“I thought you might be interested in this article. What is GS doing? Maybe HFT is on the way out!”
Many thanks for a wonderful article.
I think HFT front-running should be on the way out, which would be a very good thing and about time! HFT is too often a misnomer. I wrote on Monday that it would be more accurately described as HFFR, for high-frequency front-running.
There will always be other forms of HFT because computers are not going to go away. However, front-running was far and away the most profitable aspect of HFT, being virtually risk free because it had nothing to do with the uncertainties of actual trading. Instead, clever algorithms in the fastest computers were seeing everyone else’s buy and sell orders, nanoseconds before they were executed, and jumping in front of them. A few moments later they would sell back at a higher price when buyers were required to raise their bids. The same process worked in reverse in down markets when people were trying to sell. HFT firms were also fishing by placing thousands of bids and offers to entice institutional activity. Once large buy or sell orders appeared, they would fill only a portion of it before raising offers or lowering bids.
This was leagalised stealing from other investors, on a very small basis for each individual trade but cumulatively worth billions to those with the fastest computers because the process was so pervasive. It also exacerbated the number of sudden, sharp short-term moves, further increasing hazzards for conventional traders.
However, as the number of HFT firms increased the competition inevitably became more intense, and the process of continually upgrading the speed of access to markets more expensive. Profits were falling as firms canibalised each other. Goldman Sachs is presumably hoping it made a just-in-time exit and will avoid most of the wrath of regulators who are finally waking up. Those within the financial industry who are defending HFT front-runners have lost their moral compasses.Back to top