From the February 01, 2010 issue of Futures Magazine • Subscribe!

For CTAs, more volume doesn't equal liquidity

“Take [Nymex unleaded gas], it is doing more volume than it has ever done, but right now I am looking at the book and you go between $1.8928 and $1.8900,” Thompson says. “If you sold 10 lots based on the depth that you see in the market, you would bring down the market 30 ticks. I see a one lot, a seven lot, a one lot, a one lot, a one lot and a one lot.”

Linn says, “The little trader is deceived by that. They look at these ladders and they think they have some sort of feeling of what is there. [That’s] baloney. You don’t.”

Crow has found it harder to execute size. “If you wanted to do size in the pit, you had more tools at your disposal.” He says locals knew they could take the opposite side and spread it off in another month. “You could lay it off simultaneously [with people spreading] in the pits. [Now], if you want to do size, hit the button and say go get it, it could run 30¢. Before, the spreaders in the pits were there to do some form of differential. You don’t have the mechanisms readily available to you to execute size.”

CTA Stanley Haar says, “It is tough to measure. In theory, in the electronic pits there shouldn’t be slippage. The market is what it is and the orders are all out there and you are not subject to the whims or mercy of the
execution broker.”

But the problem is that traders don’t like to show their hand. “It makes a situation where you have to be more cautious because if you don’t have the volume on the electronic trade, the market will rush into these vacuums and move more substantially than they seemed to back in the old days of the pit,” Haar says.


Managers have gone to some defensive measures like placing iceberg orders. Icebergs show a certain amount of size at a time, but not the whole order. If a manager has 100 to buy, the algorithm will show 10 at a time until the whole 100 are filled. This can be entered as a limit or stop.

It is debatable whether icebergs are a solution or part of the problem. “We use icebergs, but at the end of the day it is hard to tell how much they help or don’t,” says Thompson.

Fall River executes some of its orders themselves and uses brokers for others. “We do have an electronic platform in house and we do execute through FCMs. The people working those trading desks are convinced [that icebergs work]. They are all afraid to show size. Every time I give them size they say ‘I’ll iceberg it, I’ll iceberg it.’ Part of me thinks it is their way to make you think that they are really working for you. At the end of the day I am not sure how much it really helps.”

And Thompson is not the only one. “I would almost prefer not having icebergs,” Linn says. “You should have to show your size. Show it. If you want to move it, move it. The floor to some degree showed its size all of the time. It made for a more fluid market.”

Haar agrees. “The locals, or so it seemed, acted a little more as a buffer. They knew who else had size in the pits. They would work the orders, meaning they would take a little discretion. That seemed to make for a smoother
trading environment.”

While icebergs prevent the rest of the world from knowing your intentions, sometimes that is not a good thing. “I think you mislead a big part of the market that does not understand the price ladders and the depth of market,” Linn says. “You don’t have a clue if you are looking down on 10 numbers — one lots or five lots — you don’t know if there are 50 there or 5,000.”

He says someone could be looking to do similar size a couple ticks away and would be willing to move towards your price, but they don’t see it. They won’t move towards you if they aren’t confident there is sufficient size to get filled. They are afraid of chasing
the market.

Managers also worry that short-term algorithmic trading programs can spot icebergs and take advantage of them. “We know that there are so many other programs and algorithms that people are trying to pick up and identify when there are orders out there and that is why all these programs have icebergs,” Haar says.

He is skeptical that such order types can prevent algorithmic traders from spotting orders.

“There are people out there that are using more sophisticated systems than we are who might be able to identify something like [icebergs], so in general it would be a disadvantage,” he says.

It is a sort of algorithmic arms race, with different types of traders looking for ways to hide orders and others trying to detect them. It turns the price discovery purpose of markets around.

Thompson has explored using TWOP (time weighted average price) and VWOP (volume weighted average price) trades to get away from using stops and protecting against pockets of illiquidity. In the end he is not convinced they are the way to go.

Haar says, “Part of the problem is electronic trading systems operate at a higher speed than the human brain operates, so you have no time to react to anything.”

Linn says, “I found it interesting that one of the big guys told me that they had 40 quant guys working on various programs and implementing them within their trading scheme; 40 mathematicians writing their
programs, many of which were 15-minute programs. The point is, things are changing and you have to adapt to them.”

It is not that managers are pining for the old days. Most wouldn’t go back, even if they could. “I am not under any illusion that the old days were totally better in terms of costs. The pit brokers took their tick or two out of the market to get an order filled so it wasn’t necessarily a lower cost environment,” Haar says. “But it is trickier if you want to do size and if you want to do it quickly you can run into problems in the pure electronic trading. And therefore if you try and push it too much when the volume is not there you can get screwed.”

<< Page 2 of 3 >>
comments powered by Disqus
Check out Futures Magazine - Polls on LockerDome on LockerDome