Line Your Own Pockets

Michael: 00:00

Hello, everyone. Welcome to another episode of Line Your Own Pockets. Last one, we talked a little bit about what to optimize for, and it was such a cool conversation. We had so much talk about, we decided to do another one. So if you missed it, it'll be the last one back there on the channel.

Michael: 00:16

We'll put it in the show notes as well. But the main premise is that there's just so many different things when you're building a strategy that you can put your time and attention in saying, this is the thing that I want, and this is the thing I optimize for, that, you know, it's kind of a a strategy in and of itself just to find the the thing to optimize for. And in the last one, remember, you know, we talked a lot about how, you know, it's gonna be individual to everyone. I think in this one, we wanna talk a little bit more about concrete things to do. But I had one thing that I I had thought about, and I actually had talked to another trader about, and I wanted your take on Dave.

Michael: 00:57

And it was kind of... We had poo pooed a little bit about just looking at like win rate. You know, he said in the cruncher, you have this ability to optimize a win rate, but really mathematically, it doesn't make much sense to do. And when I was talking to this trader, he's a new guy, right, he's just just starting this kind of a thing. And he made an argument to me that the only thing that he matters, he's he's a day trader, is he wants to optimize for the maximum amount of winning days.

Michael: 01:30

And my instant reaction was, well, that's dumb. That makes no sense. But then I kind of put myself in his shoes as someone who's brand new at this, probably doesn't even truly believe that maybe it's super possible to... You know, he's still in that, can I do this? Is this something that I can do where I can create a bot to trade for me and go to work and come back and and make money?

Michael: 01:51

And the more I thought about that, the more I'm like, I don't think that that's a bad idea for newer traders to start out with that. As kind of like a kickoff to get success, even if it's not optimal success, but get some success, and then build on that, like is, I don't know. What do think?

Dave: 02:10

I don't think it's a bad idea. I... So the the thing that sticks out to me about that is he's beyond the point of just optimizing for for profit. Right? That's...

Dave: 02:22

Which I think is a good thing. And, yeah, I think think about it that way is kind of a nice way to think about it. I think that's I think that's what a lot of day traders would like and sort of... To put it so succinctly like you did, I think is good. But, you know, other kind of things sort of talk around that.

Dave: 02:46

But, yeah, I think that's a good way to look at it.

Michael: 02:49

Well, and and that was the main kind of point that we we had talked to and got to last time, is that it's gonna be different for everyone. There probably is some wrong. Like, I know we we crapped on Sharpe ratio a fair amount, so we don't need to rehash that. But, you know, when I was talking to this guy, again, my instant reaction was, that doesn't make any sense. Because what if you're optimizing for maximum winning days, and you're making $50 a day?

Michael: 03:13

So, you're making, what's that, like a thousand bucks at the end of the month, if you hit all days. And I'm like, but you could optimize for profit, and then even if you're wrong on half or more of your days, you're right on half more of your days. If you make $2,000, well well, mathematically, that's the correct move. Right? If if the average end of the month, you're making double what you were.

Michael: 03:33

But, you know, going back to where this guy was in his trading career, I'm like, you know what? This is... Maybe I was premature on thinking that that was as as crazy as it is. Because for that guy, he just wants to know at the end of most days, there's some money in the pocket. The account has grown slowly and surely.

Michael: 03:55

And then five years from now, he probably won't be optimizing on that anymore. He'll probably be more optimizing on, you know, what what is the the smoothest return for the most money and and things like that. But it just really shows that, you know, where you are in your stage of trading could completely change the things that we're gonna talk about for optimization. And at the end of the day, whatever feels good might not be optimal, but, you know, it's your money at

Dave: 04:24

the end

Michael: 04:24

of the day. So it's gotta gotta feel good in some way.

Dave: 04:28

Yeah. Mean, I think that's a big part of why people have different preferences. Like, you're gonna have some different preferences based on your risk profile, based on your experience, based on how much money you had when you were growing up. But there's just so much that goes into it. And I think a big part of what you're optimizing for is sort of what is resonating with you.

Dave: 04:54

What are you... You know, what is... Like, I say all the time, what's your path to confidence? This is a big part of it. Right?

Dave: 05:03

You know, take his strategy, you could come up with one version that I'm sure is optimized by profit, and there's gonna be way more losing days in there. Mhmm. But way more profit at the end. But the way he's talking about it is he's probably gonna end up with more profitable days, but less total profit. But that's absolutely and totally a rational thing to optimize for, and it's it's just highly dependent on the person.

Dave: 05:39

So yeah, I think that's totally decent way to think about it, and you're probably right. That's gonna change over time. The other thing I think will change is, as you get more strategies, that kind of happens automatically anyway. So you actually have more flexibility when you have more strategies to to optimize on different things that sort of end up with what he's shooting for.

