Line Your Own Pockets

Michael: 00:00

Hello, everyone. Welcome to another episode of Line Your Own Pockets. Today, we're gonna

Michael: 00:03

talk about stops and back testing. A way why you might not want to actually put a stop loss when you're when you're running your back test, which I know is counterintuitive to a lot of people. So I think it'll be interesting to hear when to and and when not to put stops and and kind of for what reason. So this was this was your idea, Dave. So I'll let you lead it off here.

Dave: 00:26

Yeah. So part of the process of backtesting and coming up with a strategy, one of the things I do is run a backtest without a stop. Now I don't mean trade without a stop. You're not going to end up with a strategy that doesn't have a stop, but an important part of coming up with a strategy is doing some sort of clever techniques to figure out what the stop should be. And part of that is backtesting without a stop.

Dave: 01:00

Mhmm. So I was just talking about this with one of the traders I'm coaching and, you know, talking through why this is important and what you know, why you would want to do it, what the output of that is. And I think there's two main reasons for doing this. So we can get we can go into so so Michael, is this something you ever do?

Michael: 01:28

Yeah. And I it's funny. When I talk to other traders, it was one of those things that I was surprised that it was rare to to do that most people thought so. Because I always looked at it as step one was you call like a signal test. Right?

Michael: 01:46

I wasn't so much testing the whole trade idea. I was saying, okay, does this event indicate that the stock will be up or down for me like a week later or or something like that, you know, for you for people might just be by the end of the day. Right? And just doing that kind of raw without a stop loss or anything just allows me to understand whether or not the the signal is kind of predictive at all. If it was 50% of the time it was up, 50% of the time it was down, and by roughly the same amount on either direction, then I'd go, okay.

Michael: 02:18

Well, maybe I could make the strategy better by implementing some funny risk management rules, but that's that's for me my first step is to see is like is the signal actually significant or is it my I think you get fooled pretty easily by, you know, taking almost a random entry and just doing nice, you know, risk management around it and turning something that's a random entry profitable, which again, good for you. But, yeah, that was always kind of step one for me is to run it. I just call it naked. Run it just completely naked and see what came out of it there.

Dave: 02:53

Yeah. I think that's another good reason to do it is exactly that. And, and you're right. It's, I realized at some point that it is pretty rare that people do it, and it's so fundamental into what I do that it's almost second nature, and and I was just putting some thoughts together and thought it'd be a good good topic for the episode. So I think there's I think there's a couple different reasons to do it, and and one of them is super important.

Dave: 03:28

Another one is sort of a nice to have. It's kind of interesting, but maybe not you know, I don't from from the number two, I don't really get a whole lot out of it all the time, but it is there are some situations where it's super interesting. So Uh-huh. Let's go over the first one first. And that's like figuring out what your optimal stop should be.

Michael: 03:52

Yep.

Dave: 03:53

How do you do that? I mean, you could, you know, just come up with a bunch of random things and like sequentially figure out which one's best, but a much smarter way to do it is to use MAE, maximum adverse excursion, which is basically the least profitable point in the lifetime of the trade. Most, or a lot of software, you know, backtesting software automatically include this number and the backtest trade ideas includes it. I was make I'd make sure to include that in the backtest when I was there. Amber broker does this.

Michael: 04:33

Let's just define that just a little bit more just cause it's, it's a concept for some people, but you know, let's say your entry price of a stock was $10. And then before the exit, in most cases, a profitable exit, it dropped to $9 and then turned around and went up. From from you're basically starting the time stamp at your entry point saying, how far in in for longs did it drop before it, you know, turned and and however it ended the trade. What was that lowest potential point? And by doing that, that you're basically just saying what was the most I was ever underwater in any given trade.

Michael: 05:13

And as soon as you have that data, that becomes very interesting because now you can do it like on aggregate. And you can do, you can normalize it by ATR, you can normalize it by a lot of different things, but you can say, okay, how much am I able to after my entry point, how much pain do I kind of take? And that's a good starting point for a stop point because obviously, you'd want to make the stop point a little bit lower than what the average was and and knowing that it's gonna cut off a bunch of trades that may have dipped lower and then recovered and all of that. But that is a very interesting point and you know we're talking about stop losses, but of course you can do that on the other side, which is how much from your entry point do stocks normally run. And that's always an interesting stat to keep an eye on is how much does the stock run versus what was your exit price.

