Quantifying the High-Frequency Trading “Arms Race”

Abstract We use stock exchange message data to quantify the negative aspect of high-frequency trading, known as “latency arbitrage.” The key difference between and widely familiar limit order book is that contain attempts trade or cancel fail. This allows researcher observe both winners losers in a race, whereas you cannot see losers, so directly races. find latency arbitrage races are very frequent (about one per minute symbol for FTSE 100 stocks), extremely fast (the modal race lasts 5–10 millionths second), account remarkably large portion overall trading volume 20%). Race participation concentrated, with top six firms accounting over 80% all wins losses. average worth just small amount half price tick), but because volumes stakes add up. Our main estimates suggest constitute roughly one-third impact effective spread (key microstructure measures cost liquidity), imposes 0.5 basis point tax on market designs eliminate would reduce market’s liquidity by 17%, total sums at stake $5 billion year global equity markets alone.