The Bank of England has published a Working Paper (PDF 946 KB) on algorithmic trading.
The paper analyses the role of algorithmic trading (AT) in foreign exchange (FX) markets when the Swiss National Bank announced in January 2015 that it had discontinued its policy of capping the value of the Swiss franc (CHF) against the euro.
The paper finds that:
- In reaction to the Swiss franc event, AT tended to consume liquidity and reinforce price disruption, thereby contributing a lot of uninformative `noise' to the EUR/CHF market on the event day.
- Human traders did the opposite - supporting market quality by providing liquidity and aiding price discovery. Computer traders traded `with the wind', buying the franc as it appreciated, while humans `leaned against the wind'.
- However, human traders did not make net sales in the key 20-minute period immediately after the announcement. So the CHF appreciated sharply initially, with much of these gains reversed in the subsequent hour.
- This reduction in market quality was concentrated in the shocked FX rate (EUR/CHF) and, to a lesser extent, in USD/CHF. Other currency pairs were essentially unaffected, suggesting that AT models were compartmentalised.
It is less clear what drove these behaviours. The paper speculates that many of the computer trades may have been driven by liquidity needs and capital requirements rather than the new information in the market.
To discuss this further, please contact a member of the EMA Financial Services Risk & Regulatory Insight Centre.