I've said before that I'm a big fan of the Wall Street Journal's series of investigations into the fact that professional traders can pay to get better information than amateurs can, but I can't really get behind today's installment, titled "Speed Traders Get an Edge." The edge is speed! No, I mean, the edge is that they "have been paying to get direct access to market-moving news releases, a practice that can give firms the ability to trade fractions of a second ahead of less fleet-footed investors." And the news is from Business Wire and Marketwired, which are platforms for distributing press releases.
By paying for direct feeds from the distributors and using high-speed algorithms to crunch data and enter orders, traders can get a fleeting — but lucrative — edge over other investors, according to traders and people familiar with the practice. The reason: tiny lags between the time the distributors release the news and when media outlets send them out to the public, including other investors.
So if you get your press releases from a news service, you have to wait for them to receive it from Business Wire and then send it out again, and while you're sitting around waiting for that to happen precious, I don't know, thousandths of a second are ticking by. Thousandths of a second during which you could be trading, instead of biting your nails and hitting "refresh" and generally doing things that can't actually be done in thousandths of a second.
Of course, you can easily go pay for a Business Wire feed yourself. And then you're on a level playing field with the high-frequency traders. You and their computers get the press releases at the same time, and read it at the same time, and when you're done reading you put in your trades at ... oh. Different times, huh? Those computers read that press release in 50 milliseconds, which is how long it took you to read the "T" at the beginning of this sentence.
The thing is, speed traders have numerous edges over you, and all of those edges are "speed." There is nothing new or surprising or scandalous about the fact that high-speed traders trade at higher speeds than low-speed traders. It's right there in the name. If your trading isn't automated using computer algorithms, you have no ability to profit from the 90-millisecond trading advantage that you get from a direct Business Wire feed. The computers are faster than you because they are computers! The direct feed makes them 90 milliseconds faster than other computers! Who cares?
Not you, I hope, but maybe the other computers. The other computers have to either pay for a direct feed themselves, or occasionally miss out on rapid price moves just after press releases come out. This, again, is hard to care about: Firms with faster computers also trade faster than firms with slower computers. You want faster computers, you pay for them. Similarly, you want faster data feeds, you pay for them.
That said, there is an arms-race quality to high-speed trading that a lot of people sensibly dislike: People are going around drilling through the earth's crust to shave a few milliseconds off their trading times. That is expensive and seems sort of wasteful: Presumably those dollars and engineers could be put to some higher social use than just competing for zero-sum stock trading profits.
But this isn't that. Nobody's drilling through mountains here; they're paying for press releases. Business Wire is mostly in the business of sending press releases from companies to journalists. It also puts those releases on its website so anyone can go read them for free. That is nice of Business Wire! And, while I guess you could have your doubts about the social value of a lot of press releases, on balance it's probably a useful service to be able to get a stream of press releases in one place.
But someone's gotta pay for it, and I guess mostly companies and media companies do. But the fact that high frequency trading firms also do means that Business Wire can charge a bit less to the media companies. High frequency traders are subsidizing journalism. The system works, or whatever.
The point is, this stuff seems fine. I give it my blessing. "But wait," you say, "aren't these high frequency traders buying the answer sheet before the exam? Isn't that the sort of thing that gets you sent to prison for 15 years?" Well, only sometimes, is the answer. Sometimes it's fine.
The example in the article is a trade where Ulta Salon Cosmetics & Fragrance Inc., which made the mistake of releasing its earnings at about 4 p.m. and 150 milliseconds one day. According to the Journal, people with the direct feed were trading by about 4:00:00.200, while Bloomberg got it out at 4:00:00.242. (Beating Dow Jones and Thomson Reuters at 4:00:00.464 and about 4:00:01, respectively, so go team.) So they had a head start of somewhere between 42 and 92 milliseconds.
The stock closed at around 4:00:00.700, so those trades counted. One real lesson here is that, if you're releasing earnings after the close, you should aim for like 4:05. Do it at 4:00 exactly and you might get them out before the close. But that's just Ulta's dumb fault, not Business Wire's.
I mean, not easily easily: "The company doesn't advertise the service," and you can't just drop your credit card on the website, you have to fill out a "Find Out More" form and wait for them to get back to you. And in fact a Business Wire spokesperson told me that there are some criteria for getting a feed, and that you in your garage might not qualify, though another spokesperson told the Journal that, "Anyone can get a direct data feed if they want."
Still, I bet if you called them up and waved around a stack of bills you'd be able to get the feed. "Business Wire's clients include Chicago's Chopper Trading LLC and Spano Trading LLC, a Miami Beach, Fla., high-speed trader," and what do they have that you don't?
I am making this up, I don't care what your actual reading speed is, and you don't read letter-by-letter, I hope. But you're not reading a press release, developing a trading thesis, and pushing the button on it in 50 milliseconds, is the point.
Again, not really, they're just talking about it. But Spread Networks really did spend "hundreds of millions of dollars to build a straighter fiber optic cable between Chicago and New York so that signals that once took 15.5 milliseconds to bounce back and forth now took only 13.3."
To contact the author on this story:
Matt Levine at firstname.lastname@example.org
To contact the editor on this story:
Tobin Harshaw at email@example.com