Line Your Own Pockets

Michael: 00:00

Okay, everyone. Welcome to another episode of Align Your Own Pockets. In this episode, we're actually taking a conversation that me and Dave was having offline and just one of those, Hey, we should turn on the cameras and the microphone and talk to everybody about this because I proposed a question to to Dave, and it'll be interesting to talk through the solution. So what I do with my own trading and because it's swing trading and investing and all of that, I know all of the entry prices and what I'm going to trade well in advance. Usually on the weekend, I run the systems, it gives a whole bunch of prices for the next week.

Michael: 00:37

One thing that I've always kind of been struggling with is that because I'm a human, I only have so much buying power. So say, for example, I'm looking to take five entries next week, but the system says here is 30 symbols to trade. There's a couple different ways that you can go about that. And one of them that I've always just done is try to find some sorting mechanism for these are the trades that are interesting, and these are the trades that aren't. But I think Dave's got some interesting ideas as well.

Michael: 01:06

So I guess the main topic is what happens if your system outputs way more, even any more symbols than you have buying power for? What do you do? How do you sort it? So let's get into that, Dave.

Dave: 01:20

Yeah. So I think this is it's related to another topic that I don't want to go into completely now, and that's just running out of buying power in general. This is more like you're trading a strategy, It's going well. All of a sudden, you have way more signals than you ever thought you would. And I've the the the the first time this happened to me, I was like, oh my gosh.

Dave: 01:50

This is crazy. So here here's let me set the stage. This was probably 02/2012, somewhere around then. I was trading this gap strategy. I still trade a version of this strategy today.

Dave: 02:03

So anything above a certain gap level would go on my radar. I would whittle it down with some rules that I had, and then I would look for a pattern and take trades. So it was usually it was usually in the direction of the gap. Not always, but had systems around that. So if you think about it, there's a certain, you know, a a modest number of trades that would meet that criteria or modest number of candidates that would meet that gap criteria every day.

Dave: 02:40

So, you know, I traded this for years, And it was one particular day where the market gapped up big time. So all of a sudden, instead of, I don't know, 20 candidates I would have that I would be looking for this pattern for and trade. Hundreds. There was like 200, two hundred 50, maybe more. I don't remember how many it was, but it was like I was just completely overwhelmed.

Dave: 03:08

Mhmm. And I was a little bit deer in the headlights, Scott, just because I thought I had no idea this something like this could ever happen. I'm totally unprepared. I was trying as hard as I could to, for my tools to run fast enough to keep up with all the data I needed to ingest to be able to make decisions about it. So do you remember, you know, the I Love Lucy show, the old show I Love Lucy?

Dave: 03:41

Yep. And there's this famous clip where she's in this chocolate factory in this conveyor belt, and the chocolate's coming out at a certain rate, everything's fine. Then all of a sudden, like it increases, increases. All of a sudden, she's just completely overwhelmed. So that's exactly how I felt.

Dave: 03:58

So I realized in an instant I needed to do something about that. And it's not exactly the same thing as you described, but it's real close. Yeah. So the first thing I did was I realized that needed to have some other criteria to be able to whittle these trades down in this situation, in these special situations. Like, normally, like, there was no reason reason to change my rules on an average day, but I needed to have a plan for these big days.

Dave: 04:35

Because if you think about it, those are, like those days could make or break your month, your year even.

Michael: 04:43

So I was about to say if the market was was moving like that and there's some sort of macro event and you have all these gap ups and they just start ripping off the open, then yeah, you're not only ish, you know, it's not just missing stuff. I know people to say, Oh, well, you know, you just miss a couple of trades. But the problem is just the way that life works, if it's if you're randomly taking whatever trades you can you can kind of find because everything's moving so fast, those will be the worst. And then the ones all the ones that you missed would have been, you know, a month maker or even more. Yeah.

