Impact and Recovery Process of Mini Flash Crashes: An Empirical Study

TL;DR

This empirical study analyzes the causes and recovery patterns of Mini Flash Crashes in stock markets, revealing that most are caused by large market orders rather than high-frequency trading, with partial recovery observed in many cases.

Contribution

It provides a detailed empirical analysis distinguishing causes of Mini Flash Crashes and examines their short-term recovery dynamics, challenging common assumptions about high-frequency trading's role.

Findings

Most Mini Flash Crashes are caused by large market orders, not high-frequency trading.

During financial crises, the frequency of these events increases.

Less than 40% of crashes fully recover within the observed trades.

Abstract

In an Ultrafast Extreme Event (or Mini Flash Crash), the price of a traded stock increases or decreases strongly within milliseconds. We present a detailed study of Ultrafast Extreme Events in stock market data. In contrast to popular belief, our analysis suggests that most of the Ultrafast Extreme Events are not primarily due to High Frequency Trading. In at least 60 percent of the observed Ultrafast Extreme Events, the main cause for the events are large market orders. In times of financial crisis, large market orders are more likely which can be linked to the significant increase of Ultrafast Extreme Events occurrences. Furthermore, we analyze the 100 trades following each Ultrafast Extreme Events. While we observe a tendency of the prices to partially recover, less than 40 percent recover completely. On the other hand we find 25 percent of the Ultrafast Extreme Events to be almost recovered after only one trade which differs from the usually found price impact of market orders.