Michael: 06:07

Yeah. And that's, you know, that's what we always preach about with the multiple strategy things is when you're on strategy 10, all of a sudden, you're... The amount of days or the amount of weeks or months if you're a swing trader that are profitable just tend to increase because of, you know, the the fact that you're running different things that are gonna be hopefully not super correlated anyway. And and one thing that I I like that you talked about is that you could in theory, and and I I kind of approach this to the person as well, with the same idea, run multiple strategies optimized for different things. So, you know, for for example, right, just a simple kind of gap and go, opening range, breakout, whatever.

Michael: 06:48

One way to optimize for consistency of gains might be a pretty tight profit target. You wanna buy the thing, you wanna you wanna flip it out just a little bit later. And that's gonna increase your your hit rate and and probably smooth out the returns. But then you could also test the same thing with like no profit target, just holding until the end of the day. And that's likely gonna do the opposite in which you're going to, you know, probably make more money, but you're gonna be right less often.

Michael: 07:13

And then... So what I told him is that, you know, optimize for the thing that that makes you the most comfortable, but just keep in mind that the same strategy potentially could be both. So may... You know, maybe the same... Then you start running both.

Michael: 07:25

Right? Then you start running the one that has frequent small winners, and that creates one equity curve. And then in the same strategy, maybe even with the same tickers, you know, a completely different strategy is running on the other side of things that allows that kind of lottery ticket, you know, thing just blows up and goes all day, where you're making a bunch of money just more concentrated around a number of trades. And then and then I was like, then play with weightings. Right?

Michael: 07:54

Maybe in the beginning, you're weighting a lot of the risk to the consistent one, and then as you get more and more confident, you're you're kind of flipping the weights over so that it's it's more on the other side of things. And, you know, just to go back to the whole what to optimize for, that would be the exact same strategy, potentially trading the exact same tickers, optimized in two separate ways to show you that there's not just one thing to optimize for. There could be different strategies that you optimize for different outcomes, and those outcomes kind of make sense and and kinda gel together.

Dave: 08:29

Yeah. I love thinking about it that way, and that's... I think that's a great suggestion. Like, you know, go through your entire optimization process for this strategy using two different targets or two different, you know, fitness functions. So...

Dave: 08:47

And then add two different versions, and you're gonna learn a lot doing that. You're gonna see the differences. There's gonna be some similarities, but there's gonna be some differences. And anytime you go through the entire process that way and come up with a different version, you're gonna learn things, you're gonna gain some experience, you know, you're gonna have some things in the back of your head going forward that that stick out at you as a result of doing it. So, yeah, I think that's a good approach.

Dave: 09:12

And then, I mean, just having, I think, probably two versions at most for this. Like, it's it's easy to it's easy to get stuck in a rabbit hole going through all different sorts of ways you can optimize things. But I would just, for a given strategy, sort of step back and think about maybe a couple things that resonate, maybe that are pretty different, and take that approach rather than, you know, shotgunning every single fitness function that you could come up with, to come up with, you know, 20 different versions of the same idea. Mhmm.

Michael: 09:50

I I do think that's a good road to go down though, is just thinking about not just this... Not the same strategy anymore, but when you're optimizing for things, do you sometimes optimize different strategies for different outcomes? Like, this is something that that I do personally, but I wanted your kind of feedback on. So like, quick hit strategy is maybe optimized for consistency or even something like a mean reversion strategy is optimized for win rate because those generally have pretty high win rate anyways. Whereas a a more, you know, breakout or trend following strategy is optimized differently?

Michael: 10:28

Or are you looking at the same metrics for each individual strategy?

Dave: 10:33

That's a good question. I'm looking at it. I'm looking at basically the same metrics. There's some strategies that have, you know, really high tail risk. So I'll do something a little bit differently with those.

Dave: 10:45

I'll look I'll look harder at MAE. And, you know, optimizing on MAE is super interesting because that's you know, I was thinking back to your... The trader you talked to who wants to optimize on, okay, at the end of the day, I wanna have a profit. That's sort of the goal, like, in as many days as I can have with that. And, I mean, just think about the range of time you're in a trade.

Dave: 11:07

Like, he's... And you're trading in general, so you could have... You could optimize for months of profit, right? You could optimize for years of profit. You could optimize for days of profit, like he said.

Dave: 11:22

You could also optimize for, like, each individual trade. I wanna have, I wanna be profitable as many one minute increments during the day as I can. Right? That would be another way. It's like that's just a smaller time frame.

Dave: 11:35

So so I think think about the lifetime of a trade, and the ones where you're really uncomfortable. Well, it's when you're in a big loss. Right, and you don't know whether the stock is actually gonna end up profitable, but you have to tolerate a big loss during the lifetime of that trade. Well, that's another thing you could optimize on. Like, I want to minimize the pain, the intra trade pain that I go through, so you could optimize on that.

Dave: 12:11

So, like, not even... It's gonna be highly correlated with profit, but it's not the same.

Michael: 12:19

Yeah. See, and it's funny, I just... I had the same visceral gut reaction to that as I did to the guy initially telling me he was optimizing for the number of days, but but maybe I'm I'm wrong there as well, where, you know, yeah, now you're talking about feeling good every second of the trade that you'd ever want to, you know, if you're if you're doing like on a a maximum adverse excursion or something like that, you're not even optimizing for realized profits per gain. Now you're off to... On unrealized profits per trade.

