Appropriate Account Sizing for Trading Strangles | Skinny on Options: Data Science

we're recovering today okay so we're looking at a way for that big farms used to assess risk I'm in Montecarlo simulations so it's something we talked about a couple weeks a few weeks ago and when my segments kind of did an introduction the Montecarlo simulations yes this is getting a little bit more exact and putting you know a solution to a question that people have which is how much money do you need per one lot when you're trading a strangle for example and this is just it's a way of simulating how well the the position would work out over time how much money you need in order to trade successfully because the problem is is it if you don't have enough money kind of set aside for a particular position you then have to do one of two things you know you have to put more money in the account or you have to close your position down and then you're not allowing the probabilities to work in your favor and so placing kind of a data science e-type way of solving this problem we have what one Monte Carlo simulation already built into dough which shows you that chances the statistical chances of you getting to a certain profit level mm-hmm which is cool we're gonna do risk we're actually gonna do losses coming up on the upcoming release so they'll be cool too and some of that math is just you know it's pretty wild it's really cool yeah very cool so so let's start off and see where you're going with this and and remember don't this is one lots and that's why the account size is small that's we're using small numbers chair but you can extrapolate that to whatever you know absolutely no percent okay so without a properly sized account you will not have enough staying power to take advantage of the option markets propensity to overstate actual volatility lot going on for you right now but how much capital is needed per one lot here's your Monte Carlo approach absolutely all right cool so the Monte Carlo recap Monte Carlo simulation is a mathematical technique that allows researchers to account for the risks by displaying a range of possible outcomes from the possible outcomes we can arrive at probabilities the ability of losing half or doubling account size exactly yeah so instead of just saying instead of just coming out with an average or coming out with you know one number a solution I can say okay we can place probabilities around it so instead of just saying okay your average customer makes fifty seven dollars on a strangle I can then say I can instead say okay well 10% of the people make this amount twenty percent of the people tend to make this amount and you know on and on and on and so you can place an entire distribution around something rather than just a single solution okay so how much how much should we set aside for one standard deviation strangles so here's your dad go through the data science take us through the data size oh sure so when you're setting up one of these problems right you you kind of have to set it up and set up your simulation in this way so what I did is I did a back test looking at ten years worth of options data and so and I'm at answering a lot of questions too that people had and a few weeks ago yeah in past studies so one is okay let's back test to see how well it works when you're doing a one standard iation strangle closest to 45 days you're managing that position of 50 percent of max profit so if you make a $2.00 credit on the particular position you get that much credit in you take it off when you're at $1 the reason why we do that is because you tend to have a much higher P&L per day and then if the loss gets too big I'm just gonna say okay you do something with it in this particular case I'm saying let's just take it off what we tend to do is we tend to roll those positions but we need to kind of set some boundaries when you're doing a simulation so I'm saying hey let's start with a $4,000 margin account okay now when I say a margin account I'm meaning like you said we can extrapolate it so it's used $4,000 make it similar and they're easy and $4,000 is because 20% which is the which is just used a 200 on a strike price yeah 20% of that for a-1 Lada sport are exactly spiders yeah and so if you get down to $3,500 then you have to do something you know then you get kind of effect on margin call so you have to take that position off or not so I'm kind of simulating the whole process then I'm gonna compare everything to a by an old approach so I almost have a small move by the way exactly okay yeah yeah I mean so when you're doing this kind of closer to 10% move so it's a very small approach yeah I move so when you're doing a Montecarlo simulation you're sampling from past distributions so in this particular case let's say you start with $4,000 count okay and you you kind of close your eyes you sample from a distribution and you look okay I made $200 on this trade okay now we have $4,200 okay you reach in you sample another thing okay you lost $1,000 in that particular trade now you got $3,200 on that you know overall and and so it shows the progression of one account over the course of a year but we do that times a thousand we're simulating a thousand individuals putting on one station strangles over the course of one year or comparing everything to just a buy-and-hold approach which is very similar we get ten years worth of returns in the sp500 we start with a $4,000 account the same we go okay what's a monthly return okay let's close our eyes take out the the monthly return okay we have a 5% up move okay now we end up with x over the course of one year also and we do that a thousand times and so we're comparing so this is this is your back test for selling one standard VA shoe strangles yeah this is kind of a cool have some really cool that's really cool visual here yeah but you can see that you know we have far more green dots meaning that when you manage these positions you tend to have which should be like 92 percent I don't know anywhere from 85 to the low 90s probability of success and as you can see dryers soo however exactly however when we have losses they tend to be pretty you know somewhat large yeah the nice thing about it though is we have far more winners and we do losers in those winners tend to compensate for all those losers that we had the few losers that we do have and so like I said I'm sampling from this distribution that's and I'm reeling it that's a grid view of it over the course of one year somebody doing 45 day things as soon as they take it off they implement another one I'm doing this for simulating a thousand vigils doing this transaction sure and then on the next slide took a large large numbers right exactly on the next slide okay this is some ten years worth of sp500 by an old approach okay dude yeah so on the first slide exactly you know this is just much more of a yeah you know it was boring yeah it is freakin boring but in 2008 the results are actually even worse I mean if you look at him if you line them up together I mean they're not even you know the do you have - three meltdowns yeah yes and and and two or three mill towns verse there's a lot more volatility this is just a very passive by an old approaches but a lot of people do I really don't get it no I get it I mean I get it you don't get it that you get that I get it whatever but this is not that this is not the simulation now this is just kind of setting up the problem though and this is the the distributions