Two days ago, this column welcomed a move by Eurex, the German futures exchange, to explain unusual activity on its market.

In the wake of the ECB decision on February 6, its Dax futures market dropped by 2 per cent before rebounding. Some had put the decline down to a “fat finger” error. Eleven days later and the exchange said it was not so. Using a chart that covered the mere 414 milliseconds in question, the exchange attempted to show what happened.

It described trading as “orderly” with “a substantial rise in liquidity as reflected in the high number of traded contracts (1,488 in 414 milliseconds compared with an average of less than five contracts per second).”

As noted by this column, “what Eurex’s findings revealed was not only the millisecond speed with which interacting algorithms could affect the market but also the safeguards the exchange had in place to prevent things spiralling out of control”.

It further remarked: “While not fully solving the mystery, Eurex’s openness has shone light into what an exchange sees among the millisecond trades.”

A shame that just two days later Nanex, the financial markets consultancy, discovered Eurex had quietly altered its explanatory chart from this:

to this more smoothed out version:

Eurex said the alterations were minor, for things such as the format of the chart. “Substantially there is nothing changed,” a spokesperson said. Readers are invited to draw their own conclusions on that.

But if Eurex cannot produce a final, conclusive chart after 11 days, it raises the question of how much oversight and control it really has over its market. It really should have a far better grasp on presenting its data.

Such incidents do little to inspire confidence. As we have seen with outages around the world, too often the complex, subsecond interlinked markets feel very brittle. One of the great unknowns of the market infrastructure world is whether enough resources are being devoted to the technology to withstand shocks. Furthermore, what would have happened to Eurex’s systems if the selling pressure had been far greater? How robust is that Volatility Interruption period?

It is ironic that in trying to be open, Eurex has brought this on itself. Faced with this example, it would be a tremendous disappointment if others decided simply not to bother.

That would do little to encourage trust between venue and users, especially when it appears more transparency by exchanges on their workings is needed more than ever. Then again, the only thing worse would be offering up to the world inaccurate data.

Get alerts on Trading technology when a new story is published

Copyright The Financial Times Limited 2019. All rights reserved.
Reuse this content (opens in new window)

Follow the topics in this article