Victor Niederhoffer is the author of The Education of a Speculator, one of the oddest-and most popular-books about finance ever published. He is chairman of Niederhoffer & Niederhoffer, a top-ranked futures trading firm with $100 million under management that boasts a 32 percent compound annual return since 1982. Niederhoffer, whose clients include George Soros and others, uses a number of statistical models to discover pricing anomalies.

While explaining the principles of speculation he learned while growing up in Brighton Beach, Brooklyn, he regales his readers with discussions of sociology, music, economic theory, sex, squash, statistics and poker. The book is also full of trading advice about mechanical trading systems, self-reliance, and protecting yourself from tactical deceptions and institutional broker/dealers.

Derivatives Strategy: You have a unique vision of how the food chain works in the financial markets...

VN: Each futures and foreign exchange market is held together by a web of public, dealer, large speculators and broker players interacting with each other and their market environment. Food and energy, in the form of losses, is created by "the public" and other slow-moving participants.

DS: Who is a member of the public in this scheme?

VN: There are many publics, but most of them are shy about wearing their badges. They can be found among futures traders, stock flippers, market timers, system clubs and tape watchers. But because of the high morbidity and mortality within these groups, membership is always in flux. The best way to identify the species is by behavior-inflexibility, ignorance, arrogance, myopia, hesitancy, undercapitalization, overconfidence, spendthrift ways and hopefulness.

I'm not ashamed to admit that I am occasionally part of the public. I often trade with dealers in foreign exchange paying a 5-basis-point bid/asked spread when my exit point is a 25-basis-point profit or loss. I have also been known to accept a derivatives or options trade where, if I exited instantaneously after entering, my loss would be 50 percent of my cost.

My life expectancy in such activities would be lower than that of a mouse playing with a lion. Yet I persist. The life-enhancing factor is that I usually try to take the opposite side of public trades. The role of the public is to be eaten.

DS: And who eats the publics?

VN: The primary consumer of the "publics" are dealers, banks and brokerage houses. The dealing banks are very good at going about their trade. Year after year, publicly held banks and brokerages report annual profits in the $10 billion range from their foreign exchange and fixed-income trading.

DS: And how do you explain these kind of profits?

VN: Professor Robert Aliber of the University of Chicago has analyzed the abnormal profits of banks in the foreign exchange markets in recent years. One day, he arrived in my office late with a poignant reason: The chairman of a money center bank he had been visiting had kept him waiting while he, the chairman, negotiated the year-end bonuses of three foreign exchange traders who had made the bank $250 million over the year. Professor Aliber concluded that the foreign exchange profit of money-center banks is inordinately large relative to the capital they employ in these activities. He believes these profits compensate dealers for the persistent needs of importers to buy foreign exchange and the external desires of hedgers to reduce risk.

Paul DeRosa [a hedge fund manager] takes another track. He notes that bonuses to foreign exchange dealers are typically one-fourth the level paid to counterparts in the fixed-income field. He attributes the low ratio to the knowledge of the bank managers that it's the system that creates their profits. The two key contributing factors are the wide bid/asked spread and the use of limit and stop orders by slow moving participants.

I believe dealers play the role of the house, in imitation of the casinos in Las Vegas and Atlantic City. In 1995, cash foreign exchange trading amounted to about $1.2 trillion a day in volume-of which approximately one-third involves the public with banks and two-thirds is directly between the banks in the interbank market.

DS: How do the profits break down?

VN: In a typical foreign exchange transaction, a bank will quote a market of $/¥ 100.05/100.15 on 1 billion yen. The banks spread amounts to 0.10 percent, or 10 cents per $100 of each trade. Assuming the orders are balanced, the bank's profit would be one-half the spread on each transaction.

Multiplying the figures out, the average amount gained by the bank per year would be $50 billion per year. (1/2 x. 0.10 percent x 1/3 x $1.2 trillion x 250 trading days = $50 billion.)

The amazing thing, then, is that reporting banks make only $8 billion to $10 billion a year from their foreign exchange business. Assuming that those amounts come to 80 percent of total revenues, the profits come back to a figure close to the amount estimated from the bid/asked spreads. No wonder many holders of foreign exchange don't bother to hedge unless they expect a 3 percent move against their holdings.

DS: What advantages do they have in this game?

VN: One factor that tends to enhance banks' profits is the lack of a central clearing mechanism for foreign exchange. A customer must, for all intents and purposes, exit a position at the same house where the trade was entered. Consequently, the bank always knows which way a customer will go. Because money must be deposited in advance, most customers desire to leave the bank just enough funds to cover the size of their positions. This enables the bank to know which way the customer will go once it has used up its available margin funds.

For example, assume a customer with $10 million on deposit holds, as a 5 percent margin against the value of foreign exchange, long positions of $200 million. If the customer asks for a quote, the bank will know he has to sell dollars and will be inclined to quote a bid/asked spread that is biased to a low dollar bid. Customers placed in this position remind me of ads for the Roach Motel-easy to check in, very tough to check out.

No matter how hard I try to reduce this edge, the banks are always one step ahead of me in closing down the pass. Just before I extricate myself from a trade I like to call the banks and ask how much I can add to my position.

