Quarterly Reports: Systems II: the Sequel

Last Quarter, we were searching for systems--magic ways to play the market without necessarily having to think a great deal. This quarter, the search goes on.

"Don't tell anybody about this," a radio financial commentator tells me, "but I've found one of the keys to the market." This radio stuff he just does as a side line. Full time, he is a stockbroker "with the best record in the country." I ask him just what that means and whether he is very, very rich. He says he would be very, very rich if only his partners didn't keep stealing from him. But he's got a photographic memory, studies 40 pages of computer print-outs every morning, and he's better than anybody at psyching out stocks and commodities.

I'd voice my skepticism, except that he's bigger than me, we are about to go on the air (he controls the mikes) and, frankly, I am dying to know about this key to the market he's found. So I ask him what it is and he says it's simple. In the first minutes of trading, he can always tell--well, nine times out of ten--what the market will do that day. A short-term orientation, to be sure, but one that could nonetheless be leveraged to considerable advantage. How can he tell? It's a secret he wouldn't want to have get around, because why give away something so valuable for nothing, but "Just check out the opening on AMR." He doesn't know why--does it matter?--but if American Airlines (ticker symbol: AMR) opens up, the market will have an up day. If it opens down, the market will have a down day. Think you can remember that?

Naturally, since this guy has the best record of any broker in the country and would be very, very rich if only he didn't always have partners stealing from him, I got set to multiply my own considerable fortune by going long or short stock-market futures each morning as soon as I saw how AMR opened and then closing out my position with a fat little profit each afternoon.

But I figured I'd try a dry run first. And do you know what? Over 20 consecutive trading days it worked not nine times out of ten, as advertised, but only a little better than hall the time, as you'd expect. (You'd expect better than 50 percent accuracy because AMR is most likely to open up if the market as a whole opens up; and if the market as a whole opens up, it's already got a head start on finishing up.) On the first day of my test, AMR opened up an eighth and the market finished up three points. Score one for the supernatural. But the second day, AMR opened down a quarter and the market finished up 15. The day the market soared 21 points, AMR opened unchanged. On its biggest losing day, AMR opened up a quarter. Damn!

Anyhow, it's tough to find systems that work--but tempting. And, at least in hindsight, it's possible to find systems that have. Whether they'll continue to is your job to decide.

The Stop-Loss Strategy

One tenet of Wall Street, widely accepted by folks far richer than I (which has to make 'em pretty filthy goddamn rich, let me tell you), runs like this: Cut your losses and let your profits run.

So a system many investors use is to place "tight stops" on all their positions. When their brokers call to confirm that they've just bought a stock at $50.75 a share, for example, they may say, "Good. Now put in a stop-loss at 49."

A stop-loss order assures that even if you're off in Africa and not paying attention, you will be "stopped out" of your stock--your shares will be sold--if it ever touches 49. The less room you give your stock to fall, the tighter your slop. What's more, if and as your stock starts to rise, you can continually "raise your stop" right along with it, like a ratchet wrench, to lock in your gains.

I'll grant two parts wisdom in this. The first and most obvious is that if you arbitrarily set a ten-percent limit, say, on the loss you'll accept, you'll never lose more than ten percent. The second is that it's a big world out there, and you can't know all its secrets. If your stock is falling, there may be a reason. Rather (continued on page 156)Systems II(continued from page 133) than wait to find out what it is (by then, the stock could be down 40 percent), it may be better to scram.

Even so, there are some problems with this strategy.

First, it's not entirely true that by setting a ten-percent stop you'll never lose more than ten percent, because (A) you have to pay commissions, which can easily bump your ten-percent loss up to 14 percent or more; and (B) if the stock should encounter some major bad news, closing at 50-3/4 one night and opening the next day at 33-1/2, you may not have a chance to limit your loss to ten percent (which is to say, 14 percent). You will be off shooting emus, or their photographs, and your broker will be dutifully selling you out on the opening at 33-1/2. So you're not entirely protected.

Second, what often happens is that you do, indeed, get stopped out at 49, or whatever price you specified, only to see the stock bounce back to 50-1/2 and beyond. This is called getting whipsawed. The floor of the New York Stock Exchange is ankle-deep in whipsawdust.

So if you do use stop-loss orders, you might be best served using them selectively. They're fine if you're jumping on board a rising stock simply because everyone else is buying it and you hope it may have a way yet to climb (the musical-chairs theory of stock selection); it's less obviously useful if you're buying a stock you feel represents great value.

