Update 2022/01/04: Backtest is a desktop Windows app I developed in the late 2000s. A couple of colleagues and I were studying futures day trading and we wanted to test strategies against historical data. We collaborated for a couple of years while experimenting, trading and evolving Backtest.
Some screen shots of my Trading Strategy Backtest application which I’ve been developing for the past couple of years. Two books by Robert Pardo provide some of the basis of my automated trading methodology; 1) Design, Testing, and Optimization of Trading Systems and 2) The Evaluation and Optimization of Trading Strategies.
Here’s what led me to develop it
Tick data are read in from ascii csv files
The Strategy tab is the business end of it. The trading strategies have numerous adjustable parameters. The Backtest engine iterates through all specified parameter values as specified here. The product of the sizes of the parameter sets is the total # of cases to be iterated through. This is a pretty standard way of defining/representing this data set. In this example I am testing a “Stochastics with EMA filter” by Robert Colby.
Results filtering is an optimization of mine. Rather than save the entire result set from the run, we can specify the n best and m worst cases to save where best and worst are the best and worst from specified columns of the result set. The #Threads allows us to take advantage of multi core/CPU architecture. Run time can take anywhere from minutes to days and more. In practice I typically choose strategy parameter sets that generate a few hundred thousand cases which takes 10 or 20 hours to run.
The results from a run are presented in a spreadsheet like window.
Chart shows the price, trades and indicators for the particular iteration.
Pardo presents the correlation between equity curve and “Perfect Profit” as a method of choosing a strategy parameter set for walking forward as a means of guaging the relative robustness of a trading strategy. Perfect Profit is a guage of the maximum possible (theoretically) profit a market can yield for a given time frame. Basically a sum of the price differences between each and every candle for the period. The idea is to choose the strategy that more closely mirrors the market’s ups and downs.
And the Profit Report.
BackTest Profit Report 37994-Prep420080728.1_0-ES-1-1yr-20080725.txt
| BackTest version | 1.1.2 |
| Date/Time of run | 080728 08:37:02 |
| Project | Prep420080728 |
| Instrument file | C:mkmtickdataES-1-1yr-20080725.txt |
| Date Range in file | 2007/07/24 – 2008/07/25 |
| Date Range loaded | 2007/10/19 – 2008/07/25 |
| Date Range of Trading | 2007/10/19 – 2008/07/25 |
| Time Range | 0 – 2400 |
| Analysis | 37994-Prep420080728.1_0-ES-1-1yr-20080725.txt |
| Iteration | 37994 |
Instrument
| Instrument | ES |
| Tick size | 0.25 |
| Tick value | 12.50 |
| Contracts/trade | 1.00 |
| Slippage (ticks per side) | 1.00 |
| Commission ($.$$ per contract round trip) | 5.00 |
| Margin | 100.00 |
| Initial Acct Size | 10000.00 |
Strategy ColbyStochEma
Strategy Parameters
|
###### BackTest Strategy Parameters Begin ###### Strategy: ColbyStochEma Parameter Begin End Increment Begin: 830 0 0 Exit: 1500 0 0 Stop: 0.00 0 0 TrailStop: 0 0 0 Target: n/a 0 0 PerfProfMethod: 0 0 0 RevPos: 1 0 0 MaType: 0 0 0 Ema: 225 0 0 KPeriod: 60 0 0 KSmooth: 5 0 0 UpperBand: 90 0 0 LowerBand: 10 0 0 ###### BackTest Strategy Parameters End ######## |
The Numbers