Michael: 05:15

So it is a it's a big problem to tackle, one that I don't think will kind of hit people right away. But it really is if you're not if you're, you know, you got 50 different things to watch and you're not being able to pick which ones of those are most important, just through random luck, you're probably gonna pick the worst ones, and then the best ones are gonna are gonna work or go the Yeah. Without you, right?

Dave: 05:42

And, yeah, no matter, even if you pick some good ones, or, you know, like you said, you're gonna pick the worst ones. It it's gonna maximize frustration for sure that day. Yes. Because you're gonna look back and realize, oh my gosh. Why didn't I see that one?

Dave: 05:57

Why did I see this one that turned out poorly when these other ones are right there? How do I find those? And it's, it's just really important to have a plan for these days. And you're, you're right. Like it may be months or years before you experience one of these days.

Dave: 06:13

So you may think, well, not really a big deal. It only happens every so often, but these are huge difference makers. So it's good to have a plan for them. Also, it kind of puts you in the right mindset for improving your strategy in general. So, you know, we talked to the, I think, it was last week's episode, trade your best trades with bigger size, how to do that.

Dave: 06:37

It's kinda the same. This is sort of the same concept. So so here here's how I changed after that day. There's there was just so much to learn that day, like, just about my the tools I was using, how I can make them more efficient, how I could weed out stuff so that I wasn't looking at all these signals that I had to process through? Like, how can I make my entire process more efficient?

Dave: 07:07

And it starts with the back test. So you want your back test to show you when these days occur. A lot of backtest software will try to you know, they have, like, portfolio stuff in there where they try to be smart about you can specify your account size and it, you know, it will reject trades when you get to a certain buying power based on that. The first thing I do when I use any backtesting software is turn that off and put like infinite buying power. Because if you have that in place, you're not gonna see these days.

Dave: 07:41

Like by design, you're not gonna see these days. So you're gonna be really flying blind on these days.

Michael: 07:46

Yeah. And just that's a that's a kind of tip that I've gotten from you that I've started to use recently, right? And that's actually one of the things that that caused this question. So we'll get to that, I'm sure a little bit later. But yeah, you know, you go into for me real test, you set the maximum positions you can hold at like 10,000, you give yourself, you know, a trillion dollars worth of buying power, and then you hit go, and then you know that everything that you can trade will will get in there.

Michael: 08:13

And then it does do a very good job, especially when you're doing things like optimizing, but it then creates other problems. So then, again, that's what we'll get into down the road here.

Dave: 08:23

Yeah. So, yeah, like I said, it's it's pretty I I get where they're coming from to put that in and make it the default. I mean, I get why they're doing that. I just think the first thing I like, the first thing I do is turn it off because there's so many problems with doing it anyway, and it hides a lot of details that you want to be able to see. And sometimes it'll be, sometimes it'll be a long time before traders even realize that the portfolio rejections are even happening.

Dave: 09:01

Like, it's kinda some sometimes it's hard to see that in the results. You're you're seeing some things that are strange. You're missing some trades, but it's not clear why. You have to go in and debug, and then you're like, why? This one matched my rules.

Dave: 09:16

Why didn't it trade? Well, it's this portfolio thing that they're trying to be smart about. So, yeah, like I said, first thing you turn off.

Michael: 09:25

I do have a funny story about that too, because when I first started to get in deeper into systematic and testing and everything, I fell into the same trap. But what real test does by default is if you don't give it a sorting mechanism, it sorts it by alphabet. And the so part of the problem, I was doing these tests and some of these trend following tests were coming out amazing. And then I was going through and looking at the trades, which is again why it's important, right, take go through every single trade, well not most of the trades, and go through. And I'm like, Oh, this got Amazon, and it got Apple for these amazing trend following moves.

Michael: 10:03

And then, you know, it's just one of those that like hits you in the head and you go, Oh, you know, I of course it did because it's just sorting it. I noticed that there was never a point in where a ticker with the letter c ended up getting in there. And yeah, it just hit me in the head and it's like, of course, these long term trend following strategies look great. Cause it just accidentally got some of

Dave: 10:25

the best movers over the past twenty five years in them. Yeah. And the other problem with this is when you go to optimize your strategy, you're not looking at the full picture. You're looking at a sample of the trades in your strategy, and you don't realize it. You think it's a comprehensive list, but it's not because of these the software is trying to be smart and filter these things out.