Michael: 12:52

And my brain instantly went to, man, how how much of a free ride are you trying to make make trading as a whole? But it... Just like just like I... With the other guy, it doesn't necessarily mean it's wrong. It's it's likely suboptimal, but suboptimal doesn't mean wrong because it's, you know, it's something that you're you're looking to do.

Michael: 13:14

Unless you've noticed some sort of correlation with the, know, you are right on the trades. The trades that work best work right away, which is this old trading adage that I don't think really carries any any weight to it, but it's something that a lot of people seem to say is that, I I wanna be right right away on the trade. That's basically what you're saying is that you buy it and you're just never underwater for any period of time. And it's like, at some point, do you just have to close your eyes and just say just... And take a little a little heat if you're talking about, I don't wanna I don't wanna be red for a second.

Michael: 13:47

Like, at what point does it get a little too extreme?

Dave: 13:50

Okay. Well, there's a couple things. Some of the strategies that people trade have really high tail risk. So the ones where you... It's not just a little bit of pain, so it can sometimes be a lot of pain.

Dave: 14:05

So yesterday, I believe it was just yesterday Mhmm. There was a trade BIYA. It went from two to 30, over 30 in less than a minute. But that could be not just a little bit of pain, that could be a catastrophic loss for traders. And optimizing on maximum adverse excursion can help, you know, reduce reliance on some of those traits.

Dave: 14:35

It's not... And that's that's not true for every trait or every type of system, but for some systems, that's ultimate... That's really important. Right?

Michael: 14:43

Yeah. No. I I agree there. Especially if... And this is something that I think would probably make more sense for swing traders than day traders, because I've seen a lot of swing trading strategies that essentially, you're only looking...

Michael: 15:00

Because you're not looking at maximum adverse excursion at all. So all they're looking at is, this is the entry price, and then maybe two or three days later, this is the exit price, and okay, most of them are are profitable, you know, say it has a 65% win rate, and the person looks at that and goes, oh, it's great, it's smooth. And then their equity curve is, you know, doing that every two, three day update on a trade, and the equity curve looks incredibly smooth. And then the the gentleman started trading it, and then all of a sudden he realizes, well, you know, I'm buying it on day one, and then for most of day one and some of day two, I'm underwater. And then day three, I end up getting out for a profit.

Michael: 15:38

And, you know, that just made things just a little bit more uncomfortable. But, you know, like we always say, my advice to him was like that, you just opened up another door there, because maybe you just figured out how buy it on day two, and you get a better price, and then you spend less time underwater, and you turn over a little faster. So you're buying it on day two and selling it on day three. So that's another thing I wanted to talk about as well, is that optimizing for different things can sometimes create the right questions to answer. So for what we're talking about now for the maximum adverse excursion, if you notice that all of your trades spend a whole lot of time underwater before turning profitable, That's not something that you just go, okay, that's great, I can deal with that.

Michael: 16:24

That should start asking you questions. You should start saying, well, man, could could I get a better price on a lot of these things? Should I be putting limits below the market and and the ones that work come in and and fill those limits? And wow, now, not only am I I'm doing less pain, but my maximum profit is gonna work just as well as well.

Dave: 16:45

So. Oh, yeah. I mean, no question. You hit the nail on the head. I mean, there's so much potential with what you could do with MAE.

Michael: 16:54

Mhmm.

Dave: 16:54

It's just a really powerful thing that most people don't... Most traders don't think about. As you were talking about the intra trade drawdown, it reminded me, so on the equity curve, most equity curves you look at are closed trade equity curves. They're not capturing the amount of pain or loss you're in during the trade. It's just that the close of the trade.

Dave: 17:22

So I thought about adding to the cruncher, like, a little shadow underneath the curves based on the MA. And I've... It's... I had... In my mind, it looked prettier than it ended up being, but it looked...

Dave: 17:38

It's... There's something there, and... But it's... It is useful to think about and and see the... Sort of the shadow of how much you're underwater during the lifetime of the trade.

Dave: 17:50

And you're totally right. If you see a bunch of these, the first thing you should think is, okay, maybe I should delay this entry and get a better price. Some of them are not gonna... You're not gonna get filled, but it's worth the experiment. Right?

Dave: 18:04

Look and see how many aren't gonna get filled by you using a limit price lower and a delay. And then look and see how much how much do get filled, and how much of a better price it is.

Michael: 18:17

So Well, and also goes into the the two strategies potentially on one setup is that, you know, maybe... And this is the whole... I I love things that go against everything we're taught about in trading, which is you never average down. It's like, well, I've got a couple strategies that, you know, if I put the... Like you mentioned, if I put the limit at where kind of the MFE is, I'm wasting a lot of money in buying power because not many of them are getting filled.

Michael: 18:49

And... But when they do, they end up being really, really good trades. And then I have the same strategy that ends up, you know, buying them at more shallow pullbacks, and that triggers way more frequently with okay trades. And it's like, okay, so I just split the account and I trade both. So, right, one limit goes here and one limit goes there.