Michael: 06:04

And you know, if the stock say normally ran that same $10 entrance ran to $12, but you're always exiting at 11, then that's a way that you can kinda look at and say, oh, well, maybe I can grab a little bit more, you know, juice for the squeeze here if I put my stop loss that around that $12 area because, during the day for a day trade, it means at some point it spiked up and it got really, really high and then it came down a bit before you're, you're at the close or, or exit like that.

Dave: 06:33

Exactly. So, to, to figure out what the optimal stop should be, you, sort of ironically, the best way to do it is to create a backtest without a stop and just let it play out. Know, so you've got an entry, you've got a timed exit and that's it. So, and then you can do an analysis. In fact, I just added this to the cruncher just yesterday.

Dave: 07:02

So with each cruncher optimization report, it now does the stop analysis based on this MAE number. If the MAE number is included, it's going be at the bottom of that report. So you can look at it and see at what level you should be using a stop. And and you know, when you think about when I first started doing this, it's been many years since I've been doing this analysis, there's a there's a really popular blog post I wrote called How I Choose Stops, and it's got these charts in there for how I look at it. And when you first look at it, when I first started doing this, I thought, okay, I'm gonna just be able to optimally pick the stop.

Dave: 07:47

It's gonna be easy. Like, I'll be able to just look and see. But when you actually look at the numbers, there's a bit of subjectivity when you look at it to see, okay, how much, like, how much total profit do I wanna give up? How much win rate do I really want? It's not quite cut and dry, like figuring out exactly what stop you should be using, but this report sort of lays it out for you so you can see, okay, if you put the stop here, you're gonna have, you know, the win rate would be this, total profit would be this, and you can sort of make your judgment call about which one you think you should, use.

Michael: 08:31

Yeah. I I like that you brought up the subjectivity because although there might be a optimal number for total profit that it kind of depends on what you're optimizing for. Right?

Dave: 08:44

Yeah.

Michael: 08:44

Some people might look at it and say, that's the total amount of profit, but it also increased max drawdown and it also increased, you know, the variability of the trading and it also, you know, increased a lot of these things. So understanding when you're when you're looking for what your optimal stop loss is, I I think the first question you always have to ask is kind of what are you optimizing for. Right? If you're optimizing for just win rate, if your signal is actually predictive, it's like a very loose stop loss, very wide stop loss. Yeah.

Michael: 09:16

If you're if you're optimizing for, you know, more of a kind of profit per trade, it might be a little bit tighter and you know, that's always question one is, you know, what am I trying to accomplish? Which is gonna be different for everyone. Some people might not, you know, they they might not be able to deal with low win rate strategies. So they just they have to keep wide stop losses even though it means they're, you know, they're taking less shares per trade and and they're not making as much money, but they just cannot deal with something where they're wrong more often than they're right. And they just, know, they would look at that every day and say, well, I can't be red more often than I'm green, so I need to optimize for this, which is a completely different question than it's the opposite of that.

Dave: 10:02

Yeah. So it's for sure that, you know, it's oftentimes the or most strategies, I'd say, no stop loss at all is probably going to be optimal. If you're looking at a total profit, if you're, if that's the only thing you're looking at, then an infinite stop loss is going to look pretty good. But in reality, you're not going to want to actually trade that or it's going be very difficult to trade something like that.

Michael: 10:35

Very hard. And then also it comes it becomes harder to position size as well. Right? If if you're someone, you know, that I do a lot of and and I don't think it's as bad as a lot of people do where I just take the same dollar amount per trade. Some people who just like incredibly cringe out by that.

Michael: 10:54

It's gonna depend on a lot of factors. But if you're someone who's, you know, I wanna risk a thousand bucks on every trade or a $100 on every trade or whatever it ends up being, you that your stop loss is way more important and that finding that optimal number is way more important because it's going to completely change your position size every every tick you go by. And, you know, if that's how you're sizing your trades and you you better find a stop loss that makes sense or like with your infinite position sizing. Great, you win all the time, but you're only allowed to take two shares for for trade. It's what does it matter?