The Ex-Dividend System

This is a good one, so bear with me. Unfortunately, it's for fairly heavy hitters--the rich get richer--but if you can't afford it just yet, listen anyway. You can use it at the Labor Day picnic to impress your uncle Phil.

OK. Are you listening?

Stocks pay dividends--if they pay dividends--four times a year. The "record dates" for those dividends are known well in advance. Those are the dates used to determine, for dividend-paying purposes, who is a shareholder. If you own the stock that day, you are eligible for the dividend even if you sell before they actually get around to mailing out the checks. You'll find record dates in The Wall Street Journal.

Say General Motors is selling at 80 and, by buying on margin, you can afford to buy 1000 shares. (Dream a little.) You do that Monday, five business days before the record date. Tuesday, you sell it back at 80. (To own a stock on the record date, you must actually have purchased it five business days earlier, because it takes five days for a transaction to settle. You can sell it the very next day, because the sale, too, takes five days. It's like the world's most out-of-sync movie: The lips move five days before you hear the words.)

You paid $80,000 for the stock and got $80,000 for its sale, so that's a wash. You also paid--if you've got the right broker--a modest $250 in commissions. And you paid interest for one day on the $40,000 your broker lent you to buy on margin (half the purchase price). That's another $14 or so.

But you got the dividend, which in G.M.'s case might be $1.25 a share, or $1250.

After your expenses, you made almost $1000--in one day!--on $40,000 of your own money (plus $40,000 you borrowed), which is two and a half percent. In one day!

And tomorrow, you can take your $40,000--correction, your $41,000--and use it to buy some other stock that, like a mare in labor, is about to bear a dividend.

Do that every day and, compounded, you'll soon own the world. The annualized rate of return comes to around 14,000 percent (before taxes).

Here are the problems.

The first is that it's even more crucial than usual, in a scheme like this, to keep your commissions low. To do that, you've got to be playing with grown-up money. If you tried to do this with just 50 shares of G.M., say--which is still $4000--the dividend you'd get, $62.50, would be less than the commissions you'd pay to do the trades.

But it's little or no more difficult for a broker to trade 1000 shares of G.M. than to trade 50, and with this system you'd generate monumental commissions, so it wouldn't be hard to find one who agreed to take, say, a straight eighth of a point per share. That's 12.5 cents for each share you buy, 12.5 cents for each share you sell, $250 in all for buying and selling 1000 shares of stock. If you really got going, you might be able to do even a little better.

The second problem is a lot stickier. Who says, having bought G.M. for 80, you'll be able to sell it back at that price the next day? What if you can get only 77?

And, in fact, four days before a stock's record date, it "goes ex-dividend." That means three things. First, somebody makes sure that a little x appears next to it in the stock pages of your newspaper. Second, anyone ignoring that x and buying the shares anyway is buying them ex--without--the dividend, because when the trade finally settles, five days later, the record date will have passed. Third, the value of the dividend will automatically be subtracted from the quoted price of the stock the next morning--at least until it opens for trading.

If you're lucky, investors will be little troubled by the lack of this dividend and will immediately bid the stock back up to the $80 you paid for it. (In which case the newspaper will report it as having gone up one and a quarter that day.)

But if you're not lucky, it will not only be quoted at 78-3/4 (80, less the dividend), it will minutes later crash to 63 on news that Subaru has invented a V8 engine that gets 140 miles to the gallon on a mixture of alcohol and V-8. In Detroit, they'll be screaming bloody Mary, and in the meantime, in your quest for a lousy $1250 dividend, you'll have lost $17,000.

Enter a Jupiter, Florida, firm called Dividend Management, Inc., in the person of attorney and accountant Floyd J. Marenda. Marenda, 54, has done a lot of looking into these things and has found that "certain stocks have established patterns or trends in the periods immediately prior and subsequent to ex-dividend day." With American Cyanamid, for example, "we can logically conclude that one should buy it on the fifth day preceding ex-dividend day" and drop it like herpes the moment it opens ex-dividend. With IBM, ex-dividend day is usually so bad that the way to play it is to eschew the dividend altogether: Buy the stock a week before the dividend, but unload it on the day prior to ex-dividend day--letting some other sucker get the dividend (and a sharp drop in price). With International Paper, you'd also forget about the dividend--buy the stock the day after it goes ex-dividend for a quick two bucks a share in the week that follows ... all this assuming, of course, that past trends hold.

