July 15 Webinar Recording Part 1:
Curve Fitting Indicators to Data – Bad Idea
When you use the Atlas Line, Trade Scalper, ATO 2, or any other method we teach, rest assured, there is only one way to trade each method. There are either winning trades or there aren’t. There is no curve fitting. Over the 100s of videos we’ve produced throughout the years, you may be able to get the sense that we like to show our methods working under real-time conditions, often in an online webinar room full of real people, as seen in the webinar video above.
If you ever see a trading system claiming a 100% win rate, there is either a lack of truth, data, or curve fitting was used to make the claim.
When it comes to evaluating trading system performance, curve fitting refers to a process that optimizes a trading system against a data set. What is really happening is that the system adjusts itself to provide (potentially misleading) indications of optimal performance based on the given data.
July 15 Webinar Recording Part 2:
Live Signals & Explanations
In the trading world, systems are often backtested against historical data. During this process, the system’s parameters can be curve fitted to produce the greatest profit potential. Then the creator/owner of the system makes big claims about performance. On the contrary, with any trading system, there is no guarantee of lucrative future performance.
Here are some questions a critical trader may ask:
- Were the selected data points indicative of common and realistic, consecutive, or day-to-day trading?
- How many modes or changes did the system use to produce the optimal results?
- Can a human really trade the system as well as the automated backtesting?
- What types of orders were used?
- Did the backtesting account for slippage and other real-world variables?
- We prefer a straightforward, human approach. Our trading systems were made by real people to be traded by real people.
