The ECB thinks that someone could be leaking US economic data
According to a white paper released by the European Central Bank on Monday, there is statistical evidence that a large swath of US economic data is getting leaked early, leading to price shifts in markets before the releases come out.
Here's the key bit from the ECB (emphasis ours):
We find evidence of substantial pre-announcement informed trading in equity index and Treasury futures markets for seven out of 21 market-moving U.S. macroeconomic announcements. About 30 minutes before the release time, prices begin to drift in the direction of the market's subsequent reaction to the news. This drift accounts for 49 percent and 39 percent of the overall price adjustments in the E-mini S&P 500 and 10-year Treasury note futures markets, respectively, and the estimated magnitude of profits of informed traders underscores the economic significance of these price moves.
The ECB team - Alexander Kurov, Alessio Sancetta, Georg Strasser and Marketa Halova Wolfe - analyzed market movements around the releases of 30 data points ranging from the ISM Manufacturing Index to the jobs report to Existing Home Sales.
The biggest market movements, the ECB argued, are those that occur when a data point "surprises" investors. So, using the Bloomberg consensus forecast, the researchers looked at price drifts on the second-by-second S&P 500 market leading up to the release of significant "surprises" or data points that varied greatly from those consensus forecasts.
According to their data, nearly half of the market move due to an economic surprise occurs before the release of that data.
The ECB paper suggests two possible causes for early access to data: information leakage and superior forecasting.
The information leakage hypothesis suggests that people who receive the information during a "lock-up" period before the official release are getting the information out early. This is essentially the nefarious option.
On the other hand, said the paper's authors, it may just be that some market watchers are better than others at predicting the data and are making large, correct trades on publicly available information.
"Some of the superior forecasting ability may be based on smart reprocessing of publicly available data," said the researchers.
"Superior forecasts of the announcement surprises may also be generated by 'digging deeper' into pre-packaged information products, for example, by using forecasts by individual professional forecasters instead of the Bloomberg consensus forecast. Further improvements in forecasting may be due to resource-intensive legwork creating original proprietary datasets that proxy the data underlying public announcements."
Either way, the ECB research shows that much of the profit to be made by a surprising data release is captured prior to the actual data coming out. Whether this is caused by intentional leaks or just better forecasters beating the market is up for debate.