Recently, however, we have come up with a system that will allow you to monitor key stats for a single backtestable system over a large basket of securities. Not quite true portfolio backtesting, but a significant improvement over the status quo.
To do this, we took the existing trading systems that we offer in our product catalog and rewrote them to include the inline calculation of a number of performance statistics. These performance stats, which are a small but important subset of those provided in the eSignal backtester, are then displayed in the cursor window when the system is loaded into a chart.
More importantly, if the system is loaded into a watchlist then the current performance stats are displayed for each symbol in the watchlist. You can then sort on any of the stats and isolate those symbols showing good performance for further investigation. Normally, to accomplish this kind of thing, you would need to pull up a chart for each symbol, run a standard eSignal backtest, and then either export or print the results for further analysis. With this method you can simply load the system into a watchlist populated with all of the symbols that you care to track and instantly see the performance results.
Obviously to perform a meaningful backtest a large number of bars will need to be available in both the chart and the watchlist, so be sure to apply a Time Template that preloads enough bars. You will need to apply this Time Template to both the chart and the watchlist. See our Time Template article for more information on that topic.
The inline performance statistics provided are:
| Stat | Description |
|---|---|
| Trades | Total number of trades in the sequence. |
| Equity | Current equity. Each system allows you to specify a starting account size as well as the risk-per-trade, and a dollar value per one-point move. |
| Winners | The total number of winning trades. |
| Losers | The total number of losing trades. |
| WinRatio | The ratio of winners to total trades expressed as a percentage. |
| AvgWin | The average winning trade expressed in dollars. |
| AvgLoss | The average losing trade expressed in dollars. |
| AvgTrade | The average trade expressed in dollars. |
| ProfitFactor | Gross profit divided by gross loss. |
| Expectancy | The expectancy ratio, calculated as ((AvgWin/AvgLoss)*WinRatio)-(1.0-WinRatio) |
| RiskRuin | The risk of ruin is a probability (between 0 and 1) that the equity will be depleted to the point where you are no longer able to trade the system. |
| Drawdown | This is the close-to-close drawdown which displays the loss from the most recent high equity point. |
| Bars | The total number of bars considered in the inline backtest. |
| InMarket | Percent of the time (bars) that the system was in the market. |
| PointValue | The dollar value of a 1-point move in the security being tested. |
This "inline-stats" logic has been integrated into a number of the backtestable trading systems that we currently offer on our site:
- Bollinger Bounce System
- Bollinger Breakout System
- CCI Crossover System
- Hilbert Sine System
- MA Crossover System
- MSRM System
- Opening Range System
- RMV System
- Stochastic System
- RSI System
- StoCyberCycle System
To view more information on any of these systems, just go to our Products page and select Systems from the Category view.
If you have an existing EFS system or have designs in mind for a new one, we can integrate this kind of logic for you at a reasonable cost.
Divergence Software, Inc.
www.sr-analyst.com
support@sr-analyst.com
Contact Us/Subscribe
No comments:
Post a Comment