Quite a few things have been learned along the way:
1- There isn't a single solution for back-testing options strategies with enough flexibility out there for the retail guy, where you can specify all sorts of entry conditions, adjustment conditions and exits, plus flexible position sizes. Yes, QuantyCarlo used to fill a void, but not anymore. They are now closed to the retail guy. Apparently, there is no market, or the average individual trader cares about using such platforms as much as he cares about agricultural problems in Burkina Faso.
2- Some platforms for this purpose do exist, but they are either custom-built and kept private, or they are available to institutions at $93,209,403,940,324.59 dollars per month.
3- There is an amazing business built around selling historical Data. Mind boggling! Remember that, even if you build your own custom software, you still need the data to feed it. Otherwise you have an empty carcass. Well, we have officially peaked as a species. Take a peek at these prices from historicaloptiondata.com and these other prices from quandl with shiny developer access at $1,000 per month. Holy mother of Santa Barbara!! Folks, what are we doing trading? All we need in this life is to get our hands on some historical financial data and start selling it left and right.
I have always done my tests using the "ThinkBack" feature on ThinkOrSwim. But it is a tedious and manual process where you have to go back day by day, looking at prices and stats and mentally evaluating your entry/exit conditions, then keeping a manual record on your end. This makes it extremely tedious, time-consuming and obviously prone to errors. Same thing with OptionsVue.
With the aid of George, reader of this site, I have looked at countless tools and resources, among which I can list:
QuantyCarlo - left the retail space
getVolatility.com - Still in testing mode and doesn't do all I need
alpha-5 - Not really built for this purpose, in that it does not allow to configure strategies exactly as you want (instead, they have their predefined templates)
Quantopian - can't download data yourself. Powerful but you have to program. So, not suitable for the average person without coding skills
http://www.optionstack.com/learn/
powerful solution but user needs to have programming knowledge to create his trading systems.
http://www.optionnetexplorer.com/Pricing.aspx
- Apparently performs back-tests. Not sure if this is TOS like (position by position, user manually going day by day) or fully automated entry conditions. High recurring cost
http://www.oscreener.com/options_strategy_backtester
- Seems limited as to the number of options strategies to back-test, diversity of entry conditions, lack of adjustments to positions.
http://samoasky.software.informer.com/
Options Oracle, seems to be limited to a particular position, no diversity of option entry conditions for wider testing. Last supported in 2007.
http://getvolatility.com
Not all entry conditions + adjustment conditions
The common aspect in most, seems to be that back-testing a single spread position is possible, but when it comes to adjustments, there is just not enough flexibility to define specific adjustment rules, and even dynamic and flexible position sizes for the adjustments.
George has also shared much more: a lot of valuable information about developing Options Back-testing solutions. Special thanks to him for all the information.
In addition to this, I have also reached out to a couple of bloggers who have developed their own tools but which they have decided to keep for private use. Which is totally their right, and I can understand.
So where does this leave us?
The answer is clear.
If you want a flexible enough tool, with total control of its features and enhancements; with the ability to define multiple rules, not just for entries but also for adjustments, exits, position sizes, even different commissions schemes, as a retail guy, you can either suck it up, or build it yourself.
I wanted to find one out there so badly. I was more than willing to pay up for it, rather than spending time re-inventing the wheel. But the choice was limited, extremely costly, or lacking vital features I need.
I'll have to build one myself. Allah knows very well that I didn't want to arrive at this conclusion. Consequently, I have spent four months in denial, because, well, I know once I get involved in a project like this one, it will become an obsession that will absorb the little time that I have left nowadays, negatively impacting my girl's sexual life ,........and eventually mine too after she dumps the options trading lunatic.
If you want a flexible enough tool, with total control of its features and enhancements; with the ability to define multiple rules, not just for entries but also for adjustments, exits, position sizes, even different commissions schemes, as a retail guy, you can either suck it up, or build it yourself.
I wanted to find one out there so badly. I was more than willing to pay up for it, rather than spending time re-inventing the wheel. But the choice was limited, extremely costly, or lacking vital features I need.
I'll have to build one myself. Allah knows very well that I didn't want to arrive at this conclusion. Consequently, I have spent four months in denial, because, well, I know once I get involved in a project like this one, it will become an obsession that will absorb the little time that I have left nowadays, negatively impacting my girl's sexual life ,........and eventually mine too after she dumps the options trading lunatic.
I will build a custom Options Trading Back-testing tool.
There's no way around this issue. The decision is firm.
Now, this thing can be built as complex as anyone wants, but the more complex, the longest it will take to have some piece of functioning software. So far, I have thought of a few vital initial features, and it will take me 4 - 6 months to build a first prototype. I haven't even thought about the cost of the data at this point. I may end up asking interested traders to contribute, if they are willing to help with purchasing the data, in exchange for free life-time access to the back-testing tool or something like that. Or I may instead, embed the tool inside LTOptions.com which will support itself in there overtime. Not sure about that yet, but one thing is clear: the cost is significant this time. We'll have our own automated Options back-testing tool, as good Lazy Traders that we are, for quick, effortless evaluation of ideas and no more misery manually going back through years of options chains, analyzing positions one by one, day by day.
There's no way around this issue. The decision is firm.
Now, this thing can be built as complex as anyone wants, but the more complex, the longest it will take to have some piece of functioning software. So far, I have thought of a few vital initial features, and it will take me 4 - 6 months to build a first prototype. I haven't even thought about the cost of the data at this point. I may end up asking interested traders to contribute, if they are willing to help with purchasing the data, in exchange for free life-time access to the back-testing tool or something like that. Or I may instead, embed the tool inside LTOptions.com which will support itself in there overtime. Not sure about that yet, but one thing is clear: the cost is significant this time. We'll have our own automated Options back-testing tool, as good Lazy Traders that we are, for quick, effortless evaluation of ideas and no more misery manually going back through years of options chains, analyzing positions one by one, day by day.
I don't want this new project to affect the rhythm of articles published on this narcotic site. In particular, I don't want it to affect the mid-week articles that we've all come to love. But, I may have to miss one here and there. I'll do my best.
Let this all be for the betterment of humanity.
Go to the bottom of this page in order to see the Legal Stuff
No comments:
Post a Comment