Option trading strategy backtesting platforms recommendation
- Thread starter Jeff1228
- Start date
Matt_ORATS
Sponsor
Thanks taowave. I appreciate your ideas and feedback.Hey guys,FWIW,I am taking a trial with Orats,and Matt is by far the most responsive developer/software guy I have come across.I am no walk in the park,and for the record,I have no affiliation with Orats
A few general questions:
1. If I want to backtest day trade equity options, do you provide intraday option data and if so going back how many historical days?
2. If I want to backtest swing trade individual equity options, how far back do you provide historical option data? As an example, I have GE historical stock prices going back to 1993. I want to see what 30 day ATM GE IV looks like going back to 1993 and compare it to 30 day HV.
Thanks.
Matt_ORATS
Sponsor
ironchef 1) We provide 2-minute intraday snapshots back to 2015. However, we do not offer that data in our backtester.
2) We provide near end of day snapshots 14-minutes before the close back to 2007 in our backtester. We cover all US equity options including indexes. We have the ability to backtest using ratios as triggers, for example, enter a 15 day 30 delta strangle when the HV-20dayExEarnings / IV-20dayExEarnings > 1.0.
We also have a charting package included that can show the HV/IV ratio and do a stock price return test. For your example GE HV is 58.6% and IV is 71.4% or 0.82. Our chart tests around that level setting a min of 0.69 and max of 0.95 to enter trades historically and exit when the ratio is below 0.5 or above 1.15. The result of the test was terrible, indicating that shorting GE would have worked out historically with these settings.
AVG DAILY RETURN:
-0.057%
DAILY ANNUAL RETURN:
-20.94%
LONGEST HOLD:
158d
NUM TRADES:
83
WIN PCT:
52%
I like your backtester based on what I see, though started working on my own. How much do you charge for the full history of the near-EOD snapshots? And is the data somehow adjusted or consolidated with stock splits?
There was this SciFi story I read as a kid which wasn't simply possible. Jeff1228, 17 posts, 2 likes, eventually beating Goldman Sachs and Black Rock to their own game.
And if you like SciFi stories like me you realize:
1) It's impossible.
2) You are like me.
3) It's entirely possible.
A joke from Soviet times says the tallest building in Moscow is the Lubianka prison. If you're lucky, you may see Siberia from it's basement.
I code my own backtest programs using VBA excel. I have underlying data going back years that I can download from my broker for free. However, my broker only provides 6 months of historical options and I need to go back at least a decade to try things out. I computed my own option chains using HV but they were not real, and often led to wrong outcomes.ironchef 1) We provide 2-minute intraday snapshots back to 2015. However, we do not offer that data in our backtester.
2) We provide near end of day snapshots 14-minutes before the close back to 2007 in our backtester. We cover all US equity options including indexes. We have the ability to backtest using ratios as triggers, for example, enter a 15 day 30 delta strangle when the HV-20dayExEarnings / IV-20dayExEarnings > 1.0.
We also have a charting package included that can show the HV/IV ratio and do a stock price return test. For your example GE HV is 58.6% and IV is 71.4% or 0.82. Our chart tests around that level setting a min of 0.69 and max of 0.95 to enter trades historically and exit when the ratio is below 0.5 or above 1.15. The result of the test was terrible, indicating that shorting GE would have worked out historically with these settings.
AVG DAILY RETURN:
-0.057%
DAILY ANNUAL RETURN:
-20.94%
LONGEST HOLD:
158d
NUM TRADES:
83
WIN PCT:
52%
If you can provide real historical option prices and greeks in excel format, it can save me lots of time and trouble.
I can go to your website and order?
The idea with my website is to provide curated marketdata but I haven't worked in that direction much, instead concentrated on a few underliers that have reliable data and I'm interested in.
I'll give you directly the address of the guys I get my data from. As I may have already said on this site, it's sometimes inaccurate, but from out of =~ 10,000 underliers over =~ 20 years daily data, you can always find something that works. Major indices are fine, for instance.