Often times when comparing trading system performance, the natural inclination, is to compare returns. If you have two trading systems that traded a $100,000 account, and made a 30% return, and produced a 15% drawdown, and had the same length of drawdown, they would seem pretty much the same, right? I would say the answer to that is “not always”.

What if one trading system only used on average of $30,000 in margin while the other used an average of $70,000? Theoretically you could have funded the first account with less money and invested the remaining dollars elsewhere. As a result, I am suggesting that you consider viewing returns as based of off the account size required as opposed to an arbitrary number. This could allow you to specify an amount by which to fund an account. In other words, you may trade the account as though it were a 100k account, but only put up $30,000 in trading funds because that is all you think you need. Many trading systems, money managers and CTA’s allow for such a notional funding approach. It can be an efficient use of capital.

The commercial testing software Trading Recipes has an intriguing way of computing these numbers. It can run a worse-case analysis. What this does is looks at every possible start date over time (for example 10 years). If there were 1000 different trades you could have started trading from, then it (Trading Recipes) tests the trading system from every single one of those 1000 trades. It then sums the worst drawdown and the required margin starting from each one of those 1000 trades, and it goes on to show how the system performed over the next 12 months. This allows you to create a frequency distribution of yearly returns verses account size required.

For the sake of this example, I am going to test 4 different trading systems and compare the results. The portfolio is 15 diversified markets that are all reasonably high in liquidity. Also, I have approximated margin based on 2 times the 5 day average true range multiplied by the point value, and then took the average of that over a period of 5 years. I’ve done it this way because margins change quite dramatically over time. Changing margins are typically caused by changes in a market’s underlying volatility. Computing this way will cause the margin to increase during higher volatility periods, and decrease during lower volatility periods. This risk adjusted method is similar to real life margin requirements. I think this is far more robust than to use the current margin values because past margin were likely different. Even though these may not be the exact margin amounts the above formula does seem to be close to the current margin levels in many commodities. You could always use a higher multiple if you wanted a greater “cushion”. This is just for demonstration purposes; decide the best way for you to simulate past and future margins in your testing. It would be terrific if data vendors like CSI sold data files for trading systems with the exact exchange minimum margin requirements throughout history, although I’ve not seen it.

Trading Systems Comparisons

For all tests:

Period tested was: 1/1/90 through 12/31/2003
Data: CSI backadjusted contracts
Slippage and commissions: $75
Starting Capital: $100,000

Money Management: Risk 2% of equity a trade or one contract if risk is less than $3000 (whichever is greater).

15 market portfolio: Euro Currency, Corn, Kansas City Wheat, Cotton, Sugar, Coffee, Crude Oil, Natural Gas, Japanese Yen, Swiss Franc, Five Year Notes, Thirty Year Bonds, Nikkei Index, London Nickel and London Copper

In the first test, we will use a Channel Breakout System similar to the “Turtle” method of trading. Specifically, this system buys at the highest price of the last 20 days and sells at the lowest price of the last 20 days. It then exits at the lowest price of the last 10 days for long positions and the highest price of the last 10 days for short positions. Risk computations are as a multiple of average true range, and protective stops get placed at those same levels.

Channel Breakout (20/10)

Starting periods available to test since 1990: 2080
Average required account size: $62,026.00**
Average first year profit: $39,086
Ratio of average account size required to average first year profit: 0.63

In this next test we use the same exact rules as above except the input values change to 50 and 20 (From 20 and 10)

Channel Breakout (50/20)
Starting periods available to test since 1990: 1017
Average required account size: $35,009.00**
Average first year profit: $52,341.00
Ratio of average account size required to average first year profit: 1.49

Aberration Trading System:
Starting periods available to test since 1990: 472
Average required account size: $12,502.00**
Average first year profit: $23,148.00
Ratio of average account size required to average first year profit: 1.85

Checkmate Trading System
Starting periods available to test since 1990: 551
Average required account Size: $15,922.00**
Average first year profit: $39,659.00
Ratio of average account size required to average first year profit: 2.49

Synergy Trading System
Starting periods available to test since 1990: 536
Average required account size: $17,358.00**
Average first year profit: $52,196
Ratio of average account size required to average first year profit: 3.00

**(Margins were approximated, they could be significantly higher or lower)

Trading Systems Results Summary

Here, you can see an intriguing phenomenon. The average first year profits for the Channel Breakout (20/10) were almost the same as Checkmate. Yet the average required account size for Checkmate was less than half. Similarly, the average first year profits for Channel Breakout (50/20) were almost the same as Synergy. Yet the average required account size for Synergy was again about half.

This can be valuable information for somebody who wants a notionally fund an account. If nothing else, it is an eye-opening perspective of how two trading systems can produce almost the same profit in a year given the same account size and money management, yet one of those trading systems can boast a much lower historically required account size. A reversal system or a short-term system like the (20/10) Channel Breakout will likely have higher requirements because of a greater number of simultaneous trades.

We have done these tests on many other trading systems. If you would like to see those reports, or if you would like the complete spreadsheet reports used in generating these tables please contact us.

www.RelativityTradingSystem.com

Dean Hoffman

CFTC REQUIRED RISK DISCLOSURE

HYPOTHETICAL PERFORMANCE RESULTS HAVE MANY INHERENT LIMITATIONS, SOME OF WHICH ARE DESCRIBED BELOW. NO REPRESENTATION IS BEING MADE THAT ANY ACCOUNT WILL OR IS LIKELY TO ACHIEVE PROFITS OR LOSSES SIMILAR TO THOSE SHOWN. IN FACT, THERE ARE FREQUENTLY SHARP DIFFERENCES BETWEEN HYPOTHETICAL PERFORMANCE RESULTS AND THE ACTUAL RESULTS SUBSEQUENTLY ACHIEVED BY ANY PARTICULAR TRADING PROGRAM.

ONE OF THE LIMITATIONS OF HYPOTHETICAL PERFORMANCE RESULTS IS THAT THEY ARE GENERALLY PREPARED WITH THE BENEFIT OF HINDSIGHT. IN ADDITION, HYPOTHETICAL TRADING DOES NOT INVOLVE FINANCIAL RISK, AND NO HYPOTHETICAL TRADING RECORD CAN COMPLETELY ACCOUNT FOR THE IMPACT OF FINANCIAL RISK IN ACTUAL TRADING. FOR EXAMPLE, THE ABILITY TO WITHSTAND LOSSES OR TO ADHERE TO A PARTICULAR TRADING PROGRAM IN SPITE OF TRADING LOSSES ARE MATERIAL POINTS WHICH CAN ALSO ADVERSELY AFFECT ACTUAL TRADING RESULTS. THERE ARE NUMEROUS OTHER FACTORS RELATED TO THE MARKETS IN GENERAL OR TO THE IMPLEMENTATION OF ANY SPECIFIC TRADING PROGRAM WHICH CANNOT BE FULLY ACCOUNTED FOR IN THE PREPARATION OF HYPOTHETICAL PERFORMANCE RESULTS AND ALL OF WHICH CAN ADVERSELY AFFECT ACTUAL TRADING RESULTS.

DH Trading Systems