Michael: 11:33

Right? You're you're getting a 90 plus percent win rate, not making any money. Okay.

Dave: 11:39

Yeah. Yeah. So, I mean, that's a good I'm glad you brought this up because a difficulty some people have when they trade a certain strategy where they're sizing their positions, not based on money at work or, you know, dollars at work, something else based on a stop level. So, you know, a candle pattern, they're sizing it based on like a natural stop distance. Then to trade it, you know, to, you know, how do you size the position if you're not including a stop in your backtest?

Dave: 12:19

So I have a mode in my systems where it sizes the position based on the stop, but then doesn't actually enforce the stop as it plays out. So then that gives you another look into this where, that that allows you to do this type of analysis and sort of separate the two. The other thing that this is important to do is when you have a stop in your strategy, what you're essentially doing, and then you go to optimize after that. What you're saying is, and think about the different think about two different, trades here that stop out. Let's say one trade stops out, but then continues in your favor, and it would have been very profitable had you not been stopped out.

Dave: 13:24

Compare that to another type of trade that hits a stop and then tanks, like goes down even further, would have been way worse by the end of the day or by the time your timed exit hit. When you're including a stop, you're saying essentially that those trades are equivalent. Right? Where in reality, they're very different trades. Right?

Dave: 13:51

When you do an analysis without a stop in place, and then you optimize on that, you're actually allowing that trade that went well against you after your stop to be penalized even further and allowing the other one that hit your stop, but then continued in your favor, you're rewarding that one in your analysis. So it's a bit of a subtle thing, and it's really clear to see in the data when you look at this why that happens, and it actually makes a lot of sense to look at some of those trades that go against you and then keep going. Like, how could you penalize those further? And to to do it backtest without a stop, it allows you to do that.

Michael: 14:47

Yeah. So what you're saying is, you know, say you're using, I don't know, one ATR for a stop. You're sizing it based off of if that stop existed. So, you know, the larger the ATR of the company, the less position size you're taking. But you're just saying to the system, just ignore the stops, hold the trade until the end of the day or whatever your your time is.

Michael: 15:10

So you're already doing a slight amount of like volatility normalization without without worrying about the stop loss, which is good because I could see if you didn't do that and you were running it without a stop, it's gonna be like, everything's gonna get really skewed because you could just have a couple positions that you would have taken a ton of shares in that that have a really bad time and that just makes up your entire like return base or we have a really good time and that just makes up. So you'd be heavily weighted to the ones that are like say the cheapest stocks, for example, because they're gonna have the biggest percentage return if you don't end up normalizing this. Why all of this stuff and I always talk about I think ATR is probably the most important indicator for for technical traders because or systematic traders because we just use it it so much in so many different things even though we're not using it to make decisions. It just creates it's I envy, I think, people who just trade like futures or one instrument because you can always compare it to itself, but as soon as you get into equities, the dollar amount doesn't normalize things.

Michael: 16:20

There's so many things that you can't do. You just have to take how much does it normally move or how much has it moved recently and then compare that to to other things or else everything just gets like way in a whack really, really quickly.

Dave: 16:33

Yeah. Totally true. You're right. People that come from the future side into equities do have a hard time figuring that out or, you know, grokking that. So

Michael: 16:45

Mhmm.

Dave: 16:45

Yeah. So the other thing I like to do so so so to back up, the primary thing and and you pointed out a really a second thing that I think is really good, which is so there's two primary things. One of reasons for backtesting without a stop. One is to choose a stop that's optimal across a large number of trades. The other is like your your sort of your smoke test, essentially what you said.

Dave: 17:14

Like, is your signal strong, or is it skewed by using a tight stop? You know, want the tray you want your signal to work on its own without having the help of a stop or a target. So those are two really good reasons. And essential reasons for, like almost every strategy I do and backtest with, I'm doing that analysis for that reason. There's another reason that's super interesting, but not absolutely critical, but in certain situations, it can be very enlightening to about what's going on.

