Brooks, thank you for pointing out this back testing service for us. It seems to have a large array of stock/options strategies capabilities .. do you know if it includes testing of delta-based strategies (i.e.: sell 16-delta put every 22 days, say) ? ...and pricing? Many thanks Stam
In their tutorial section, I actually found some links to YouTube videos demonstrating the software — and the answer to my question above (about delta-based strategy) is a resounding yes!! BTW: I did like their rule/logic/strategy interface as demonstrated there... still, curious about the price— their free version is limited to a one-year of back data— but is would be a great place for me to start, I suppose..
A review of OptionStack ... see the discrepancies section: https://spintwig.com/options-backtesting-with-optionstack/
Those problems with some of the statistics calculations may be valid. I'd report them to OptionStack and see what they had to say. There may be explanations, but if they are bugs, I'd expect they would fix them. OS does not have the depth of test reporting that you would find with Tradestation or Amibroker, which is why I wrote my own option backtester. Another issue is that you're working with EOD quotes, so that constrains the type of exits you can test for and the option values are at the EOD when you might have exited during the day, at a different price. With Tradestation or Amibroker you can test with intraday (i.e. N minute) quotes, but they can't handle options with expiration dates, strike prices, Greeks, etc. Sometimes you have to work with the best tool available, knowing it has limitations.
Brooks and Adam too... Clearly there is still much to be desired with public (non-proprietary) backtesting software, especially in context of strategies involving options. There seem to be several outfits out there geared towards that, such as: ORATS, TradeStation, OptionStack, eDeltaPro, etc.. Perhaps, either one of you can take the lead on a new thread dedicated exclusively to a comparative discussion of he available Backtesting outfits (for the retail guy)....keeping in mind that the quality of the available data is paramount. BTW, I personally see no problem in working with EOD data, as the backtesting is meant to evaluate performance (a posteriori) on a 'frozen' time-point of the data-- make it to be EOD, or Noon time or at 9:35am--that should not matter much, as long as it is consistently preserved throughout the study.. Stam
Stam My firm, ORATS, uses near end-of-day market data where we snapshot all bid-ask options and stock prices 14 minutes before the close. I was a market maker and saw what happened to the prices as the close approached, widening out. We apply a smoothing algorithm that produces good theoretical values and greeks. Using a filter, you can avoid backest trading in markets where the bid-ask width divided by the stock price is greater than a certain percentage, I usually use 0.5% however in VIX it's 1.5%. You can also stagger trading, for example, trading a 90 day strategy every 30 days. You can set the study start date and end date to test out of sample. You can also control the dates to enter and exit a trade. If you use technical indicators, you can use those Entry Indicator Triggers along with volatility measurements like only trading a strategy when the IV Rank < 50, VIX price < 30 and MACD > 0. You can combine strategies in various ratios. For example, for hedging a stock portfolio I use long term, way out of the money puts on all the time and exit when the profit > 300% when I roll out again to a certain Spread Yield Percentage (options price / stock price) so I don't spend too much when the market tanks and IV goes up. In addition, during times of market calm (contango > .25 and VIX < 20), I use a put 1x2 ratio - buying 1 OTM put and selling 2 farther OTM puts. I combine these strategies together to over hedge using the puts by using a combine ratio > 1 and have the put 1x2 at 1. Our scanner uses the same filters as the backtest and adds assessing trades based on a distribution forecast, historical volatility forecast, and smoothing forecast of options prices. There are many more features. Let me know if you want to use something that you don't see here.
@matt, thanks for all your posts related to options & back testing. ^^ that comment about last 14 minutes before close, I've noticed option trades/volume on the index or ETF's even underlying equities pick up in the last 5 minutes before close. Is it me that sees this or is it common place for option trades/traders to do end of day buy/sell options on the index's or underlying equities, if so is it a predictor or indicator of the next day open or next day movements? if my assessment is all bunk, please say so or ignore the post Thanks
Thanks never2old Yes, volume picks up near the close. I have a client that trades millions of dollars in equities on the last print daily. I am not sure if it is a predictor of the next day open or movement.