Dave: 10:50

So it's it has all these implications that you might not be aware of. So that first thing to do is turn that off and, give yourself infinite buying power just so you can see the full that's one of the beauties about trading and backtesting is the data is plentiful and easily accessible. So you can actually in data analytics, one of the big problems is sometimes you have to get a sample of the set. You can't get a comprehensive list of everything that could happen. Where in trading, you could it's it's easy to do that.

Dave: 11:32

You don't have to look at just a sample of the thing. You can look at the entire thing. So it's really powerful to be able to do. Mhmm. And that's what makes trading unique and fun or one of the many reasons.

Dave: 11:45

But alright. So back to the back test. Yeah. I think that's it's a little it is a little bit counterintuitive some traders because they see that setting and they're like, okay. I'll just put my account size in there, and this will be even a better simulation, what I'm doing.

Dave: 12:02

But the very first stage of a backtest, I turn it off and take every trade. Even though you know you're not gonna take all those trades actually in real life.

Michael: 12:14

Mhmm.

Dave: 12:14

To start off with a backtest, you need to have the entire set. So what I'll do is I'll have the entire set of trades that match, and my plan is always filter it down using the optimization techniques that I use. And what I end up with is something that trades maybe 10%, twenty five %, some small portion of those trades that are in the original backtest. And that's the that's the right way to do it and the best way to do it. But it does it's it's a little counterintuitive for some traders when they start.

Michael: 12:50

Yeah. And, you know, that's how I'm doing it as well. The the the kind of what we're going for here is then the distribution of trades becomes an issue. So say you start with 100,000 trades and you backtest. Just to use extreme examples, you're all the way down to 1,000 trades over your backtest windows from 100,000.

Michael: 13:09

What if, again, an extreme example, 500 of those happened on a single day? And then the rest of them are spread throughout. And it's, you know, credit to you building the trade ideas back tester, because that's something that, you know, when I was helping clients, always told them to take a look because it would show the distribution of trades per day and say, hey, you know, if they all happen on one day, you probably just did something that lined up with the market on that one day. You know, it's not a true representation of what you're going for. But, you know, that's an extreme example.

Michael: 13:40

But in the case of a lot of the things that I'm trying to do in the systems I'm trying to build, it will be more clustered like that. And it's it's not necessarily a bad thing. It's just the way the distribution of trades comes out. So it then creates this problem of, Okay, right now what? So I have periods of time where there's no trades.

Michael: 14:02

That's fine. And then I have periods of time where there are 2,030 positions. And especially, you know, someone who's who's swing trading being in 30 trades at once or 40 trades at once doesn't make a lot of sense because when you're doing that for an extended period of time, you're just you're getting market returns at that point. You might as well just, you know, if you want, if I'm in 40 trades at once, I might as well just bought the SPY, right? Because if the SPY goes up, I'll be successful, if it doesn't.

Michael: 14:28

So not a unique problem for swing traders, but again, just a different problem when it comes to that distribution and then how to manage. I've probably already owned some things. I only have buying power for X amount left trying to figure out which of those to take and how to filter that down. So the distribution's a little bit more even.

Dave: 14:49

Yeah. Yeah. So when you start optimizing stuff, you'll notice that, like you said, some some rules will create situations exactly like that where everything happened on a single day. I have a in my reporting and and tools that I use for backtesting, one of the reports I have is trades by day. So it's a report that shows, like, a column chart that shows all the days that the trade, you know, across the entire backtest and just a column chart for how many trades occurred on any given day.

Dave: 15:26

I always look at that because when I'm applying a rule before I do that, I get more I get more and more discerning about applying rules to strategies. Like, if you apply a rule to a strategy, it's an important thing. You need to look at a lot of things. It's it's a big deal to apply a rule. Yeah.

