Automated Trading

by

Coffee trades are swift and the trader’s toolbox is more sophisticated but fundamentals still matter.

By Dan Bolton

Trading at the speed of light is not without peril.

A fat-finger (keyboard input error) initiating the wrong buy or sell order in the coffee commodity market can be disastrous, says commodity manager Steve Hawrylik, but on the whole, the transformation from the open cry pit to the cloud has proven itself.

Hawrylik, commodity manager at NDCP in Duluth, Georgia, illustrated his point with a timeline showing how technology has transformed trading at an accelerated pace, beginning in the early 1970s when NASDAQ became the first all e-exchange.

Twenty years elapsed before E-Trade was founded in 1992. Sell-side retail matured in the late 1990s. By October 2007 one billion CME Globex contracts were traded in a single year. A decade later, on Jan. 12, 2017, CME cleared a record 1 trillion Mexican peso interest rate swaps in a single day (equivalent to $47 billion in notional value).

The influx of cash from the extraordinary increase in trades led directly to the development of sophisticated metrics that, in turn, encourage trading. Online trades are very profitable for the exchanges, he noted. The market cap for Intercontinental Exchange (ICE) has increased 260% since ICE moved away from the floor, he said.

Hawrylik puts a good measure of faith in the remarkably agile and sophisticated technicals that inform market decisions, but he cautioned that coffee traders must also understand basic human nature to avoid blindly following the herd.

“The majority of market participants are reactive, not proactive,” he said. When mistakes are made successful traders act decisively to find a way out. Machines don’t.

ALSO: Changes to ICE Futures Robusta

Hawrylik began his career before automated trading became prevalent. He tells the story of his uncle, a successful trader who painstakingly plotted trends on stacks of graph paper that covered an entire table when unfolded. No one has the arithmetic ability in their head to calculate technicals widely used today.

Charts and technicals are far more accessible and big data is leveraged to facilitate trades. Momentum indicators like RSI (relative strength index) measure when a market is overbought or oversold. Yet tools like the popular 9-day moving average are not obsolete.

“Electronic trading has made the 9-day moving average more effective over the last 10 years,” he said. New rules to limit the ability of one trader from viewing details about competing trade known as “icebergs” are beneficial. Stop/limit market orders automated trade exits and auto-spreading allows more efficient spread execution, he explained. Front-end execution is instantaneous and audited to reduce trade breaks leading some traders to rely on technicals, he says.

One benefit of open-cry trading that is missing today is “the constant sharing of market information and discussion,” he recalls. “News was vetted amongst all the market participants and contrarian trades moved on the spot,” he said.

In contrast, with enormous amounts of information on screen, “things can be more opaque.” IM (internet messaging), phones, and email are all relied on heavily which means rumors spread faster and moves can be more extreme,” he said.

He cited news scrapers as an example of technology that offers huge benefits–but not without serious drawbacks.

Scrapers are a category of software  that acts like a specialized version of Google by continuously scanning the web for the occurrence of terms like “severe” “destroyed” paired with keywords such as “frost” and “Brazil.”

During his presentation, Hawrylik displayed a word cloud with frost-related terms to illustrate what the machines “see” based on June weather reports.

In this instance, transactions were triggered by fears that were unfounded. Temperatures in Brazil remained above 34 degrees Fahrenheit (0 Centigrade).

“They were forgetting the fundamentals,” he said.

Some traders rely almost exclusively on “black box algorithms.” These “rise of the machine” transactions, known as “algo-trades”, can easily exceed the number of trades initiated by retail investors with the ability and experience to gauge economic factors.

They may be slower to act, but “patience in considering a trade rarely comes back to bite you,” he said.

All this points to the fact algorithms are arguably the biggest players in the market today, he said. Event the best algorithms must be monitored and constantly modified, but automation has brought “huge benefits,” he concluded.