Transcription
Okay, welcome everyone. Um, here we are on another episode of the podcast with Justin K, aka Special K. Um, I had a few courses on Option Omega. I've known him in Discords for a while. Uh, I value him as a very good source of a contrarian view. Um, I often, there are very few people that I think are more intelligent than me in certain aspects, and Justin is definitely one of them. Um, I think he brings a very good take based on his trading and also unique life view, being that he used to be an attorney or lawyer. I don't know what the difference is between the two. Um, so I appreciate that detail. We basically brought him on. So, if you just want to give a little intro of who you are, how you got to options, and another thing I normally ask people is, it's like, what was kind of one of the last aha moments where you're like, "Wow, this was a poor assumption," and then you eventually work through it, and how that kind of helped you improve your trading?
Sure. So, um, I am a practicing attorney. Um, not currently practicing, as weird as that sounds. But, um, I got into finance law, did a lot of bankruptcy, tax, civil litigation, um, asset planning, tax planning, a lot of that kind of stuff where I eventually moved, uh, in-house to a couple different companies. Ended up making it up the, um, corporate ladder, uh, getting into the C-suite type stuff. Um, I started trading in equities, uh, in the late 90s, I think 1999. Um, things were very, very different back then. Options weren't really, weren't very, um, prevalent. It was also very expensive to trade. There wasn't, you know, commission-free trading. There wasn't, uh, fractional share trading. I mean, there wasn't any of that stuff. Uh, a lot of the ETFs really weren't available, uh, especially to retail. Uh, and I spent a lot of time figuring out what didn't work for me. I found a lot of quote-unquote gurus, none of the ones you see right now that are hanging out there, but, you know, the ones that were on TV back then, uh, Jim Cramer, Phil Town, tried a bunch of those. None of those really ended up working for me. Tried, um, uh, you know, uh, the CAN SLIM method, a variety of different stuff. And sometimes it would work for a little bit. It really kind of wasn't my style. Uh, did a lot of YOLO type of trades in options. You know, hey, I think, you know, uh, Google's earnings are going to come out, you know, back in the, you know, 2000s, and I think it's going to go through the roof, and you buy a way out-of-the-money call thinking that it's going to, you know, end up working, and I'd find that I'd be right about the direction, and I'd be right about the time, and I'd still lose money. And, you know, it was mainly because I didn't really understand options, and I just thought it was a leverage type play, and that was the end of it. Um, and then it was about 2015 that I kind of said, "Okay, like, let's actually learn what this is all about." Uh, started reading a bunch of books, trying different strategies. Uh, ended up going more into the wheel originally to try to learn how they work and the pricing and all those sorts of things. And eventually, it kind of ballooned into starting 1DTE in 2018, uh, starting zero DTE in 2019. And then I've slowly been building portfolios of various strategies. Uh, the information available back then was really pretty poor. Uh, you had to backtest through Thinkorswim's OnDemand or their ThinkBack. Uh, it would take you two, three weeks to backtest a strategy going back two years. Um, and then E Delta Pro came out, uh, which had end-of-day option pricing, and you could just click a button, which was amazing. You had no idea the value of the option in between, but you knew what the price was at the end of every day. So you could backtest a lot quicker, and it changed some of the strategies. And then eventually Option Omega, which caused you to see intraday pricing and intraday Greeks as well, which again accelerated a lot of the strategies.
About 14 months ago, the company that I was working for decided to, uh, well, didn't really choose, was forced to close their doors. Um, and I kind of stepped away from that environment. It was, wasn't the best time. It was a very stressful time. Uh, being the one that went from, you know, uh, 800 employees, I was the last one of the last eight there, and we padlocked the doors and handed the keys over to the trustee. And I just kind of decided I was going to take some time off. Uh, I didn't know if that was going to be two weeks or, you know, what that was going to be. And I had built all these strategies, so I was trading full-time. Um, de facto, kind of, I was doing what I was doing before, but now it was my sole source of income, and that's been, you know, 14 months ago. So that's what I've been doing since. Uh, it continues to kind of work for today. I don't know if it'll continue to work for the next, you know, 50 years or whatever. We'll see. Uh, it, I have all the faith in it. All the models appear to line up, but we'll see how it goes. Uh, so right now, I'm a full-time trader. Maybe getting back into the legal world. Maybe, who knows, uh, take a lead from from you and start a a fund. I don't know. So that's who I, um, that's where I came from.
Yeah, it's a very interesting journey. And like, I also find that a lot of people, um, like, I think the saying is it's like, uh, you know, losing your job can be the best thing that happened to you. Just don't make a habit out of it. And, uh, I really kind of believe, um, in that. But, uh, a few things that you've kind of said that I found are pretty profound is, we'll kind of start with this one. I made a video on it, and it was in, uh, you know, one of your resources as well. But you kind of talked about, um, basically the VIX regime and how it changes because it's a very, so like, big fallacies with traders out there is like they become a slave to their models because this is kind of what people say on Twitter, but then they don't really think rationally about it. So, a very popular filter is you're using an arbitrary VIX number of 18, 27, whatever, to make trades. But I think what a lot of people miss is like, VIX is always relative to movement. So, like, it's static throughout time. However, different things happen in the world that basically change it. So, like, you know, the average VIX in 2008 was like a 35 handle, and then these times trades. So, if you kind of want to talk a little bit about what your findings were with, you know, VIX and how that relates to regimes, and that it may not actually be the best barometer for volatility in general or a proxy for taking risk.
Sure. So, I see a lot in the Discord where people share, um, backtests, particularly Option Omega, but I mean, other ones where they say, "Hey, look, you know, if you filter for this VIX, you know, if, uh, you only trade when VIX is under this number, or you only trade when VIX is above this number, the returns improve." And maybe over that time period, it kind of does. But what it really leaves out of a lot of that analysis is what are you excluding, and what will you exclude moving forward? So, if you ended up picking a number like, you take a look at the VIX today and you say, "Okay, well, you know, VIX 30, for example, might be really high, right? So, okay, so let me not trade this in really risky environments above VIX 30," and you exclude it, and you see, "Okay, you know, the returns look better." That's great. What you're not really accounting for is what number of trades are you really removing? Are you removing five trades out of 500 because VIX was really high for two days within that time period that you had, in which case that may not really be statistically significant? You're just picking five random days that ended up being losers, and it improved your returns. That's one thing to look at. So, instead of picking random numbers, you might want to take a look at, "Okay, let me put it into five different chunks, you know, where I end up finding out, you know, 20% of the time VIX was is within this range, 20% within this range." So, if I remove the top 10% of trades, or top 20% of trades, or top 40% of trades, what does it end up looking like? Does it change? It may not improve things, but it might change the metrics enough to say, "Okay, well, that's kind of what I'm looking at." Not straight VIX numbers, but, "Okay, moving, removing the top 40% of trades within that backtest timeframe." Then when you take that, you want to actually back up a little bit and take a look at it a little bit differently. So, VIX is mean-reverting in that when VIX is low, it has a tendency of hanging there for a period of time until it spikes higher. Eventually, it will spike higher. When VIX is at eight, it can't stay there forever. Eventually, it'll go to 15. Eventually, it'll go to 30. When VIX is high, it can't stay there forever. Eventually, it will come down, right? When VIX is at 70 or 50, eventually it'll come down. You don't know when, right? For, you know, five years straight, VIX could be in a range of 8 to 10. And in theory, for five years straight, VIX could be in a range of 40 to 70. I mean, that is theoretically possible. But what you need to look at, not just is VIX mean-reverting, but is the mean staying the same over that extended time frame? And if you zoom out on 15 or 20 years worth of VIX data, you'll see it's not. So the mean changes over various time frames. So the mean may be 12 during a five, eight-year time frame, and then it may be 30 over a different five, 10-year time frame, and then it may drop, it may go up. So when I look at it and I say, "Okay, I want to remove everything that's over VIX 25 because that's too risky based on my my parameters that I set up." When VIX is at a mean of 15, removing everything over VIX 25 is very risky. So you're removing that. When the mean is 30 or 35, now you're actually accounting for 60% of your potential trades that are happening. You're not just removing the most risky, you're removing the average and under. So now you're removing so many more different possibilities that could happen based on where VIX is today. So taking a look at different time frames will help with that, but also having a logical understanding of like, "Okay, what is VIX at during this time frame? What's the average? What's the range? What's the percentages of different directions that VIX is at?" And then let me take a look at, you know, "Okay, we're moving the higher, the lower, 50%, 40%, whatever that number ends up being." I'm not a huge fan of filtering with VIX. It's great to get a picture and an understanding, but, you know, some strategies, you know, zero DTE in particular, you get paid a lot of premium when VIX ends up going high. Which means that you could be wrong a significantly larger number of times and still be profitable. That's not really accounted for when you just say, "Okay, let me slap a VIX number on on the table and see what it looks like."