Dave: 17:55

And that's when you run your optimization with the stop in place and then the exact same optimization from a backtest without the stop in place.

Michael: 18:09

Okay.

Dave: 18:10

And for exactly the reason that I just described those two trades that are equivalent when you have a stop in place, but completely different when you don't have the stop in place. So looking at the difference in the optimization report, what comes up in the cruncher for the suggestions that are different in those scenarios? Because a lot of times what you'll see is some of the differences, like something, you know, relative volume shows up in one report, but it shows up as in a very different value in another report, or it doesn't show up in the other one at all, say. Or something else is important based on that. A lot of times that can, in certain situations, it can lead to different lines of thinking, different lines of thought that you might want to test something differently, or you may want prefer one rule over another, or it might make you want to use a wider stop in a situation based on what you see there.

Dave: 19:15

So it's, like I said, it's not a cut and dry thing for every strategy, but it often is good for generating different ideas or different strains of thought that you hadn't come across before.

Michael: 19:31

Yeah. Like, I didn't I like that. I didn't think of them. It's it's funny that you were you were talking about the different ones that you'd find. For me, I was more interested right when you said that about what would been the same.

Michael: 19:44

Like, think I would have you know, if the same setting, the exact same, like, filter and the the same amount roughly, it came through on both. I think I would take that as like a a much stronger indication that that's something that's predictive. Right? If we, you know, if relative volume of two came through and didn't matter whether or not I was just holding it till the end of the day or didn't matter whether I had stop losses and profit targets and all that stuff in take in place, then I'm like, okay, well, that's really has to be, I guess, that point, like very predictive to this particular strategy because it's just like, doesn't matter what you do. Right?

Michael: 20:24

If you're buying this particular stock at this particular time, the relative volume better be too, and you're gonna you're gonna make a lot more money.

Dave: 20:32

Yeah, that's a great point. Yeah, so that type of analysis is interesting to do, and like I said, it's not something I do with every single strategy, but certain ones, especially where I know like, for example, shorting low floats where a really wide stop is what's gonna work. Mhmm. Those it it could be interesting to see those because, you know, you typically have MAE numbers that are really large. And unlike other strategies where MAE might not be that important, with some strategies, it's super important.

Dave: 21:14

It could be super enlightening about what, you know, what action to take and what's gonna be profitable with a certain strategy.

Michael: 21:22

So let's get let's get real nerdy. We're what, twenty minutes in, so everyone who's left is is real nerdy with us. But so when you're doing this analysis on MAE, are you what looking at like standard deviation of it or you looking at averages or so in my head, I'm picturing essentially a bar chart of all of the stocks and how much on like an ATR basis they move against you. Right? So some are gonna move against you by one ATR, some are gonna move against you every now and then you're gonna get one that moves against you by 10.

Michael: 21:58

That's going to really mess up the data. How is it that you're are you just kind of sliding that number up and down and saying this is the profit at these different levels? Or are you doing some sort of, right, like removing outliers? Like, I just get into the the real nerdy side of the math.

Dave: 22:19

Yeah. So, you know, imagine you've got the MA number. Some of them some of the some trades, the number's gonna be zero. That's really good, right? It right in your favor right from the beginning.

Dave: 22:31

You never never win a loss, the whole trade.

Michael: 22:34

Yeah. Find me a bunch of those, please.

Dave: 22:35

Yeah. And then some are gonna be really large. So, the idea is, you're basically what the cruncher does, it's showing you a report, and there's four different charts you can look at. There's total profit. I've got one in there for the percentage of trades where the stop is hit.

Michael: 23:02

Okay.

Dave: 23:03

So let's first talk about that one. It's going to simulate a very tight stop and then all the way to a very loose stop. And there's going to be several stops. There's going to be a chart of all those. So at the very tight stop, the percentage of trades where the stop is hit is gonna be very high.

Dave: 23:24

And as that stop gets looser, that number goes down because of course, it's not as many trades in the set are gonna have the stop hit. So that number is going down, as that stop gets wider. And so typically, there's another chart for the win rate. So similar to the amount of number of stops that are going to get hit, the win rate is going to be, smaller the tighter the stop, right? Because more trades are going to be stopped out with a tighter stop.