Michael: 19:08

And most of the time, just take half size and and get out and everything's fine. But then every now and then, I get both trades and then get out, and those end up generally being a little bit better. So, yeah, it's one of the main things I wanted to talk about with this one too is just that it's just we're not gonna give you... So if you're searching for it, I'm sorry. But we're not gonna give you this is the thing you optimize for to make the most amount of money.

Michael: 19:32

But it's understanding what each bit of data you're optimizing for tells you, and then getting more questions from there. Because, you know, if for example, you had a strategy that was right 70, 80% of the time, like an insanely high win rate, it's almost to the point where it's like, okay, you've optimized on that. But at that time, I would almost back off and say, okay, well, if I'm being right 80% of the time, should I be extending my profit target? Should I be holding for longer? Should I be doing something that's going to drop that win rate a little bit, but allow more of these kind of outside gains?

Michael: 20:11

And just like we're talking about, if you're if you're noticing with maximum adverse excursion that you're getting a lot of it, well, that's one thing. But I would even say it'd be the complete opposite if you're getting like none. If you're entering the trades and they're going with you right away, maybe you wanna enter them earlier because maybe there is more gain to be made there. You know, if you're buying a breakout, maybe you should be buying the breakout sooner or something like that to end up getting more out of those those trades.

Dave: 20:39

Yeah. So it reminds me of a thought I had recently about the system that I'm trading now. So I use a stop limit to get in. So it needs to go in my direction a little bit first, it's confirmation. And then once it goes enough in that direction, I get in.

Dave: 21:00

But I don't wanna get too... Like, I'm putting a limit on the price, so my price isn't too crazy. So I've traded that for a while, I have a certain distance between the stop and the limit there. Now, what I've noticed over the past probably three months of trading this strategy, there hasn't been a single missed trade. I haven't...

Dave: 21:25

Because you would expect some to jump through that and not get a fill or get a partial fill. So that makes me think that the distance between my stop and my limit is too far. I can probably... I I kinda want to see some mistrades every once in a while because that would mean that my limit is set probably at a pretty good distance away. I know some of these come back and would hit a a tighter limit there, and I could get a better price.

Dave: 21:57

So the first thing I think is, know, first thing, trader that I was ten years ago, I would have thought, hey, I'm getting all the trade, no missed trades, that's great. The trader I am now, I think, okay, I should have some missed trades here. Like, if something's not right, like, I could get... This is telling me I could be better.

Michael: 22:21

Yeah. No. And I I I think we do need to address the elephant in the room sometime with optimization, is that at what point at what point do you kind of pull the trigger of over optimization, I guess? Like, if you are, you know, if you're using MAE and and you're optimizing for it, and then you end up buying the exact low of all of these setups, is there is there any flag that happens in your mind to say, okay, I need to really, I know there's no definitive answer. There's no, okay, this is...

Michael: 22:53

The only real test is time and actually trading it. But since we're talking about optimization, is there any point along this way of optimizing for all these different things, and then looking back and learning from all of them, and then doing more optimization, do you go, okay, I've I've gone too far here. I've just, you know, made the greatest looking equity curve in the world, and and nothing else. Well,

Dave: 23:16

The part that I mentioned there about the stop and the limit, I don't look at that as over optimizing at all. I look at that as pure execution. How can I get that to match the back chest more optimally? And I don't think I don't think of overfitting and execution necessarily in the same conversation. As far as designing the strategy, yeah, I think that's...

Dave: 23:43

It's something that should always be in the back of your mind. Am I overfitting? And, you know, go back to the episodes we did on curve fitting to understand more about it, but I think you you should be challenging yourself every time you add a rule. Is this curve fitting? Have I already added a rule for a similar feature that I'm sort of duplicating now?

Dave: 24:10

I mean, that's another big thing. You should... Each individual rule that you add should add something unique. If there's a lot of overlap, there's more of a chance of curve fitting. So, yeah, think should always be in the back of your mind.

Michael: 24:26

I guess what I was asking, and I think I probably word worded it poorly, was that... So if you build a setup, and in backtest world, it looks perfect, but... Or not perfect, but like really really good. And then in the real world, you find that the trades deviate quite from it. You know, slippage or or like you were mentioning, you know, you're trying to mess with your stop limit to make sure you're as as close as you can.

Michael: 24:53

At what point do you think... At what point do you stop messing with the execution side of things and say, okay, the original strategy may have been too forgiving or unrealistic on fills, as opposed to constantly chasing. I think that's more... So over optimization, I think, or curve fitting might not have been the right thing, but just saying the original strategy looks great, but it's just not viable in the real world. Because I I do have...

Michael: 25:19

I've talked to a lot of traders that I think are in this scenario right now, where they have created something that looks amazing, but I just know they're never going to get those particular fills that way. And, you know, trying to find out at what point do you look at it and say, even though they may not have messed with the back test too much, which is usually what happens when it looks... When the execution is the problem, usually, you're not adding a ton of filters to optimize the back test. It just looks great right away because it is slightly unachievable in the real world. So like, where do you pull the chute there and just say, I've messed with this for too long.

