Occasional Paper No. 37: Flash Crash in an OTC Market

This Occasional Paper contributes to the research on flash crashes which are high-profile episodes that can be thought of as short-lived malfunctions of capital markets typically involving a substantial price change and a drying up of liquidity followed by a price reversal.

Occasional Paper No. 37 (PDF)


The foreign exchange (‘FX’) spot market is the biggest and most liquid in financial services. Over 5 trillion US Dollars’ worth of currencies are traded in this market every day. Corporates, financial institutions and private individuals can exchange currencies immediately in the spot market, enabling them to buy products in foreign countries, or speculate on currency price movements. Alongside the spot market, there is an over-the-counter (‘OTC’) FX derivatives market. This large market mainly consists of forward contracts, which are non-standardised contracts between 2 parties to buy or sell a specified amount of currency at a pre-determined price in the future. This market is used by large corporates or financial institutions to hedge foreign exchange risk, for arbitrage, or for speculation.

In this paper, using proprietary data reported to the FCA under EMIR (the ‘European Market Infrastructure Regulation’), we examine the underlying drivers of the flash crash in the spot rate for Pound Sterling vs US Dollar (GBP/USD) in October 2016. To our knowledge, this is the first paper to take use these trade reports to analyse how different market participants react in times of market stress and their impact on the liquidity dry-up in a flash crash. Our research is the first study to examine the underlying drivers of a flash crash in an opaque OTC market; previous analyses have focused exclusively on flash crashes in exchange-traded markets. Finally, our paper is also the first to investigate the impact of derivatives on the underlying spot market during a flash crash. This allows us to test 3 competing theories of flash crashes in an OTC market: (i) order flow toxicity; (ii) limited risk-bearing capacity of market makers; (iii) developments in a related derivative market.

We do not investigate the initial trigger of the GBP/USD flash crash in this paper as it has been discussed comprehensively in previous literature (BIS, 2017). Furthermore, a recently published report by the Bank of England provides an in-depth analysis of the liquidity deterioration in GBP/USD using a number of different liquidity metrics (Noss et al., 2017).    

However, there is no research on the behaviour of OTC market participants during flash crash periods. This is largely due to the lack of publicly available transaction data. The behaviour of market participants in OTC markets is of special interest given the significant market volumes executed for particular asset classes (eg foreign exchange) and the different trading structure involved. In contrast to an asset traded on a centralised multilateral trading platform, an asset traded in an OTC market may trade simultaneously at different prices. This is because trades are agreed bilaterally between 2 counterparties such as an investor and a dealer. Also, dealers have no obligation to provide liquidity on OTC markets. These unique features of the trading architecture mean it is particularly interesting to study how a sudden price movement propagates through the market.


Florian Schroeder, Matthew Allan, Andrew Lepone, Henry Leung and Stephen Satchell.  


Occasional Papers contribute to the work of the FCA by providing rigorous research results and stimulating debate. While they may not necessarily represent the position of the FCA, they are one source of evidence that the FCA may use while discharging its functions and to inform its views. The FCA endeavours to ensure that research outputs are correct, through checks including independent referee reports, but the nature of such research and choice of research methods is a matter for the authors using their expert judgement. To the extent that Occasional Papers contain any errors or omissions, they should be attributed to the individual authors, rather than to the FCA.