Dave: 24:02

And that's going to grow over time as that stop gets looser as

Michael: 24:06

you,

Dave: 24:07

and the idea is you're, you're, it's basically a shortcut to not have to do 25 backtests. It's doing you can get your answer here with one backtest, one optimization in the cruncher, and you can see exactly what your, you know, you can choose what your stop should be based on these numbers. And one, I need to add some more documentation because I know that that I can tell just by looking at this report that some traders are gonna be drawn to some tighter stops when some of those are they shouldn't be. Right? There's gonna be a lot of slippage that comes with that.

Dave: 24:53

And a lot of those are illusory. So

Michael: 24:57

I I think we've all done that when we've built a strategy with like a super mega tight stop and the thing just just rips and you just don't realize that when it's saying you're gonna be stopped out maybe 60% of the time, but you're gonna get some really good winners every now and then. You're really getting stopped out like 90% of the time because it's just so tight that you're not including just bid ask spread and like regular jiggles and all and all of that. I've done that before where it's like, if I put like a 5¢ stop on all my trades, look at the amount of money I make even though I've got a 30% win rate. And then you run it and your win rate actually drops like 10%. And you're, you know, it's it's just not working because you're just getting stopped out so much more than any simulation would ever tell you.

Dave: 25:42

Yeah. So, the other thing I'm adding, and will be in the cruncher by the time this airs, the cruncher is optimizing on average profit by default. You can do actually do two things. You can optimize on profit or you can optimize on win rate. Now I'm adding a third option for optimizing on MAE.

Dave: 26:12

Now, why would you want to do that and why would that be interesting? Well, think about the during the lifetime of a trade, think about the most painful part of the trade. It's when it's the least profitable. And lots of trades, certainly for some strategies, you'll have really big MAE numbers, but by the end of the day, you've got a break even trade or even a profitable trade. So some strategies have that characteristic.

Dave: 26:43

And when you optimize on profit, you're missing out on this, the most painful part of the trade.

Michael: 26:51

So

Dave: 26:54

ideally That's

Michael: 26:55

why I always like I always like the MAR ratio, which is just the, you know, your CAGR, your your annual return over that max drawdown is because, there's one thing looking at the number and just saying, oh, this strategy makes 60% a year. It's like, yeah, but Right? How much how much pain do I gotta go through to get that 60% a year? Right?

Dave: 27:17

Exactly. So the same way you would look at that across an entire strategy, you could optimize that per trade. So, you know, in aggregate, looking at that number instead of profit. You know, like I said, profit, you're missing, like it's not capturing the pain that you went through to get that in each individual trade. So that, you know, just like we said, comparing an optimization of a backtest with stops versus without stops, and looking at the results, looking at what's similar, like you said, looking at what's different, you could it's also gonna be interesting to look at an optimization based on profit versus an optimization based on MAE.

Dave: 28:01

I think you'll get some very different

Michael: 28:04

Yeah.

Dave: 28:05

Results there, and it might inform it could it could have a big impact. Like some of these strategies for shorting low floats, I mean, if you could figure out a way to to reduce your stop a certain amount, I mean, that would be enormous. It'd have an enormous difference in your bottom line. Yeah. So and this is the way to do that.

Michael: 28:30

So what's your what's your thought process? Might be a little off topic, but I've heard people use MAE and they they study that as a potential way to get a better price on their trade. And and you know what I mean by so go back to the example of, you know, you buy a stock at ten and then, you know, it normally drops a dollar before it turns around and makes you a profit. A lot of people will look at that and saying, well, instead of buying the stop at ten, I, when it hits ten, put it a limit, you know, somewhere in the middle to get kind of more of those trades to take place. Have you ever thought of using MAE as that?

Michael: 29:09

I it doesn't sound right to me when when people say it. I don't I haven't given it so much thought. But essentially using a you know, if I always if I'm always in a drawdown or or 90% of the time, I'm in some sort of drawdown when I enter the trade, well, that means my limit would have gotten filled and I would have gotten filled at a better price. I think it's about better way to to say that.

