Option Pricing Efficiency

[tomk96]

for the OP, if you are looking the hedge a stock position with long puts, isn't your return just your net premium lost over time? wouldn't you be reducing that more if you exited your long put when you exited the position it was protecting?

as a mm, my theo values are used to price options relative to one another. there is no expect value i expect at expiration based on my tv. options are worth whatever somebody is will to pay for it or sell it at. you seem too concerned about Black-Scholes rather than how to use them for your needs.

if you pay $2 for 10 puts, and you expect them to go to zero, you are going to be out $2000. Annualize each attempt to get your return.

fyi, if you need to buy an put 1 month away for a stock you only plan to hold 10 days, you should be selling the put that still has value.

 

[jazzguysoca]

Quote from atticus:

It means that the excess return is a function of realized volatility.

Best practice would involve modeling atm call purchases (married put synthetic) and test it empirically.

Nobody knows what the cost will be.

Thanks, atticus - appreciate the insights.

In the context of the system I mentioned, I'm not sure that a synthetic call is exactly equal to an actual call. I know when I first came across the article I mentioned, I wondered why the guy didn't just trade calls and save himself some hassle and commish.

But I think the trick is that there is the possibility of exiting the long stock early if it hits its target, while leaving the put open for the possibility of closing it out for a profit. Somewhat akin to legging out, but in a less discretionary way.

Obviously I need to backtest this concept, but I'd really rather do it by modeling option prices rather than using historical data.

Which brings me to my next question: Would generating option prices via the BSM model during backtesting give me reasonably realistic results to what would happen in the real world?

 

[jazzguysoca]

Quote from tomk96:

for the OP, if you are looking the hedge a stock position with long puts, isn't your return just your net premium lost over time? wouldn't you be reducing that more if you exited your long put when you exited the position it was protecting?

Actually, I think my return should be equal to the return of the underlying stock system without the puts, less the average cost of the puts and transaction costs.

The problem with exiting the puts at the same time I exit the long stock position is that I'd be giving up the cases where the stock hits its target for a profit, then the price pulls back and leaves the put with a profit at expiration.

as a mm, my theo values are used to price options relative to one another. there is no expect value i expect at expiration based on my tv. options are worth whatever somebody is will to pay for it or sell it at. you seem too concerned about Black-Scholes rather than how to use them for your needs.

I agree that they're worth whatever the market will bear, except at expiration, no? At that point they're worth exactly their instrinsic value, with no interpretation or gaming by the market.

if you pay $2 for 10 puts, and you expect them to go to zero, you are going to be out $2000. Annualize each attempt to get your return.

That was how I looked at this whole thing it at first, but remember you are assuming that the ATM puts will always expire worthless. That won't always be the case. Just because your stock hits its target doesn't mean that you will eat 100% of the TV. If you sell the puts when you exit the stock, you'll still have some TV left. If you hold them until expiration as I'm suggesting, your costs should be further reduced, since sometime the puts will expire in the green as well. This should lower the average cost of the puts below its original premium (hopefully significantly so).

fyi, if you need to buy an put 1 month away for a stock you only plan to hold 10 days, you should be selling the put that still has value.

My original instinct was to always close the puts when the stock positions are exited (while they still had some TV), as you suggest. But now I'm not so sure.

Assuming the BSM model has any relationship to reality (and that I'm understanding it correctly), the odds of a put exiting in the green are more or less equal to it exiting the red, hence the need for a TV premium. So more then just a few times the puts will pay for themselves or even generate additional profit, which reduces their overall cost to the system.

 

[sonoma]

Quote from jazzguysoca:

... Which brings me to my next question: Would generating option prices via the BSM model during backtesting give me reasonably realistic results to what would happen in the real world?

No. You'll have to approach this heuristically, as per atticus.

For what it's worth, as a partial answer to a related issue, the CBOE put index (consistently shorting near month ATM puts on the SPX) results in ~12%/yr return.

 

[Martinghoul]

Quote from jazzguysoca:
So my question is: How much performance will I give up by adding puts? Will the negative expectation of the puts be so large that they will cancel the benefits of a smoother equity curve?

