400ms of insider information

High frequency (insider) trading.

Nanex Research has found what looks like insider trading in the natural gas market prior to the release of the US Energy Information Administration's natural gas report. The quirk is that that insider trading happened just 400 milliseconds before the report was released:

On January 31, 2013, approximately 400 milliseconds before the official release of the EIA Natural Gas Report, trading activity exploded in Natural Gas Futures and ETFs such as UGZ, UNG and BOIL. Now that the Feds have stated (as claimed by a recent WSJ article) that they don't think there is merit in prosecuting people who get news information earlier than others by milliseconds, is it any wonder?

It is worth pointing out that the EIA Natural Gas Report comes out weekly (every Thursday at 10:30) and the market reacts within a few milliseconds. This is because the report centers on one number which makes it easy for machines to process and take action.

As Nanex points out, a recent SEC investigation into whether some news organisations gave investors access to economic data "a fraction of a second before the official release time" resulted in no charges being brought. At the time, it was speculated that there were two reasons for that: the first being that such a prosecution would stretch the definition of insider trading, and the second being that it was difficult to conceive of such a head start leading to any measurable advantage.

Insider trading is typically defined as acting on information which has not yet been made public (the legal definitions are far more complex than that, but that's largely owing to the byzantine nature of financial regulation). The problem with prosecuting news organisations for that is that typically, information hitting the newswires is the definition of it being made public. This has caused problems before: last year, Netflix's CEO faced trouble from the regulators for announcing on his Facebook page that the company had had over a billion cumulative viewing hours in one month. Facebook is not, apparently, "public" enough for the SEC.

If one of the newswires publishes information a fraction of a second before the others, that might constitute a broken embargo, or an undesirable leak; but it probably doesn't constitute insider trading, because the very act of publishing made formerly private information public (even the etymology's the same! "Publish" literally means "to make public").

But the second argument was that, in the seconds leading up to a potentially market-moving data release, trading slowed down and waited for the news. After all, there's no one — not even an algorithm — which isn't going to think a trade a fraction of a second before a data release offered at a markedly different price isn't a tiny bit suspicious.

That argument might not hold as much water if Nanex's data is accurate, though. It shows a definite collapse in the price of a natural gas exchange-traded fund (ETF) over the course of a hundred milliseconds. A fall of one per cent — even one which is then followed by an even greater fall once the actual data is released — is not to be sniffed at.

It's not clear who the counterparties in these trades were — who, that is, was convinced to make trades milliseconds before a major data release — but it's pretty likely that they were also algobots. Insofar as this represents a transfer of income from one set of computer-owners to another set, it's not the most concerning news. But it does raise further questions about how the market for information is shaped in the near future, and whether the simple dichotomy between public and not public information can hold up in that new world.

"UNG showing trades color coded by exchange between 10:29:59 and 10:30:04." Chart: Nanex Research

Alex Hern is a technology reporter for the Guardian. He was formerly staff writer at the New Statesman. You should follow Alex on Twitter.

Photo: Getty
Show Hide image

Who will win in Stoke-on-Trent?

Labour are the favourites, but they could fall victim to a shock in the Midlands constituency.  

The resignation of Tristram Hunt as MP for Stoke-on-Central has triggered a by-election in the safe Labour seat of Stoke on Trent Central. That had Westminster speculating about the possibility of a victory for Ukip, which only intensified once Paul Nuttall, the party’s leader, was installed as the candidate.

If Nuttall’s message that the Labour Party has lost touch with its small-town and post-industrial heartlands is going to pay dividends at the ballot box, there can hardly be a better set of circumstances than this: the sitting MP has quit to take up a well-paid job in London, and although  the overwhelming majority of Labour MPs voted to block Brexit, the well-advertised divisions in that party over the vote should help Ukip.

But Labour started with a solid lead – it is always more useful to talk about percentages, not raw vote totals – of 16 points in 2015, with the two parties of the right effectively tied in second and third place. Just 33 votes separated Ukip in second from the third-placed Conservatives.

There was a possible – but narrow – path to victory for Ukip that involved swallowing up the Conservative vote, while Labour shed votes in three directions: to the Liberal Democrats, to Ukip, and to abstention.

But as I wrote at the start of the contest, Ukip were, in my view, overwritten in their chances of winning the seat. We talk a lot about Labour’s problem appealing to “aspirational” voters in Westminster, but less covered, and equally important, is Ukip’s aspiration problem.

For some people, a vote for Ukip is effectively a declaration that you live in a dump. You can have an interesting debate about whether it was particularly sympathetic of Ken Clarke to brand that party’s voters as “elderly male people who have had disappointing lives”, but that view is not just confined to pro-European Conservatives. A great number of people, in Stoke and elsewhere, who are sympathetic to Ukip’s positions on immigration, international development and the European Union also think that voting Ukip is for losers.

That always made making inroads into the Conservative vote harder than it looks. At the risk of looking very, very foolish in six days time, I found it difficult to imagine why Tory voters in Hanley would take the risk of voting Ukip. As I wrote when Nuttall announced his candidacy, the Conservatives were, in my view, a bigger threat to Labour than Ukip.

Under Theresa May, almost every move the party has made has been designed around making inroads into the Ukip vote and that part of the Labour vote that is sympathetic to Ukip. If the polls are to be believed, she’s succeeding nationally, though even on current polling, the Conservatives wouldn’t have enough to take Stoke on Trent Central.

Now Theresa May has made a visit to the constituency. Well, seeing as the government has a comfortable majority in the House of Commons, it’s not as if the Prime Minister needs to find time to visit the seat, particularly when there is another, easier battle down the road in the shape of the West Midlands mayoral election.

But one thing is certain: the Conservatives wouldn’t be sending May down if they thought that they were going to do worse than they did in 2015.

Parties can be wrong of course. The Conservatives knew that they had found a vulnerable spot in the last election as far as a Labour deal with the SNP was concerned. They thought that vulnerable spot was worth 15 to 20 seats. They gained 27 from the Liberal Democrats and a further eight from Labour.  Labour knew they would underperform public expectations and thought they’d end up with around 260 to 280 seats. They ended up with 232.

Nevertheless, Theresa May wouldn’t be coming down to Stoke if CCHQ thought that four days later, her party was going to finish fourth. And if the Conservatives don’t collapse, anyone betting on Ukip is liable to lose their shirt. 

Stephen Bush is special correspondent at the New Statesman. His daily briefing, Morning Call, provides a quick and essential guide to British politics.