Dave: 29:30

Yeah, I think it's an interesting question. And the way I would think about it is, there's two ways I would think about it. One is, if you're seeing that, then waiting and getting in later is really a different strategy. Yeah. So just backtests getting in a little bit later.

Dave: 29:51

Right? That that

Michael: 29:51

was my thought. Right? So, you know, say say just to make it simple, but say the first strategy is an opening range breakout. Right? And you do the opening range breakout and you notice that you, you know, you take a a 2% pullback every time or most times 90% of the time before it goes.

Michael: 30:08

And I'm like, okay, well, you're right. That was my initial thought too is that's a that's a separate strategy. It's like, let's do an opening range break, but then, you know, we buy a pullback and test that as a completely different strategy and to see if it and that's gonna have its own MAE or MFE on on top of that. Right?

Dave: 30:27

Yeah. So I think that it strikes me as the way the trader mentioned it to you, is just adding a little bit too much complexity when there's a simpler way to think about it.

Michael: 30:40

But

Dave: 30:43

I think it's valid to think about. And what I would think about here is I would I would separate those into different entry tactics that you would apply to a different universe. So you might have two different entry tactics for this one universe there. Now, the other thing the other very interesting thing that you reminded me of here is alright, let's say that you have a good strategy, some of the trades go against you, and some and if the trades go against you, and if and you've got a pretty wide stop, if your strategy if those trades are still profitable and you're confident in that, you have an opportunity to get in and make your add size and get a better price, which I don't recommend for everybody. But for certain types of strategies, that's essentially what's happening, is you're getting an opportunity for a better price where you could be adding size at that point.

Dave: 31:57

So like I said, it's an advanced move. You need to really think through the ramifications of that. You don't want to use it as an excuse to add size when you shouldn't be, or add size to a loser that's just going to get stopped out soon. But it is a valid way to look at it, and if your strategy has edge, then that's definitely a way to add size.

Michael: 32:23

Yeah. I just I thought that was an interesting use of MAE is is it always just an interesting concept of using backtest data to help refine itself. If that make kind of like the column library works where Yeah. It's like, okay, you have this crap ton of data on, you know, hopefully thousands of different trades and how they work. How can you use that data to potentially create another strategy?

Michael: 32:54

And that was always interesting and and, you know, I thought about the same way you did when he brought this up. It's like, well, you're just creating another strategy from the same signal set that you were before. And it just but it leads you down that road of, okay, what other data points am I am I capturing without know I'm capturing that could not only increase my current returns, but could also then potentially create other strategies, which if you miss it, that's a podcast we did a couple ago where we talked about how to, you know, create your second strategy from your first one. And that might be a way to look at now that you have these numbers. And it also, I I've noticed it can start to it should do a better job of making you understand how you how the strategy will feel.

Michael: 33:42

Like there's this old trading adage that the best trades work right away. You can prove that if if whether or not that's the case for your particular strategy. It could be that, it could be the complete opposite, but it'll give you a really good understanding of of how it is to feel, which is why I'm like, I like that you gave people the multiple options because that's gonna be different for everyone. For me, I I'm fine with the drawdown. It doesn't it doesn't bother.

Michael: 34:05

I'm just old. It doesn't bother me. But some people might not be okay with that and they just want to cut their trades as as quickly as possible even if it's suboptimal. At least they're gonna know how suboptimal it is. Right?

Michael: 34:19

They're just they they'll know, okay, I'm leaving this much money on the table in order to cut my losers really quick and make myself feel better. And could be a good trade off. It might not be the trade off I would make, but it could be a great trade off for them to say, yeah, I'm totally okay with that.

Dave: 34:35

Well, yeah, and it may not be just to make yourself feel better. Like, if you can smooth out your returns, make your drawdowns much shallower, even at the cost of total profit, that's, that could be potentially a very good approach because you can scale something that, like that. You know, you, the drawdowns are what are gonna prevent you from scaling. And if you could minimize the drawdowns in whatever way you can, adding multiple strategies, you know, making a tighter stop, of different ways to do it, even if the cost of total profit, that's an important thing, Because you can scale something like that. People will give you money to trade a strategy like that.