Well, ultimately it depends not so much on the puts, but on your long equity strategy. You can calculate the returns on your protective puts easily during different regimes (regardless of a very long-term return, which isn't too interesting a number). Then, depending on the drawdown characteristics of your main equity strategy, you can make a decision.

My point is that it's really immaterial what the really big picture, really long-term return of of your protective put portfolio is, since ultimately you don't really care about the really big picture, really long-term returns of your main strategy. If, as you said, you care about leverage and drawdowns, you need to do the work and determine if your idea is viable empirically. I think this is pretty much what everyone's telling you. You don't need data for every single stock, every single expiry and strike. Just use ATM index puts and I believe you'll get your answer.

 

[galvinlee888]

Quote from sonoma:


For what it's worth, as a partial answer to a related issue, the CBOE put index (consistently shorting near month ATM puts on the SPX) results in ~12%/yr return.

Where you get your result from ? Are you talking about long term or only specific years ? Obviously you are getting a good return to short puts in 2009 and 2010, not in 2008 and etc..

 

[sonoma]

Quote from galvinlee888:

Where you get your result from ? Are you talking about long term or only specific years ? Obviously you are getting a good return to short puts in 2009 and 2010, not in 2008 and etc..

http://www.cboe.com/micro/put/

The data is long-term, more than 20 yrs, just like their BXM index. My point was not to advocate for put-writing, but rather to address the query of the OP re: real-time data about consistently selling (buying) ATM options. It's just food for thought.

 

[jazzguysoca]

Quote from sonoma:

No. You'll have to approach this heuristically, as per atticus.

For what it's worth, as a partial answer to a related issue, the CBOE put index (consistently shorting near month ATM puts on the SPX) results in ~12%/yr return.

Yeah, I was afraid of that. I think I might start with just the BSM model to see if there's any daylight at all in the strategy, then (as much as I'd like to avoid it) move on to using historical options quotes.

Would using other models like Cox-Ross do any better than BSM?

I'm curious: Do you guys use historical price data or do you have proprietary models you've created to backtest ideas?


Thanks for the CBOE info; 12% is more than I'd like (I was hoping for single digits), but its not prohibitive.

Still, I'm amazed that option sellers get such a large payday (relative to T-Bills) for what they do.

 

[jazzguysoca]

Quote from Martinghoul:

Well, ultimately it depends not so much on the puts, but on your long equity strategy. You can calculate the returns on your protective puts easily during different regimes (regardless of a very long-term return, which isn't too interesting a number). Then, depending on the drawdown characteristics of your main equity strategy, you can make a decision.

My point is that it's really immaterial what the really big picture, really long-term return of of your protective put portfolio is, since ultimately you don't really care about the really big picture, really long-term returns of your main strategy. If, as you said, you care about leverage and drawdowns, you need to do the work and determine if your idea is viable empirically. I think this is pretty much what everyone's telling you. You don't need data for every single stock, every single expiry and strike. Just use ATM index puts and I believe you'll get your answer.

Thanks for the info, Martinghoul. I like the idea of using index puts as a litmus test. Perhaps I could further reduce the data burden by just Monte Carlo sampling a subset of one month segments from a 5-year sample to validate the idea and/or build a model.

Considering your (much) greater experience trading options, I'm curious about your opinion regarding the put-marrying strategy's viability; Am I wasting my time chasing fools gold, or do you feel there might be some daylight here?

 

[Martinghoul]

Quote from jazzguysoca:
Thanks for the info, Martinghoul. I like the idea of using index puts as a litmus test. Perhaps I could further reduce the data burden by just Monte Carlo sampling a subset of one month segments from a 5-year sample to validate the idea and/or build a model.

Considering your (much) greater experience trading options, I'm curious about your opinion regarding the put-marrying strategy's viability; Am I wasting my time chasing fools gold, or do you feel there might be some daylight here?

I can give you a general idea, but if you want ideas specific to equities, you'll have to ask someone wiser in the ways of stocks. like atticus. I believe the answer (similar to most of my answers) is actually "It depends". There may be something there worth exploring, IMHO, and, even if the specific approach doesn't work, it will provide sufficient food for thought for you to explore. As a first step, maybe you want to look at the academic literature that examines the returns of these strategies in equities. My Z$2c.