To many observers, the bond market operates much like a souk. There is no formal exchange, no central posting of bids and offers, no registry of trades once they occur. Hundreds of bond dealers trade more than $400 billion worth of securities a day over the telephone and by fax machine, and they thrive on the lack of public information. Investors, who have no way of knowing if dealers are quoting prices that have been marked up steeply, are at a disadvantage.
So when Berwyn (Pa.) entrepreneur Larry E. Fondren offered an electronic service that took some of the mystery out of the bond market and cut trading costs, he thought it would be a roaring success. Some players, though, didn't think it was such a good idea. Now, Fondren is out $10 million. His story could trigger closer inspection--and possibly major reforms--of the largely opaque bond market by Congress and regulators.
BASIC INFO. Fondren began small in 1985. He started the Exchange, a computer matching service for insurance companies looking to dispose of junk bonds that regulators wanted sold from their investment portfolios. The limited foray into the bond market convinced Fondren that there was a need for an electronic service that carried basic pricing information, similar to Instinet, the equity trading system of Reuters Holdings PLC. What information did exist was anecdotal, and it came from a small number of mostly New York-based bond dealers, such as Bear Stearns, C.S. First Boston, Cowen & Co., and Merrill Lynch, rather than from a national market system such as the New York Stock Exchange or NASDAQ. Today, there are more than a dozen online services for the bond market, but all have limitations.
After several years of research and trials, Fondren in 1993 persuaded a small number of investors to back a radical new product: InterVest Financial Services, an electronic system for the estimated $100 billion-a-day secondary market, in which bonds that are held in investor portfolios or in professional dealers' inventories are traded. And that's when his troubles began.
With Securities & Exchange Commission approval, he offered what no other service had: electronic posting of bids and offers, the ability to make trades at the push of a button, and anonymous transactions between institutions. The system, in theory, should reduce the cost of trading bonds by making the market more efficient. The average commission for an InterVest trade was 36 cents per $1,000 transaction, split between buyer and seller, according to Fondren. By comparison, the average markup on a corporate bond trade from 1995 to 1997 was $1.875. During that period, corporate bond buyers were overcharged $41 billion, says Fondren.
Fondren then asked Bloomberg Financial Markets, which carries other third-party services, to carry his system. Bloomberg agreed, but six months before InterVest's scheduled debut in 1996, it backed out. While Bloomberg eventually relented, Fondren believes it did so only after the Justice Dept. and the SEC began an informal investigation into whether dealers were colluding to keep InterVest from growing. Bloomberg refuses to comment on the episode.
FEW TAKERS. In December, 1996, InterVest began appearing on Bloomberg. But it never really caught on. Problem was, InterVest made dealers unhappy. It reduced the need for their services as middlemen between buyers and sellers. Indeed, only a handful of bond dealers posted their prices on InterVest. Says Thomas A. Price, a senior vice-president at Bank of New York's BondNet, an electronic trading service for BONY clients: "The Wall Street community didn't like what InterVest was about and boycotted it. They put [Fondren] in the penalty box." Adds James W. Toffey, president and CEO of TradeWeb, a new Internet-based service formed for four major investment banks: "Dealers are market makers and they work for their customers. That's the way the world works. Maybe [the bond markets] weren't ready to be mixed up."
Fondren says some 130 firms used the service. But most were on the "buy" side of the market--institutions such as life insurers, money managers, and pension funds that buy securities for their portfolios. Only a small number actually used InterVest to transact business with one another--about 25 orders per day on average--and the typical size of each trade was relatively small, about $2 million. By the end of 1997, Fondren didn't have enough subscribers to keep the system going, and he had run out of money.
Thinking he had nothing to lose, Fondren in late January of this year advertised that InterVest was available for corporate bond issuers to sell directly to institutions. "It caused an uproar," says Fondren, "because dealers didn't like that I was cutting them out of new issues."
A week later, Bloomberg dropped InterVest. Fondren believes Bloomberg did so because of either real or perceived pressure from dealers, who are important Bloomberg customers and valuable sources of bond-market information. A Bloomberg spokeswoman denied this was why InterVest was cut off.
Dealers claim there was never a boycott. They speculate that Bloomberg wasn't convinced InterVest would ever become a commercial success. If dealers don't like a service, it has no hope of getting the critical mass of transactions that make a market functional. InterVest failed because "nobody likes going to a party until there are lots of other people there," says Christopher J. Carroll, managing director of global electronic trading for DMG.
HIGH COSTS. Washington authorities are watching Fondren's experience. Lawmakers who oversee securities trading could begin an inquiry--somewhat like the 1993 congressional effort to create a reporting system for government bonds, which was beaten back by dealers. "We've long known that the bond market's a big mystery to most investors," says one GOP aide. "It would be in the public interest if we could make it more transparent."
SEC officials also would like the bond market to be more accessible and to offer more information on prices and trades. But the commission prefers to let the private sector do the job. While the SEC since 1975 has had the power to require corporate bond trades to be reported, for instance, the agency has never used it.
University of Chicago finance professor Paul Schultz, one of the authors of a landmark 1994 study of the NASDAQ market that found widespread collusion among dealers, has been studying the bond market, and he doubts that he'll find NASDAQ-type collusion. A Justice source says so far they've not seen evidence of that sort of behavior. Still, Schultz expects his study will find that trading costs are "fairly high."
So what is Fondren's fate? He currently is trying to raise $10 million to switch InterVest to an Internet-based bond service. Otherwise, InterVest will go out of business. "I mortgaged my house, my wife sold her jewelry, and even my employees have lent me their savings," says Fondren. He may have a better mousetrap, but so far the world isn't beating a path to his door.