The Striker Report - gold standard for objective trading research

Mike Barna

Founder of Aaron Asset Management, LLC

Program(s) Developed: Mesa software and Alfaranda systems

Interviewed by John F. Gallwas, Founder of Striker Securities, Inc.

An interview with Michael L. Barna, who is associated with system vendors MESA Software and Alfaranda Systems. In addition to being a captain for a major airline and a nationally ranked handball player, Mr. Barna is also the founder of Aaron Asset Management LLC (AAM), a registered Commodity Trading Advisor (CTA).

John Gallwas: How does a graduate of Arizona State in Mathematics, with a M.S. from Stanford in Astronautical and Aeronautical Engineering, who was working for space defense contractors in California's Silicon Valley, get involved with the futures markets as a system developer, and most recently an emerging CTA?

Mike Barna: Before college and throughout undergraduate and graduate school I maintained my interest in the financial markets while pursuing my studies in science, mathematics and engineering. When the Intel 386 chip came out I became even more involved with technical analysis because of the demand for programmers that could write code for statistical research into financial markets.

Missile and spacecraft guidance and control systems are similar in many ways to trading systems, so the career switch from defense systems, as the Soviet threat abated, to developing trading systems for the financial sector, was a natural progression. My first commercial system RMESA-2 was released in 1996 and was rapidly accepted as a standard day trading system for many traders and brokerage firms.

John Gallwas: Tell our readers the unique features of your trading systems, and how they are designed to provide the same basic function as an autopilot on the planes you fly.

Mike Barna: Large jet aircraft have autopilots for several basic reasons, but mainly for safety and efficiency. Using an autopilot to perform basic functions of an aircraft allows the pilots to become, as they drill into our heads during training, "master monitors". In the trading systems world, the system becomes the autopilot and the trader the pilot. Trading systems are programmed to react to certain market information to make trading decisions in a non-emotional and consistent manner, which offers an increased level of safety to the trader, and enhances his capabilities to make profits.

Through MESA Software and Alfaranda Systems, I provide more top rated trading systems listed in Futures Truth Magazine ™, than any other developer in the country. When you enter a trade using a researched system you do so believing that the payoff edge is on your side. You will have drawdowns and strings of losses, but over time, with a robust system, you will be profitable.

John Gallwas: Explain how diversification works in your Trading Program I, offered through your managed account program, which uses individual "Market-Model" combinations to smooth performance.

Mike Barna: Program I is a portfolio-based product providing a mixture of markets and trading systems, termed "Market-Models". I have found that it is not enough to have one system trading ten different markets, or ten different systems trading one market. You need multiple systems and multiple markets to really reduce equity curve correlation. Proprietary systems, commercial systems and enhanced systems from MESA and Alfaranda provide ample choices for trading systems to be used within my managed program, Aaron Asset Management. Taking dozens of different systems and applying them to 60 markets or so produces hundreds of top Market-Model combinations. Running portfolio simulations will then produce efficient frontiers of mixtures of Market-Models that exhibit superior performance. Only the best systems are used. A weak performer is not included in to the portfolio just because it "smoothes out the equity curve". Curve fitting portfolios is just as possible as curve fitting individual systems, and a Market-Model must stand on its own as a viable combination or it is not included at all. Diversification thus occurs in FOUR dimensions: Time frames (day trading to daily systems), markets chosen, systems chosen and the portfolio assembly schemes used. Larger portfolios will allow advanced portfolio allocation schemes to be implemented that go well beyond "Markowitz" allocation, in that the portfolio is dynamically adapted in a fashion similar to the way a trading system enters and exits the market. Typical portfolio allocation schemes assume that return variance is a constant throughout the equity curve and we know that not to be true.

John Gallwas: Why are the systems for your AAM Program II, which incorporates active stock index trading, different from most other day-trading systems you have developed since 1997?

Mike Barna: Day trading has been and continues to be popular due to the elimination of overnight uncertainty and the fact that margin requirements are so much lower if positions are not held overnight. Brokers and clients repeatedly asked me to provide a day trading alternative within my managed product, over and above RMESA5. As a result, I introduced Program II, which uses RMESA-8 as its primary trading system. RMESA-8, only available in my managed account program, uses several different intraday patterns, plus some found in RMESA5, which have been down scaled to operate on a lower time frame, all of which tests well back to 1985 (approximately 20 years of data). Very few other developers can or will do a 20-year back test of their day trading system.

John Gallwas: Do you have any projects in the "what's new" department that you can share with us at this time?

Mike Barna: This industry has been "static" for about 10 years in that the only enhancements have been in nicer charts, graphics, increased processor speed, and some improvements in trading systems and indicators, etc. We still develop analytical tools, namely systems and indicators, the same old way. Small "evolutionary" improvements have come around, but "revolutionary" enhancements have not been developed. In other words, traders and developers still use essentially the same old technology to trade the markets or develop trading systems as were used 10 years ago. We as developers use Genetic Algorithms or search engines to help us determine the best indicator or pattern to use in our trading system. Yet with all this power available in the Artificial Intelligence (AI) arena applied to the financial markets, there are few if any AI based trading systems populating the top rankings of Futures Truth™.

For years I have been using various elements of what is now called an "Automatic System Generator" to help me with system development. Recent advancements with this generator, that has been accomplished by myself and others, have created what I think is a true "breakthrough" in the way systems are developed and maintained. Genetic Algorithm search routines help in the system development, but they have significant limitations. My latest "alpha" product engines are self-producing 100,000 trading systems per hour, out of nearly a vacuum. The new alpha engine is showing promise of producing results that far exceed the performance of the best existing systems I have now. However, my work is not complete, yet. I should say that major program enhancements would be available only through Aaron Asset Management, LLC.

John Gallwas: What should an investor who is interested in "alternative investments" but has never traded futures know about the pros and cons of futures system trading?

Mike Barna: There are two critical areas that I look for in futures trading systems that a newcomer to trading systems should know:

Back testing: Testing a day trading system over 5 or 10 years of data is not enough. For example, the S&P market is trading now like it did in 1992-1994, not like it did in 1996-1998. If you are wondering why index day trading systems are in a drawdown now, take a look at the returns for system developers during 1992-1994. Exhaustive testing is very important.
Number of trades: Anyone can curve fit market data using parameters, logic or portfolio combinations. If someone asked you to determine if a coin was loaded, would you flip it only 10 times or would you flip it hundreds of times to be sure you simply are not having just a long string of heads or tails?

This interview is for informational purposes only and is not intended to be a solicitation of any kind. Trade only with risk capital. The risk of trading can be substantial and each investor and/or trader must consider whether trading systems are a suitable investment.