Denver Colorado/USA
Experience: Intermediate
Platform: SierraChart
Broker: AMP
Trading: ES
Posts: 116 since May 2017
Thanks Given: 132
Thanks Received: 145
|
The SC site has all the info for backtesting with their software. Everything I know about SC I learned from the documentation they provide and asking the occasional question on their support forum as well as searching through the posts for things already answered. I would recommend using google to search the SC site, because it works a little better than the site's built in search. Ex: type into google search: "site: sierrachart.com backtesting". Before you start coding and getting fancy, you can simply replay the chart and make simulated trades to test your ideas. As far as using good methods to gather accurate statistics, Kevin Davey has good recommendations and has a free Monte Carlo simulation spreadsheet that you can download from his site. He has many webinars you can watch on here.
Also they have a spreadsheet system that allows you to write automated strategies using the calculation format, cell references, and formulas that work in excel. So if you already can do basic excel stuff that is a good way to start. It is not as efficient or capable as using the C++ based ACSIL, but it's really fun to play with to test out some basic ideas and have an automated test run with a statistics report generated afterwards.
Another thing I do sometimes is load the spreadsheet study and throw whatever indicators are interesting to me on the chart. Then set the spreadsheet study to use a whole lot of rows so that it populates the spreadsheet with the values of the indicators as far back as you want. If you have your chart set to 5m bars then you will have readings for all of your indicators for every 5 minutes as far back as it goes with the amount of rows that you populate. You can then copy all of that data from the SC spreadsheet inside trhe program and paste it to an actual excel spreadheet to do analysis and calculation on.
Example:
Say you want to know the average distance of price (in ticks) from the 200 period moving average of a 1 hour chart on Fridays between 1-2pm. You get all the price and moving average readings from the SC spreadsheet and in excel you simply filter for the day and time period you want. Find the difference for each row, then get the average. Create an average for when the price is below the 200 and an average for when it is above by splittign up the positive and negative values. before averging. Now you have new information and can start building a database of things like this for the instruments you trade. It gets very tedious at a certain point when you have a lot of data, but you can take this pretty far and excel can do a lot. There are free statistical add-on packages you can use in excel that are great for strategy development.
This is very basic for long time excel users so if I am over-explaining I apologize. I find it to be extremely useful and some never give it a chance so I thought I'd share
|