The Speed of the Equity Markets
U.S. Securities and Exchange Commission (October 9, 2013)
Analysis
Each day hundreds of millions of quotes and orders to buy or sell equity securities are posted to public exchanges. The trade-to-order data series reveals that at most, only a few percent of all orders and quotes for shares posted by market participants result in trade executions. The majority of displayed quotes are canceled. The chart below presents such a summary of quote lifetimes over the period of Q2 2013 for corporate stocks traded across all exchanges for which there is sufficient data to produce these metrics (see Quote Lifetime Methodology for a complete discussion). Each bar represents the cumulative percentages of full cancellations, partial cancellations, full trades and partial trades for specific periods of time.
As shown, nearly 100% of all orders are either canceled or result in a trade within 10 minutes of order placement. More than one third of all orders stay in force for at least 5 seconds, and over 60% of all orders that are canceled are in force for more than half of one second.
On the other side of the spectrum we see that no more than one quarter of all cancelations occur in less than 50 milliseconds and less than 8% of all cancelations are faster than 500 microseconds. Though this is a substantial number of events, the data do not show that the market for corporate stocks is, at least presently, dominated by orders that are canceled in milliseconds. Rather, it shows a much broader distribution with many orders lasting many seconds.
When compared to the time frames during which human decision making occurs, the cancellation of an order or quote in less than a second appears very fast. But since most market participants use (either directly or indirectly) execution algorithms to fill their buying and selling requirements, it may be more appropriate to compare the speed at which quotes are canceled to the speed at which they are hit by incoming market orders and result in an execution.
As shown in the chart, the distribution of quote lifetimes for quotes that result in trade executions seems to follow a similar pattern as for quotes that are canceled, at least down to 500 microseconds or less. For example, though 38.7% of all canceled orders were in force for half of one second or less, 27.2% of executed trades were the result of some participant hitting posted orders within that same time period. Similarly, though one quarter of all cancellations occur within 50 milliseconds, approximately 20% of all trades occur within that same 50 millisecond window.
Why This is of Interest
Quote lifetime analyses shed light on the overall speed of the market. Monitoring the distribution of quote lifetimes as the markets continue to evolve provides a good indicator of whether or not cancellation speeds are increasing, and if so, by what proportions.
Comparing the distribution of quote lifetimes for quotes that are canceled versus those that result in trade executions reveals the extent to which the market is dominated by quotes that are canceled before they can ever be accessed. At present, the data show that the vast majority of quotes can be accessed by at least some market participants before they are canceled. The data does not show a market that is currently dominated by quotes that are canceled so fast that they cannot be accessed.
It is important to note that the data does not inform on the pros and cons of the absolute speed of the market. But the data show that liquidity-takers have an ability to access quotes approximately as fast as liquidity-providers might cancel them. Slowing the ability of liquidity-providers to cancel their quotes without similarly slowing the ability of liquidity-takers to access those quotes would not necessarily slow the market itself, but could disadvantage those who provide liquidity compared to those who take liquidity.