Large price jumps often make headlines. But what actually happens when prices move that much, that fast?
In a new paper, with coauthors Lucas Saru and Shihao Yu, I examine these “price dislocation” events in EuroSTOXX 50 futures. It is the result of a project that was carried out as part of the Alberto Giovannini Programme for Data Science at the European Systemic Risk Board (ESRB). We combine public limit order book data – which show how prices evolve – with confidential EMIR trade repository data, which reveal who trades.1
Markets always move, so we need a way to detect the truly exceptional moves. We use a statistical process that flags large returns only when they are extreme relative to predicted volatility. Between 2020 and 2023, we identify 2,637 such dislocations, with returns ranging from −96 to +99 basis points in just 10 seconds. In short: brief, sharp, and worth studying.
Trading volume jumps during dislocations, which is no surprise. The more interesting question is: What is the nature of this high volume? Using EMIR data, we can tell whether trading becomes more concentrated (fewer participants who trade more), or more dispersed (more participants). We find evidence of the latter; the number of buyers and sellers increases, and the Herfindahl-Hirschman index declines. Even traders who had not traded earlier that day, enter the market during these bursts. So, rather than fragility, we see resilience – a broad set of participants absorbing shocks.
Can private trade repository data explain jumps, beyond what can be explained using standard analysis based on public order book data alone? Our results show that they can, but the increase in explanatory power is modest. Still, the private data add nuance: Dislocations are smaller when liquidity is supplied by a more diverse set of sellers. When only a few supply liquidity, jumps are larger.
We also try to predict dislocations using both types of data: public and private. The results are humbling: very little predictive power.
So what do we learn? When prices dislocate, markets do not freeze, they breathe. More participants step in and concentration falls. For all the noise of sudden price jumps, volatility comes with vitality. That, in my view, is a reassuring sign of market resilience.
Please find the paper here.
The data had been anonymized before we received them to do our analysis. ↩︎