How to create a strategy based on my price forecast

[xappppp]

Let me through a concrete example first:

Below is a price forecast from my simulation model which shows possible trajectories. I think the mean price(green line) is basically not very useful, but the individual forecast might be helpful to understand the possibility of future movement. What's the best way that I can create a option spread trading strategy and either reduce the risk/increase profit.

 

[taowave]

IB has an option optimiser that "supposedly" finds the optimal strategy per your chosen inputs..I say supposedly as I have never traded off of it,only played around with it.I believe Orats may have a tool as well.Matt (owner of Orats)is on the board,ask him.Hes got a great product

 

[stepandfetchit]

You may want to "re-think" your objective, to insure you are not "getting the cart before the horse"!
1st Hurdle: Are your projections accurate? Have they been actionable? And if so, any reason to think they will not remain actionable?
2nd Hurdle: If and only iff you pass 1st Hurdle, then you need to add a similar set of projections for ATM Volatility.
3rd) With both hurdles behind you, evaluation of which option strategy/trade per underlying is fairly mechanical, assuming you have access to timely option data. {I have implemented this step, and continue to refine it.}

The 3rd item below should not be attempted prior to completion of the other two.

--------------------
BTW: I like your approach, and hope you may be on to something! I hope to hear more, as this may have legs!

 

[xappppp]

Thanks for the notes! And the that's very encouraging. And hope you can provide more challenge and question.

My basic thoughts is I want to separate the price forecast and the strategy construction and optimize both of them separately assuming either one of them might not be accurate/ or wrong.

Currently, what I can come up with is some directional spread bet. For example, the forecast result shows that TLT will be very likely above 160 for the next 6 weeks, then I will short a 155-160 call spread, and earn about 10% over 6 weeks of exposure. This construction seems fairly easy. I want to post here to jog some advises is this idea work or not, and whether there is any room for improve the risk/benefit of this strategy.

As for other two tick (GLD SPY), how to profit might not be very straight forward to me, but hope someone can advise if they can see something profitable and how to make a strategy.

 

[stochastix]

Any forecast not based on stochastic process theory is doomed to failure. I see no mention of probability theory

 

[xappppp]

Interesting for your second comment. Can you explain why I need project ATM Volatility? My thought is I will just trade and wait for expiration.

You may want to "re-think" your objective, to insure you are not "getting the cart before the horse"!
1st Hurdle: Are your projections accurate? Have they been actionable? And if so, any reason to think they will not remain actionable?
2nd Hurdle: If and only iff you pass 1st Hurdle, then you need to add a similar set of projections for ATM Volatility.
3rd) With both hurdles behind you, evaluation of which option strategy/trade per underlying is fairly mechanical, assuming you have access to timely option data. {I have implemented this step, and continue to refine it.}

The 3rd item below should not be attempted prior to completion of the other two.

--------------------
BTW: I like your approach, and hope you may be on to something! I hope to hear more, as this may have legs!

 

[stepandfetchit]

The three variables that must be known for precision are Time, Price, and IV!
Typically each of these will have some uncertainty (as are your plotted examples above for price). However if your projections for time are to the instant (to the second) AND all occur precisely at Expiry, then you do not need IV as it will be zero!

IMO: if you limit your trades to only set and forget (no target pricing), your results will be Binary and may be difficult to justify increasing the trade size.
My approach is to give each variable "breathing space" commensurate to my confidence in the variable range! However, each of us need to understand and be comfortable with our processes!

 

[xappppp]

Okay, my forecast are in days/weeks, and I don't have a good framework how to trade it actively through out the day yet.

 

[stepandfetchit]

Okay, my forecast are in days/weeks, and I don't have a good framework how to trade it actively through out the day yet.

I don't think you understood what I meant. There is no correlation to what I posted to "actively trading"! For example, if your reliable projections indicate a specific debit spread will be an ideal trade candidate, then you could purchase that spread, then place a GTC limit order for your target price! -- No need to be staring at a screen every day!

 

[xappppp]

That's what I do for now. I do price the option spread based on BS model with those input, but I'm just taking a 50 IV as a assumption without forecasting it.