Dave: 15:41

And one of the things I look at is, does this rule cause all the signals to bunch up on just a few days?

Michael: 15:51

Mhmm.

Dave: 15:51

If so, then I'm gonna discount that. I'm gonna take that into account, and I'm not gonna choose that rule. I want I want rules that sort of spread things out. And I value having more trades on you know, spread out across the days relatively evenly. You're not gonna be able to get it exact, of course, but I wanna penalize the ones that bunch up trades on on specific days.

Michael: 16:18

Yeah. And then but at, you know, at there's gonna always become a time. Right? And Yes. This is kinda what we're what we're marching our way to.

Michael: 16:28

There's always a time where there's there's a binary choice. And I think it's easier to think from my example as opposed to an intraday one, simply because there's there's a frequency difference. Right. So for example, right, say you're using I'm going to buy things as they break down into the gap after a huge gap or short things as they break down the gap after a huge gap up. There's times where, yes, there might be a couple come through, but probably not the same second.

Michael: 16:59

So if you're using some sort of automation, there'll be some ordering that happens just based off time, right? Because on time one, you short something, because you don't know that on time two, there'll be something else that come in short. So that's that's happening there, but when I'm doing it again on, recording this on Friday, on Saturday, I'm gonna sit down and it's gonna spit out 50 some odd trades. And right now I'm doing a a sort mechanism myself just to say these are the ones that that kind of test the best. But this is where I I really think that there's a lot of edge.

Michael: 17:35

This is why, again, I asked Dave this question immediately, and then we said, hey, we should turn this into an episode, is because I think there's a lot of edge because you can take the same backtest that I've run and sort it by two completely different metrics and the returns are wildly different. Right? You know, it's not from a winning strategy to a losing strategy, but sometimes it's double the year over year growth just by choosing a different sorting metric. So that's what I'm trying to really figure out is how to efficiently and systematically be able to go through and say, the end of the day, when I get to the point that the trade level is is good and the distribution is okay, there has to be that binary decision. And then how is it best to kind of make that?

Dave: 18:17

Yeah. So I've got two different ideas for that. One is a was a really simple one I put in for the following day. You remember the day that 02/2012 where where everything gapped it up, and I was overwhelmed. I was like, well, it's probably gonna be similar tomorrow, I was thinking at the time.

Dave: 18:37

So I need to come up with a plan real fast to do something, you know, sort of quick and dirty. So the thing I did then was and this is something you should be able to put into place right away. So I looked at the SPY, and I looked at the relative volume for the SPY that day. And it was a certain amount. So it was a lot compared to what it normally does.

Dave: 19:02

Say it was 1.5 or two compared to what it normally does. So

Michael: 19:07

Just for those who don't know, just in case you're someone who hasn't come across relative volume, basically just meaning the spy that day did a 50 to double what what it did normal in normal volume. Right? Just Right. Just a way to compare the volume on day one versus an average of of generally what happens.

Dave: 19:26

Right. So I'm use that as a reference. And for my entire candidate set for the next day, I use that. So let's say it was 1.5. I would say, okay, any one of my candidates, it has to be relative volume higher than that for it to make my list.

Dave: 19:43

Otherwise, it's sort of underperforming the market in relative volume terms. So I I was still looking for stuff that relative to the market was very high relative volume. So I wanted it to be so I used that as a gauge to get rid of, like, half of the candidates right away. Mhmm. So that's one way to do it, and that's what I put in place immediately the following day.

Dave: 20:14

Once you step back, though, I think there's a better way to do it, but it takes a little bit of time that I didn't have quite have time to prepare for at that time. The better way to do it is well, there's actually two different ways that are are better. So one is pretty much what we talked about in last week's episode, which is come up with sort of a tighter set of criteria using your optimization process to sort of come up with a, like, set within the strategy of trades that you're highly confident with. So this is what I always do for lots of traders and for any strategy I trade. I get to the optimized set that I'm gonna trade day in and day out, but then I go a step further and have a couple more rules held out, sort of held in my back pocket for days like this that I can put in place that is gonna have fewer trades, but they're more they're more profitable.

