Gerald Loeb, the Wall Street financier, may have never bet on a horse, but his general strategy of money management is one that every horse player might consider.

In "The Battle for Investment Survival," Loeb wrote, "I am personally convinced of the inevitability of loss when attempting to secure a safe income of small return. I constantly suggest speculation rather than investment as the policy less apt to show a loss and more apt to show a profit."

The experience of most race trackers confirms Loeb's observation. The bettors who are surest to lose in the long run are the ones who regularly go to the track with modest aims: "I'll be happy to win 20 bucks today." "I'll just find a couple of solid favorites and bet them to place."

The rare people who beat the races do it not by grinding out a slow, steady daily profit, but by making big scores that compensate for their losses and which provides -- over a year's time -- a reasonable net return on their capital.

I can offer firsthand testimony on the virtures of aggressive betting. Over the past few years my profits at the track have increased sharply, but during this time my handicapping methods have changed little.

What has changed is my betting strategy. Instead of trying to hit singles, I am swinging for the fences.

I rarely bet horses to win any more. When I find a horse I especially like, I analyze the race anew and try to find one or two or three horses I can play in the exacta or triple underneath my top choice. Instead of settling for a solid 8-5, I am looking for a relatively large return for a relatively small investment -- the objective in any type of money management.

Whenever I venture into exactas or triples, I try to observe one maxim: The virtue of exotic wagers is that they offer great rewards for small risks, and they should be used for that purpose.

I know many horse players who abuse this principle. They beat exotics in shotgun fashion, scattering $1,000 on a number of combinations in the triple and congratulate themselves when they hit a $2,000 payoff. This has to be a losing strategy in the long term; certainly there are sounder ways of getting even money or 2 to 1 on one's money. If I am not shooting for a return of at least 5 to 1 on my total investment in an exotic wager, I shouldn't be making the bet.

For instance: I like a solid favorite and see three horses in the field who plausibly could finish second.The exacta coupling the favorite with Horse A is returning $11; with B, $14; with C, $20. If I bet all three in equal strength, I will be getting only a 2-to-1 return on my money, even if the longest-priced combination hits. That is not an adequate return.

If I could not resist playing the race, and wanted to invest a total of $100, I would prefer to bet $60 on the exacta with C and $20 each on A and B as insurance. At least that way I would be taking a chance for a healthy score.

In all these cases, when I put my money in exotics rather than the win pool, I am willing to run the risk that my favorite horse will win and my exacta bets will lose anyway. Only when I cannot face that possibility -- either because I love a horse so much or because his odds are so enormous -- will I bet him only to win.

But even when a horse player is making a conventional win bet, he should remember that exactas can be useful as a hedging device. Many horse players protect themselves on a big wager by betting some of their money to place.

I consider this a bad policy. If I am so unsure of my selection that I fear any of the horse's rivals could wake up and beat him, I probably should not be betting much in the first place. Usually, however, I will be most afraid of a specific horse. By betting that horse on top of my selection in an exacta, I can hedge my bet cheaply.

Exotic bets are most useful not when they are hedging devices, not when they are substitutes for a win bet, but when they give a horse player the chance to do something that never used to be possible: turn a profit from a negative opinion about a horse.

There are many times when a handicapper will declare, "I don't know who's going to win this race, but I hate so-and-so." I felt that way about this year's Preakness. I wasn't sure who was going to win it (and ultimately guessed wrong) but I was totally convinced that cofavorite Colonel Moran didn't have a prayer. This kind of opinion is a perfectly legitimate basis for a big bet.

If I cannot isolate a likely winner, I will box as many as three horses in an exacta or four horses in a triple while throwing out a short-priced horse whom I don't like. This is the one intelligent use for boxing: playing all combinations of a number of horses in an exotic wager.

Many gamblers box three or four horses simply because they have no real opinion on the race. This is a mistake, as is any bet not firmly rooted in a sound handicapping judgment.

Handicapping is, and will always be, the foundation for success at the track. But many horse players who think they lose because of their handicapping deficiencies actually are being hurt more by their bad betting habits. This game, after all, does not directly reward handicapping skills; there are no special bonuses for picking a lot of winners. What counts for a bettor is maximizing the profits on the winners he does pick.