Dave: 35:31

You know, people will give you a lot of money to trade a strategy like that. You can scale a strategy like that in a way that the drawdowns are gonna prevent you from scaling.

Michael: 35:41

Yeah. And that's the, you know, back to the whole path of confidence thing. That is the main purpose behind all of this is which you can if your stop losses are a place that makes sense to you and and get to your confidence level of how much drawdown you can potentially achieve, then things are going to kind of work way better for you where, you know, for example, if if right now you're just doing a stop loss. If it shows you a stop loss of something that you know you wouldn't be able to stomach, then you just might you might even wanna just pass on that particular trading strategy. The the the equity curve might look good all zoomed out and the trades might look fine all zoomed out.

Michael: 36:30

But at the end of the day, if you if you look at it and you say, no, I wouldn't be able to I wouldn't be able to deal with that. I wouldn't be able to stomach that. Then, you know, it's just best it's just best not to to trade that trade. And again, that's just a a very personal thing. It's gonna be different for everyone.

Dave: 36:47

Yeah. Yeah. And another important thing to to understand about MAE. That there's no time associated with it. So you don't know whether, you know, with that one number, whether that happened right in the first five minutes of the trade or for a day trade, you know, right before the end of the, you know, you exited.

Dave: 37:14

So that's an important thing to understand and think about. Now, what are the ramifications of that? Well, you know, take the flip side, take MFE, which you mentioned, which is the amount of, you know, the point of the trade where it was most profitable. You might think that, you might naively think that you can do an optimization report for both sides of the the the you know, figuring out what the stop is based on the MAE and figure out what a target is using MFE. But you can only choose one of those sides because you don't know which one came first without additional information.

Dave: 37:58

So the way to do this is what I always do is choose the stop first. So run a back test without stops, use MAE, figure out what the stop should be, run another back test with that stop, and then use the MFE to figure out what target I should use, if any. It's not gonna be valid if you try to do those both with one backtest because you don't know which one came first. Now, there are some ways to so as I get into this further, there are some columns you can add that would allow you to get to that analysis and actually do some quite interesting things, picking optimal sets of stop and target levels, which, would be pretty interesting all in one single backtest. So there are some, I anticipate going down this road and figuring out and and adding some more stuff to the cruncher that would light it in some really cool stuff with, you know, optimal combos of stops and targets.

Michael: 39:16

It was funny that when you said m a MAE doesn't have a time component with it, It's not where I thought you're going. So I thought I'd go, you know, you need to remember that the not having a time zone also means it doesn't have a time drawdown. And and that's less important I think for you guys as day traders, but it is important for us as swing traders. I thought you were gonna go, you know how long you're gonna be underwater for and that might be uncomfortable to other people. Right?

Michael: 39:48

If you if you gave both people the same number and say, hey, you're gonna be in drawdown of of a thousand dollars and you went to one person, but only be five minutes. And then the other one's like, yeah, it's gonna be a month and a half. That could be a different. Right, that could be a different thing to some people because one, you know, you might be able to look at a drawdown and be okay with it for a short period of time, but not a long one. But that's, you know, more or less I was thinking about with with what you're thinking about.

Michael: 40:16

There is no time side of things. But you're right, the way you put it is that you don't know in which order they happen, so you've got to do one and then you've got to do the the other after that.

Dave: 40:27

Yeah. So so there there is another approach you can take. So imagine, you know, mentioned, you know, waiting to get in later because the MA suggests, like the trader you mentioned. The other ramification of not having the time is, like, do you want to scale in or add size right before the timed exit if it reaches like, at some point, you're not gonna wanna be adding size the closer you get to your timed exit. So by default, these platforms do MAE across the lifetime of the trade.

Dave: 41:10

But you can add a column for whatever you want and you can compute it yourself. So one of the things I've done is have like a scale in window, the first X minutes of the trade, and I'll capture MAE and MFE data sort of manually during that window. And then I can see, you can get some interesting things because that might be the window for which you are willing to add size, after which you're not willing to add size. So you can get some more information based on that. So you can have another number and sort of get creative and use that data to figure out different things about, you know, what you might want to do during the trade in that situation.