Yeah. And the thing that I find kind of really profound like this, and we can even relate it to other things. So these were some of the thoughts that kind of popped into my head is like, you're changing VIX regimes basically constantly, but no one knows when that VIX regime changes, and no one will know when that is priced in. So, like, even some examples, it's like, one, Donald Trump was originally president, 2016. I think we're in a market of kind of strong Fed, weak president, ala 2018 with the rate cycle. And then basically starting in, you know, 2020 with Joe Biden, we transitioned to strong Fed, weak politicians. And then it looks like now we may be transitioning back to, you know, strong politicians, weak Fed. However, bond vigilantes could basically change that. And, you know, what's popular out there is the back-of-the-napkin form when we talk about Tastytrade, and they talk about IVR, implied volatility rank. And basically what they're saying with IVR is they're saying the last year is the VIX regime. They're just arbitrarily assigning that, you know, time frame, and they're saying when it's high compared to the last year's data, it benefits. Which you can obviously see is a very non-nuanced approach to what that is, almost as much as just using a random VIX parameter. Or another example is, you know, the lowest 5% and the highest 5% of the year in VIX ranges tends to probably not do that well. Um, and then in like 2024, when we were below a 15 VIX, it got absolutely murdered. If you look at the data in 2017, when we were in an extremely low VIX regime, like VIX 15 to 11, it absolutely printed, and the market didn't move. And then when VIX went below 11, it got basically like completely, um, kind of tightened in that. And the other concept here that I kind of realized from looking at that is that when this regime changes, we don't know when it is. We don't know until hindsight. Is that it takes time for that risk premium to get repriced in? And one thing that kind of makes sense to me is that it tends to appear like these shorter duration options get that priced in in a shorter period of time. Granted, it's also getting repriced more. So, there's always some type of trade-off with that. And it's not necessarily the end-all, be-all, because if you can't handle the change from a low volatility regime to a high volatility regime with your trading, you're obviously going to get absolutely destroyed in zero DTE because it happens five times more than it does in monthlies. But if you've gotten to the point in your trading where you can handle that, you know, after you get over that weak hump of getting smoked out from those volatility changes, all of a sudden you were in a 2% drawdown. You can go back to making money. So these are all kind of like interesting concepts. The one thing about notional VIX, and I'd be curious of your thought on this, is like there's a large amount of structural flows from ETF option-based products, and these are only going to grow more because what a lot of people don't know is that when you trade options ETFs, they're making 100% return on your account for the margin of treasury. So with the 6% rate, they're basically in the return of the stock market plus any type of options overlay. And a lot of these V-based funds or leveraged funds do have things in their prospectuses at VIX 30, they have to do some type of rebalancing, some type of risk parameters. So like that is one notional thing that like I would say, you know, holding something overnight or there could be some type of bigger move, it may make sense to do. So personally, I haven't really adjusted any of my trading for what I do, but generally intraday long volatility, which is zero DTE, which I run a decent amount of, tends to appear like it really drops off of the VIX 30. And my thought with that is just from an empirical sense, open to close is generally mean-reverting over a very long period of time. All sustained market movements happen overnight. And it also appears that when we get above that handle, even though long volatility works very good in ZDT because you're only holding it for the day, so even though over time it does nothing, there's lots of ups and downs, it seems to mean-revert on a very short basis, at least from price for premium to buy, and those occurrences above 30. So that's one thing I've noticed with the actual barometer, but we have based it on some type of theory, and I just think that's very interesting around the VIX and kind of what I got from your initial idea.
Yeah, I mean, I completely agree. I mean, part of the questions you need to ask yourself is when you're trading, and, you know, you're used to being 15, and it jumps to 35, is it jumping to 35 temporarily? Temporarily meaning a day, you know, a couple days, a week, or is it on its way up to having that be your new mean, you know, your new average on an extended period of time? And you'll never know until you're in the future. But how long it takes to get to that new mean number really matters a lot in the mentality and a lot of other traders and how they, you know, dictate their trades. You're right. There's some ETFs that, you know, hey, they have rules in place that, you know, once VIX hits a certain level, they need to shut their trades off, or they need to change their trades, or they need to go long vol instead of short vol, or the other way around. But also, some of those funds, they don't really meet that frequently. They don't meet daily. They don't have automatic rules in place that say, "Hey, if VIX is above 30 at noon, stop." Some of them, they meet every Friday, or they meet, you know, twice a week, or once a month. So, and it depends on their strategy. So, if VIX jumps to say 35, premiums inflate, and then suddenly news comes out and it comes back down, it may not change it the same way as if it extends for three days. And if it extends for three days into a Friday that one of these funds meet, it might be different than it extends from Monday to Wednesday. And none of those anybody really pays attention to. Instead, they just say, "Hey, look, VIX is high. I'm out. I'm going to sit it out," when all of these funds are just plugging along and doing the same thing that they were doing because their rules didn't actually implement the change until it was VIX 32 or 35, and you didn't really notice that. What you do get the option to do within 1DTE and ZDTE, like you mentioned, you get to see the, um, option chain expand when one of these events ends up happening, and then see the pricing that occurs within it, and then actually observe how it's pricing that in. You can see a lot of that with the, you know, Liberation Day that occurred, especially with 1DTE. 1DTE got absolutely hammered, and in large part, you had one of these, um, change in potential implied volatility events as the markets just jumping up and down and going crazy because it can't average any of this stuff in. If you took a look at what a 15 delta put, for example, was on the 1DTE basis, you know, before Liberation Day, it might have been, you know, for example, you know, 150 points out of the money. Well, right afterwards, normally when the market drops 3%, it now estimates a 15 delta would be closer to four, four and a half, 5%. Now, if it's still pricing it in there, it has priced that into, um, the option chain automatically. What we saw the last time was, you know, market dropped 3%, and a 15 delta put was 2.2% out of the money, which is insane. You know, I mean, 15% chance of occurring. You know, we saw it happen today. We saw it happen three of the past four days, for example. So, you can look at that and say, "Okay, it's not really pricing it in as much." It prices things in much faster on zero DTE and 1DTE basis than it does on 90 and 180-day because a lot of other events can happen moving forward. But you also get the added benefit of compounding choices. So, you can have five losing trades with zero DTE, and then it prices in on the fourth, and you're good to go. If you traded once a week or once a month, you might enter your trade at the time that it wasn't priced in, and you're screwed. I mean, that was your trade for the week or the month. Whereas for zero DTE, many small bets, 1DTE, many small bets get you closer to the average of the return, whether it's priced in or it's not.
