When trying to collect one’s thoughts about a topic, it sometimes helps to work on the phrasing of the question for which you think you are trying to get an answer. To twist the viewpoint a bit, what can option probability calculations tell one about stock-price probabilities? (Here’s a source (Power Options) I like for knowing about options probabilities.) For most options positions you might contemplate you can find out what the probability is that the stock price is above, below, or for spreads between certain prices.

One has to always keep in mind that options probability calculations exclude essential practical risks, such as event risk and market direction (just for starters). So those probabilities for real purposes are suspect from the get-go. With that caution in mind, one might ask the question “how reliably can options forecast a stock’s price level?”  Or phrased differently, how reliably can options calculations predict a stock’s price will be above, below or between specified levels by a certain date?

It turns things upside down, but in the context of RE it starts you thinking along these lines: suppose I am running a strategy where at some bar three different linear strategies all signal sell at price X, but the options probability calculations say there is a 90% chance the price will be above that point a month from now. The tough but practical question is, what action (sell or hold) has the best expectation?

The corollary to that is, it is good that RE has options calculations capabilities, yet to be brought to bear.

-Pete
(See also rightedge-ats Yahoo group )

If you've ever read Options Trader, there was a backtest simulation selling iron butterflies on the SPX 1 sigma out.  1 sigma is one standard deviation or 68%, so in theory, the price between the time he opened the positions and when they closed should be within that standard deviation 68% of the time.  Amazingly *cough*, the win rate was 65%.    Did I need a simulation to tell me this?  Point being, I think the pricing for the most part is pretty damn accurate.  Far in the money and far out of the money options are skewed (vol smile) because of fat tails events, or so the theory goes.  I think if you truly wanted some predictive powers here, perhaps you could take the prices along the chain and match it against the underlying.  See if the FOTM option skews bend more before something big happens.

Anyway, the rough calculation I use is delta.  I think this is common and a pretty good rule of thumb.  If the delta of the option or spread is .50, it has a 50% chance of landing in the money by expiration.

To twist the viewpoint back to strategies, if all options are priced against a lognormal distribution based on the volatility of the underlying, it seems like to gain the theoretical edge, you'd need to take all of the prices and construct various strategies to determine the mispriced options.  This is where I go back to the machine to help me with this.

>> I don't think there are any predictive powers with options.  It's all probabilities. 
Doesn’t that come close to a contradiction in terms? Note also the commonly used terminology: for a given strike “the probability of the stock price being above X at expiration is y% probability”. That is surely a stock price forecast (prediction), phrased as a likelihood of a future state. If those odds are good enough to justify an option buy or sell, how could it possibly be the case that those odds become no longer applicable simply because one contemplates a buy or sell of the stock instead?

The pedigree of the numbers and the computations for options is pretty good. Aren’t they good enough to justify some experimentation? Yahoo will provide information about strikes free and, I assert, is adequately timely to make the derived stock-price-level prediction (probability) useable in practice. Deriving that probability is where the RE Options class may come in handy. A key point is, wouldn’t this be one good way to leverage RE’s capabilities?

(I am aware of the other elements you touch on in your post. All of them are useful things to have in the back of one’s mind. They are useful criteria when looking at individual cases, but hard to scan for reliably.)

-Pete
(See also rightedge-ats Yahoo group )

I, personally, don't think so.  Given that delta works both ways and is continuous, I think it's the opposite of predictive.  Let's say the option is at the money.  It's like saying, there's a 50% chance it will be in the money at expiration and a 50% chance it will be out of the money at expiration.  Take an option with a delta of .10, saying there is a 10% chance it will be in the money and continually revising that statement as the stock moves upward and downward isn't really going out on a limb.    Predictive power to me is someone selling me an option with a delta of .10 that is 100% certain to be in the money.

I'm not sure I completely understand what you mean here.

-Pete

-Pete
(See also rightedge-ats Yahoo group )

-Pete
(See also rightedge-ats Yahoo group )

Is it really just a smooth equity curve with compounding that will help you beat the averages?

Only 40% of people think there are more gains to be obtained from options than from stocks is a bit surprising.  Why would you trade something with a "less positive" expectancy?  I'd like to know the answer to that (and risk management is an acceptable answer).

-Pete
(See also rightedge-ats Yahoo group )