from water sampling do you know in order to kind of solve the the problem or get to an ideal amount okay so here's the thing if you start with okay remember you need about four thousand dollars to generate this and then one station angle we're not talk about four thousand dollar counts just know no one's clear about this for once we're just yeah we're just talking about one lots using four thousand dollars we're not really talking about a fourth oh here's the thing you know sometimes people think okay I got a $40,000 account I should trade ten Lots all right okay maybe so maybe so maybe that's that's right for you however keep in mind that if you do do that twenty according to simulation twenty one percent of the time it dips below thirty five hundred and therefore you one or two things you got to either put more money into the account or you have to close down a few positions this is not allowing this is a spectacular piece of research based on you know the just a pure number of occurrences because that's the question that we get eight billion times I want people succeed at this you know and and if you're not funding your account and you're not trading appropriate-sized you'll be forced to take some action which you may not be comfortable in this answers the question of why we always talk about kind of the linear nature of capital and why we always talk about you know those who have more capital can stay small and really do whatever they want to do because as you see here you know you're you're barely ever affected yes not touched so when you're when you when you have a four thousand dollar account you're using like ninety percent of your account for that particular position and so you only have a small move before you basically have to to do something you have to liquidate or not necessarily liquidate but do some action when you do we have five thousand dollars in that account or per that one lot okay now all of a sudden you use and about 70 percents you know you know that account so therefore you have more room to play so that in that case only 7% of of accounts dip below thirty five hundred dollars okay and then when you get to six thousand dollars okay that's more appropriate per one lot that you should kind of set aside because not only about one percent of the time do you reach below you go below thirty five hundred dollars and so you have less action required reason why we're saying this though is that we want the probabilities work out in your favor and if you're having to do some action you're not allowing those probabilities work out the the fact that a volatility tends to overstate actual volatility I find it fascinating that that and this makes the argument I think for active for the for the act of trading because the difference when you get to a reasonable amount of capital in the way you're trading is so it's so minuscule that that yeah you know you have to go for the higher returns just a sense not to go for the higher returns but keep in mind that 20% of accounts or twenty percent of the time when you have a four thousand dollars in the account you're having to do some action whereas with a six thousand dollar account only about one percent of the time right when you look at it on a on a passive level to you know it doesn't it's only half a percent so there's nothing much difference exactly but that the whole fear everybody has is that you know the active will kill you when when the difference is minuscule exactly so this just shows now that okay if you're able to stay within your trade yeah this is where the probabilities start working out in your favor so if you just look at the 50 percentile meaning you give a thousand people in or of the course of one year right and then right in the middle you're a kind of typical person there you know will tend to make about twenty-two percent if they're trading a four thousand dollar account keep them on though that about twenty percent of people have to take some action and have to put more money into it so that may is probably not all that appropriate to you now if you put six thousand dollars in the account that's what I would recommend you know put six thousand dollar in your account per one lot and so in that case the 50 percentile makes fifteen percent at the end of the year compare that to a buy-and-hold approach now keep in mind that you're using a hundred percent of that account when you're doing a buy and hold approach so that's why the numbers 8% 8% 8% or all across so you tend to make about 8% when you're doing a passive buy an old approach now look at the bottom 25% you know that's what I think is kind of cool the bottom 25% of a typical buy and hold approach you're losing two percent whereas the bottom 25% when you're doing a one sanitation strangle and you're trading with a six thousand dollar account you're making four percent that's really cool that shows why while the method works you know you're allowing their abilities to play on your favor it's really powerful it's great this is a great piece of research I mean it's just you know it's heavy it is heavy you have certainly seen the numbers plays heavy but it's technical and this is a two for one job man it's it's two for one it's not like we get two for one anywhere else in the world it's jump all over we jump all over but if you're sitting at a blackjack table and you say listen you know what this has gonna be two one for you maybe we can double down if you have the right cards you go sure sure why wouldn't I right absolutely I mean like a lot of times this is a cool thing about like data science right we're using empirical data the back of our decision-making process instead of me just saying well you know okay based on my historical performance okay I am able to see to five hundred dollars and of course it's gonna pull it on the button you hear I'm actually I'm backing it up by data you know backing up by information backing up by simulation and the simulation is based on actual data yeah that's great stuff I think like I said I think it's heavy I think it's an awesome study takeaway so there's more advantageous to have enough to cover potential losses that's capital it makes it possible to take advantage of implied volatility prosperity and to overstate actual volatility and the simulation points to having six thousand dollars for each strangle or to having one and a half times the amount of capital required as a minimum forget about that we just we just reduced everything down to the lowest common denominator sure sure did not trade just because you got a six thousand dollar store doesn't mean that you have to use a hundred percent of it we're starting to show you the math a lot of the math behind the logic of trade small trade often which is which is so important that nobody gets of course good job by you did a good job by the team team we take a quick break with the e-mini S&P is trading their lows down fourteen we got Wow Mergen measure next the state real life thanks for watching if you found this video useful give it a thumbs up or share it with a friend click below to watch more videos on this subject or subscribe to our channel for even more data and research to check out tastytrade.com

Channel: tastytrade

Views: 21,920

Rating: 4.9388647 out of 5

Keywords: trading, markets, Michael Rechenthin, Dr. Data, data science, implied volatility, expectations, monte carlo, strangles, account sizing

Id: pDz10RKJdp0

Length: 14min 36sec (876 seconds)

Published: Fri Aug 05 2016