Felix Salmon

Twitter datapoint of the day

Felix Salmon
Nov 17, 2010 22:17 UTC

I work for a global information company which makes billions of dollars a year selling valuable data to banks, hedge funds, and other people in the financial markets, often at very high prices: $2,000 a month or even more.

And then there’s Twitter, which jealously guards access to its full stream of tweets (roughly 1,000 per second, these days). As of now, however, it’s signed a deal with Gnip whereby you can get a randomly-selected 50% of those tweets for $360,000 a year, which works out at $30,000 a month. You’re not allowed to republish them, but that’s OK—the people willing to spend that kind of money are likely to be high-frequency trading shops who want to keep the data as private as possible in any case.

I don’t have a problem with Twitter monetizing my public tweets in this manner; as I understand it, DMs aren’t included, and neither are any tweets from protected accounts. But it’s quite astonishing how much those tweets are worth, when they’re aggregated into a fat pipe. And it’s also interesting to me how much more 50% of the full stream is worth than 5%, which you can get for just $5,000 a month. Given the rapidly-diminishing marginal returns of each additional Twitter stream, I wonder where the added value comes from. I’d imagine that if a topic starts trending on the 50% feed, it will almost certainly be trending on the 5% feed as well.

I do, on the other hand, have a problem with other sites—Facebook in particular—monetizing my private information. I worried that Mint might be doing that kind of thing back in March, and in general if any website wants to sell any information of mine which isn’t public, I want them to ask my permission first. As Twitter shows, aggregated user data can be very valuable indeed. And with that kind of money on the table, there’s a lot of incentive to be ethically flexible.


You are in fact incorrect in claiming that 5% of the stream is roughly equivalent to 100% of the stream for capturing trends.

Secondly, trends are just one facet of all the interesting things that can be accomplished with the Twitter. For example, if you wanted to – given an arbitrary Twitter id – find out their topics of interest, good luck doing that with 5% of the stream.

Similarly, if you want to build a social media monitoring service (of the kind that Sysomos built and sold successfully last year) and then sell the service to large brands, once again, good luck doing that with 5% of the overall stream.

Lastly, the folks who license the firehose – and that list of companies is easily available via a google search – are inherently uninterested in being a reseller. They are not high-frequency trading shops but are mostly Silicon Valley companies trying to build innovative apps and services on top of this mass volume of data.

Posted by saumil07 | Report as abusive

Trillium wasn’t quote-stuffing

Felix Salmon
Sep 14, 2010 14:11 UTC

Chris Nicholson is right, where Courtney Comstock and Marcy Gordon are wrong: the $2.26 million fine slapped on Trillium is not for quote-stuffing. Instead, it’s for layering.

The distinction is an important one. Quote-stuffing, if it exists, is a destructive attack on an entire stock market. Layering, by contrast, is relatively benign, and the only people who get damaged by it are high-frequency traders who are looking to sniff out where the market is going and place trades attempting to front-run that move.

What Trillium did is market manipulation, to be sure, and it deserves a fine. But it’s a bit of a stretch to paint this as the first battle in the war against high-frequency traders — not least because there isn’t actually anything particularly high-frequency about what Trillium was doing.

Yes, Finra does say that Trillium’s layering was an “improper high frequency trading strategy”. But fundamentally it was about misdirection, rather than speed.

Layering is a way of getting good execution on a trade you already know that you want to do: it’s not some kind of market arbitrage where you’re trying to make immediate profits in a fraction of a second. The profits calculated by Finra constitute the difference between the prices that Trillium got and the prices that it would have got if it hadn’t been moving the market: in other words, Trillium could easily have made a loss on the trades, while still making illicit profits in the eyes of Finra.

An example would probably help. Let’s say that XYZ stock is trading at a bid of $24.50 and an offer of $24.55. And let’s say that a trader at Trillium wants to sell XYZ stock. He could simply hit the bid, and receive $24.50. Or he could put in a sell order at $24.54, and hope that someone wanting to buy will take him out there. That would be a much better outcome, because not only does he get a better price, but he also counts as the liquidity provider, and so gets a small rebate from the stock exchange. On the other hand, there’s no guarantee that anybody will take him up on that offer.

But let’s say that he puts in a hidden offer at $24.54, in a dark pool. That means that the offer is there, but no one can see it. (This is perfectly legal, by the way.) Then comes the sneaky part: he puts in a large number of public and visible bids at prices like $24.48 and $24.47. He doesn’t actually want to buy XYZ stock at those prices: in fact, he doesn’t want to buy XYZ stock at all. But to other traders, looking at the public order book, it looks as though there’s a large amount of buying interest in XYZ. So they start putting in their own bids: at $24.51, $24.52, $24.53. They’re all trying to get in front of the big new buyer they see in the market.

And presto, when they bid $24.54, they find that hidden sell order from Trillium, which gets a very good price on its trade and gets to count as a liquidity provider. And as soon as that happens, all those fake bids at $24.48 and $24.47 suddenly disappear. But the victims are the people (or algorithms) who thought there was a naive trader posting public buy orders, and wanted to trade against that order. It’s hard to feel a lot of sympathy for them. As Kid Dynamite says,

There’s a big problem with Trillium’s strategy: it’s illegal. It’s called market manipulation, and they got bagged for it. Ironically, most of the same people who are/will be happy that Trillium is getting punished for this behavior would also be happy that the participants who Trillium was trying to fool got fooled. It’s been made quite clear that the Populace At Large does not like the trend of traders, be they high frequency (more so, lately, obviously) or low frequency, reacting to publicly displayed quotations by changing their own quotes. Or lifting bids / hitting offers in response to publicly displayed quotes. Those are the very traders Trillium was trying to (illegally) take advantage of.

And so it’s important to distinguish what Trillium was doing from quote-stuffing. Quote-stuffing is essentially a denial-of-service attack, aimed at trying to slow down a market in an environment where milliseconds matter.

With quote-stuffing, high-frequency traders enter an enormous number of bids and offers, significantly outside the current bid-offer spread, just to introduce a vast amount of noise into the quote feed. All that noise takes time (maybe just a few extra nanoseconds) for rival HFT shops to process, giving the quote stuffer a crucial time advantage.

The SEC is looking into whether quote-stuffing exists, and whether it’s a strategy that anybody has actually used to make money. Opinions differ on that front. But let’s not confuse the SEC investigation into quote-stuffing with the Finra fine on Trillium. They’re two very different things.


This and other topics that are relevant for speed traders and institutional investors will be discussed at High-Frequency Trading Leaders Forum 2013 London, next Thursday March 21.

Posted by EllieKim | Report as abusive