Defender Capital Management, Inc.

From MarketsWiki
Jump to: navigation, search
Not impressed? Tell us how to improve it or sign up to edit.
Occ-logo-2.gif
Defender's corporate logo, 'Troutman'... both an inside joke for the company, and a reminder that you don't have to be stuffy to take investing seriously.
Defender Capital Management, Inc. was a corporation founded in the 1990s to provide professional futures trading services. The experiment in providing large-fund risk models to ultra-small accounts brought both success and failure during the company’s six-year existence.

Startup

Defender was founded in 1997 by Jonathan Matte. The company first established itself as an Introducing Broker, offering its services in Usenet discussion groups (back in the days before web-based chat forums) in a very low-key way. Matte believed strongly at the time that high-pressure sales techniques were offensive and that a business could be successfully grown with a more passive approach to marketing.

With retail electronic futures trading just on the horizon and with discount high-volume brokerages becoming the norm, the company faced unlikely odds as a small, full-service brokerage. Defender reduced its rates to stay competitive with larger full-service outfits and offered advice-free discount trading as well. Nevertheless, it became clear after a year that the trend toward drastically cheaper trading would accelerate, not fade away, and that personal futures trading via the internet would become a common reality.

Faced with the prospect of an unacceptable profit margin on transactions and an unwillingness to become a part of a larger and more aggressive brokerage, Defender shifted business direction and registered as a Commodity Trading Advisor in order to manage trading accounts.

CTA Operation and Trading Programs

Defender's first trading program opened in the second half of 1998, buying and selling 30-year Treasury Bond futures for sub-$25K accounts. That first foray into trading was an unqualified disaster, with starting accounts losing approximately half of their value in less than a year. The risk controls in the system were inadequate, and early substantial gains were sharply eliminated in a volatile environment.

In the face of heavy losses, Defender quickly shut down trading and returned to research for a few months. The result was the launch in early 1999 of a second, much-improved mechanical trading program, again for 30-year Treasury Bond futures. The second program was volatile in its results, but risk was far better controlled, and the program returned nearly 100% across its three-year run.

With the three-year anniversary of its trading program approaching, two anomalous events created problems. In September 2001, the World Trade Center was demolished in New York, throwing the investment community into a panic. Defender's performance through this period was unaffected; as a short-term trader, the company simply stayed out of the market through the uncertainty. Nevertheless, the company was hurt by substantial redemptions in funds as some clients feared a greater calamity to come.

On Halloween 2001, appropriately enough, the US government issued a statement that they would no longer issue the 30-year Treasury Bond, the one commodity that Defender traded. Though the company did not suffer major losses, the announcement prompted a multi-month explosion in volatility. The risk controls in Defender's trading program shut down trading for so many months that eventually the program was simply retired.

In 2002, Defender launched two new programs based on the original model but with additions and improvements designed to make them more adaptable to different market conditions. The first program, Maple Sugar, traded a combination of Canadian Dollar and Sugar #11 futures; the second, Big ED, traded only in the Eurodollar futures market in a nod to the old T-Bond program.

The company hired an employee to take over marketing, web site development, and create relationships with brokerages. By the end of 2003, however, personal problems between the owners spread rapidly to Defender's operations, and in 2004 the decision was made to simply shut down the company.

The intellectual property, including the trading system rules and the method used to develop them, was sold to John J. Lothian & Company, Inc., at that time a newly-established CTA. Matte continues to provide trading support and research to JJL as part of that sales agreement.

Trading Style

Defender's trading style developed from a simple and absolute principle: No one can reliably determine ahead of time what any market will do. Every trading decision, no matter how carefully made, is just a guess. This runs contrary to the beliefs of many (and possibly most) in the financial industry, who often expend enormous effort trying to divine the future and rationalize their failures when that effort fails them.

Defender therefore created a trading method that did not require being right all that often. If selected trades were unusually accurate, then extra profit was returned. If the trading guesses were worse than desired, the losses did not immediately bury the trading accounts. When the market handed out months of "win some, lose some", the system returned very respectable numbers.

Though the method was trend-following in nature, it incorporated loss-management rules more typically found in short-term trading systems. Initial stops were tighter than usual for trend followers. Defender's reasoning was that if a direction were chosen and initially proven wrong, there wasn't much point in holding the bad trade in case it got better in a few weeks.

Once in a trend, the method would widen and trail stops, but depending on market motion the stops could halt their trailing, widen or trail more closely.

Typically, the method was a simple reversal system, following the market long and short. However, risk controls on volatility were also used to guard against taking positions that were toxic to the trading account (reflecting the hard lessons that were learned in Defender's first trading program).