Dave: 21:24

So I can I know that I have a quick way to determine if they're in the high confidence bucket, and then I can put that into play on on days like it? So it's already built in. The other thing that that helps you do is it gets you in the mindset of improving your strategy and sort of holding out these rules for big days like this is also very valuable for holding out rules. You're doing the same thing to prepare yourself for a potential drawdown and an adjustment you wanna make to your strategy in the event that happens. So it's kind of the same thing, and it's I think that's really valuable.

Dave: 22:14

Now there's another way. I want to hear your thoughts on

Michael: 22:17

that first.

Dave: 22:17

Yeah. There's another way. I got another one that I want to suggest after that.

Michael: 22:21

Yeah. So so kind of the way I heard it and what the idea would be is let's just use a mean reversion strategy that I have, right? Just looking for stuff that's overall strong and pulling back kind of really harshly now for for buys. Say, you know, I want 10 symbols to get output and there's 100 this time. So what you would is you would have somewhere kind of written down for this strategy.

Michael: 22:48

Hey, if you, you know, if this indicator is a little bit more oversold, there's some evidence that the backtest works better. And then if, you know, say, you know, the range is really high on the prior candle, there's more evidence that that works. And then you would start to re kind of feed those back in and continue to run the report of what it would output until you got to kind of roughly that that 10 number that you were looking for. So you're refining it the day of that you're running the test based off of obviously work that you've done before.

Dave: 23:23

So here so I do all this well ahead of time as part of my normal optimization process. So I get my set that I'm gonna trade in and day out, and then I have my high confidence set that I sort of hold in my back pocket. Now how do I implement that? Well, what you can do is you can set up a custom formula, do that in trade ideas. I've done this in Amber Broker also, just like another column that says, hey.

Dave: 23:51

Is this one in the high confidence bucket? So it matches my rules, but I've got another column that says, okay, this one's high confidence. Now maybe I don't use that except for on days like this. So I've got it there. I know that I've done a lot of research to be able to populate that in a very valuable way, and I can instantly use that as a filter on that particular day in the event that this happens.

Dave: 24:21

And I'm and I'm, like, overwhelmed.

Michael: 24:23

I like that because, yeah, you have your your strategy for normal conditions, and and that would work because I'm always just going back to kind of my workflow when I when I'm thinking through this. So it's not like I think any of these are better or worse, just how I'm going to implement it. I would have a quote unquote, like a base strategy that would, you know, allow in a lot of trades. And then I would have my my really tight strategy and on weeks where I'm running it and the base strategy is just overwhelmed, then I can go check with, as you've mentioned, that column that said, Okay, these are the ones that if if I were to put another filter on and then back test again, these are the only ones that probably would have come through and then use that as well. So that's interesting.

Michael: 25:06

Yeah, give me the other idea and then I'll talk about what I'm using and potentially if that makes sense to you as well.

Dave: 25:14

All right, so you might not have sophisticated enough optimization process to be able to come up with this other set. Right? There's sort of a quicker way to do it which I would look at, which would be you you probably have some sense of what you should be using for an additional filter. Like you said, you know, some extreme condition, extreme relative volume, unextreme relative volume, something like that.

Michael: 25:49

Mhmm.

Dave: 25:50

In your backtest, you can do just a simple correlation. One thing you could do is just correlate all the features that you're looking at, all the indicator values with profit, and just sort it by the correlation number. So it's not gonna be perfect. There's better ways to do an optimization than this, but that's gonna give you some sense of which ones to focus on. So maybe you come up with maybe maybe relative volume is the most correlated thing with profit in your backtest.

Dave: 26:24

Then you've got one filter that you can look at. It's easier to come up with than the, you know, the high confidence that that I mentioned before.

