Trading vertical spreads using only market-implied probabilities?

E

ensemble

I wrote a 'textbook BSM' program to analyze all of the possible spreads on all liquid options expiring <1 week, and filtered by the highest return/risk and probability of profit computed from the normal CDF. This of course assumes a log-normal distribution.

On $6500 Reg-T margin, the return for the week was $545 (8%) at a 5/5 win rate. Not much, but it would be easy to scale up with additional size and underlyings.

Anyone currently doing this in practice?

Capture.PNG
 
A weakness in your process (as inferred from your post) seems to be your management and projections for volatility, however your short term view does help to limit your exposure some.
Am I correct in assuming your trades are taken to expiration?
5 of 5 is too small a sample size to make inferences from. One would expect, your process, if done correctly would be break-even with some losses due to slippage & fees.
 
There are a handful of markets like interbank foreign exchange that don't immediate quote price - they quote implied on the options.
 
ensemble said:
Anyone currently doing this in practice?
More...

What exactly are you doing? How is the BSM model helping you decide which stock, what strikes, what trade type, what DTE to chose? When to buy and when to close? When and how to adjust?

No disrespect, but it seems as you entered a lot of effectively long positions, and since the market had a up week, so all your trades made a profit. If the SPX had fallen a hundred points, most of the above trades would have been losers.

Not trying to be negative, but I just want to understand what was the rational behind these trades.
 
@ensemble, interesting OP as well as the concept that you are coming up with, good luck with the project & keep posting any results or strategies.

appreciate if you could expand on your OP with more information & detail?

as others have posted, there is little to see or know about the success rate as well as based on your 'picks' the probability of success is only if the picks are on something that is moving in the direction of your plays.

Q.

how did you pick the ones that you did 5/5?

what if your sample rate was 20 picks over 3 mths, have you figured out the probability?

I looked at barchart to get a sense of what's what.

from options link, 'market pulse' tab, side bar looked at highest activity, volatility, traders cheatsheet, then looked at options expiring 1 May.

https://www.barchart.com/options

I dont know if simple analysis vs technical analysis is any different, or stocks/ETF's picks.
 
ensemble said:
I wrote a 'textbook BSM' program to analyze all of the possible spreads on all liquid options expiring <1 week, and filtered by the highest return/risk and probability of profit computed from the normal CDF. This of course assumes a log-normal distribution.

On $6500 Reg-T margin, the return for the week was $545 (8%) at a 5/5 win rate. Not much, but it would be easy to scale up with additional size and underlyings.

Anyone currently doing this in practice?

View attachment 225976
More...
Very interesting approach. :thumbsup: I think I understand what you are doing.

I do have a couple of questions:

1. How you download all liquid options expiring < 1 week? It is an enormous set for all tradable options for NYSE, Nasdaq....., and to also screen all possible combination spreads after you compute the CDF of each? What computer/processor do you use?

2. Isn't N(d2) essentially the CDF for each option?

Thank you for sharing.
 
ffs1001 said:
What exactly are you doing? How is the BSM model helping you decide which stock, what strikes, what trade type, what DTE to chose? When to buy and when to close? When and how to adjust?

No disrespect, but it seems as you entered a lot of effectively long positions, and since the market had a up week, so all your trades made a profit. If the SPX had fallen a hundred points, most of the above trades would have been losers.

Not trying to be negative, but I just want to understand what was the rational behind these trades.
More...
OP specifically said spreads, not long positions.
 
I wasn't clear if you do this, but I would recommend that you takes trades based on maximizing payoff, which is:

payoff = probability*(return/risk)
 
ironchef said:
OP specifically said spreads, not long positions.
More...

Ironchef, not sure if you trade options - if you do, then you will be aware that spreads can be long positions (or short ones). A position can be a long or short and it has nothing to do with it being a spread or not. If I buy a call, I'm 'long', if I sell-to-open a call, then I'm short.
Same with a call or put spread.

I'm sure you knew all this and you didn't need me to point out the basics. :-)
 
ffs1001 said:
Ironchef, not sure if you trade options - if you do, then you will be aware that spreads can be long positions (or short ones). A position can be a long or short and it has nothing to do with it being a spread or not. If I buy a call, I'm 'long', if I sell-to-open a call, then I'm short.
Same with a call or put spread.

I'm sure you knew all this and you didn't need me to point out the basics. :)
More...
Thank you. I only trade single leg, so appreciate the coaching.
 
You must log in or register to reply here.
Back
Top