| Net | Long | Short | |
| Total Net Profit | 15355.00 | 4950.00 | 10405.00 |
| Total Gross Profit | 26517.50 | 10780.00 | 15737.50 |
| Total Gross Loss | -11162.50 | -5830.00 | -5332.50 |
| Total Trades | 79 | 40 | 39 |
| Num targets | 0 | 0 | 0 |
| Num stops | 0 | 0 | 0 |
| Num MOCs | 76 | 38 | 38 |
| Total Pct Profitable | 62.03 | 60.00 | 64.10 |
| Win | 49 | 24 | 25 |
| Loss | 30 | 16 | 14 |
| Largest Win | 1232.50 | 1232.50 | 1207.50 |
| Largest Loss | -892.50 | -892.50 | -880.00 |
| Avg Win | 541.17 | 449.17 | 629.50 |
| Avg Loss | -372.08 | -364.38 | -380.89 |
| Avg Win to Avg Loss | 1.45 | 1.23 | 1.65 |
| Avg Trade | 194.37 | 123.75 | 266.79 |
| Max Con Win | 9 | 2 | 3 |
| Max Con Loss | 3 | 2 | 3 |
| Avg Win Streak | 190.00 | 194.00 | 185.00 |
| Avg Loss Streak | 223.00 | 225.00 | 220.00 |
| Max Drawdown | -1297.50 | -1375.00 | -1287.50 |
| Drawdown Dates | 2008/02/11 13:32 | 2007/11/20 14:59 | 2008/02/07 14:59 |
| Sterling | 11.83 | 3.60 | 8.08 |
| Profit Factor | 4.80 | 1.85 | 2.95 |
| Max Contracts Held | 1.00 | 1.00 | 1.00 |
| Acct Size Reqd | 1297.50 | 1375.00 | 1287.50 |
| Return on Acct | 153.55 | 360.00 | 808.16 |
| PRR | 1.72 | 1.18 | 1.86 |
| CC Eq PP | 0.9843 | 0.0000 | 0.0000 |
Holidays skipped
(2007/11/09) (2007/11/12) (2007/11/21) (2007/11/22) (2007/11/23) (2007/12/24) (2007/12/30) (2008/07/04)
Trades
| Date Time | Misc | Name | Price | Volume | Points | Profit | Total | Total Weekly | Total Monthly | Equity |
| 10000.00 | ||||||||||
| Fri 2007/10/19 11:19:01 | S short entry | 1546.00 | 1 | |||||||
| 2007/10/19 14:59:00 | B MOC | 1523.25 | 1 | 22.75 | 1107.50 | 1107.50 | 1107.50 | 11107.50 | ||
| Mon 2007/10/22 13:04:00 | B long entry | 1524.00 | 1 | |||||||
| 2007/10/22 14:59:00 | S MOC | 1527.50 | 1 | 3.50 | 145.00 | 1252.50 | 11252.50 | |||
| Wed 2007/10/24 11:16:01 | S short entry | 1520.50 | 1 | |||||||
| 2007/10/24 14:59:00 | B MOC | 1537.50 | 1 | -17.00 | -880.00 | 372.50 | 10372.50 | |||
| Fri 2007/10/26 08:51:01 | B long entry | 1547.50 | 1 | |||||||
| 2007/10/26 14:59:00 | S MOC | 1554.75 | 1 | 7.25 | 332.50 | 705.00 | -402.50 | 705.00 | 10705.00 | |
| Thu 2007/11/01 09:45:00 | S short entry | 1549.00 | 1 | |||||||
| 2007/11/01 14:59:00 | B MOC | 1526.75 | 1 | 22.25 | 1082.50 | 1787.50 | 1082.50 | 11787.50 | ||
| Wed 2007/11/07 13:20:00 | S short entry | 1514.00 | 1 | |||||||
| 2007/11/07 14:59:00 | B MOC | 1495.50 | 1 | 18.50 | 895.00 | 2682.50 | 12682.50 | |||
| Thu 2007/11/08 13:02:01 | S short entry | 1482.25 | 1 | |||||||
| 2007/11/08 14:59:00 | B MOC | 1494.00 | 1 | -11.75 | -617.50 | 2065.00 | 277.50 | 12065.00 | ||
| Tue 2007/11/13 11:36:00 | B long entry | 1475.25 | 1 | |||||||
| 2007/11/13 14:59:00 | S MOC | 1500.50 | 1 | 25.25 | 1232.50 | 3297.50 | 13297.50 | |||
| Wed 2007/11/14 08:55:01 | B long entry | 1503.50 | 1 | |||||||
| 2007/11/14 14:59:00 | S MOC | 1487.00 | 1 | -16.50 | -855.00 | 2442.50 | 377.50 | 12442.50 | ||
| Mon 2007/11/19 11:26:01 | S short entry | 1459.50 | 1 | |||||||
| 2007/11/19 14:59:00 | B MOC | 1452.00 | 1 | 7.50 | 345.00 | 2787.50 | 12787.50 | |||
| Mon 2008/07/21 11:42:01 | S short entry | 1261.00 | 1 | |||||||
| 2008/07/21 14:59:00 | B MOC | 1259.50 | 1 | 1.50 | 45.00 | 15355.00 | 45.00 | 2587.50 | 25355.00 |