2%	6	0
Catalina Marketing	Dec	9.1%	6	0
Guidant	Jul	8.5%	6	0
First Industrial Realty Trust	Dec	7.3%	6	0
Bank of Montreal	Oct	6.4%	6	0

Why do these patterns occur so consistently? I think at least five factors are at work, in no particular order:

• Industrial and retail product-buying cycles. Smart institutional investors are always looking ahead at sales and earnings growth for stocks over the next three to six months. HiMARQ analysis lets you pick up on their investment pattern without knowing exactly why it’s happening.
• Product announcement cycles. Companies get into the habit of making their major product announcements at the same time every year. These announcements often energize investors. HiMARQ analysis picks up on that pattern as well.
• Annual shareholder and analyst meetings and quarterly earnings reports. Companies hold their stock-moving shareholder or analyst meetings at the same time every year. And, of course, they announce earnings reports at the same time each year. HiMARQ analysis allows you to pick up on that pattern without knowing the details.
• Insider selling habits. The fate of many great growth stocks, like Microsoft and data-storage products maker EMC, is still largely controlled by their companies’ founders. They appear to sell their very large blocks of stock regularly at certain times of year that coincide with federal securities rules and their own habits.
• Analyst upgrade cycles. Veteran market-moving brokerage analysts who have researched and commented on individual stocks over a long period probably pick up on the four cycles above and regularly get in the habit of upgrading or downgrading stock at the same times each year.

HiMARQ analysis should be used only on stocks with at least four years’ price history, a ratio of positive to negative results in the target month of at least 2:1, and a standard deviation for the month’s returns that is lower than the average return. Ken Moss, chief software architect at MSN MoneyCentral, devised a formula for the best possible combination of HiMARQ factors: (Average monthly return times number of positive months) minus (standard deviation for the month’s returns times number of negative months), the result of which is multiplied by 100. In his honor, I call this figure the KenMARQ Index—and the higher the figure, the better.

For example, payroll-processing powerhouse Paychex has scored a 14.1% gain in Septembers over the past 11 years, its standard deviation for those returns is 8.6%, and it has racked up 11 positive Septembers vs. no negative Septembers. Thus, ((0.141 * 11) - (0.86*0)) * 100 equals a KenMARQ Index score of 154. In general, a KenMARQ score of 25 or better is preferred. Scores greater than 75 are terrific; stocks with negative scores should be avoided. As you'll see in Appendix C, the scores vary widely by month: September has two dozen $1-billion market cap stocks with KenMARQ scores greater than 75, while August has fewer than 10. In September 2000, while the S&P 500 Index sank 5.5%, Paychex was again up 20.0% in the month.