For a minimum of $2000 a year, which won't be much if he can do for your $100,000 what he thinks he can do (but you never know), Marenda will drive your broker crazy with instructions to buy and sell stocks and collect dividends. Currently, he does this for about 50 well-heeled investors and a collective $2,000,000--never touching their money himself, just directing the activity--and the results have been good. In 1984, he says, a typical $100,000 account involved about 100 separate plays, $28,000 in commissions, $3900 in interest charges, $34,000 in dividends and $26,600 in capital gains. Net return: 29 percent.

Conceptually, what's appealing about this system--apart from your never having to do any work and Mr. Marenda's never actually touching any of your money--is that the dividends can, indeed, handily exceed the commissions on any given trade, so that with any decent luck at all, you should at least not lose much with this system. Which, believe me, is a lot to say for any system.

The biggest drawback, apart from having to watch your broker rise month by month to a higher standard of living than he or she could ever possibly deserve, may be that any income you do earn from the scheme will be fully taxed. (Doctors, lawyers, electricians, take note: This may be something to consider for a portion of that $1,500,000 tax-sheltered retirement fund of yours. And Marenda does run a sort of mini mutual fund that allows participations as small as $25,000.)

Relative Strength

Have you ever noticed how, when the market is going down, some issues seem to avoid most of the loss? They're stronger than their brothers and sisters, and many believe that this relative strength will translate, when the market starts back up, into superior gains.

Dan Sullivan, who has more than quadrupled his own portfolio since 1969 (though, amazingly, to do so in 16 years is merely to compound one's money at nine percent), has made a science of calculating stocks' relative strength and a business out of selling his advice to others (The Chartist, Box 3160, Long Beach, California 90803). He says he really doesn't care what the companies he invests in do; he cares only how their stocks move relative to others and which way the market is headed. Into his computer go stock data and out come relative-strength ratings based on a formula he has developed. Lately, his performance has been near the head of the class, but (we all have our good years) has he really found a way to beat the market? Time will tell. But, as usual, not until it's too late to do us much good.

Making a Fortune in Teensy-Weensy Increments

In one sense, there's no question that systems work. The ones I have in mind are highly quantitative systems that allow pros with big computers and a lot of time and capital--and minimal transaction costs--to exploit the little gaps of good sense in the market. If Treasury bonds are selling even a few pennies per $1000 cheaper than they should relative to T-bill futures, why, there's a golden opportunity!

There's a whole new breed in the trading rooms of Wall Street brokerage houses. They're called quants--quantitative types, good with numbers--and they sit in front of computer screens all day, looking for those little gaps and wondering whether radiation from video-display terminals will really fry their genes.

One Wall Street quant, according to Institutional Investor, recently brushed aside a potential $850,000-a-year job offer because she was happy where she was, at perhaps half the pay, supervising 35 Ph.D. mathematicians (and a world crossword-puzzle champion) in applying higher mathematics to the bond-trading strategies at Merrill Lynch.

By and large, these 35 geniuses are not figuring out ways to make clients of Merrill Lynch richer--at least not clients like you or me--they're employed to make Merrill Lynch richer. And you've got to believe that these academic wonders and their counterparts around town, earning mere pittances at graduate schools just a few years earlier, wouldn't be getting paid so much today if they didn't know what they were doing.

It's easy to earn big money on Wall Street without knowing much if you're in sales--you make money from your clients. But these folks have shown their firms how to make money from the market.

There are many ways they do this, some extraordinarily complex and having to do, say, with buying and selling tens or hundreds of millions of dollars of Treasury-bill futures from different months. Don't even try to understand this--if I did, I'd be putting my own hundreds of millions into the deal. I just want to suggest how it smells.

Market Timing

With real estate, everyone knows it's location, location, location. With investing, writes author/investor Dick A. Stoken, the three crucial elements are timing, timing and timing. It is vastly easier to make money in stocks or precious metals or real estate when the general trend of stocks or metals or real estate is up. Just which stocks you choose, or which metal, or which rental property, is secondary.

That much is sure. Whether, in addition, you should buy stocks "when either short- or long-term interest rates have fallen to a 15-month low" and hold them until both long and short rates have risen to seven-year highs, I cannot say. But if you like this kind of thing, Stoken's book, Strategic Investment Timing, is excellent. Certainly, with the perspective it provides on political and economic cycles and its rules for interpreting four commonly available indicators (interest rates, the political cycle, the Producers' Price Index and the Dow), when you lose money in the market, your losses will be based on much more sophisticated misjudgments.