Michael: 25:59

I I can't seem to get the fills from the the back desk.

Dave: 26:03

I think you should be able to get that... You should get to that point pretty quickly. And I think you you should go back... Listeners, if you haven't listened to the episode we did about Sridhar's trading a few weeks ago, it's worth listening to because he went live with some strategies, and in my opinion, didn't have as good a process for reconciling back to the back test that he should. You should really...

Dave: 26:31

You should pretty quickly be able to see how the trades measure after the back test and extrapolate back from that, you know, for that whole backtest pretty quickly. So if you're going several days or lots of days and you're... You haven't figured that out yet, then I would look at those trades each day, look at the reconcile each day, where is the slippage coming from? Is it coming on the exit side? Is it coming on the entry side?

Dave: 27:03

Is there something you can do about it? There's just so much value in doing that. You're gonna get so many to do items, like stuff you can actually hang your hat on. It's okay. If I if I can go figure this out, then that's gonna concretely improve the strategy.

Dave: 27:17

I've just proved that to myself. If you don't have that process, then you're kind of guessing. You're not really, like, you've got a back test, you're trading it live, you're like, well, it didn't match up, I don't really know why. It's, you're in this world where you... There's nothing you can really do about it until you dig in, and that's where the real learning is gonna come from, and the lessons you learn from doing that.

Dave: 27:40

You can go back and modify your practice in in a way to handle some of those things and improve it. So rather than give up on a strategy because, the backtest is crap, or this this whole process doesn't work, or whatever Mhmm. You can actually see, okay. I see why the backtest is too optimistic here, and I... Let me go back and fix that.

Dave: 28:02

And maybe the backtest is not as... Yeah. The the equity curve doesn't look as good, but I know it's, like, more in line with what reality is gonna be, and that, you know, that sort of dance that you do over time is really very valuable and productive.

Michael: 28:21

Yeah. And we've talked about this a bunch, but I think it's always worth hammering on it more. That is the real work. Right? The...

Michael: 28:30

It's it's not the fun work, at least not for me. The fun work is coming up with new ideas and testing them and, you know, seeing the equity curve and saying, oh, that'll be, you know, employing that. But the real work is tracking the fills and Yeah. You know, making sure that you're narrowing that as much as you can. And the beauty of it is eventually, once you have all these problems sorted out, it's it's not a lot of work.

Michael: 28:56

It can be automated, things like that, but especially in the beginning, in like the virgin parts of a strategy, it is hugely important. You know, I've... I'm spending a lot of time on some of the strategies I'm running now, Not because the fills aren't making sense, but more just I'm putting out a shit ton of orders for not a lot of fills. And Mhmm. On paper, that looked fine until I'm sending, you know, 30 orders to buy and getting like two or three fills a day and just saying this is just horribly inefficient use of buying power.

Michael: 29:29

I... You know? And so it's it's not the fun part, but it's it's the part that's gonna make me the most amount of money. If I can make it so that I'm getting filled on 50% of trades as opposed to right now, which is like 10% of trades Yeah. That's going to dramatically increase the profit per buying power used of the of the strategy.

Michael: 29:52

It's just not the point I I... The part that I think people come to systematic trading for, but it's the part that's gonna change. It's the difference between profitability. It's a difference between this strategy being a good strategy and a great strategy, if I can shrink that number down where I'm putting out less useless amount of buying power and having it sit all day all day waiting for for orders to be filled.

Dave: 30:13

Yeah. So it used to be in Interactive Brokers that those unfilled orders wouldn't count against your buying power.

Michael: 30:22

Mhmm.

Dave: 30:23

But now they do. I I have a strategy that sends a bunch of trades at the open, and it was... I rem... It must have been... I don't know how many years ago now, but at some point, they said no.

Dave: 30:35

Like, a lot of them started getting rejected. And I was like, what's going on here? This is worth forever. But, yeah, they started dinging your buying power based on the unfilled orders. So yeah, you're right.

Dave: 30:48

All of a sudden, you've got a big optimization problem there to figure out, okay, I can't send all these orders. Let me see what I can do to filter them out so I can only send the ones that are likely to fill, or like the best ones that are gonna fill

Michael: 31:02

Yeah.

Dave: 31:02

To be to be able to make it work.

Michael: 31:05

Well, what what... I bet you, and I'm I'm trying to remember when, I bet you it was around 2020. Because I bet you what happened is there was a lot of systematic traders with mean reversion setups, and then a couple times in 2020, the market just locked limit down. So all of a sudden, the trades that were getting filled maybe 10% of the time, and they're like, you know, these other ones are fine. You know, we'll with the buying power as they get filled, and then we'll we'll cancel the orders or or stop them from filling new ones when that happens.

Michael: 31:34

All of a sudden, you have one surge down in the market, and dude gets filled at like a, you know, a thousand percent is normal buying power, and they go, oh shit. We're we're really on the hook for, you know, we just let this guy with a million dollar account buy $20,000,000 worth of stock. We should figure this out. And I I would imagine that's what happened to the rules. They, you know, a couple times of that going on, they're like, okay, we we gotta change this thing.