Yeah. And kind of the other thing with that too is like, a way to think of it is the, you know, it's always priced in, but we have no long, no idea how long it will take to get priced in. And even though the volatility risk premium is always present, how we harvest that volatility risk premium will basically change over time. So, you know, this is, I think, a good way to segue to hedging. Is it like the, you know, perfect hedge should be close to random, like a near expected even return, because the purpose of a hedge is to offset your risk. And obviously, the idea of a profitable volatility selling trade would be as close to basically statistically significant as you can, 99% certainty or whatever that is. So, when those basically cross from those regime changes, that randomness can play in. Like this move, you know, historically for holding overnight vol, VIX was pretty terrible in relation to other things. But, you know, the awful beauty with holding overnight VIX is like, if the VIX rips up a bunch overnight, that has to come out during the day. So, like, you just sold something that you thought was priced in. It clearly was not yet. The market opened up the next day, but now the volatility is richer than what it was the night before, where there's an illiquidity premium because you can't trade overnight. And then that vol has to expire by expiration. So, like, these are always the trade-offs, and it comes down to sizing and understanding what you're doing. Because, like, if one part of your portfolio or VRP selling has fallen out of favor, there should be another part that benefits from that, and it doesn't necessarily just have to be a hedge. That this is something that I think we may have a different view on, but if you could kind of talk about your thought on hedging, what you kind of view that, um, one that I agree, I think we both hate is like the less than one delta put that you're somehow able to magically liquidate at the open. Um, you know, yeah, so I, I don't like that one, but I'd be curious on kind of what your thoughts are around that, and then we get a more tactile about it.
So, for me, when I'm building a portfolio of various strategies, you know, each strategy kind of has its own, uh, metrics on what's, I consider acceptable, what the target is, what it's trying to do, right? So, if I'm selling, uh, premium on zero DTE, you know, hey, I'm kind of expecting implied volatility to be overstated compared to what realized volatility will be. What I don't want to do is add a hedge that works exactly opposite to that. Right? So, if I want to sell premium on, you know, 1DTE, 7DTE, 30DTE, 90DTE, whatever you want to end up viewing as selling as far as your premium. I don't want to turn around and then say, "Okay, I also want to buy volatility at the same time." Because volatility will both be, you know, overstated and understated at the same time. Logically, that kind of doesn't make sense. And you'll see that sometimes with people that share some strategies on some of the Discords are like, "Hey, look, this worked." You're like, "Yeah, that's right." Because it's a random opposite side of the trade that you're currently taking. So, while, you know, yes, you get a dollar, you're going to turn around and hand that dollar out as you're buying it, and you're going to lose in both scenarios. Hedging, though, in that scenario, really requires your sizing to be on point. That's the most important thing because you can have short, really difficult that I think people miss is like, you basically have to use SPX because it's kind of settled, and that's what you're selling. And SPX is a $600,000 instrument. And like holding half SPX notional in a straddle, we're talking about 5% type portfolio drawdowns. And then XSP for as much as people are like, "Oh, it's one-tenth the size, you're paying like 50x the bid-ask slippage on that." Which, so that's just one thing where I think everyone really gets missing with hedging. It's like hedging is like having a will and inheriting a bunch of money from your parents. It is very much a luxury for larger enactors in the market. Like, we'll get into how you should look at credit trades, but continue to go.
The, the other thing that's challenging about a lot of hedges is, I mean, you've got like the straddle, for example. If I want a hedge, I actually want it to reduce volatility. I don't want it to increase volatility. And that's why I mean, not that the straddle's not a hedge, it's not necessarily for me because the entire point of a hedge to me is to make things more certain moving forward and remove outlier events, not increase outlier events. You know, there are other strategies you can do. I mean, you know, buying the teensies 90, 60 days out. The issue that I have with those, while you can run backtests and they cash out, you know, sometimes once every five years, sometimes three times a year, statistically it's complete randomness. I mean, you're just guessing on what's going to happen. I mean, you know, the probability that it's going to repeat itself is as close to zero as statistics and estimate. But at the same time, you got to understand that the whole point of this game is not to make money today, but to survive long enough to continue to trade 40 years from now. So, because of that, you know, sometimes you do need to implement some form of hedge. Even if you don't expect it to make money, as long as your expectation is the cost of this moving forward is enough that I can absorb it in the profit that my other strategies that I think are going to work can afford. Another reason why I don't like the straddle, it's really expensive as far as a hedge. So that again makes that sizing so very important. But, you know, it, the price of the straddle ends up increasing as volatility increases, which isn't necessarily what I want. It's great if you had one on. You don't have the ability to continue to put a hedge on at that same price point as volatility continues to expand, but that's like that with most hedges. So, yeah.
