One of the benefits of systematic trading is the ability to do a full and rigorous backtest. Those of us without a programming background may feel very intimidated by the prospect of coding our own systems to be tested, and may also feel that the cost of hiring a programmer to code it for us is too high. Not so…
The familiar saying “Plan Your Trade, Trade Your Plan” can only truly be followed if the plan is reduced to a set of rules. I believe that one of the biggest mistakes an aspiring trader can make is to begin live trading on the strength of a “visual backtest”.
I’ve heard certain FX gurus suggest that we can perform a “backtest” of our strategy by eyeballing a chart and finding places in the price history where our strategy would have worked. WRONG! A true systematic backtest will expose the situations where our system WILL NOT work, and what type of drawdown and/or losing streak would have been experienced. This is far more valuable. Remember that amateurs will focus on the reward, while professionals will focus on the risk!
Visual backtesting is a perfect example of two system development biases.
From Van Tharp:
”
Law-of-small-numbers bias: People tend to see patterns where none exist, and it takes only a few well-chosen examples to convince someone that a pattern has meaning. When you combine this bias with the conservatism bias (below), you have a very dangerous situation.
Conservatism bias: One you believe you have found such a pattern and become convinced that it works (by means of a few well-chosen examples), you will do everything you can to avoid evidence that it does not work.
”
Visual backtesting fits this description to the letter. We find a entry signal or setup which believe should be profitable, look at the price history to find a few good examples of it working, and consciously or subconsciously ignore the periods in the price history where following said strategy would have been disastrous.
NO VISUAL BACKTESTING allowed! Learn to program, or learn to write down your rules and pass them to a competent programmer for rigorous testing. This is a research and development expense, and it is far more cost effective than letting the market show you that your strategy is not a winner by draining your trading account!
Mental Reset
With the Cable Glider mired in an ongoing drawdown, and my former means of professional income now absent, my psychology is at an all time low. I am in need of a mental reboot. The following points attempt to address this issue.
New System
As always, I have been tinkering with variations of the Cable Glider system, and as yet, I’ve found no satisfactory way to revise the stop loss positioning or trade sizing algorithm. I have, however, come up with a new entry signal that materially changes the nature of the Cable Glider system. Previously, the system could be described purely as a channel breakout system. After re-reading the chapter in Van Tharp’s Trade Your Way to Financial Freedom, I discovered a form of volatility breakout entry that works well in conjunction with the channel breakout entrys that were previously used.
New capital allocation
At present, my total capital allocated for trading the FX market is $40,000. I have decided to take $10,000 of that off the table as a living expense buffer, and to immediately withdraw enough profits to cover an additional month of expenses any time the system logs a gain that is twice the average gain. Such gains have occurred 15 times in a 20 month backtest. My current living expense buffer is approximately 6 months. If the buffer falls below 3 months, I will immediately reallocate capital to the buffer to push it back to the 6 month mark. If for any reason I am unable to refresh the buffer, I will look for other sources of income.
Reaffirmation of devotion to pure systematic trading
Within the context of the system’s recent drawdown, I have compounded the problem by manually taking trades that have lost me more money. I hereby reaffirm my committment to pure systematic trading. In order to accomplish this, I have set up separate machines for demo trading and live trading, and I intend to merely monitor the results once per day to ensure that trades were properly entered and exited. No discretionary trading is permitted. The system may succeed, and the system may fail, but damnit, I’m not gonna wreck it.
Recognition of large stochastic element
There is a large degree of randomness in the returns of the Cable Glider system. As an example, I was stopped out of a long position this past Wednesday moments before the GBP/USD skyrocketed nearly 200 points. My stop was penetrated by a mere 2 points. This represents the difference between taking a 9% loss v. taking a 30% profit from the trade. Lest you think I am just complaining, I recognize that there are several times where I come within a pip or two of entering or exiting a trade that would have proven to be unprofitable. For this reason, I believe that my chances of long term success given my current capital level are a coin flip.
New return chart
Everyone likes performance charts, so I’ll be starting a new one dated from June 1 to represent this new phase of my trading journey. For the record, the former version of the Cable Glider returned 73% over an 8 month period of live trading. Given my current level of capitalization, a return like this is now actually required in order for my trading to be self-sufficient. Yikes, I’d better keep my computer programming skills sharp in the meantime. 
Optimism
Every trade taken in accordance with the system’s rules, win or lose, is a good trade.