• 
• 
• 

• Performance
• Services
• Company
• Learn About
• My Attain

The future is now for electronic trading

October 8, 2007

 

The news out of the CME Group last week that over 1 Billion contracts have been traded on the CME's electronic Globex exchange so far this year (with three months to go, no less) was an impressive accomplishment in our eyes, and caused us to go back and re-visit an old newsletter we wrote titled "Will the Pits Last?"

The gist of that article in September 2004 was to ask the question of whether futures trading would go 100% electronic in the US as it is in the rest of the world, and if so - when that transition would finally be complete. (we had been hearing about it happening "really soon" for almost 10 years without it actually happening)

The conclusion was that the move to 100% electronic trading of futures markets was well underway in many markets, while we theorized that we were at least 5 years away from 100% electronic trading in some of the biggest and most successful "pit traded" markets - like S&P 500 futures.

Well, what's changed in the past three years? A lot, as roughly 75% of all US futures trading are now traded electronically versus in the pits. It seems as if the future is finally now for electronic trading.

Markets already there

There are several market sectors which are already close to being 100% electronic. Nearly all interest rate futures (bonds, 10 yr notes, euro dollars, etc.), currency futures (Euro, Yen, Pound, etc) , and energy futures are now traded on the "screen" - as traders call electronic trading - with over 90% of the volume in the interest rate and currency futures electronic, and close to that in the energy markets. This wasn't much of a surprise with the interest rate and currency contracts, which started the switch nearly 5 years ago, but is a big switch for the energy futures, which only started offering side by side trading with the electronic Globex platform in 2006.

How do we measure the percentage traded electronically? The exchanges make that somewhat easy, because they have not merely shut down the trading pits in favor of the electronic pits, but rather set things up so both the "pit traded" and electronically traded forms of the same market trade side by side.

This was done mainly to smooth the transition for those not wanting to give up the pit traded life, but also to enable arbitrage opportunities between the two markets, as well as provide for a back up in case the electronic exchange goes down. The result for the rest of us is an easy comparison between the "pit traded" and electronic.

Markets moving that way quickly:

The traditionally pit traded soft markets such as Cocoa, Sugar, Cotton, and Coffee are making a big move to the screen lately. After ICE (an all electronic exchange out of Atlanta better known for its competing energy contracts) bought the New York Board of Trade about a year ago and gained access to the softs, they launched side by side trading electronically in February of this year. Just 3 months later, ICE reported having 70% of the volume in the soft commodities coming from electronic trading instead of the pits. That's an incredible move - from nearly 0% screen trading to 70% in just a few months, and in our opinion shows just how bad the New York pits (legendary for poor fills, lousy communication, and shadiness) were regarded by the trading community.

This same move to the screen is now happening in one of the last places we expected it to - the usually staid grain markets. These trading pits were long thought of as the ones which would be the last to go electronic, given the relatively older age of the locals who do their trading in those pits (and have been doing so since before there were computers). But that theory has been proven dead wrong, as the Grain markets have jumped significantly to the "screen"in the past two years.

Consider the stats in this table, which show electronic trading in the Corn, Soybeans, and Wheat markets as a percentage of the total volume in those markets. In 2005, there was less than 4% done electronically, while in 2007 we see that over 55% of the volume is now done on the "screen". Seeing the writing on the wall, MF Global (formerly Man and one of the largest futures clearing firms) sent out the following notice last week:

"In observing the total volume of side-by-side trading (electronic execution is now the vast majority of the grain futures volume) and the current issues that the CBOT trading floor is having with their order routing devices, starting Thursday October 4th MTrade will be directing (routing) ALL Grain futures (C, S, SM, BO, W, O, RR) and calendar spreads to the electronic E-cbot platform."

This is just the latest cost savings move by the large clearing firms to push orders away from the trading pits, where they have to employ clerks and runners to take, relay, report, and manage the orders. After the set up, the computers don't cost a thing, and this has to be one of the main reasons the biggest clearing firms (MF Global, etc…) are currently routing orders in bonds, currencies, metals, energies and now grains to the electronic contract. The only access to the pit contracts are to go straight to the brokers over the phone.