Michael: 31:59

This is

Dave: 31:59

gonna Yeah. That timing does sound right. I don't remember exactly the date, but that that sounds about right.

Michael: 32:07

Well, I I guess it kinda shows... First of all, it shows that even the big brokers out there, multi billion... I don't even know how much I... An interaction brokers worth, but it's hundreds of billions of dollars of market cap. Still, every now and then, looks at stuff like that and says, we need to optimize our strategy some, and then goes back and and take some change.

Michael: 32:25

So just know that if if they're having to do it, then you will probably always have to do it as well. Right?

Dave: 32:33

Yep. So I wanna talk about, you know, we talked a long... Like, sort of about the whole range of a lifetime of a strategy. And I... When I think about optimizing, I really think more about designing the strategy itself, like, before you kinda go live with it.

Dave: 32:56

And there are different fitness functions that you can use to optimize on. That's sort of the technical term for these for for what you're optimizing on. We most often think about profit. I mean, that's obviously sort of the default. That's what most people think about.

Dave: 33:15

But there are other ones you can come up with. Like, we've talked about using MAE as a fitness function to minimize that. And then, you know, we talked about the... Looking at the equity curve and seeing how far underwater you are. One of the things I thought about, and I haven't I've implemented this a little bit, but I haven't done a whole lot of research into it.

Dave: 33:41

I've thought about another fitness function called... That I'm calling efficiency. Okay. And it would be, instead of optimizing on profit or MAE, like, think about the perfect trade. The perfect trade is you get in, it goes immediately up, and it never goes back in any way.

Dave: 34:02

Right? You end up at the high of the the day or the, you know, your your high watermark for profit. Yeah. Right? And you get out.

Michael: 34:08

The the the fantasy trade. Every time you look down at your your blotter, you're up another $100 and you're... You never... That never downticks, at the end of the day, it's at the highest equity that it's been that day. The the the fantasy trade that we we just dream that they could all be.

Michael: 34:25

Now, every every trade you

Dave: 34:27

make is somewhere in between that fantasy trade, and then like the nightmare trade, which you get in and it goes immediately down, and you get stopped out. No variation. Right? So every trade you make is somewhere in between that. Well, you can quantify it by a metric that I have come up with called efficiency, which is basically rewarding trades that look more like the dream trade, and penalizing trades that look more like the bad trade.

Dave: 35:04

Now think about the different scenarios. Some could go up and be really profitable, but end up at zero profit at the end of the day. Some could go really negative and come back and be zero profit by the end of the day. Well, you're just looking at profit, you're ignoring those two situations. Mhmm.

Dave: 35:27

So the efficiency basically is gonna be... It's it's a metric for, like, the the difference between MFE, the point where it was most profit profitable, and the ultimate profit at the the closed trade profit. Yep. And then the difference between the... Your, you know, zero and your MAE.

Dave: 35:52

So basically, the smaller the m... The the MAE, the better. And you can actually divide profit by those two distances. So maybe I'll put this in the show notes, it'll be easier to understand. But basically, what you're doing is you're rewarding trades that go more like the dream trade

Michael: 36:13

Mhmm.

Dave: 36:13

And penalizing the ones that maybe go up really profitable, and then end up of half your profit. Right? Those get penalized. If you were just looking at profit, they wouldn't get penalized that much at all. Right.

Dave: 36:27

So it's a... I think that's gonna be a good way to look at things, and I've I've I've not looked closely at it yet, but I'm excited about looking in and seeing exactly what you can do with that. And, you know, so you... This would be an option in the cruncher where you could say, instead of optimizing just on profit, you could optimize on efficiency, which sort of is I what you mean, you're sort of minimizing pain by doing that.

Michael: 36:56

I I think it it shows a lot too how... I always love to dispel the oh, systematic traders just, you know, go to the beach because they don't need to be there to have the trades happen for them? It's like, no, we don't. The whole, you know, like, I always joke with all the AI optimizations, everything I have, I could get hit by a bus and all my trading would exist for probably a month or so before there'd be any problem. But this is the kind of stuff that we're looking to always do.

Michael: 37:23

Right? Is because... Just because you don't have to hit the buttons to place the trades, doesn't mean you're not you're not spending time thinking about and and building this stuff out. Because that is an entirely different thing to optimize for and and could lead to different results. It reminds me, I don't know if you studied it too closely, but actually reminds me a little bit of the ulcer index.

Michael: 37:43

Mhmm. Might be something to to look up.

Dave: 37:45

Yeah. Of it, but can't... Don't know the definition of it.

Michael: 37:48

It's essentially that. It's it's trying to add value the more the trade is in profit, and during any sort of drawdown, it starts to kind of subtract value from there. So the idea is that that same type of thing that if you if you optimize for this particular metric, you're optimizing for return without drawdown. So not necessarily absolute return and not necessarily, you know, smoothness of equity, but as soon as a drawdown either, you know, happens at all, it subtracts value from this. So the higher value kind of ulcer index...