Michael: 41:55

And that's, you know, again, we've we've come to the spot where we have to remind people that systematic trading is a lot of work and it's just this just goes and shows because we haven't talked about anything about creating a strategy or, you know, discovering an edge in the market or or anything like that. This is just purely you have a system and here's all of the different ways that you can do one thing which is the stop loss for this system. Right? It's not the time stop, it's not it's not any of that. This is just a pure, right, stop loss for that system.

Michael: 42:32

And here's all of the different ways you can think about it and all of the options and all of the millions of things that you can do. And it yeah. This is this is one number. And sadly enough, a something that I don't think people think about enough. Right?

Michael: 42:49

That's people I'd say people are kind of they look at a stop loss and they'll put something in and then they'll test their system and just see if it see if it works with that stop loss. But you can just see the audience is listening. There's a million decisions to be made just on the stop loss of one single system alone and some people think that, right, this kind of stuff is is easy. Right?

Dave: 43:13

Yeah. Well, I think it's it's a little I look at it a little bit differently than that. Yeah. There's like almost infinite ways you could like knobs you can create to modify a given system. But the way I look at it is there's so many ways to improve things, and and it doesn't have to the improvement doesn't have to be a lot to make a massive difference in your strategy.

Michael: 43:39

Well, because you're taking the exact same. So you had an idea and your idea was x y z, it doesn't matter. I'm gonna buy support. Right? I like it when stocks come to a prior day's low, whatever it is.

Michael: 43:52

And from that, you can create almost an in from that one basic idea, you now have an infinite amount of ways you try to optimize that idea to get the most amount of money out of it. And this the stop loss is just kind of one part of it, but you can see how deep you can go just on this one this one topic.

Dave: 44:12

Yeah. I'm glad you used the word deep because that's where I was going. I mean, referencing the podcast we did a while ago. I mean, there's so much depth with these when you find something that works. And you're the way you mentioned it earlier where you run this, a backtest without a stop just to see how strong the signal is, sort of the smoke test, once you get that, there's lots of different you have lots of different knobs, lots of different levers you can pull to improve the strategy.

Dave: 44:40

You can add trades, you can remove trades, you can change the stop, you could do I mean, there's all sorts of things you can do, and the bar is not high. Like, just small changes, if they're consistent over time, can make a dramatic difference. I mean, that could be the difference in whether you can scale the strategy or not. I mean, it's really that that's what excites me about trading is just you're so close to being able to have a dramatic difference in your strategy. It's it's it's kinda crazy to think about.

Michael: 45:13

Yeah. And it's just from that. Right? Just from that one thing, not only you could see where you could expand it out, like we're just talking MAE. And I'm sure, hopefully, maybe this is something that you're looking for in the cruncher where it's that we didn't talk too much about MFE.

Michael: 45:32

Right? That seems like a next logical step to take a look at is if MAE is a simple question of where should I put my stop, MFE is a simple question of where should I put my profit target. You can see based off our conversation how that simple question has like a million different answers to it that you could kind of expand out from there. Right? There's tons of different ways that you could go on is is your stop loss a percentage of your profit target or variable to your profit target or a completely different number based off whatever and and yeah, all kinds of things that you can go and and take a look at.

Dave: 46:10

Yeah, I mean profit targets are fun too because it's fun when they get hit. Right? Yes. It's not fun when stops get hit. So it's it's it's fun to think about those.

Michael: 46:21

Yeah. So as always, you know, a concept that I thought that was might take two minutes to explain. We're forty five minutes in and could probably even continue to go longer if we if we had to, but I think we covered a good way there. And you know, again, check out the cruncher. It's it's it's in there.

Michael: 46:40

You can you can start playing with it. And if you don't want to, just check out MAE in in general. That's we've opened up a whole rabbit hole that you can fall down and do all kinds of cool stuff with it. But as always, I'm Michael Nauss.

Dave: 46:53

And I'm Dave Mabe. Talk to you next week on Line Your Own Pockets.