The Commerce Department said it fired three employes yesterday for using their advance knowledge of forthcoming economic statistics to play the market.
The employes' actions came to light during an investigation of leaks of Commerce figures for the gross national product in July and September of last year. The figures became common knowledge among stock, bond and commodity traders at least a day before they were to be released.
The source of the leak to the financial markets was not found, a report by the department's inspector general, Sherman M. Funk said. But the investigation by the Inspector General's Office and the Federal Bureau of Investigation turned up the alleged use of inside information for personal gain.
Unlike insider trading on Wall Street, which can result in civil and criminal penalties, use of the government economic data for personal profit is not illegal. Commerce Secretary Malcolm Baldrige said the department has drafted legislation, now under review at the Office of Management and Budget, that would make such disclosure a crime.
"There was a leak, but the inspector general cannot identify a leaker," Baldrige said. "There was no smoking gun, but there were some hammers cocked."
"I cannot guarantee that there will never be another leak," Baldrige said. "But I will guarantee that the steps that have been taken will help us to better track down any leak, and any future occurrence will be dealt with as severely as possible."
The employes were not identified because they have 30 days in which to appeal their dismissal.
The FBI gave lie-detector tests to 14 employes out of a group of 262 who were asked if they would voluntarily agree to be tested. Seven employes refused, the report said.
The FBI examiners believed that eight of the 14 were completely candid, while five had been at least "partly deceptive," the report said. Results on the remaining employe were inconclusive.
Officials would not say whether the three fired employes were among the five believed to be "partly deceptive" or the seven who refused to be tested.
Two of the employes used or attempted to use the confidential GNP data last September, and on earlier occasions, to deal in bond futures contracts, the report said. Baldrige said that the third employe passed on the information to someone who profited from its use.
A bond futures contract commits someone to buy or sell a specific quantity of government bonds at a predetermined price at a set date in the future. If the value of the bonds rises, the holders of the futures contract can make a large profit on a small down payment.
The figure leaked last September was the so-called flash estimate by the department's Bureau of Economic Analysis, or BEA, for the increase in GNP in the third quarter. The estimate, that the economy was growing at a 2.8 percent annual rate after adjustment for inflation, was lower than many financial analysts had been expecting.
The leak of the figure helped spark a rally in bond prices, because it indicated that the economy was more sluggish than had been thought. That caused interest rates to fall and bond prices to rise, both because of an expectation that there would be less demand for credit in a weaker economy and because the Federal Reserve would be encouraged to make money more readily available to the banking system.
Both employes who traded personally on the information made money, though Commerce officials declined to say how much. "It was a relatively minor gain in the great scheme of life," said Commerce spokesman B. Jay Cooper.
In the investigation report, Funk was sharply critical of the security procedures used by BEA to protect the confidentiality of its estimates between the time they were completed and their release date.
"Although only 16 employes should have had access to the final aggregate number prior to its public release, we found that at least 60 individuals knew the September flash GNP number beforehand," the report said.
"Indeed, the number was so readily available through word of mouth and review of unsecured documents that it was impossible to determine precisely how many employes did know it."
The report said BEA let too much time lapse between completing the estimates and releasing them, did not restrict access to work areas where the estimates were available, had no way of telling who looked up the data on the BEA computer system and had inadequate control over "sensitive" computer printouts.
As the investigation progressed, BEA took a number of actions to reduce the possibility of leaks, including allowing far fewer employes access to the numbers and beefing up the agency's computer security.
The investigation also found that many employes required to file current financial disclosure statements, including those with access to sensitive data, had not done so.
Another element of the investigation was an analysis of more than 13,500 long-distance telephone calls from all 638 BEA phones between July and October 1985. "Telephone calls were placed to major brokerage houses, financial institutions, accounting and law firms, and a variety of industry and commercial concerns," the report said. "While some contact with such organizations is expected, the number and frequency of these contacts by so many BEA employes were questionable."
A detailed review of 5,567 of the calls determined that 32 percent of them lasting a total of 11,030 minutes were not for official business. BEA officials, presented with the findings, collected $2,500 from employes within a month, and more collections are expected, the report said.