Yeah. And the other thing that you bring up a really good concept, which is like, even though it's random, anything could happen. And the whole point is staying alive because the compounding is really the benefit. So, like, this is kind of the concept of erodicity, which means we have no 10% average return over 20 years. We have zero return for a decade, and then 20% return for a decade, and that averages to 10%. However, and this is like the classic example where I just like bash financial advisors. It's like, 1968 to 1988, the average return was 10%. From 1968 to 1978, it was 0%. From 1978 to 1988, it was basically 20%. So, if you took the same couple with a million dollars and assumed like a 4% withdrawal rate, if one couple got the 0% return from '68 to '78, they would have $32,000 a year for retirement. That same couple that got the 10% average return, but got the 20% the first 10 years and then the 0% the following 10 years ends up having like $140,000 a year for retirement. So, we need to understand what those confines are. And this is what I assume you do is like, you could just trade smaller, and you would accomplish significantly more than holding the straddle. Also, and this is a guaranteed way that you can offset your risk. And I think this is kind of the next good segue is like, when we're hedging, we always think about kind of like correlation of those things. So, the dumbest hedge is a correlation of one to minus one. Every dollar you make, you lose a dollar. So, like, you're always going to have to take risk with a hedge. And the one risk that you can take off with no trade-off is just sizing smaller. And you're just adding like all this complexity. And kind of the other example here is like, if you have a correlation of 0.8. So, for everything that goes up, you should relatively experience 80% of that. And kind of what this means is that when one asset goes up by one standard deviation, whatever that is, the second asset at 0.8 should go up 0.8 of its standard deviation. And the real key here is assets that are 0.8 correlated basically have about 28% of their risk that is unexplained. So, like, whatever one does is not going to help you out with the other. And this is kind of the concept of idiosyncratic risk, or it's known as risk remaining. So, the one-to-one correlation is basically systematic risk. You know what you're going to get. It's basically only thing in credit market. But then the more uncorrelated you get, the more of this risk remaining, or basically randomized risk of clustering happen. We've all heard things cluster to one. And this adds so much more complexity. But I see people often come at it at way the wrong way. So, they think if they have, they use the word uncorrelated. And I hate this word because sure, uncorrelated means like, 0.3 correlation to negative 0.3 correlation and everything in between. However, in actual trading, there's only inversely correlated and correlated. So, like, think about minus 0.01 is inversely correlated, and then 0.01 is correlated. And what this does is now you've started to introduce a massive amount of risk remaining. So, to optimize basically that concept of risk remaining, we need to optimize basically smooshing our standard deviation. So, like, big mistakes that people do out there is they'll choose very random trades. So, like, let's say someone's running seven credit trades, and they add the straddle. To your point, this is not a very good choice because you just added a massive standard deviation uncorrelated trade to your portfolio, which is just like quadrupled your risk of clustering. Whereas if you had 25 credit trades and you added a long straddle that was such a small standard deviation, it was like the fourth largest position of beta that can make sense in your portfolio. So, these are the real nuance things when we come in to basically the hedges. And you could even say with that random 120-day out teeny put, it's like that standard deviation, at least on a one-daily basis, is like astronomically larger than your rest of your portfolio. So, it's a way to look at it from a risk of execution perspective. So, basically my thought, and I kind of be curious to this, and we can go through a live example, is it's like, okay, you run your backtest, you say my max drawdown is minus 3.5%. You never exclude anything. You have 20 trades, you're all good, you think you're a genius. You run it through a Monte Carlo analysis, and you say over 40 years, my worst drawdown is 6%. So, huge problems with this is like, the correlations of those were optimized based on past data, and since those were by definition random or uncorrelated, 0.1, 0.2, 0, those are going to astronomically break down over time because they're basically 100% unexplained phenomenon. And this is where I would also push on AI, that it's like, no one can predict that because it's already 100% unexplained. Then you go trade it. And since you're using this past data, all of a sudden you start clustering, and you've hit your max drawdown. So, now, how do you know if you're off or if you're on? And like, I know one thing you talk about is the p-value, statistical significance. So, it's like, when do we know we have statistical significance? If we have hedges that are random, that's obviously going to bring our statistical significance down. So, like, what's the trade-off? I know I've kind of been rambling, but from that kind of perspective with hedging, it's like, how do you look at any type of objective metric and then think really critically, um, from an assumption standpoint of like, where is this flawed, and obviously what can I anticipate, you know, this deviating from? And then when I'm lost and have lost my conviction because I've hit this max drawdown that the numbers don't say, like, what can I now bank my conviction on to try to get back on course?
Yeah. And and part of it, though, you know, like you said, you know, you might run a backtest and it says a drawdown of 3%, and you run a Monte Carlo, and it actually ends up being, you know, potentially six, seven, 8% because of, you know, sequence of different returns just ended up lining up differently. That's just based on the backtest data that you have. You know, your actual drawdown moving forward could be multiples of what was in the Monte Carlo simulation. But not only that, that's one strategy. So, let's say you take two strategies and you run a backtest, and you say, "Okay, one has a max drawdown of 3%. The other one has a max drawdown of 5%." And when I combine them together, I have a max drawdown of 2%. That's phenomenal, right? Well, you take a look at the correlation between the two strategies, and let's say you say it's fairly uncorrelated at 0.2, you know, 0.2 for example. Well, what happens if they end up being strongly correlated together at a certain point in time, and that certain point in time happens when both of them hit a drawdown, which can definitely happen, you know, when guaranteed because the correlation from the past mathematically is basically 100% random. Exactly. So, now you look at it and say, "Okay, well, I was seeing, you know, 3% drawdown and 5% drawdown together equals 2% drawdown." Actually, they can compound on top of each other, you know, five, eight, 9%, and that's just based on the historical data. So, by adding those together, what you thought was better actually exposes you to a larger potential loss moving forward if they operate or can operate very, very similarly. Some of your hedges kind of won't. You know, if you're buying vol and you're selling vol, the odds that on today both of them end up exploding and losing large amounts of money and having the max back, theoretically possible, not very high probability, but theoretically possible. But if you have two different vol selling strategies, for example, you're selling in the morning, and you're selling in the afternoon, and then in that scenario, both of them, you know, didn't appear to be very correlated based on the historical data. It's very possible that both of them become highly correlated at a certain point in time, and you hit your several multiples of what your drawdown actually ended up being.
Yeah. And also like when people compare correlation, what they'll do is they'll have a 932 trade, and they'll have like five toners of a condor through the day, and they trade each other one lot, and they say this is a 0.1 correlation. But it's like, what they miss out when they're calculating this is like, you need to compare, like, you can use an R value. So, it's like 932 may have one R trading it, and then the five TR condor could have six R from a max loss. So, then you need to add six lots to the initial one because you're comparing standard deviation to basically standard deviation. And like, this is where I really see that like people lean on numbers and they don't lean on assumptions. So, it's like, if you're trading trend in a condor, it's actually more likely that that could potentially cluster at the same time because if you're trading a condor with a stop-loss, you are in turn trend following because you're saying, "I'm taking this side off when that gets challenged because I don't believe this trend will go." And then if it reverses, you've lost on your trend following condor, which you thought was uncorrelated, and your trend following put. So, like, this is critical to where kind of the options knowledge comes into play of how you're building your portfolio. In the example I was kind of talking with Justin earlier, is initially, if we try to simplify this, if we chose a hundred people off the street, including, you know, monkeys and dogs and Warren Buffett and the best hedge fund managers on Earth, and we told them to choose 300 stocks, everyone would get the return of the S&P 500 because it's just so much data that it's going to get, it's going to revert to the mean. Then, if we chose them to trade one stock, we'd basically be guaranteed that it's pure luck. There'd be no predictive if the dog, the monkey, or the hedge fund managers would basically pick the best. But then as we go from 2 to 4, 4 to 6, 8 to 10, we see that Warren Buffett and the hedge fund managers start to really separate themselves from the dogs and the monkeys and then the random people. And basically, this starts around 8 to 12, and then we don't see much benefit past 20, let's say strategies. However, all these strategies are not created equal. Like you're talking about vol regimes. It's like if you have a hedge that's held for 7 days, that's a lot different than zero DTE basically trading a dozen times a day. Regimes shift a lot faster and more often on zero DTE, and then, you know, vice versa on the other duration. So, I'd kind of be curious with that little facet. It's like, how many strategies do you have, and then how do you view having them diversified? So, it's like we talked about long vol, you know, obviously shows some skewness, fat tail potential distributions, which is going to be like, guarantee a preventive clustering at the extreme ends. So, like, credit trades are obviously going to suffer with massive expansions in volatility, even though you may be only 0.2 to, you know, inversely uncorrelated with the long vol, if we move 5%, like that should make, you know.
