Great information and exchange throughout this thread. A continuos Macro Position. How does a trader balance model implementation costs when volatility deviates from modeled range and still maintain a consistent monthly return ? Say for example you are running an options pricing model for over 19+ months and it/you generate(s) a consistent monthly return. Then you find that implementing the model as-is is difficult because, say, volume is cut by 2/3 in your universe or costs go up by 30%. Perhaps you run a reversal strategy with underlying and options and open interest dries up. You can give a better example i'm sure but i hope i am getting my question across. How do seasoned option writers adjust to and judge the marketplace? Thanks. Apologies ahead of time if i revived any old subjects.
I think I agree with this. The term 'expectancy' is of no use when describing a series of taken positions, it only refers to a single position, at least as I understand it. Expectancy is the expected value of a position in the future, given the actual current situation and a statistically modeled distribution of possible moves away from it. If a strategy is a series of subsequently taken positions, in a if-this-then-that actions system, there is no way to give the expectancy of the strategy since the taken actions are not known at this moment. If you'd strictly prescribe every action and its conditions, you could run a MonteCarlo sim of your strategy, a lot of times (10000 or more) and get an average expected outcome. I would be really surprised if this outcome would be a positive one if every single step has a negative expectancy. In case of a legged-in butterfly: after the favorable move you can choose to either take the profit or to complete the fly. In the latter case it may seem that you got the fly cheaper than a new one would cost, but this is a delusion; you also used the profit on the initial half-fly to 'buy' the position. Thinking that something is free because you use unrealized profits is as silly as thinking that an unrealized loss is not a real one. So, I would conclude that it is only possible to live from writing options by either prediction of direction or of volatility. I must admit that I find this a difficult conclusion, since my own strategy was based on an adaptive method, that most of the times seems to succeed in tapping time-value, but I guess I was paid for the sporadic shake-out in advance. Ursa..
when mark to market your position every day, your position already resulted in profit and loss, and then you count only the profit of course the expectancy is positive. ---------------------------------------------------------- http://www.fengshui-123.com
I'm not going to continue in this merry go around. Apparently riskarb was smarter then me and jumped ship early. I waited too long with all the other stragglers. Everything that needs to be said about expectancy and selling premium has been said over the previous 100+ pages. Just use the search button and look for your answers. In the meantime, happy trading!
Totally agree with this argument. Maverick and riskarb, with all due respect, your reputation is really at risk here for both your interpretation of expectancy. I think you two have it all wrong. I am surprised that with all your knowledge, you can be so wrong with your interpretation of expectancy. What is important is your "initial" expectancy at the initiation of the position, not subsequent changes in expectancy which is a sure thing when price changes. By the way, I am a newbie in options but I know what expectancy is. Maybe you meant other things but use the word "expectancy" wrongly.
this is so painful to follow... Maverick and riskarb have always been helpful. i admit, the stuff they deal with are beyond my options knowledge, but just because i don't understand them, i don't try to pretend they don't exist. people ask them technical questions about options and they always come back with answers. lately, i get the impression that people are asking them to lay down step by step how they trade. and when a step is not understood, they get challenged /insulted as if they don't understand the simple meaning of a "word" such as expectation. expectation of what? of the movement of the underlying? of the direction of the IV? of the profitability of the position? Maverick has practically given the farm when he answered you on 8/3, loosenup. he said he puts on a position and whatever the market does, he puts on another position, all the while, keeping track of his exposures. one position's certain greek exposure may be offset by another. who knows? only he does, coz he's the one putting them on. maybe he's not always locking in a profit, maybe sometimes he's cutting losses. i don't know and i don't think he should be teaching everyone how he handles every situation in the market. he has amassed a great deal of experience over the years and when a certain thing happens, he knows what to do (or what not to do). Riskarb has a journal and puts on trades everyday. many times, he has short explanations on what he's doing which i think is hedging his exotic binaries to protect the loss of his premium. i don't understand what he does and i'm certainly not about to go there and start questioning his trades on the simplest of concepts. i would look like an elementary student that got lost and ended up in a physics classroom at MIT. Maverick mentioned a few posts back that options are not like stocks/futures. because they're multi-dimensional, there's opportunities that exist to capitalize on certain situations. recent example was when chiguy found a mispricing? of the straddle and the strangle on Ebay prior to earnings. it didn't gurantee that it would be a profitable trade, but it provided a way to pick the "better" position. think of google prior to the last earnings announcement. you could have gone short or long the stock prior to earnings and if you were short, you'd have made money as goog went down after earnings. in the options, i think if you were a buyer (even of puts) that most likely you'd have lost money. if you can, try to have access to the last wizards book by schwager called stock market wizards. read the section on john bender and how he "exploited" the B/S formula assumptions about price movement expectations (specifically on gold). questioning the obvious... i could be wrong on many things on this post, but all i'm saying is that let's give respect to the people that have been more than generous with their knowledge. that's just in my humble opinion... Peace.
Ursa -- I never stated that the fly-basket was +expectancy. The rationale has everything to do with greek-mag. under RegT haircut. I owned a 20lot fly in the GOOG example mentioned. I had no interest in taking 2000 shares notional risk to expiration. I reduced my exposure through the wing puchase, but it hasn't altered the, gasp! expectancy. There is more to it, but the rationale has nothing to do with a determination of +expectancy. The only positive expectancy trades within the model are +skew and lock-arbitrage, period. Anyone trading outside of those positions should stop what they're doing. [tongue in cheek]
Selling index skew has a +expectancy. Now, should we sell all the index puts we can and hedge our discrete-delta? Expectancy w.r.t. a gaussian distribution is critical to the model, and the model is fatally-flawed. All data since 1987 has proven the case. Expectancy is model-dependent and discounted by the market-making function. In reality, it's complete bullshit. BS comments like this[loosenup] is why I will limit posts to my journal.
No offence riskarb but you are answering to another question. No one claimed that expectancy isn’t model dependent. Most people just argued that is impossible to increase the expectancy of a position by adding zero expectancy trades. I think this is a very simple concept. Expectancy is an arithmetic average of all possible outcomes. If a position today has 100$ expectancy then adding today a 0$ expectancy trade to this position will not change the 100$ expectancy. You will change the shape of your payoff function but not the expectancy (average payoff). I think this is obvious. Expectancy isn't complete bs because every market participant will use a model to price and enter a trade. Black Scholes model is certainly flawed but no one uses a simple Black Scholes without adjustments for these flaws. So unless someone is a market maker, he will not enter a position without having a positive expectancy by his model (except of cource if he is hedging a risk). Again sorry for my bad English Regards
I agree, and I've never stated otherwise. I've been aware for some time that 1 + 0 = 1. I don't see a need to contribute anything further[beyond the following] on my end, but I stand by my contention that expectancy, in the context being expressed here, is GIGO. I believe the argument being made is indeed model-independence. Using any option model for a determination of expectancy is building on mud. Quote from gbos:So unless someone is a market maker, he will not enter a position without having a positive expectancy by his model (except of cource if he is hedging a risk). And the process of hedging involves discounting this model-expectancy through replication via hedging. Can't be avoided. It's moot as the "model" employed produces "expectancy fallacy" just as if the successful ES futures trader believes his model over skill produces +expectancy. Relying on stated-expectancy is illusory.