One of the annoying things about the debate over HFT, particularly related to the quality of liquidity supplied by HFT traders, is the lack of any historical context. HFT firms are criticized for pulling liquidity suddenly, particularly when volatility ticks up, thereby exacerbating price moves and the impact of order flow on prices.
The thing is, that market makers/liquidity suppliers have been doing this since time immemorial. Liquidity suppliers have always chosen flight over fight.
I’ve written about that in the past (focusing, for instance, on Black Monday). This recent blog post provides some excellent historical color that illustrates the point quite well.
Having traded the markets since 1985, I would like to explain what market liquidity actually meant in the pre-electronic days and what it means in the world of electronic trading.
During the 1987 October crash, Black Monday, I was a junior trader in government bonds and futures. Black Monday was not a crash which was over in seconds or minutes; no, the world stopped turning actually for days. The mini crash of 1989 had a similar pattern. Chernobyl 1986, Gorbachev crisis in 1991 and UK election day in 1992, were all similar liquidity gap events seen by market traders.
The beauty of those days (the late 80s) for a trader, was that markets were very often in a ‘fast market’ condition. This meant that one was able to trade at any price and all rules were thrown overboard until the board officials were able to control the pit again. Without computers, we simply could not deal with the enormous amount of activity in the markets at those times and simply stopped (or delayed) sending price information out, not just for seconds, but minutes and even (during the week of Black Monday) days. Clearers had backlogs to clients for days.
What investors, customers, institutions looked at in those days were screen prices on systems like Reuters or Bloomberg: feeds fed by the voice of the official. If the official did not yell a price in the mic, the data group would not enter a price, and the world would never see this price. Under fast markets no prices were entered and the screen would just read ‘fast’ (if you were lucky). Future and option prices on screen were seen as tradeable prices and in slow markets they may have been pretty close to the truth. In fast markets they were inaccurate and it was not possible to trade the price indicated on the Reuters screen. Yet the world still thought this was the correct price.
Being in the pit, one cannot keep buying or selling an instrument, so traders would hedge their positions. Hundreds of times I have been in situations where there was simply no bid or offer in the instrument to hedge in. The signal from a broker out of the futures pit simply stated, ‘no bid, no bid’, yet all the screens where showing a bid or an offer. Was this liquidity? The world clearly thought it was, but we all knew there was panic on the floor and you were lucky to trade one lot.
Yes, I know Mr. Spanbroek is talking his book. (The EPTA represents HFT firms.) But what he writes about the good old days on the floor are accurate. They were good days mainly for the guys on the floor, who would supply liquidity when it was profitable, but would head for the exits when the order flow was toxic. Ditto traders upstairs in the OTC markets. This is a characteristic of liquidity supply and liquidity suppliers pretty much everywhere and always.
This subject always brings to mind a legendary-and I mean legendary-trader, who told me in the early-00s he was all in favor of the markets going electronic because he was “sick of getting raped by the floor.” About 2 years ago he told me that he couldn’t compete with HFT. I guess there was a Goldilocks “just right” era in between, say 2002-2007 or so, but the criticisms of liquidity suppliers is a hardy perennial. And there is some justice to the charges. The point is that it is not new, and it is not unique to HFT. It is inherent in the economics of liquidity supply.
And there is no easy policy solution. A policy intended to fix one problem will just create others. Again, the problems inhere in the economics of liquidity supply.