Markets still trading in the Pits -

There are a few markets which are holding onto their "pit traded" roots as long as possible. Among them are the meats markets (Hogs & Cattle) and probably the most famous and by far busiest trading pit in the world, the S&P 500 futures pit. The meats remain pit traded for no good reason rather than there hasn't been enough interest to switch them yet. The CME has bigger fish to fry, and the locals in the meat pits seem content.

The S&P 500 futures are a different story, however. The volume in the S&P futures has remained steady for the past few years, and it remains the market of choice for many larger investors to hedge equity exposure or speculate on stock market movement. There is of course the e-mini S&P 500, which is arguably the most successful futures product of all time - likely to surpass 400 Million contracts traded this year - and definitely takes trading volume from the full size S&P.

But unlike markets like the grains and energies, where the decision is whether to trade a market in the pits or on the screen, with S&Ps and eminis - the decision seems to be more about which market you want to trade instead of how you want to trade it. This is likely because of the contract size difference, where it takes 5 emini S&Ps to equal the exposure of just 1 full size S&P. Whether knowingly or not, the CME created a Golden Goose with that sizing - as investors who wish to trade the stock index futures via the "screen" have to do 5 times the volume to get the same exposure. For this reason, I believe larger investors and those not looking to scalp or do arbitrage choose the full size contract - so they don't have to pay 5 times the cost to get the same exposure. That same reason is why I believe the "pit traded" S&P will be around indefinitely. If the CME puts it side by side on the screen, people can do 1 SP instead of 5 minis, while getting the same great liquidity and execution of Globex. That would mean 5 times less business for the CME, something they can't be crazy about.

What does this mean for systematic investors:

The easy answer is nothing. A systematic model doesn't care what market its executing in, and there could be some benefits to the better efficiency and lower cost structure of electronic markets. For those investing with a CTA, a move to all electronic could arguably be a good thing, as they will find it easier to execute.

But, the answer is not all that simply, especially if you're trading a system or designing one. For one, systematic traders who rely on historical market data to base their buy and sell decisions must wrestle with the fact that the electronic bonds and currencies only have 5 years worth of data - as compared to close to twenty years of pit traded data. "Can't you test on the pit traded and execute on the electronic" I hear some of you asking. Of course you can, but while the results may be close to those you tested - they are guaranteed to be different. This small discrepancy can be worrisome for systematic traders who prefer to leave nothing to chance.

Beyond the length of data available, there is also the logistical problem that the electronic markets have different hours (8 AM through 4 PM the next day versus 7:20 AM – 2:00 PM). The lack of a standard beginning and end to the day will require some systems to be re-coded to fit the new schedule and possibly change the market dynamics around the normal market hours.

Lastly, many proponents of open outcry argue that the human element is essential in keeping orderly markets, and that without "locals" in the pit willing to buy into sell offs and sell into rallies, extremely volatile market action could ensue. There is a reason the NYSE curbs program trading once the market is down a certain percent. A human trader may say, "I don't care how much the market is down today, I haven't seen any big banks selling, so I think S&Ps are a buy these levels" In letting that emotional decision into her trading, she has a stabilizing effect on the market. Without that "pit sense", traders watching the screen may base their decisions on the numbers only.

This is a possible reason we have seen poor results for systematic trend followers over the past two years. It's possible that a more "screen based" trading community reacts more quickly to breakouts, end trends by reversing sooner, and so on. Basically - the markets on the "screen" may be more efficient than they were when they were "pit traded", and therefore do a better job of price discovery. If the market is closer to its true value, and reacts more quickly to when its not at its true value - than there is something to the argument that systematic approaches designed to identify and profit from market inefficiencies will be less successful.

How this all plays out remains to be seen, but the so called future where the pits are dead and all the trading is electronic is in all respects here now.