"Victor, why are you going to all this trouble to jerk us around? We know you never add to a winning trade," they say. Somehow they always know.

DS: Are these advantages changing?

VN: In all systems where one species preys on another, there is a constant evolution of the techniques that both groups use to maintain their margin of benefit. One technique the banks utilize with good effect requires that all customers deposit money in advance of a trade, purportedly so the banks will not have any credit risk on the transaction. The banks justify this deposit by alluding to the differences in settlement times and banking hours in various countries when the transactions are settled.

DS: The so-called Herstatt risk.

VN: Yes. That's been used ever since as an example of the great risks banks face. Payment up front results in major benefits for the bank. In addition to the Roach Motel bit, the customer's funds can be invested at a profit by the bank, to the extent that they earn more in the Fed funds market then they pay us in interest.

DS: What do traders like yourself do to balance the equation?

VN: One strategy is to unload positions on several banks simultaneously, so that the banks are forced to get out at a loss when the trades start coursing through the system. To prevent this from happening, banks frequently delay making a quote until the other counterparties the customer might be trading with have indicated a direction. "You're in line," I often hear when I have a trade of some real size to do.

On the rare occasions when I am able to get out of a trade with a bank without an immediate loss, I invariably get a call from the sales manager at the bank the next day. "Vic, my dealer complained. We thought we were seeing your whole business. But whenever you call us for a quote, we hear your activity with others over the squawk box. We're making you five-point spreads. And we're beginning to wonder why."

Well, I tell them, the reason is that you make $1 billion a year on transactions like that. You don't hear me complaining about the 90 percent of the trades I have a loss on, do you?"

DS: Are the fixed-income markets different?

VN: Dealers in fixed-income make a comparable bid/asked spread profit. Trading again is on the order of $1 trillion a day. The spread on a typical transaction would be 100 1/32 or a 2/32 spread, which is .0625 percent per $100 item. This is of the same order of magnitude as the 0.10 percent spread on the foreign exchange trade.

Unlike their foreign exchange counterparts, the government dealers do not insist on a security deposit, and customers are not locked into a particular position with a bank. Government securities can be delivered against each other at various clearing banks or brokers. Perhaps because of this lack of edge, government dealers typically refuse to quote a two-way market.

The standard patter goes like this:

"Linda, Niederhoffer here; kindly offer $10 million on the long bond."

"We're at 100 3/32."

"I buy."

"Bob [the dealer sitting next to her], you can sell $10 million of the long bond to Niederhoffer."

"That's agreed."

The broker protects himself from being exposed the wrong way on a trade by knowing the customer's direction in advance. Finally, the time between the salesperson's relay of the offer of a trade and the dealer's confirmation that a trade has been made enables the dealer to guard against a move against him during the 10 seconds or so that the trade is open.

DS: And what role do hedge funds play?

VN: They're the highest link in the web, the secondary consumers, who feed on the banks.

DS: What's the hedge fund's secret? They can't make it on the bid/asked spread the way dealers do...

VN: Well, certain naive commentators attribute the return to a constant interchange between government officials and traders, which seems to mark all successful funds. But that doesn't account for how all that information enables the predators to overcome the bid/asked spreads.

The returns have been abetted in recent years by the tendency of hedge funds to borrow short and invest long in equities and debt. During a rising market such leverage works great, but something deeper is going on.

Dealers have uncertainty as to the likely variations in the value of the inventory they hold to conduct their business. The uncertainties surround the optimum quantities to hold and the prices to pay and charge. To maintain access to credit, it is helpful for dealers to lay off risk by trading with speculators who are willing to facilitate their hedging. Like insurance companies, some large hedge funds are willing to accept these risks in exchange for a profit. Because dealers generally maintain an inventory of goods to conduct their business, on balance speculators buy goods from dealers. For this, dealers charge a price.

The ability to conduct their business while remaining insulated from fluctuations in the value of their inventories enables dealers to specialize in the relatively secure activity of capturing spreads rather than speculating on value. And banks and suppliers are willing to provide more credit to dealers who are able to insulate themselves from price fluctuations. Frequently the dealers can engage in a higher volume and a more profitable business than they could handle if they were not able to lay off their inventory risk. The dealer's total reward as well as the reward per unit of risk is enhanced by the activities of the large hedge funds.

DS: How do spreads in the derivatives markets compare with the foreign exchange and fixed-income markets we've talked about?

VN: The derivatives markets have no compunction about quoting spreads with an implicit house take of 25 percent. A quote of 0.3 to 0.4 percent for an at-the-money straddle, with three days until expiration is standard. One of the biggest dealers routinely quotes a 100 percent markup on such markets.

I try to educate the dealers with some comeback like, "Please don't put that in writing. If my customers found out I considered such a quote, they would have you and me locked up." If they don't shape up, I close the account.

A typical bid/asked spread in futures such as silver or soybeans is 1/2 cent on a $5 item, which comes to 0.1 percent. For bonds the most liquid market in the world, the bid/asked is 1/32 on a $100 item, or 0.03 percent. These small spreads, plus a comparable amount for commission, don't look like large hurdles to overcome in isolation; they come to less than 1/10 of the percentage level for stocks. What an illusion. The bid/asked spread plus commission plus bad execution quickly adds up to a staggering load.