Michael: 26:34

Mhmm.

Dave: 26:34

And you can use that to determine you know, that's gonna be your go to when there's a day like this. You could say, okay. I'm gonna sort it by relative volume in days like this, and I'm just gonna take the top ones and just go down the list from the top. Yeah. So that's

Michael: 26:53

more or less kind of what I'm doing currently, right, is RealTest has a feature that just ranks, creates a, you know, cross sectional, I think he calls ranking of every stock. And then, yeah, you try to. So the correlational chart is something that I I haven't thought about, and I think that's very interesting as well. And that's probably something that would be fairly easy to do in Excel. But then using that to to rank the symbols, that I think is is kind of the crux, and that's where I've always been pushing towards.

Michael: 27:27

So that might actually be the missing piece of the pie for me, because at the end of the day, what it would be nice to have is, okay, here is you can take 10 symbols at a time, and then you're going to decide what those 10 are just based off of relative volume was your example, right? The higher the relative volume we're going to go. So when I'm running them on the weekend, if there's 100, you're going take the top 10 based off relative volume and then move on. To me, that seems like, you know, to dive into simplicity, that seems like it's going to be the ultimate simplistic way to do it. So my question was always of the infinite things that you could rank a system on, you know, how to come up with it.

Michael: 28:06

You know, a lot of this is has to be done just logically, right? If you're looking for a mean reverting system, it's going to probably be the extent at which it's it's it's down or stretched. If it's a trend falling system, it's probably going be the strength of the thing over years. But yeah, it's really delving into what to rank it by. So, you know, when it comes to doing that correlation stuff, is that just something you just build something in Excel and you just pop the trades in and it runs based off the filter types or?

Dave: 28:33

Yeah. You could do it in Excel pretty easily. The tools I have, you can do it. You know, it just looks at all of them. Mean, there's hundreds of filters I have, and it just just looks at all of them and ranks them by or correlation, so it takes very little time.

Dave: 28:51

Yeah. So but there's ways to do it, you can certainly do it in Excel, and that's a great way to start. So it reminds me as as you're talking, this reminds me of another approach. So, really, what you're talking about here is some way that you can be confident in ranking the trades, just like you said.

Michael: 29:14

Yeah.

Dave: 29:14

And it's gonna be hard to come up with one particular feature to rank it on. It's probably gonna be a combination of multiple features, multiple indicators. Right? So Okay. How do you deal with that?

Dave: 29:31

Well, you use some of the tools that I've created to do that. That's exactly I mean, that's the whole name of the game. Mhmm. Another way to do that, and this is what I've done with with several strategies, is I've used a a data analysis tool called Random Forest. Okay.

Dave: 29:51

And it basically takes a you feed it you you you tell it the features to look at, the indicator values.

Michael: 30:02

Mhmm.

Dave: 30:02

You say, okay. Look at these. Here are the ones that I'm giving you permission to look at. Create a model based on those, based on historical data. And then going forward, take a look at those same that same list of features for future trades and come up with a score.

Dave: 30:25

And the score is you know, you can make that one to a hundred. And that what that allows you to do is is instead of just looking at a single variable, you can look at multiple variables and distill that down to a single rank variable that you can be really highly confident in based on this optimization process using Random Forest. And then that's really very, very valuable because it's not think about what I was saying before where you got a high confidence column that's either one or zero. Now you've got a rank column that is between one and a hundred. That's a lot more valuable than this binary one or zero.

Dave: 31:08

There's a lot you can do with that. And if that if if you can prove to yourself that that's correlated with profit, that's super valuable. Because on days like this, you could say what that allows you to do is on normal days, you could say, okay, I'm only gonna take take trades with rank above 30, say, some some threshold. But then on the busy days, you could say, okay, I'm only gonna take trades higher than 90. And you can know that that is the, like, the best statistical item or data point to be looking at.

Dave: 31:43

So you can really have high confidence that you're gonna be doing the right thing on these days.