Michael: 38:25

I'm trying to remember the exact math, but I think I'm doing this right. The higher value ulcer index it is, then the least amount of of essentially red candles against you during that period of time Mhmm. You would have you would have experienced for a long trade. And, yeah, doing this kind of stuff just will completely sometimes change the exact same trading strategy for what trades you decide to let in. Because, yes, this trade might have given you the absolute most amount of profit, but it could have been incredibly volatile along the way as opposed to, you're right, the one that just ticks up kind of slowly over time.

Michael: 39:02

Those are are a, probably going to be way easier to size up in over the time, so give you the most absolute profit that way. But also, it's just gonna make it way easier to stick with the strategy if most of your trades just just go up. Right. Right? It's just up and

Dave: 39:19

That's right, you an important point. I mean, back to the guy you were talking to at the beginning, who wanted to optimize on, you know, number of days profitable.

Michael: 39:31

Mhmm.

Dave: 39:32

That's probably gonna end up with something that has less total profit, but it's probably gonna be a smoother curve, and that might be the exact way to go. I mean, if you have a smoother curve, you've got a better path to confidence, even if it doesn't make as much money. Like, you you can scale that up in a way that you can't if something has a lot of tail risk, or has drawdowns that are gonna be uncomfortable that lasts a long time.

Michael: 40:02

Yeah. And that was the last kind of discussion point I want to to hit on is that all of these metrics we've talked about are either per trade basis or per strategy return. So, like, profit would be a per strategy. Strategy starts out with a 100,000. What does it end up at after six months or a year or whatever?

Michael: 40:24

You know? Or per trade basis, what does each individual trade look like? None of those incorporate this aspect. So I was wondering how you kind of look at that. The aspect of the smoother the profile, the more you can size up the trade.

Michael: 40:41

So if you go from, originally risking $50 a trade, depending on the size of your account, that's going to be harder to do. You know, say someone's got a $10,000 account or whatever, and they're they're they're risking $50 per trade. If their win rate is really high, and the amount of, you know, it's like 80% win rate or something like that, it's gonna be way easier for that person to slowly increase the risk size to $100 and maybe $200 on the exact same trading strategy than somebody who, you know, they're risking $50 per trade. Even though they make more money at the end of it, they're only right 30% of the time. So you're you're getting just more of those $50 paper cuts as as kind of time goes on.

Michael: 41:28

Do you consider that at all, the amount of equity drag that comes on or the, I guess, the ease of sizing up in the trade? Because that would be the way to overcome something that is smoother, but generates less profit, is if it's just you can look at that and say, yeah, it generates less profit, but it'd be way more easy to to size up on this one, so I'm just gonna go with this instead.

Dave: 41:53

Yeah. I mean, that's... I'm thinking about that from the very beginning. And I don't think about... Like, see a lot of traders that will, you know, abandon a strategy, or try to think ahead about, okay, how much percent is this gonna make my account if I go live with this?

Michael: 42:10

Mhmm.

Dave: 42:10

I don't think about that at all. All I'm thinking about is, how can I get the smoothest curve? Is it gonna make a decent amount of profit per trade? I I wait until I actually start trading it before I start making those decisions about, you know, how to scale this up. And I think I think a lot of traders can get stuck trying to anticipate how these...

Dave: 42:35

You know, how a strategy might work with other strategies you're you're trading, or, you know, how... What the percent turn... Return is gonna be for this one. Like, that's completely irrelevant, really, because you can scale up after you start trading it, and it's it's Mhmm. And that is it's gonna be highly dependent on the size you use, and the only way you're gonna be able to increase size is to have a smooth curve that you're comfortable trading.

Dave: 43:01

I mean, even if... Even when you got an enormous account, even the the the amount of money that you're making and losing each day in raw dollar terms has an impact on your psyche. Right? It's not all about just percentages. It's the amount of money, and a lot of people really have trouble risking lots of money, even if their account is really big.

Dave: 43:28

So it's really important. I mean, the other thing to think about is, if you've got smoother returns, you can't... Your... Other people would give you money to trade, the smoother the the returns are. Right?

Dave: 43:43

There's lots of different ways to scale as long as the the curves are smooth, and that's really the important thing.

Michael: 43:51

Yeah. And that's that's a good point there to to talk about as well is that, you know, if you had a strategy that made money most weeks and most months, you look at it and say, well, this isn't it, you know, I I could never, in my current situation, size up big enough to make this worthwhile. You don't need to use your money. Right? There's a lot of, you know, there's there's prop firms that you can go and apply at, or there's there's ways to get, you know, allocated capital, there's copy trading.

Michael: 44:24

There's like a million different ways to go and take a take a good strategy and apply more capital to it. There's... If you're really confident in it, right, there's even, you know, ways that you could probably leverage it up and use margin and things like that. So, yeah, always think about how the strategy actually performs first, and then go from there. Because I I don't know how many times I've heard people say, well, you know, I only have a 50,000 account.