Had you followed Stoken's rules from 1921 through 1983, you could have grown $1000 not into the mere $19,000 you'd have had standing pat with the Dow Jones industrials but--by side-stepping declines in the market and then coming back in at the bottom--$2,714,466. Tailoring your investments not just to the Dow but to rotating groups of stocks that perform best in various phases of an up market, you'd have done yet another four times as well. None of this takes into consideration dividends or taxes, but what it really doesn't take into consideration is that, sadly, Stoken's book appeared in 1984, not 1921. As he would surely acknowledge, it's far easier to formulate rules that fit the past than the future. That said, you're a lot less likely to buy at the top or sell at the bottom after reading this book.

Following the Insiders

A lot of study backs up the commonsense notion that, over time, insiders are likely to do better than you or I buying and selling their own stocks. Not that they don't often get carried away with their own bullish forecasts and buy when they should sell or mistake the power of the market and sell too soon--they do.

Several newsletters will help you follow and assess insider moves, which must promptly be reported to the Securities and Exchange Commission. One is The Insiders (3471 North Federal Highway, Fort Lauderdale, Florida 33306), from the same people who publish a bunch of other good newsletters. But how much can one afford to spend on newsletters? And what if you're only out to buy six or a dozen good stocks and hold them for the long term? What do you do with five or six newsletters arriving each week offering hundreds of recommendations a year? Tending your investments could quickly become a full-time occupation, with no guarantee you'll do any better than, or even as well as, you'd do in a no-load mutual fund or a bank.

The Generic-Stocks Strategy

A new book by Avner Arbel, called How to Beat the Market with High-Performance Generic Stocks, makes a strong statistical case for buying neglected stocks. Arbel's notions overlap the good cases others make for buying "doghouse" stocks (the contrarian system of investing) and for buying "low p/e" stocks (stocks with low prices relative to earnings) and for buying "small cap" stocks (stocks with small capitalizations that the big financial institutions can't be bothered with). But Arbel says a company doesn't have to be small to be neglected--almost a third of the big blue chips in the Standard & Poor's 500 are not regularly covered by Wall Street analysts--and that, in fact, some small companies are not neglected and tend to be poorer buys than those that are.

Arbel advises that we look up a potential purchase in the Standard & Poor's Slock Guide (any broker has this on his desk, if you don't) and check out the column that tells how many different financial institutions own it. As a rule, he says, a stock held by fewer than ten institutions can be considered neglected. If you further screen your choices to weed out the financially infirm, you're likely to be left with a portfolio of what Arbel calls "generic stocks"--stocks whose prices are not inflated by a brand-name premium.

"Amazingly," he writes of the results of one study, "the return of the most neglected stocks was more than 60 percent higher than for the most followed stocks even after adjustment for total risk!"

One clear advantage to this system: A stock owned by a lot of institutions can one day be sold by a lot of institutions, crushing it, while a neglected stock may some day catch their eye, causing it to soar.

The Low P/E System

Technology may be the future--how could it not be?--but it's possible that you'd do a heck of a lot better buying old shoe-machinery companies at seven times earnings than Digitalis at 50 times. For a 300-odd-page elaboration of this notion, see David Dreman's much-praised The New Contrarian Investment Strategy.

The no P/E System

Neglected is one thing, reviled another. The case for buying real losers (many of whose p/e cannot be calculated, because they have no e) is made in William J. Grace, Jr.'s, The Phoenix Approach, subtitled "The Contrarian Investor's Guide to Profiting from Out-of-Favor, Distressed and Bankrupt Companies."

The Interplanetary System

Former Merrill Lyncher and ex-encyclopedia salesman Arch Crawford counsels stock-market and commodities traders on the basis of his readings of the planets. ("The 20th and 21st of May [1985] have a complex formation involving Jupiter, Uranus, Venus and Mars, which could provide the energy for an explosive blow-off. Mars will oppose the electrical planet Uranus on the 20th.") He's made some good calls in the past, and if you believe these had anything to do with his ability to interpret the heavens, you will surely want to send him your money (Crawford Perspectives, 250 East 77th Street, New York, New York 10021). If you have any money left, that is, after the 1929-style crash he's predicted should occur by the end of this year.

Cycle Jockeys One Lap Later

We closed last quarter, particularly faithful readers of this column will recall, with Speed Sexton and Mike Jenkins, whose Harmonic Research (Rooney, Pace, Inc., New York, New York 10004) uses the cyclical patterns of the moon and the market, the Elliot and Kondratieff waves, and just about anything else that ripples, to forecast the future.

When all the various waves that have been fed into their computer are more or less canceling one another out, not much happens. But when almost everything is aligned in harmony, as sometimes happens, the computer begins to vibrate wildly and bounce around the office--if not literally, then through the enthusiasm its output inspires in Speed and Mike.

