http://qz.com/126823/wednesdays-federal ... t-analyst/
On Sept. 18, gold futures surged nearly $20 in the 30 seconds after the US Federal Reserve released the news that it would not “taper” the monetary stimulus it’s feeding the economy. Someone made a lot of money off of that—potentially by dubious means.
According to research from Nanex LLC, a Chicago-based research firm that monitors trading activity, a mountain of orders placed at exactly 2:00:00 PM ET—and the activity following those orders—indicate that someone almost definitely had bets in early. Eric Hunsader, Nanex’s founder, calls the evidence “overwhelming.”
News from the Federal Reserve is released from a lock-up room in Washington, DC, so that reporters can write stories ahead of the actual release. Special “black boxes” prevent reporters from transmitting the news until the exact millisecond the data are officially public. After that, information takes 2 milliseconds to travel from DC to New York City (really, computers in New Jersey), where stocks trade, and 7 milliseconds to travel to Chicago, where futures trade.
But Nanex finds that a large pile of trades—both in a gold ETF (GLD) traded in New York and in gold futures traded in Chicago—happened exactly at 2:00 PM ET, but not a thousandth of a second later. The news simply wouldn’t have had the chance to travel this distance in that time.
Could it just have been dumb luck? Probably not. Based on the data, Hunsader believes that the bet came from a single actor who would have had to “commit well over a billion dollars,” meaning it was too large a gamble to take lightly or blindly. Hunsader believes someone had the information early.
Could a reporter have leaked it? As we’ve previously written, the lock-ups haven’t always been as secure as they should be. But given recent publicity about leaks, Hunsader thinks that’s unlikely.
Could a news organization have sent the data in advance to its own servers in both New York and Chicago and programmed them to release it at the exact same moment? They are allowed to do that with certain kinds of data, Hunsader writes, but in the case of last week’s Fed announcement that shouldn’t have been possible. And even if they had, such releases typically happen up to 15 milliseconds too early or too late because the servers’ clocks aren’t precisely set. In this case the spike in trades began on the very millisecond of 2pm.
“This is not a bad technology case. This is a somebody’s-hand-is-in-the-cookie-jar case,” Hunsader told Quartz.
“The Fed news was leaked to, or known by, a large Wall Street Firm who made the decision to pre-program their trading machines in both New York and Chicago and wait until precisely 2 PM when they would buy everything available,” he writes. ”It is somewhat fascinating that they tried to be ‘honest’ by waiting until 2pm, but not a thousandth of a second longer.”
It’s not clear how the US government would attempt to track down the source of the leak. But it could certainly investigate the trades that took place. At the very least, the laws of physics combined with the laws of the land seem to make it impossible for a firm to have had information that it appears to have traded on at the time that it appears to have traded on it.
News organizations respond to Fed lockup questions
http://www.cnbc.com/id/101056168
The Federal Reserve says it is contacting news organizations to discuss the rules surrounding lock up procedures and the release of market moving information from the Federal Reserve's headquarters in Washington.
But the leading expert on millisecond level trading says he is focusing his attention on a certain type of news organization – those that offer so-called "low latency" services to feed market moving data at high speeds directly into computerized trading systems.
Eric Hunsader, founder of the market analysis firm Nanex, says that's because he saw simultaneous reactions to the Fed's announcement last week in trading in New York and Chicago. That would be theoretically impossible if the information was released from the Fed's headquarters in Washington. In theory, the trading reaction should have begun in New York several milliseconds before it began in Chicago, because information takes several more milliseconds to travel the longer distance.
"The very first thought I had when I looked at this closely was this is a low latency service," Hunsader said. "We have just very recently looked closely at some of these low latency releases and seen that they are indeed at the same exact millisecond. We have immediate history behind it."
(This below would make this comment terrestrial, but premoderated walking dangers - as me - are waiting a few day to appear.
I don't know why I am associating to D&C 124...
64 And they shall not receive less than fifty dollars for a share of stock in that house, and they shall be permitted to receive fifteen thousand dollars from any one man for stock in that house.
65 But they shall not be permitted to receive over fifteen thousand dollars stock from any one man.
66 And they shall not be permitted to receive under fifty dollars for a share of stock from any one man in that house.
67 And they shall not be permitted to receive any man, as a stockholder in this house, except the same shall pay his stock into their hands at the time he receives stock;
68 And in proportion to the amount of stock he pays into their hands he shall receive stock in that house; but if he pays nothing into their hands he shall not receive any stock in that house.
72 Verily I say unto you, let my servant Joseph pay stock into their hands for the building of that house, as seemeth him good; but my servant Joseph cannot pay over fifteen thousand dollars stock in that house, nor under fifty dollars; neither can any other man, saith the Lord.
125 I give unto you my servant Joseph to be a presiding elder over all my church, to be a translator, a revelator, a seer, and prophet.
Even so. Amen.
Fortunately, JS didn't know the word millisecond. There is no such word in Reformed Egyptian.