Michael: 44:51

I need a strategy that's really gonna go big in order to make it worth my while. It's like, no. You know, if you can make a couple percent every single month and do that consistently, Doesn't matter if you have a $50,000 account. Right? You can go to, you know, a prop firm and and show them the track record on that $50,000 account.

Michael: 45:10

Even though you're not feeding your children or anything with it, eventually, if you have a long enough track record and and good enough results, somebody will give you a million dollars to trade with that strategy. So don't worry too much, especially in the beginning about the absolute dollars that gets given. Look at, you know, risk adjusted return and and consistency of equities and or consistency of equity gain and and things like that. And, you know, the the money will figure it out somehow down the road.

Dave: 45:39

I mean, the other thing to think about is you're... A lot of this learning happens after you're trading something live. Like Mhmm. Like buying power. We were talking about buying power.

Dave: 45:54

I I I don't suggest thinking that much about buying power at all when you go live with a strategy. Wait until you see your account run out of buying power, and all of a sudden, you're gonna be focused right in on the problem that as it's happening. I see a lot of traders like, prematurely optimizing for this problem when they... It's it's something that's never gonna come.

Michael: 46:20

Mhmm.

Dave: 46:20

And when it does come and you run out of buying power, often it's like one, it's it's very often one trade, like a big outsized trade or something. And immediately, there's something that you'll brainstorm and think of a way to solve the problem. And that doesn't happen till you see it in real time, you see the the message come up, hey, you're out of buying power, and then you immediately can see what's going on. It's really... I mean, it's almost impossible to do that when you're backtesting and anticipate that.

Dave: 46:51

And you're... You could be laser focused on the exact thing to do, and then you can go back and modify your backtest in such a way to to optimize for this. And that's the fun part. That that... That's a really fun problem to solve, and...

Dave: 47:06

But you're not gonna be able to... You're not even in the position to be able to do it when you're backtesting.

Michael: 47:12

Yeah. That's, you know, first world first world problems, which are the best ones to have, where it's like, oh man, I'm, you know, all of these strategies I have that are making money are are making me run out of buying power. How can I more efficiently make more money?

Dave: 47:27

Exactly.

Michael: 47:27

And Yeah. Yeah. Yeah. It's much better to have that, and you know, have that, watch it happen real time, and solve it there, than worry about that where most people generally never get off the off the launch rates. Like like the the gentleman we had talked about a couple weeks ago, right, who was running...

Michael: 47:46

Trying to run like eight strategies at once, and he wasn't profitable on his first one yet. It's like, okay, just get in there, figure out something, make a couple bucks, then from there, you start to expand out. And like anything, it's it's... Trading is very much kind of running a business in this way, where there's people who will spend their entire lives planning out the best way to launch their newsletter or business or whatever, and then they never really get anything done. And by the time they do launch, they think, oh, well, I figured everything out, and then they come across a bunch of a bunch of issues anyway.

Michael: 48:19

Whereas if you just say, hey, I'm gonna launch a newsletter, I'm just gonna start writing, And then you just start writing. You... You'll figure... You'll have all the same problems the other person did. You just won't waste all of your time trying to preempt them.

Michael: 48:32

You'll you'll just figure it out as you go. You go like, oh, you know, it's not converting enough. One, work on the conversion. No, you know, it's not... Right?

Michael: 48:39

So you'll just, you know, it's better, especially in trading, is just to get started reasonably and sensibly, and then optimize for these different things that we're talking about as things happen. Like, the example that I gave where I have this strategy that's doing well, but it's just tying up a ton of buying power. So now, I'm, you know, I'm going back and I'm optimizing that. I'm not going turn off the strategy. I I don't have to do anything right now because it's a great strategy and it makes money.

Michael: 49:06

So great. That's a... You know, if I if I left it, that's fine. But it's always more fun, like Dave was saying, to have something that's working and just constantly work to make it work better than to be, you know, still in back test land trying to figure out something that works. Just to stumble upon the same problems that you don't even know you're gonna stumble upon until you get out there anyway.

Michael: 49:28

Right?

Dave: 49:28

Yeah. I think that was... You know, your problem where you're running out of buying power with the other orders, like, if you didn't try to anticipate that way back when you were backtesting, I I think you... I see traders get stuck in that rut of, like, get something live, and then figure it out once you run into the issue you're running into. I think that's just a much better way to do it.

Michael: 49:51

Well, and like we talked, well, you always wanna be working on your next strategy when your your current strategies are working. It's the same thing. I got it in there, it's it's making decent money. I'm just like, well, I, you know, I could make it better. Like, I I could make it better.

Michael: 50:05

And it's the whole perfectionism versus just kind of tweaking on everything So, over you know, two episodes, I I figure we could go on forever about different things to optimize for, but I think that covers a lot of it. This is definitely a topic I'd imagine that we're going to re approach later, maybe when Dave's done his work on his cool math equation that he's he's playing with, we'll come back and and chat about that. But really good food for thought there. As always, thank you everyone for tuning in. And as always, I'm Michael Nauss.

Dave: 50:38

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