Yeah. Yeah. So, I mean, I run on average about seven different strategies, various sizing for all of them. Some of them incredibly, incredibly small. Um, other ones of them, you know, they're they're kind of a large portion of the income generation that I do. You know, some of them have, you know, much larger risk to return than other ones, which some of them are long vol, some of them are short vol. I'm not going to say, you know, I don't counteract myself sometimes. But I'm not buying zero DTE and selling zero DTE. I won't do that. I'm not buying 1DTE and selling 1DTE. I won't do that. You know what I will do though is I'll let some of the potential profits run on trades that let other portions stop out. You know, for example, on ZDTE, I don't close wings. I buy wings. Um, I have stop losses on the shorts, and I will hold those wings till they die. And I understand that it's really just a cost of doing business. I don't expect to make any money off of it. But over the course of the past year, three times some of them ended up landing in the money and by a good margin. And, you know, that's that's the way I do it. Does that count as buying vol at the same time as selling vol? If you want to view it that way, I don't know. But I let my losses be capped by stop losses. Understand it's not really capping my losses, but it's trying to cap my losses and let my profits run and go endlessly. You know, for 1DTE, there's a, you know, um, a lot of people that will buy 7DTE, 5DTE wings instead of the, you know, 1DTE selling, you know, and they put profit targets on those. To me, I've always been a strong advocate against profit targets, trailing stops on any of those. They're hedges. Let them be hedges. You know, if the market explodes up or down, crashes, let your long vol hang out there. That's what it's there for. You don't want to try to capture 500 or 700% profit and then see the absolute market bottom out and you lost out on all of that. Instead, just let them run for the time frame that you determined was profitable. Let your profits run.
Yeah. And what I think Justin does well, and everyone else does not, is like it just comes down to sizing. Like the standard deviation of that reverting hedge, you know, 300% of his profits does not materially affect what his portfolio is trying to do. So, like, it makes it so much easier to do. Whereas if you're holding three times notional in your account on long puts, and you're negative 10,000 S&P deltas, you know, every tick, you know, could be several thousand, it's going to be really hard to hold on to this. And this is just a further example of like, the real benefit is from your sizing. It's always the easiest thing to do, and then balancing that out. And you can also hear from him as like, he's at statistical significance in terms of strategies. He runs seven or eight strategies. He's got, you know, they're actually diversified, and he's thought about critically. And this is kind of where the people, you know, miss a little bit. But I guess, kind of the next thing I wanted to do is kind of show these live assumptions with what I've done. You know, in my portfolio a little bit. And then, you know, I'd be curious to any comments that, um, you know, Special K has as well in regards to this, because I think it's pretty interesting. So, um, can you guys, it'll lag a little bit, but if you could see my screen, you should be good here. Don't say anything yet.
Yeah. Well, uh, I have a big GitHub file on here, so it can kind of crash sometimes, but generally, uh, if you give it two minutes, it will kind of be able to display it. While you're loading it, I will mention that, you know, I prefer to try to find strategies that, you know, every strategy you have will have periods of underperformance. I mean, it just, it happens. It's natural. Sometimes it's a month, sometimes it's six months, sometimes it's three years. You want to find different strategies not just for income potential, not just for standard deviation changes, not just for Sharpe or Sortino purposes or drawdown reduction, but really to give you better opportunities to say, "Okay, if I'm trading five strategies, and three of them are in that period where they're going sideways or they're going down, then the other two are more likely to go up than if I had just traded two or three strategies." And that's a little bit different when you trade long-term than you do when you trade short-term, because if, if you're a young professional and you've got, you know, $30,000 in the account and you're just trading and you're living off your paycheck, and it doesn't really matter what happens to the returns, it may not matter for you that you have an entire year, two, three years of underperformance, whatever. You, you don't need that money. It's a game for you. But when you're running a hedge fund, and when you're living off your income returns, when you need that money there, it's very different. You need to make sure that something's performing when other things aren't. Understanding that a time frame, everything's going to go to a correlation of one, you know, or negative one, you know, one or the other at a certain point in time. So, sizing really matters. Diversification of strategies really matters, especially different versions of what you're attempting to gather.
Yeah. Another thing with that, too, is it's like, I think this is kind of like the Bitcoin example. So, it's like, if you're a tech employee in Silicon Valley and you know, you're partying on the weekends and just having a good time living life, you know, it probably sounds appealing to hold Bitcoin because over five years it'll probably go up, but like the ride.
Of that is absolutely brutal from like an income perspective if you were banking on it to basically live on it. And the other thing that I think a lot of people miss with correlation and stuff also is like you your biggest risk is always counterparty risk. So, let's basically say like CBOE doesn't provide liquidity for zero DTE options. I think that's extremely unlikely. It's obviously the most liquid asset on Earth. They, you know, put a lot of effort into that.
But like holding fixed income with no volatility in a high yield savings account, has actually a lot of value from a diversification perspective, even if it's not based from math, because the CBOE failing is going to have no effect on your money market account or you paying down your mortgage is going to have no effect on your zero DTE. Whereas even if you have an amazing hedge, if the CBOE pulls liquidity, which is counterparty risk, you can't transact your hedge. So like you're just going to eat, you know, effectively whatever you have, which is the same reason why I own some gold. I don't own a large portion of gold, but I own or do you own the No, no, no, no, no. You know, GLD, those sorts of I actually had some coins and then um because it was probably like 15 years ago, I realized that the spread between buying and selling was so large that it was almost impossible for me to actually turn a profit from it. So, I didn't do that anymore.
But, you know, I I don't own gold because I think it's going to the moon. I don't own gold because I think it's going to massively increase in value. I don't actually think the opposite of that. I think gold's likely, you know, it may go up, may go down, but I think it's largely going to be fairly stable. But by having some diversification in gold, it changes the volatility of my overall portfolio significantly and gives me more options to take some more risk in different areas. Understanding that that gold portion is going to remain very stable. Same thing with, you know, money market accounts, USFR, you know, I have a portion in there, not because I think it's, you know, going to return the same amount that it would for the market, but by having that in there, I already know like, hey, everything else I can let go a little bit more risky because I'm not full tilt hedge in a different direction.
Yeah. And the thing I also think is interesting about gold is it's like, you know, for the last 30 years, uh, fixed income has been inverse to markets. So like markets go down, Fed lowers rates, fixed income basically returns money. That's obviously broke in 2022. Whereas like gold throughout basically most of time has shown also kind of like a hedge perspective that in large volatility events it kind of helps. Obviously it's financialized. So like when people get margin call they sell out of the gold. So it does have a little pull down on it but it generally kind of reverts over time. So I have the same agreement that now looking at this.
So some concepts I want to go over is we'll kind of look at you know uh live portfolio account I run and we'll we'll chat about some things. So just in general with the portfolio and you can stop at any um part here. It's like concepts on how I originally build a portfolio is like my expectation is that the past data is always going to be worse. I mean the future data is always going to be worse than the past data. That's obviously obvious because uh I'm sorry 100%. No, you're right. And then the correlation of return of the trades and or the portfolio will remain similar over time. So over the course of a year if I have a correlation of 0.1 for one trade to the other I do anticipate basically being 0.1 at the end of the year. However the path of how those will cluster over weeks months are obviously going to also astronomically vary over the time. So no matter what my Monte Carlo says, no matter what my past data says, I basically expect to hit my max drawdown within 90 days from the data set and I basically expect to double it within 6 months. So I kind of built the size of the portfolio for that basically accordingly.