Speed is a Harvard M.B.A., Mike a veteran mutual-fund manager, and their goal is to use the work they've done understanding cycles to grow their money under management at ten percent a month. I can't say for sure how well they're doing (though to grow $1,000,000 at ten percent a month is to grow it, in ten years, to 93 billion dollars, so they've got their work cut out), but, as promised, I have faithfully reviewed their fortnightly analysis.

Every two weeks it comes with a capsule of what to expect of each of the next ten trading days, like a two-week weather forecast. For each day, an arrow points either up or down, indicating the direction of the Dow, with occasional comments such as "Strong opening, weak close" or "Big decline possible!"

Since they have had the chutzpah to forecast the financial weather, I had the gall to check out their record for 12 weeks. I gave them two points for each day they were right, one point if they were sort of right, subtracted one if they were sort of wrong and subtracted two if they were ass backward. Days they starred as being particularly noteworthy I tripled in importance. For example, for Wednesday, May first, when they predicted a strong up day but the Dow was down 16 points, I subtracted two. But for the previous Thursday, when they had predicted an up day and the Dow went up, they were credited two.

Right all 60 days, they'd score 120 or so; wrong all 60, minus 120; no more right than wrong--which is exactly what you'd expect--they'd score zero.

Admittedly a crude gauge, but not without foundation. Final score: minus three.

Lest my rankings be too subjective, I also tried a more mechanical approach. On days they said the market would go up, I gave them as many points as it did go up--or subtracted the number it went down. And vice versa. On days they said it would be flat, I credited them for each point fewer than five the Dow moved and subtracted for each point more than five up or down.

Here they had a maximum possible score--if the market always went in the direction they had predicted--of around 330 points (or minus 330 if it had always foiled their predictions). Again, you'd expect a score around zero, the good calls canceling out the bad. But you'd hope these guys could do better, because they're nice guys, smart guys, hard-working guys, and it would be great to think they'd found a key to the market. They scored minus 55.

Given their remarkable level of effort, sophistication and computerization, it does make one wonder whether even the above-average Joe can beat the market with cycle analysis.

And you should know this before you send your millions to Speed and Mike: You'll be sending them just to Speed. There's been a bit of disharmony over at Harmonic Research since last I reported, and Mike has struck off on his own.

The Best System

For almost everybody, the best system for the stock market is not to try to play it yourself at all but, with the portion of your assets you can afford to risk in the market (if there is any such portion), to invest through one or more carefully selected no-load mutual funds.

Speaking of which, there's a system. Two, actually--mine and a more sophisticated one. Mine is simply to recommend a program of steady investments--$100 a month, $500 a month, whatever. The only twist I offer is that in times when the world seems sure to collapse, you should double up on your investments (if you have any money), and at times when the market is marching steadily to new highs, you should take a little vacation from the market and stash that $500 a month in a savings account, instead.

For details, I always recommend The Handbook for No-Load Fund Investors ($32, Box 283, Hastings-on-Hudson, New York 10706).

Another book of interest is Market Timing with No-Load Mutual Funds ($12.95, Backwater Books, 7438 SE 40th Street, Mercer Island, Washington 98040). By being in your funds only half the time, this book suggests, you can cut your risk in half and--if you choose the right half of the time to be in--greatly enhance your returns. Easier said than done, to be sure, but worth a look. The authors note that serious no-load fund investors can now buy their funds on margin, through Charles Schwab & Co., and, at relatively modest cost, conveniently switch around among hundreds of otherwise unrelated funds. The one caveat with this book is that it serves also as a sales brochure for the authors' money-management services--at a hefty two percent of your assets per year--so bear in mind that it is written with a point of view.

A similar system comes from the respected if not infallible Institute for Econometric Research, which has begun offering a newsletter, Mutual Fund Forecaster ($49 a year to charter subscribers; 3471 North Federal Highway, Fort Lauderdale, Florida 33306), that purports to "use sophisticated computer technology to make sure you own the right fund at the right time." When you consider that the sales commission on a $10,000 investment in a traditional-load mutual fund would be a non-tax-deductible $850--with the salesperson advising you on which fund to buy naturally advising his own--this tax-deductible $49 price tag could be worse.

Or you could save even that $49 and choose the no-load fund I've been recommending for eight years now, Mutual Shares (26 Broadway, New York, New York 10004), which, had you been foolish enough to invest your life savings in it simply on the recommendation of some clown like me, would, as it happens, have served you very well. But with Mutual Shares, as with any fund, there is no assurance that such success will continue.

a timely accounting of timeless principles of personal finance