What currencies are being traded using the algorithm and what is the trading size?

Take a look at the chart below for the level of algorithmic trading in different currencies. Not surprisingly, 94% of transactions for G7 currencies have customers using algorithmic transactions. Ticket size - The average trading volume per customer - is about 31 million USD for the group of G7 currencies.

Statistics on algorithmic trading

Note: Group of G7 currencies includes: USD, EUR, JPY, GBP, AUD, CAD, CHF

What are the consequences of this?

One effect of the increased use of algorithmic trading to talk about first is the impact on market makers (usually banks), who try to do internal execution instead of pushing them to the outside environment.

This trend has raised a number of concerns for the Bank for International Settlements - BIS (ND: This bank often publishes money market statistics), because if the internal matching happens too much then the quality of the quote may be compromised. You can briefly imagine that then, the quoted price does not reflect the actual price. This can also cause the trading volume at the venue's primary trading to decrease. It is important because prices from central trading points like Refinitiv and EBS are often used as the source of prices for other currency trading platforms.

Another impact of algorithm transactions is the proliferation of "flash" incidents. These are very strong, extremely fast and sudden price movements, surprising the whole market. It is often error-prone for algorithmic trading, however, there are also "flash" issues that happen for no apparent reason.