And then another thing is kind of optimizing for that standard deviation which you know Justin talked about in regards to if you had a straddle increase the standard deviation. So you need to think about even if some trades are profitable they could you know increase your M and increase your CAGR significantly but by increasing your standard deviation that in turns opens up the cone of what randomness can happen with the unexplained risk then unbias yeah real quick. So the one thing that um as it relates to correlation I think part of it really matters the frequency of your trades. Yes. So if there's 100 trades that you're taking on, you know, I typically when I take a look at things, I'm really unwilling to make a firm conclusion. It depends on the strategy, but if you have less than 100 to 250 trades. That's kind of the range where you know some strategies you need 250 400. Some of your hedges honestly you need a thousand before you can really make any type of determination. But most of them it's kind of in that range. Now if you're putting on you know you had mentioned it when we had talked earlier a double calendar once a week or you know a FOMC trade you know that happens quarterly for example you need significantly more of those to really feel comfortable about what it's going to actually do.
So when I look at something I say hey look it's a correlation of 0.1 over the course of a year. If it's two different trades that I'm entering daily probably yes I would agree. you know, if it's one trade that I'm entering weekly and another one that I'm trading daily, I would actually want to zoom out a little bit more to say, okay, what's the correlation over the course of three, four, five years, that I would agree overall would continue long term. It may increase up point4, decrease down to negative.3, but I should expect it to be around 0.1. But I think it it's a fallacy to take two double calendars and take a look at it since dailies that enter, you know, one on Monday, one on Friday and say, hey, the correlation is zero. it'll be zero moving forward. I don't think you have enough data points to really make that determination. So yeah, one little catch.
Yeah. And that thing with it too is like people often ask me like how far back do you backtest? It's like obviously I'd like to backtest as far as possible, but if I run 12 trades a day for like trend following, selling credits, I'm not that worried only going back to 2022. I'd kind of like to go back to 2020 to basically, you know, view shock scenarios. But when we look at that, that's obviously enough data. Whereas like a double calendar where you only run it to SPX dailies, that obviously has a huge problem and it's like just the data set in general has massive flaws with kind of um making that assumption. So when we're looking back at this portfolio, we also know that like some metrics that we can look at that are unbiased to hold ourselves accountable to is we know what our max drawdown is from past data set. We've also expected we've also stated like we expect to hit that and double that. And then the thing that I really find interesting is again this is kind of like a north star but I expect for it to cluster over time is using our whole portfolio over the average of all the days what's our average daily return and our average standard deviation which were the inputs to the Monte Carlo analysis. So obviously that can cluster different. The correlations of the trades will break down. But if we hit and surpass our max drawdown and our average daily return and standard deviation is the same as the backtest. We know that's just the unexplainedness of it and we should expect for it to mean revert over time.
And then also the overall data set is in line with our portfolio expectations which would be like correlation daily returns and then red flags that would pop out that would obviously cause more drawdowns is like if we have variance in execution. So like if we assume 20 cent slippage there were you know a couple massive days like April 9th to where your average execution now is 40 cent slippage. Obviously, there's nothing you can do about that. I would definitely expect for that to mean revert over time unless there's some type of regulatory change. And then also when you're comparing that, let's say 6% max drawdown in your backtest. Um, you know, compared you're using that same backtest metrics and you're basically comparing it to the 3-month period where you took a 6% drawdown your actual forward testing. It's like, does the backtest at least match that? Obviously, it's exceeded it, but it's there. So, this is kind of an example of one of the portfolios I run. This basically shows the metrics of my actual portfolio versus basically the expected metrics from the backtest. And just so you know like this was I basically had 10 consecutive weeks of profitability and then I had four consecutive weeks of losses and this is basically the mark at the fourth consecutive week of loss um coming down here. So you can basically see that like our Sharpe ratio is basically in line where we expect. Our Sortino is basically there. Our correlation and beta, you know, they went down a little bit because the volatility of the market was higher. However, our average daily return is right where it's supposed to be. Our standard deviation is slightly higher. Our win rates on a weekly and daily perspective are basically identical. And then also, you know, average weekly return, all these are basically right on point. So even just from looking at this and then looking at the Monte Carlo analysis over 20 years of 6% we're at a 5% return. However, our standard deviation and average return are basically identical. So like I was kind of doing an exercise with this on the concept and everything's basically matching and you can see that our p-value is here which is like our confidence it's non-random. So like I have a 96% this is non-random and that obviously may be worries some person credit trades but the fact that we're running long straddles we're buying volatility this is obviously where most of that randomness is displayed and people will obviously want to strategy hunt but like from a numbers perspective everything here is well within what is statistically expected and we're basically right on what our expectations of performance are. However, the path is basically the only problem.
So, the only thing to really check in this perspective is what is my sizing and is this sizing appropriate? Unbiased things are standard deviation, different types of trades, stuff like that. But this is just kind of the way I would encourage people to think about it's not even is your test or edge breaking down, but is it behaving as expected from an ebb and flow perspective? Like I'd be kind of curious how you view your portfolio or it's like when you are off in your portfolio, what do you look at to inspire confidence from an objective perspective and then even a subjective perspective from like a theoretical sense of worrisome with portfolios or such.
So for me, you know, I I think it's very common to take a look at the return and the drawdown, compare live to backtest and make an off-the-cuff reaction whether or not it's working or not. That tells me nothing. I I will say that if the backtest had a drawdown of, you know, 5% and live I'm looking at 20%. That's a problem. You know, I fully expect at least to hit twice the drawdown in the backtest. And I actually expect my return to be about half of what the backtest says, maybe a little bit less, you know, only because I'm optimizing for previous, you know, performance. When I find a strategy that has a 10% return and an 8% drawdown, I don't trade it, right? If I find one that had a 40 and a 2% drawdown, I might trade that. But by selecting that one as opposed to another one, I'm automatically selecting for periods of time in the past of overperformance. It's just natural in the way it's going to make. So it may be that I get 40% return and 2% drawdown, but I don't think I will. I think I'll get 15 or 20% return and I think I'm going to get four, five, six% drawdown. That's just in my head. So when I run a backtest and I compare it to live, I'm keeping that in mind as I move forward.
I also want to make sure that I have enough trades within that time frame that I think that I can actually make an assessment which is hard to do because you know there was a strategy that was going around what and what's enough you know what I'm saying exactly so there was a trade the um continuous iron fly uh 0 DTE a 1 DTE version of it where it was an iron fly with a profit target or a stop-loss and then it auto re-entered afterwards and you know after trading it for three or four weeks, you're not really making a whole lot of money and things don't really look quite the same as it did in the backtest. But you had 20 trades, you know, 20 days worth of trading. Is 20 enough to really make determination? I don't think it is. You know, I think 200 maybe. Do I have the stomach to continue this for another 10 months? That's a completely different different question other than objective analysis find it. But what I think you get much more than comparing returns, comparing drawdowns, comparing correlations with other strategies, comparing VIX regimes, comparing other types of strategies. Take a look at the distribution of your returns, both the backtest and live. I think that tells you significantly more. And you don't need to be a statistician. You don't really need to understand, you know, standard deviations, all that. Like take a look at the picture and overlay the two of them, right? So if one had a normal bell curve that looked like this and you're like okay I get a lot of returns and it's you know most of them are positive some of them are negative but most of them are clustered around here and then I give mine and it looks something like this and then a large fat tail on the right side. I can tell you that looks very different from what it should be and a lot of people will say you know if it's negative and I'm losing money I'll take a look at that. I want to look at both, right? So, I want to know if I'm making money and I'm making more money than I should be. Hold on a second. You know, maybe that's great for my bank account, don't get me wrong, but am I making the right decision here? Has something changed significantly that I might need to resize some different things or stop trading it entirely?