Michael: 31:49

Yeah. And that that to me is a little outside my pay grade, but I'm sure I could look it up and figure it out. But that to me is very interesting because, at the end of the day, you're trying to balance frequency of occurrence, but then also, right, basically, you don't want to miss the best trades because you're you're being really tight about it. But at the same time, you just there's buying power. And, you know, even if you're not fully automated with your trading, there's restrictions there as well.

Michael: 32:17

Right. So, you know, we mentioned my kind of diving and exploring in the day trading. Well, I don't currently with the system I'm using have a way to automate the trading. So this is part of this as well, where, you know, at the end of the day, there are of the setups that I'm looking for, there could be hundreds. So it's that that ranking and sorting mechanism ends up being being super interested.

Michael: 32:39

And I think this is an interesting thing to talk about too with the audience. And as we're kind of, you know, spitballing and hammering it back and forth here is because I bet you it's something that will get a couple of comments of people didn't even really think of. And like we mentioned before, if you're wouldn't it be unfortunate if, you know, as a trend follower, for example, you're sorting by the worst stocks that are down the most over the last year. I guarantee you if you take that and you flip it on its head and say, wanted the best stocks over the last year, you could take a system and, you know, double or triple the returns and lower the drawdown and everything just because what it's leaving out is it's leaving currently, you're having it leave out what could be the best trade. So it's a very interesting and, you know, this is one, I like this one because it's super nerdy, but this is the kind of thing for people who are trying to get into systematic trading that you end up taking your time thinking about.

Michael: 33:32

You know, we talk about this all the time where what you've done when you become systematic trading is you've eliminated the time that you're staring at charts and waiting for something to happen, hitting a button when you're buying, right? Hopefully, you automate all that stuff out of the way. But you're not done. You're just replacing it with, you know, questions like this that comes up is, am I using a suboptimal sorting method for a strategy? And if you like that kind of stuff, then this is the the game for you because I've been thinking about this for weeks now.

Dave: 34:03

Yeah. So it also is very, very valuable. Once you start coming up with these theories about how to sort things in your strategy like we're talking about here, you can always go back to previous days in your backtest. So remember the the report I suggested where you're looking at number of trades by day. Go back, choose one of those days, and see exactly when the trades came in and sort of simulate in your mind, maybe there's some automation you could do, to simulate what would have happened, what it would have felt like on that day given the signals that came in, given this theory you've come up with about you know, relative volume example and really piece through imagining what would have happened that day.

Dave: 34:53

It's really that that day that this happened to me in 02/2012, like, I had to drop everything about what I was doing. Like, no work, no anything. I had to get to the answer here. Book and partly it's because it's so fresh in your mind. It's easy to forget some of this stuff.

Dave: 35:07

So it's really good to like, I write a whole bunch of log files with my software because I want to be able to piece back together exactly the order of events that happened so I can get better at this kind of day. That you want to do, you want to be in a position to do this quickly when it happens because you want everything that you were all that frustration that was going on in your mind, the the the trades that you didn't take, that you remember, that you're pissed off about, that's motivation for getting to the answer. And you're you're there's never gonna be a moment where you're more motivated and you have more information right in your mind ready because it's fresh in your mind from that day where you can improve. So it's it's good to it's good to seize the moment when this happens.

Michael: 35:54

Yeah. And I I think that that's I think that's something to to talk about a bit because I think that's something that people sometimes don't take it seriously. They'll look at their P and L at the end of the day, and they'll say, okay, that was good. I made money here or whatever it is. But looking at the trades that you have missed, think it's super important that you were talking out because, you know, if you want to motivate yourself to do better, even, you know, you sit down, you make a hundred bucks that day, and you say, that's good.

Michael: 36:24

Let's get a little automated. You know, I was at work, came back, made a hundred bucks on all my systems. But go through and look at the trades that were missed, or if you had expanded out some filters, some trades that you wouldn't have got. Because if you're going through that and you say, I could have made a thousand bucks this day with the same system and the same buying power if I had just sorted my trades a little differently or done something. I guarantee you that as mean as it sounds, little high you had of, I made $100 good for me is kind of completely destroyed when you're like, oh, I could have made made that much.

