Turbulence in stocks and bonds was part of the Federal Reserve’s calculus for slowing interest rate increases in February. Now an economist with the central bank’s Dallas branch says relying on the equity market for economic signals is a mistake.

Benchmark gauges such as the S&P 500 Index are such flawed mirrors of the economy that they probably fail as predictors of gross domestic product, according to a new paper by Julieta Yung, an economist in the research department at the Fed Bank of Dallas. Half the components are manufacturers, for one thing, much more than is reflected in GDP.

Searching markets for clues about the future of employment and consumer spending became an obsession on Wall Street in the first six weeks of 2016, when the S&P 500 plummeted 11 percent for its worst start ever. But Yung says the quest is unlikely to yield any conclusions because noise is just as likely as news to be driving prices.

“There’s a definite divide between the state of the economy and any decline you might see in the equity market,” Yung in a phone interview.

Service-providing industries have accounted for more than three-quarters of U.S. GDP over the past decade, whereas more than half the S&P 500 consists of manufacturers, she said. Yung’s analysis included recalculating how much sales and profits companies in the index derive from various types of economic activity.

Yung, 29, holds a doctorate in economics from the University of Notre Dame and has authored multiple working papers on the subject of interest rates and their corresponding term structure. The research she conducts for the Dallas Fed is done independently of Fed policy makers and is intended to inform discussion among officers.

Short-term fluctuations in equity prices come too fast and furious and are caused by such a multitude of inputs that assuming they’ll directly translate into changes in real economic output is an error, Yung wrote. Michael Antonelli, a trader at Robert W. Baird & Co., says big swings often just reflect human emotions.

“The two can disconnect in the short-term because of the immediate effect sentiment has on stocks,” said Antonelli, an institutional equity sales trader and managing director at Robert W. Baird in Milwaukee. “Then that nervousness wanes, and people capitulate, and that’s when you see the market come back.”

Case in point was the S&P 500 selloff in January and February. The benchmark index erased its loss within two months thanks to virtually unprecedented rebounds in the industries that dragged it down in the first place.

From a broader lens, the stock market isn’t useless in forecasting recessions. Among the 20 percent drops that have hit American stocks 13 times since the Great Depression, 10 preceded U.S. recessions and only four recessions occurred without a bear market warning, according to data compiled by Bloomberg.