So, if you want to dive into a lot of, you know, the p-values and the standard deviations and, you know, a lot of the skew and the kurtosis, if you want to, that's great. You can learn a lot from that, too. You can learn a lot just from having ChatGPT give you a distribution curve and putting one next to the other or having them overlay the top over each other and do they look similar? You know, both historical and go forward. So, if I traded this for a year and I ran a backtest for the same year, did the backtest data give me the same type of distribution that I did in live? Because if it's not, then one of the inputs within the within the backtest is not accurate. Slippage is a good example, entry or exit. But maybe move some of those around a little bit. But also historically based on the assessment on what I had, is that accurate?
One of the other things that I think is crucial too, I never backtest to today. You see it a lot where they say, you know, hey, let me do dailies to today. Look at how awesome this is. I never do that. I always zoom out at least three months from today. Usually six months or a year. That's what I prefer to do. is difficult with zero DTE because you don't have that many dates and you kind of need to crunch it a little bit closer. So it might be three months, but I want to take a look. Okay, let me backtest for that time frame. And now that I know what optimal is over there, let me zoom forward the next 3, six months after that. Do I get comparable results? And I can guarantee you after people start optimizing, you know, do 125% stop loss instead of 150. If you graph that out, it looks like this. So going straight up and to the right in your backtest time frame, and then you end up using your out of sample data afterwards, and it goes like that. It never looks quite the same, which gives you an indication maybe I tweaked it a little bit too much.
Yeah. And it's never that easy. And like the other criminal thing I see with people, especially with smaller accounts, is like they see that up into the right curve. And then like their first response is like, well, since you can't lose money and I'm doing this 150 wide, why wouldn't I just trade four times the contract at 50 wide? And then it's like there's not enough edge for you to do that. And like that's ultimately the biggest fallacy because it's like when you buy a tighter wing, you're essentially like buying more of what you think is expensive. Or the other thing that cracks me up is people will show trades of like let's call it a 20 wide with no stop comparing it to like a 200 wide with a stop and they're like, "Oh, I don't want to pay slippage so I don't use a stop-loss." Well, it's like when you enter the stop-loss trade, you shop a $2 order and a 5-cent order, and you basically get 5-cent slippage into that. And then when you exit, let's say you pay 25 cents of slippage to exit on 30% of the losses. Then when you enter in the four-legged condor that's 20 wide, you're shopping a $6 option and a $4 option on both sides. And it's like, yeah, you just paid 45 cents of slippage to get in. So you've just made this fallacy of like, oh, I'm not paying slippage because essentially you don't see it. It's like the private equity myth. It's like private equity has the most volatile portfolio out there, but they don't have to market on like, you know, any real basis that you have no idea how volatile it is. It's like if you compared private equity to like 2x levered NASDAQ and just only took quarterly returns of the NASDAQ, you would see that like, oh wow, private equity kind of sucks. But, you know, they they did balance sheet gymnastics on basically what that is.
And another concept I like that I've recently came to, it sounds like you've came to do is just like what is the variance or volatility of those returns and kind of like as much as we've been talking about complex things, I think the most important thing that people take away is like we all learn differently and like humans can recognize patterns. So like even if you don't understand what like a Gaussian distribution or these fancy words are, you can just look at a graph from AI and be like ooh mentally I don't think that line would feel very good comparative to this line. Maybe I should just check my ego and choose a straighter line. And and kind of like a randomized example of this is options are in a population distribution and they generally exemplify an even bell curve over a long period of time. There's a little more move to the upside because the market goes up over time. But to like show that in a non-kind of mathematical way, it's like if you drew a giant circle in the sand and you just like threw spears into that circle, basically at a certain point, if you threw enough spears and measured the radius, you would get really close to pi 3.14. However, it's like how many spears do you have to throw in that? And then in in terms of trading like let's imagine you someone you loved was standing in the square I mean in the circle where people were throwing spears. So it's like now how exact do you want to be to 3.14 and that's the way you should think of it. It's like you are literally standing in the sphere of the probability distribution and people are throwing spears into that. So, like the more accurate you can be with your variance from your trades on a day-to-day basis that you know you're close to pi or 3.14 or half a percent, whatever your volatility is, the more confident you can basically stand in that circle and stand in a spot where you don't think you're going to get hit by a spear.
Yeah, I completely agree. I think that, you know, one takeaway even if you're new, even if you don't understand statistics, look at a picture. you know, you should be able to take a look at a bell curve and say, "This doesn't look right." The other thing, you know, and you had mentioned it, sizing is key. And honestly, you're really better off sizing smaller than you are sizing larger. You and a good example of this, so let's say I took, you know, a coin and I said, "Let's play a game. Uh, you have $100, right? and uh we're going to flip this coin and you bet a dollar. You you can bet whatever amount you want from your portfolio stash, but you need to play for a set period of time. 100 coin flips, 300 coin flips, whatever it and I tell you that this coin I I it's a fixed coin, right? So, it's it's going to hit heads 60% of the time. You can bet however much you want, but you're trying to maximize the value that you actually have at the end of the day. Most people when they look at this and they say, "Look, the coin's rigged against me, so why not bet as much as I can, right?" Well, if I choose to bet 50% of my, you know, pool on every trade, I the odds of loss out of that just from getting tails over a period of time, assuming I chose heads every time, your risk of ruin is very high. But if you chose like five or 10% actual account size as far as betting, your risk of ruin in that case is as close to zero as as statistically you can get. So by betting more, you have a larger chance of actually going negative. So your actual expected return is less than if you had bet less even though the coin has a 60% chance of profitability. If you're really interested in that, some people can read up Kelly Criterion and it gives you optimal betting sizes and those sorts of things. A half Kelly sizes or a quarter is usually a little bit more appropriate, but it gets it has some math examples of, you know, optimal betting sizes when you know the returns. Challenging part with options is, you know, you don't know what the odds are. You, you know, you can take a look at what it was in the past and you can assume that that's the same moving forward. It might not be. And the, you know, risk to return isn't always the same as a coin flip. You're not going to gain a dollar and lose a dollar on a coin to cost. You know, you might gain a dollar and lose $5 or lose $20 or lose a dollar. You don't really know. So, a little bit different in live trading, but it gives you some type of idea.
