What is the most statistically predictable Option strategy?

[xandman]

I was hoping we could have a high/abstract level discussion (easy to digest) with the option traders.

ET members mostly have a fair understanding of unlimited loss/limited profit and vice versa. Thus, we can probably theorize that a PREDICTABLE option strategy would have limited loss/limited profit. ( I may be worng about calling this predictable. Perhaps. more easily measurable.)

Additionally, there is large loss/small profit and vice versa. Small profit spreads being the most frequent winners.

So, I come to the conclusion that the Short fly ( long volatility ) strategy is the MOST PREDICTABLE making it the best candidate for algorithmic position sizing and stop loss schemes.

 

[panzerman]

Hmmm.... "Statistically most predictable". That would be any option strategy whose underlying value most closely follows a lognormal distribution at all times, and you select your strikes based on the computed probabilities using implied volatility.

Or are just interested in which option strategy has the largest win percent?

 

[xandman]

Thanks for asking clarification. I am not sure! I may be a little lost in the sauce and I hope we got Maverick to chime in.

I seek predictability because I think the application of position sizing and stop loss schemes can make me profitable...just like the guys who trade MA crossovers (etc, etc) in trading the underlying.

It seems that strategizing around a lognormal distribution means that I have to find this holy grail called "edge". Be it execution or arbitrage..which is mythical in retail.

So largest win percent must be IT. (Caveat: I thought you could take advantage of the natural drift by buying deep in the money, but somebody explained that to me already). Because stop loss scheme and sizing can be the source of edge.

 

[VIXTrader]

So largest win percent must be IT.

It would certainly have to be at least a big part of it, because "statistically predictable" may not actually be that profitable.

This really is an abstract thread, because there is no way to assign "the best" to anything to do with trading. High absolute returns might be the best for some, but not others. Highest win percentage might be someone's holy grail, but again need not imply profit. Best risk adjusted returns? By what measure? We could go on and on.

I don't know... Sounds like a cool thread, but having a hard time getting past the semantics.


But if you're looking for a good base for a trading system with trade sizing and stop losses being central to it, I'd say you could start with wide wing Iron Condors (or strangles if you have the account size)

 

[xandman]

Makes sense, that a good sizing and sl strategy shines in those short vol strategies.

Something is gnawing at me at the lack of long vol discussion which makes me feel that it's the place to go. Can't put my finger on it. Thanks for the cool vote. I already told a private poster that I am a horrible thread starter.

I hate brain farts as much as the next guy, but I think I can make money.

 

[Teycir]

Selling far OTM puts is the strategy that puts the most the odds in your favor and gets you the best win rate. Does not make it the best strategy in absolute, because some unexpected events could wipe out years of profits (that's what happened to Victor Niederhoffer among many others using this strategy). I think that thinking that there a strategy is superior to all other is bad logic. For each situation there is an optimal strategy. I would say selling OTM puts is the best strategy in case there is unusually high IV and a very little chance of further drop in price. This same strategy sucks if price is high, IV low and there is a risk of pullback.

 

[newwurldmn]

Predictability comes from isolating the only risk you care about.

If trading volatililty: Variance swaps and VIX futures. No path dependency issues and simple pnl attribution.

If you are trading options for delta, then you will really never find anything more "predictable" than stocks because options will have all sorts of other risks that will contribute to your pnl.

 

[xandman]

Predictability comes from isolating the only risk you care about.

If trading volatililty: Variance swaps and VIX futures. No path dependency issues and simple pnl attribution.

If you are trading options for delta, then you will really never find anything more "predictable" than stocks because options will have all sorts of other risks that will contribute to your pnl.

This will be a bullet point sticky on my notebook for quite awhile.

 

[sonoma]

I was hoping we could have a high/abstract level discussion (easy to digest) with the option traders.

ET members mostly have a fair understanding of unlimited loss/limited profit and vice versa. Thus, we can probably theorize that a PREDICTABLE option strategy would have limited loss/limited profit. ( I may be worng about calling this predictable. Perhaps. more easily measurable.)

Additionally, there is large loss/small profit and vice versa. Small profit spreads being the most frequent winners.

So, I come to the conclusion that the Short fly ( long volatility ) strategy is the MOST PREDICTABLE making it the best candidate for algorithmic position sizing and stop loss schemes.

You're best bet is to rely on the empiric observation that iv tends to be greater than realized, and following that, concentrate on the subsets of this misplacing: terminal positioning centers around the current position more frequently than estimated, and that moves further from ~ 2 sigma occur more frequently than predicted. Of course there are trading frictions that will negate the above, so you'll have to focus on the frictions you can control.

 

[VIXTrader]

Makes sense, that a good sizing and sl strategy shines in those short vol strategies.

Easy to understand, even easier to trade, and it's a pretty broad net that captures a significant portion of market movements. However as has been said, nothing works in all environments. That is literally a mathematical impossibility. The very structure underlying the markets prevent the possibility of some kind of magic bullet that you can fire at all targets and keep hitting.

So it gets back to the concept of having a tool box with multiple strategies at your disposal. Likely if you asked the average successful option trader which 2 or 3 strategies they use the most, you'll get all the same boring answers. Iron Condors or strangles if you have the account size, deep in the money put selling, put and call verticals for mean reversion, stock with collars, etc... Not likely to find anyone talking about long vol as a significant portion of their trading.

I just don't think an answer exists as to what's truly the best, or as you put it the most statistically predictable (whatever that means )