Developing a System

I am revisiting an interview I did with LTC Ken Long, a systems expert with the U.S. Army.  Here’s what Ken said about developing a system:

Define Who You  Are:  “Before you conduct any planning or system design, you must have a thorough understanding of who you are and what your objectives are.  Individual investors, private hedge fund managers, public mutual fund managers, and trust managers will have different dynamics, time frames, and risk profiles. This relates to system design in that the final product must fit the circumstances and dynamics of the group or individual.  If you jump into system design without considering these basics, you will sow the seeds of future problems.”

Objectives:  “In trading system design, the problem is to define what you want the system to accomplish.  With as many ideas, events, circumstances and adjustments that occur in system development, you have to have your objectives crystal clear in your mind.  If you don’t know where you are going, then any old road will do.” 

“Objectives give you the basis for making choices and prioritizing actions.  This is not to say that objectives are static.  In fact, they can change as you discover either unexpected limitations or advantages in your system as it matures.  But before you start you must have an initial set of goals and objectives to guide you.” 

Calibration:  “After the system is deployed and operational, part of the process of calibrating the system is checking to see if the objectives still fit the person or organization that you have become.  That’s a very exciting part of system design.  I can’t tell you how often I’ve been part of a design team that started with a limited set of objectives and discovered in the “imagineering” phase that by adjusting our sights we were able to accomplish far more for much less. But, you have to start somewhere.  If you don’t start with objectives, you are spinning your wheels.”

I posed this question to Ken:  “This section is critical.   How will you know if your system is working or not?  What are your performance benchmarks?  What are your criteria for knowing that your system is not working?  How will you make decisions when these criteria are met?  Will you scrap everything or just make position sizing adjustments?"  All of these questions are critical to developing and operating a good trading system.

How to Make Decisions Within the System

Here’s what Ken said about this critical topic:

“If you don’t work out how you will make decisions ahead of time, then you will certainly have to sort it out at the time of the first difficult decision.  If you make decisions on the spot, with no guidelines, you have two problems:  1) figuring out what to do and 2) how to do it.  And these problems must be faced under great stress and limited time.  It’s better to calmly sort out the decision making process ahead of time so that the decision mechanism is agreed to before hand.” 

“In the Army, no plan usually survives the first contact with the enemy, and so our goal in planning is to develop a range of alternatives that can apply to a number of scenarios.  Through rehearsal and analysis, we know which strategy works best for a given set of conditions.  The goal of strategy development is to provide the decision maker with a menu of choices that are robust enough to cover a wide range of contingencies.” 

“In general system development then, we look for robust, simple plans that can cover a wide range of conditions.  When you preplan like this, you don’t try to force the world to adapt to your plan.  If you fall in love with a strategy and become emotionally invested in making it work no matter what the market or the world says, you lose the ability to adapt and learn.”

“A real world example for a trading system might be a trader who decides to check his actual trading performance every month against the calculated system expectancy, and determine the statistical significance of the variation.  He might decide that any result greater than one or two standard deviations is a signal to stop trading and recalibrate the system or reconfirm the validity of the trading model and its underlying assumptions.  If the actual expectancy is close to the predicted expectancy, then the trader knows he’s on target.  In modern manufacturing systems this concept is called “Statistical Process Control.”

“It lets the system controller know when the production machines are drifting out of tolerance and degrading the quality of the output to the point where the line is stopped and the machines are retooled.”

I asked Ken about how his advice applies in view of the fact that many trading systems are automated.  Here’s how he responded: 

“It’s a general problem of the information age, which provides us with a wide range of automated decision support systems that can compile massive amounts of data, analyze and process it, and present us with decision packages for action based on criteria that we can specify.  I use a lot of these.  However, the key to making them work is to make sure that you understand the underlying business model and system logic.  When you do things automatically by computer, you need to understand what the computer is calculating and filtering.  I won’t use power tools until I know how they work and I have mastered their use in simulations.”

“If you have done all the preparation work that you outlined in your system design workshop,⁴ and you have chosen indicators that provide you the right signals for making your trading decisions, then the right thing to do is to rely on the signals to make your decisions.  Periodic calibration of the system, however, is still necessary to confirm that you have chosen the correct signals and that your actions are correct.  If you have not done that work though, it may be the case that you simply picked up the latest hot indicator and are using it regardless of how appropriate it may be for your trading system.  If it fails to work as advertised, you are likely to dump it for the next hot idea that comes along.  Then you’re not a system’s trader, you are only reacting to advertising.”

1.      We have two newsletter back issues in which we interviewed Tom Basso for those of you who would like to know more.  Call 919-466-0043 for more information.

2.      William O’Neil,  How to Make Money In Stocks.   New York : McGraw-Hill, 1987.

3.      We have an audio program on business planning for traders that takes you through the development of a business plan. 

4.      The workshop Ken is referring to is the, How to Develop a Winning Trading System That Fits You workshop, which we offer once or twice each year.

About the Author: Trading Coach Dr. Van K Tharp, is widely recognized for his best-selling book Trade Your Way to Financial Freedom and his classic Peak Performance Home Study Course for traders and investors. Visit him at www.iitm.com for a FREE trading game or to sign up for his FREE weekly newsletter.