Yeah. The other concept is like going back to the, you know, retired couples 1968 to 1988 with 10% 20% returns varying is like this comes back to the concept of ergodicity. It's like your investment path is a path continuously. So averages don't matter. Like the TP code is how do you drown in a river that's 2 feet deep? It's 2 feet deep on average. So that's like really what the danger is and the difference between optimal and I optimize for standard deviation and that potential lack of loss. And then what else is optimal? Like the other classic example of this is like imagine you could go to a casino and you went with a hundred different people every day with 100 people and you knew that your it basically cost $1,000 to go in and your expected return was $2,000 and essentially um one person out of the hundred loses their entire money. So, like from a bet perspective, if this was non-ergodic, meaning you just went with a new group of 100 people every day, um you would always do that and then let's say all the hundred people had pulled money and you just shared a cut of it. So, it would never matter if you blew up. Then, if you had the opportunity to play that same game, but you had to play it consecutively until you finished the 100 turns, you would never want to play that game because you are virtually guaranteed to get rich and then also blow up.
And the other very simple concept with this like sizing bigger versus sizing smaller, I do have a little bit of difference with optimization on what that is. So like for example, if you take a 5% drawdown, you have to get 5.25% or 5.5 to get back to even. However, if you can get a 5% drawdown with a 40% return, basically when you do a 3% monthly return towards the end of that year, it actually ends up being like 5% monthly return because it compounds on itself. So like that type of risk-reward makes sense then like I don't necessarily get why people will target like a 1% drawdown and thinking it maybe like two and a half. However, on the far flip end of that if we see something's profitable at a 10% drawdown that can obviously vary over time. You know when you have a 10% drawdown you pay 1%. If you have a 20% drawdown you have you have to pay 5% to get back. So like if you just decide to up it and this again is what I see in the communities here is they're like oh I don't mind losing 25%. Well, it's like you could lose all your edge because with a 25% drawdown, you need like you lose basically lose 8% per effectively because you're losing all the money you get. So like that risk is clearly not there. And this is where everyone screws up in zero DTE options is like the benefit of zero DTE options is I think in six to 12 months it's very doable for people to do 20 to 25% returns with 10% drawdowns on a small portion of buying power with their account. But you could just see from this example how it would be exceptionally unpalatable to double the size for potential 20% drawdowns that could vary to 40% and to use like more and your more your money of it. Whereas what you do and what I try to do is like the optimization of like how do you make 30 with a 4% drawdown like that leap is just astronomical between those two those two facets.
And the other key here that I noticed when you were talking is we talk about these non-ergodic returns like through life and trading and I think that people that leave their career and really put themselves out in the line like me or you, we are actually living our life ergodically. So like we've given up the opportunity cost of building our career. We've chosen one path that we've set on and really you either kind of leave a failure or you make it through the other side. So that psychologically and physiologically on what your behavior is lends over the most to trading possibly whereas like people that are still working they aren't behaving that way with their physical body. So, it's it's going to be much more difficult to be congruent. Whereas, they're taking more of a stance of like I work a job, I make money, and if I gamble and make money in options, I can get out of my job earlier. Completely different than I'm leaving a broken man or working a, you know, going back to being a sales development rep for $50,000 a year or I'm going to make it out like a, you know, financially free millionaire from trading options. And like that, people always talk about the mental game. It's like first of all, your psychology and your mental game doesn't mean if you don't know your numbers, which is something I think a bunch of people miss. And then secondarily, it also has everything to do with how you're behaving on a day-to-day basis. Like all these people on Twitter, I've even met some hedge fund managers. It's like they live in Vegas. It's like those type of people and they like like to drink and party and gamble and it's like they just have like a keen fixation for risk and there's a lot of bodies on these risk takers. Obviously, there's going to be successful people in that whereas like there's other people that are like, "Oh, here are my numbers. I deliver 10% a quarter or I'm going to make enough income to live it off." And then they deliver on it, but we don't champion the people because we're like, "Oh, that person had such conviction." And that's just something like I think people should look in the mirror and really ask themselves like what their behavior is and if you're trying to trade for a living like is everything in your life in congruence with the habits of trading for a living because the quality of life of making a million dollars a year completely remote through an automated trading is like much more appealing than making 50 million in my opinion and potentially risking what you have and need for what you don't have and don't need. That that's my last thought. Do you have anything else?
You bring it up, but I noticed it when you brought it up and you talked about the path and I could see that related to your life. Like I definitely that for a while. It's like I used to throw up when I wasn't sure about these numbers and this is kind of what stemmed my fixation with figuring it out. Like I would be so uncertain that I would wake up and throw up. Like my body was like rejecting my assumptions. Uh, it's fine now but like and trade long enough you're gonna do it. I mean everybody knows. I mean, you you don't know your risk tolerance level until you hit it, you know, and part of that has to do with conviction and understanding your strategies, but part of it has to do with understanding yourself. And I will say that the your risk level will change over time and it won't increase. It'll always decrease. And it may decrease just simply based on the numbers. So if you were running a strategy where you were making 100% per year on average and it had a 20% drawdown, most people would say from risk-return perspective, that's amazing. Well, when you're working a full-time job and you have $100,000 in the account and you say, you know, losing 20%, I can lose 20 grand. Like that's fine, whatever. All my bills are getting paid for. When you hit a million dollars and you're losing $200,000 is still 20%. But $200,000 sounds a lot less palatable than $20,000. But then you turn, you know, if you're a young professional and you got a W2, that's one thing. You're a young professional, you have a W2 and you have kids, suddenly that $200,000 looks a lot more real. Then when suddenly you don't have that W2, that $200,000 is very, very real. You know, so all of them are 20%. But at different times, different account values, different stages of your life, things will change. and you don't know that you hit that point that you can't take it anymore until you hit that point and you can't take it anymore. So, I would always say, you know, if you look at something and you say, you know, I think I could live with a 30% drawdown, well, why don't we try 20 or 15 and see how you feel at that point and if you feel fine, then maybe increase it from there. But don't start off at 30 at a number that you've never hit before thinking that you'll be able to stomach it because the odds are you probably won't.
Yeah, I completely agree. That's all I got. Yeah. Uh, okay. I don't have anything else. Do you have anything for me or I'm good? No, not on my end. Okay. Well, thanks everyone for tuning in. Um, you know, do you want anyone to reach out to you? If not, I'll just Yeah, no, it's fine. Uh, most easily recognizable through Discord, uh, various channels, special Kai, you can find me at, uh, Trade Busters, Speaking Greeks, Option Omega, any of those. I'm in a number of other ones as well. Not hard to find there. or maybe not to bombard you like feel free you guys can email me questions if you really like this comment we can just have them back on so it's time efficient for everyone too that works too um and if you're interested you know I go through a lot of this kind of stuff in any of the courses I have I have two 1 DTE courses through Option Omega's Academy really push it very much it was something fun to make I'm glad I did it but it's hanging out there if you're interested and you like what you hear check it out over there or hit up Mark or myself with questions you know I'm always available.
Yeah, I've taken both the courses. Um, they're good. Like I even have traded strategies and um the returns of the strategy have dragged quite a bit lately. Maybe I don't know. So, it's not it's definitely not the holy grail money maker, but very good info and I think it just kind of shows like sometimes you can do everything right and like you just do need to go through periods of those drawdowns or whatever. And other thing I say is like every drawdown is always worth it because you learn some effect type of lesson whether it be sizing or whatever do and then it's like I roughly trade that strategy one-third of the size I used to trade it. Granted, I kind of noticed it happening over a period of time and it was slowly taken down, but interesting with