When I talk to people about my options selling algorithm and what I’ve been doing for more than a year now, I usually encounter a fair amount of skepticism.

“You wrote an algorithm for options? Does it actually work?”

My answer is always yes, and I point to my returns (75% – 120% as of the past few months). But return on investment (ROI) is just a single piece of evidence–and it’s biased because it comes directly from my personal trade history. Is there a more quantitative, more general way to assess the algorithm’s performance?

The answer is YES! Every time I post the algorithm’s Weekly List, I also post a link to the previous week’s performance. That means that every week, I have 2000+ options that either “won” or “lost”. What is the aggregate (total) win/loss rate, and how does that depend on the algorithm’s weekly rankings?

*In theory*, options ranked higher on the list should have a higher win rate than lower ranked options. Look at the plot below:

The dashed line is the worst case scenario. It means that the algorithm really isn’t doing anything–higher ranked options are just as likely to win as lower-ranked options. If this were the case with my algorithm, you might as well pick an option at random because you’d be just as likely to win vs. using my Weekly List. The blue line is the “ideal” scenario. In this case, higher ranked options are in fact more likely to win than lower ranked options.

Let’s see how the algorithm actually did…

Look at that! *In general*, higher ranked options have a higher win rate than lower ranked options. The best win rate of ~90% occurs in the top 250 options. Compare this to the worst win rate of ~65% for the lowest ranked options. Hooray!

I know you’re wondering something…You’re curious about that little dip right at the beginning of the plot. Why is there a dip there? Does that mean that the algorithm is wrong for the very highest ranked options?

Well, there’s more to making money than just win rate. There is also the matter of *how much* *money* you actually win when you’re right. This is essentially ROI. And there’s also how much money you lose when you’re wrong (loss on investment, aka LOI). I’ve been keeping track of ROI and LOI for every ranked option since I started this. That allows me to look at the “average ROI” for every ranking. If the algorithm is working, then higher ranked options should have a higher average ROI than lower ranked options.

Let’s see how the algorithm did:

Wow! That makes sense. The highest ranked options had the highest average ROI. That translates into more money in your pocket–on average–if you used the Weekly List!

TL;DR–the algorithm is working, and the Weekly List really does provide value. It’s proven out in my personal ROI over the past 16 months, and now it’s proven out in an aggregate, statistical, general way.