Ultrafast Extreme Events: Empirical Analysis of Mechanisms and Recovery in a Historical Perspective

TL;DR

Ultrafast Extreme Events (UEEs) are rapid, localized price moves that can destabilize markets. This paper conducts a large-scale empirical comparison across years (2007/2008, 2014, 2021) to identify mechanisms and post-event recovery, using high-resolution TAQ data and model-free analyses. It finds that brief liquidity shortages and order-book instability—not exogenous news or trader type—drive UEEs, with clustered occurrences and robust, similar recovery patterns across regimes. The results underscore the central role of market microstructure liquidity in UEEs and suggest non-stationary but structurally consistent dynamics shaped by market sentiment.

Abstract

To understand the emergence of Ultrafast Extreme Events (UEEs), the influence of algorithmic trading or high-frequency traders is of major interest as they make it extremely difficult to intervene and to stabilize financial markets. In an empirical analysis, we compare various characteristics of UEEs over different years for the US stock market to assess the possible non-stationarity of the effects. We show that liquidity plays a dominant role in the emergence of UEEs and find a general pattern in their dynamics. We also empirically investigate the after-effects in view of the recovery rate. We find common patterns for different years. We explain changes in the recovery rate by varying market sentiments for the different years. Overall, our results hint at a certain degree of universal behavior.

Figures (17)

• Figure 1: Beginning and end of a UEE with prices $S{(t_\mathrm{start})}$ and $S{(t_\mathrm{end})}$ for Microsoft on 7 January 2014. The price at which the $0.8\%$ criterion is first fulfilled is denoted by $S{(t_\mathrm{change})}$.
• Figure 2: Total number of counts for UEEs on a logarithmic scale for every trading week in 2014 (top), in 2021 (bottom).
• Figure 3: Frequency histograms for UEEs over a whole trading day for 2014 (top) and 2021 (bottom). The dashed lines indicate the start of the main phase of trading at 9:30 a.m. and its end at 4:00 p.m.
• Figure 4: Total number of counts for the largest return in the quotes for flash crashes (left) and flash spikes (right) on a linear scale for 2007/2008 (top), 2014 (middle) and 2021 (bottom). The top figure is taken from Ref. braun2018impact.
• Figure 5: Barplots for absolute returns in the quotes larger than 0.005 (top) and 0.008 (bottom). Numbers for 2007/2008 are taken from Ref. braun2018impact.