Yes, everyone — bar the old open outcry pit traders — loves electronic trading in commodities.
Ahead of this week’s CFTC hearings on position limits and speculator influence on prices, however, have the commodity regulators perhaps forgotten to question the obvious? That is, the influence of electronic trading on commodity prices. After all, it has only been a few years since ICE and Globex screens revolutionised the way commodities are traded. Coincidentally, it has been in that time that the so-called price “anomalies” have begun to manifest themselves.
Let’s consider the following: ICE Futures closed open outcry on the International Petroleum Exchange in favour of electronic trading in April 2005. Nymex introduced electronic trading in energy futures on June 13th, 2006. ICE went completely electronic in March 2008. Here’s a 10-year chart of the WTI with the respective dates pointed out:
Now, the idea that electronic trading can be anything but a good thing is not a notion the industry will take to favourably.
Among one of the main ‘positive’ side-effects has been the elimination of “curve shavers”. The electronic screens provide an immediate calculation of where contracts further down the curve should be priced, irrespective of liquidity, quotes and trade. In the days of floor trading, these would be derived from implied calculations. Specialists who were able to generate these quotes quickly and more accurately had an immediate arbitrage opportunity. These specialists became known as curve shavers.
Incidentally, we are told, among the biggest practitioners of curve shaving were, among others, J.Aron – the commodities arm of Goldman Sachs.
Going back to electronic trading, however, it’s worth considering the way the practice has changed commodities trading more directly. Looking at the ICE screens, which traders particularly seem to love, two immediate effects become clear.
Firstly, there’s the 24-hour nature of the market. As the recent PVM incident proved, traders are capable of processing large orders at any point in a 24-hour cycle, irrespective of how much liquidity there is, there always appears to be enough to get the trade done. Meanwhile, the screens themselves provide complete anonymity as to who your prospective counterparty is.
Lastly, and most importantly, there is the algorithmic tools that exchanges like ICE provide traders with. Note the following description of ICEMaker, available to all platform holders:
ICEMaker® is an innovative tool available to all traders on the ICE platform. ICEMaker allows traders to link their proprietary front-end trading strategies in Excel to manage orders on the ICE platform without the need to write complex API code.
ICEMaker enables traders to efficiently manage multiple simultaneous orders and complex spread relationships using Excel formulas to integrate real-time data from ICE and 3rd-party data feeds within the browser-based WebICE.
ICEMaker, gives users the power to: * Manage your order book and execute orders from Excel automatically or manually * Use proprietary formulas to derive bid and offer prices * Derive ICE prices based on other external data points * Simultaneously manage multiple orders * Build Excel formulas directly within WebICE * Use sophisticated trading tools on ICE such as “Hold on Hit” function
In effect, among the features is also ICE’s own automated tool for “ice-berging” orders. As Investopedia explains:
What Does Iceberg Order Mean? A large single order that has been divided into smaller lots, usually by the use of an automated program, for the purpose of hiding the actual order quantity.
These tools are actively used by traders, especially those who are physical intermediary players and general position takers backed by physical volumes. To our knowledge these parties are not largely invested in developing high frequency trading programmes of their own, but rather depend on the automated tools provided by ICE and the like.
As written about here and here, a big debate has grown around the impact of high frequency trading programmes on equities. Among the issues is the ability for superior proprietary programmes to efficiently detect other “ice-berg orders” and via that, dark pools of liquidity. There are also algorithms focused on detecting buy and sell limits. With such knowledge, what traders might deem to be hidden orders and limits are not necessarily as hidden as they might think.
Meanwhile, there is another issue to consider. Most physical intermediaries are focused on hedging via the energy markets. They’re not in the business of trading oil equities and interest-rate products. That’s not necessarily the case for hedge funds and investment banks active in the commodities space. With clever algorithms at their disposal there’s no reason why they can’t outwit the physical intermediaries with what they know more about, especially with a latency advantage. Areas they know more about, that is, like equities and bonds.
This becomes increasingly tempting with so many commodity ETFs trading on “high frequency” friendly platforms like the NYSE Arca. The key here is detecting the arbs between the underlying price of the commodities and the actual trading price of the units. Then there’s the arbitrage with the interest-rate and Treasury markets. In the days of floor trading, simultaneously buying oil and selling Treasuries if a mis-pricing in inflation expectations opened up would be possible but not half as fluid. Algorithms can completely automate that process.
Is it any surprise then that oil is showing such an increased correlation with the dollar and equities?
Related links:
High frequency trading in Europe – FT Alphaville
A forward curve proposition - FT Alphaville
Oil, the great inflation hedge – FT Alphaville
So who says there’s no oil/dollar correlation? – FT Alphaville
Correlated returns – FT Alphaville