Selling delta 3-5 ES Puts with 40-57 days left
- Thread starter tradelosses
- Start date
Ah.. interesting to see your views. And you are correct; i do not believe in forecasting market processes. I will not argue about this point further since we all can have our opinions.
I want to leave you with one thought though.. there are differences between physical processes which exhibit stationary behaviour and hence can be modeled properly (your examples above) vs heavy tailed ones which are non stationary (which is what market data is). Also think about the number of teams within huge hedge funds and the Cargill's and Enron's of the world trying to model and predict with their huge logistic and supply chain detail advantages but still closing shop on regular beat - simply due to not being able to provide any meaningful advantage to the business as a whole. Building better models and trading just based on that and believing your models are better than all the rest of the world is exactly how i think genius fails.
That said, i am not saying its totally impossible to build a better model; just not my cup of tea. Market is always right is my stand, and i trade from there. I do still monitor a few models i build a while ago and watch their signals very regularly just for fun (i dont trade them not because they are wrong totally but because the edge is not significant for me to use in a meaningful way). Anyways, just my opinion.
Good luck trading!
-gariki
Well I think it should be pointed out that a big reason why many big firms fail is not because their models are wrong but because they are constrained by liquidity (LTCM). Big difference. This is a huge advantage for the little guy. Also, all non stationary data can be made stationary. This is a rather straight forward process. Also, it makes no sense that you simply slap on a position and say, I have no opinion about anything to do with my position therefore I will take a completely random position and consistently profit from it. That makes no sense. What you are basically doing is saying, I believe vol is over priced and you sell it with impunity month after month saying it matters not if vol is overpriced or not, I will profit either way. And if you come back and say, well if it turns out under priced and I need to manage the risk, suddenly you become a great forecaster and you can trade out of it. LOL. I mean this is the song and dance I always here, I can't predict direction so I sell premium. But when the shit hits the fan, I start predicting direction to maneuver out of it. Anyway, it's contradictory and we'll let it be.
Bottom line, funds blow up because of liquidity and leverage, data can be transformed. That is all.
If you have a model why would you even consider selling delta 3-5 garbage?
Because (unlike Maverick's model ; ) mine is not 100% accurate. It's % based so I know at certain volatility what is likelyhood of vol going down or up from there. But if I'm wrong I can still manage my "garbage" better than ATM. If you're wrong with ATM you just take the loss. I was playing around with ATM selling but I can't make the returns match up to mine (other than that I agree ATM of course is much safer because of gamma I never disputed that)
I think your notion of positive expectancy is not correct. You don't need to be right 100% of the time and you don't even need to be right 51% of the time. It's entirely based off risk/reward ratio. You should know though that with 3-5 delta puts you pretty much have to be right 95% of the time because the risk to reward is so horrible.
This game is about risk FIRST.
I think the problem is you're looking at positions as static and not active. I.e you enter and then exit at expiration or some profit/loss amount. I don't look at it this way and instead shift the positions. So I'm actually right 100% of the time : ) That produces a very good positive expectancy for me. I know how to do that with OTM, I am not sure how to do it with ATM because playing with backtesting right now is not yielding any good result.
Well you know what, I'm just gonna start selling 1 contract SPX ATM (using my models) and we'll see what works better. I have no horse here b/c if the ATM works better I'm happy. Remember the name of the game is not to be the smartest trader it's to make the most money. I will even post results for you : ) And then Maverick can also share his % gain from his strategy (that way it will be clear if my ATM strategy is a lot worse than his then I'm not doing it the way he would).
Well you know what, I'm just gonna start selling 1 contract SPX ATM (using my models) and we'll see what works better. I have no horse here b/c if the ATM works better I'm happy. Remember the name of the game is not to be the smartest trader it's to make the most money. I will even post results for you : ) And then Maverick can also share his % gain from his strategy (that way it will be clear if my ATM strategy is a lot worse than his then I'm not doing it the way he would).
I think there remains a terrible misunderstanding on this thread first of all, what a model even is. Models in finance are NOT used to predict a specific price or a specific vol. Nobody is talking about predicting the future. The problem with ET is there is a huge distribution in both intellect, experience and understanding which makes these conversations very challenging. There is no concept of right or wrong in finance. This talk about static positions and adjustments and being right speaks volumes to just the total lack of understanding here.
I'll try to spend some time later and give a 3rd grade explanation to what a model is and what a model really does and how models are really used in finance. Spoiler alert: it's not to predict things.
I'll try to spend some time later and give a 3rd grade explanation to what a model is and what a model really does and how models are really used in finance. Spoiler alert: it's not to predict things.
Mav, cmon man your smarter than this. Of course the model itself does not predict price, its just a mathematical construct that describes a situation in this case the market. But the whole point of creating models IS in fact to predict price movement. You said yourself you have a model that helps you understand volatility.
Instead of wasting time describing models, how about you discuss your edge. That is what the conversation was about initially and I'm still waiting on the answer
Instead of wasting time describing models, how about you discuss your edge. That is what the conversation was about initially and I'm still waiting on the answer
Maverick had already given us his methodology and definition and that was:
May I ask: In that case if I buy options, going OTM or ITM should be better as it is the mirror image of selling and gamma should work for me?
Appreciate your insight.
Regards,