Yes I do know that the single option probability remains unchanged in a random system Mark - thanks. But also note the price volatility changes each and every option period. This complicates the probabilities even further. So I just use averages. Observe that a very unlucky person in theory could lose 100% of the time on any option or equity he every bought or sold no matter the quality or risk of his "investment". We just do not see it that often in practise (or maybe such losers keep it to themselves and don't tell anyone and that explains why its not seen). The probability I was expressing was the probability of losing 3 or more times in a row for a static or average win probability. That is simply calculated by taking the loss rate (100% - win rate) multiplied by itself for each attempt (e.g. 90% win IC win rate [if a real valid probability] would result in the following probability of 4 consecutive chances of losing in a row: .1 x .1 x .1 x .1 = .0001 = .01% probability) . Is it possible to have 4 or 1000 or more runs of losses with an IC win rate of 90%? Yes it is. Is it probable - NO, it is very improbable. Do we see it often? No, since few have the financial balls to do it and since after too many consecutive bad luck runs there is not enough demand for the products (over supply) or enough free market money available to make a market and net credit probably trends toward zero to compensate for over supply ... TrendSailor
so , you want to double down becouse the rate win is 90%...Why not to trade 5 , or 10 times more next month ? The odds for losing next month are only 1:80 , no ? Sounds like a good deal to me.
Hey some of us tend toward pessimism and cover our shorts rather than go naked shorts and assume we might lose 1, 2 or even 3 in a row. If I bet 5x or 10x on attempt number 2 that would probably take my cash down so low I could not recover this year and be able to come back and try a 3rd or 4th time. You sound like a real gamblin man - want to lend me the money at what ever interest rate you want to try it? I'll let you hold my shorts as security if you buy the insurance/longs. I'll trust you to pay me back for my 2 cent investment... TrendSailor
landing money to strangers that I meet in chat rooms is not a gambling , its very secure and proftable biz ; I do it all the time. What is your account # and how much do you need ? PS : Riskarb is out and TS is in , sounds like a fair trade , as long I am learning here something
Trendsailor - funny how you came to this site as I was surfing the web looking for sites on SPX credit spreads. I also got burned on some DIA spreads - up until this rising knife I was up about 20K this year - now I'm down about 40 and some change. What really bothers me is that I almost closed out my spread with a small loss but decided to hang on thinking the market would drop since it was at a resistance point - instead it rocketed up another 100+ points the next day and I was caught like the dear in the headlights. Been reading through this journal off and on today and I've picked some good pointers. It is a little comforting to know that even the pros got caught in this tsunami of a market. I did have a question to the coach which I should probably ask on a seperate thread - but I would like to know if he has posted all the spreadsheets from his trades somewhere in this journal. I've seen a couple but I would rather not try to swim through all the posts to find the ones with the spreadsheet posted.
TrendSailor, I cannot help you in your trading strategy, but I probably can help you in probability. Your probability is correct for losing 3 consecutive times in the future (i.e. the 3 events haven't happened yet, and they are unknown). The conditional probability that you lose the third time given you have already lost twice is the same as the probability of losing one time if it is a pure random system. I don't want to argue if the market is a pure random system. I believe the proper position sizing is the key to trading success. Let me give you a probability question. If you have a 60% probability of winning with 1 to 1 risk/reward (a positive expectancy), do you think you'll win in the long run if you don't have unlimited capital? The key comes from the following question, "How much do you risk each time?". If you keep betting more when you lose, you actually will lose all your money. If you don't believe me, write a monte carlo simulation, and you'll see.
Any way someone can bookmark this post and have it pop up the first ten times a newbie views this thread?
Yip - thanks for your reply. This is the meat of what I wanted to get to. I also agree that the trading size is the key and that one must NOT play too large an initial position and lose early and be unable to continue and have a net profit in the future. Pragmatically, without running the proof we all probably die of old age trying to win back over committed principal that is lost early on. Unfortunately I did not glom onto this truth/concept until after I started the IC strategy and put more of my principal at risk than I wanted to or should have exposed. The TEMPTATION now is to double down and try to accelerate into getting back to "even" and then taking the lesson learned and make more moderately sized trades and go slower with accumulating wins and losses in a more tolerable way. I know its a rationalization but I only want to do it a few times to get back "even" and then from a position of normalcy defer to good statistical behavior and not challenge the odds gods with such blasphemous displays of arrogance and easy profits ever again. These probability gods just want respect and an occasional pound of flesh. Now if I can just sneak a few trades past them and trick them into taking their pound of flesh from the market makers I just might be able to sneak out with my wallet and my derriere intact... The investment strategy that keeps screaming at me from the ether of common sense tells me to have strata or hierarchy of risk profiles extending from a very stable and proportionally much larger core holding. The core holding would be in very secure margin-able securities that pay a reasonable rate of return (5-10%) and the outer band of riskier investments never exceed more than about 10%-15% of total net assets. The optimal rule I would imagine would be to capture profits along the way and only expand the higher risk side of the portfolio (e.g. S&P ICs) as the total account accumulates significant profits in occasional quantums of expansion. I'd envision it as a kind of "quantum investing" model with a hierarchical strata of investment risks that radiate out from core holdings that expand the outer rings as the core mass increased. My gosh, it may come down to a system that looks like planetary orbits or the modern physics model of the atom haha... If Avogadro's numbers, Fibonacci sequences, pi constants, natural logs, energy states and Einstein's equations start showing promise of predicting market behavior I think we may be close to a grand universal unifying theory for everything but how to beat inflation/entropy. Sorry, slapping self - late here and waxing philosophical... TrendSailor p.s. I find your question intriguing. What is the "optimal bet size" according to probability theory or Monte Carlo? I think it must scale to investing at large and should be a general rule that can be applied for an arbitrary basket of risks (even if we do not know the real risk and probabilities or the shapes are not binomial/normal.
Let them have it TS-- starting here first http://www.arbtrading.com/kelly.htm or here http://en.wikipedia.org/wiki/Kelly_criterion ~B
TS: If you anticipate that the probability of winning each month is approximately 90%, it seems to me that the BEST way to make your trading system very profitable is to be certain that the losing months don't make a significant dent in your bankroll. That means accepting losses when necessary, and locking in those losses early (by closing the position) and moving on. It does NOT mean doubling down for your next trade. If you do this, you will lose more often that your anticipated 10% because part of the time that you close the position for a nominal loss ealry, the position would have turned out to be profitbale if held to expiration. But, it's still a winning strategy because you will never (barring a black swan event) experience a large loss. Mark