Michael: 36:56

But I think it's in a good way because it's gonna motivate you to say, okay, well, how do I catch more of those trades that I miss as opposed to the trades that I ended up getting? And that kind of kick in the butt, I think it really helps, even though, you know, celebrate your little win, but, you know, understand that your goal is to try to get more. And that's that's one of the things that spurred on this conversation is as I was going through some of these testings, like you mentioned, I have a base strategy, which is completely opened up. And, you know, there's thousands of like 100,000 trades over thirty years that comes through there. And then every now and then, roughly once a month, I go back to that base strategy, and I go through, okay, what were all the trades that that would have taken this week versus my optimized strategy?

Michael: 37:40

What did it take? And there was two, three weeks in a row where the optimised strategy just missed so many winners. And so then I went back and I said, Okay, well, what should I filter by? And again, it created this this problem of I don't want to just constantly be chasing the dragon. And, you know, going back to the ones that I've missed, but I want to make sure that I'm doing this this filtering and this sorting in the best possible ways to get to the most amount of winners.

Dave: 38:05

Yeah. I think that's that's a great way to think about it, and, yeah, you should have a process for going back. And you hit the nail right on the head. You should be looking at trades that are just outside your rule set because there's a lot of there's just for what you said, like, there's winners there that you could be adding to your strategy. And I it's you've heard me say it so many times, and that's what I love about trading is there's there's small things you can do that make an enormous difference in your results.

Dave: 38:37

And they're they're just laying around everywhere. If you look if you know how to look for them, they're everywhere.

Michael: 38:43

Well, and this is it's why I'm attacking this problem so hard because I just I feel with a lot of these that, you know, they're they're optimized to a point. They're taking trades that I like. They're, you know, they're taking trades that I'm looking for, and and everything is good there. But again, as part of this process of going back to the base strategy and looking what that's did, it's it's like kinda opened my eyes to, okay, it's doing well. I'm I'm happy with it.

Michael: 39:08

I don't wanna, you know, do anything to ruin it, but it missed, you know, these five trades that would have been amazing. Then it had done so because they were filtered out. So I actually went to the report of the optimized strategy and I ran it. And right now I'm taking the top 10 based off a ranking condition. And then just a couple down, I saw those other strategies or those other trades.

Michael: 39:30

So if I had just taken that same basket of trades and ranked it in slightly different order, the returns of these things, like one of them doubled in in a week. On. So it's like one of those. And that's a that's a month making difference of a amazing month versus an okay month, getting that one or two. So yeah, it's definitely a a good problem to go down.

Michael: 39:52

Like with most things in trading, there will not be a definitive solution, and I know that, but it's trying to get as close to optimal as you possibly can dig and make sure you're getting those those winners because that's the difference between, you know, buying the wife a hot tub or for her birthday or or, you know, going out to McDonald's, right, that that kind of one trait. So you really have to

Dave: 40:13

look at it like that. Yeah. Yeah. There's I remember looking back when I first started getting really motivated about this. I remember going back and looking at my results very early on when I was trading this system manually.

Dave: 40:29

And I remember thinking, gosh. If I could just get rid of one losing trade a month, That would be significant in my results, actually. Like, it would be worth and I was thinking, gosh. You think I could get rid of just one trade a month? Seems reasonable.

Dave: 40:48

I was taking a whole bunch of trades. Seemed like a reasonable thing to do, and it seemed like a bar that I could meet and I could just adding that up over time, like, it's a it's a there's a lot of results there. So like I said, that that's what ever since then, that's that's what motivates me to keep going. I mean, this is so fun, and the bar is just so low for improvement. It's great.

Michael: 41:17

Well, yeah, and I will say it it gets way more fun if you're going down this road and you're you're you're frustrated at the time. It gets way more fun when you've got some amount of of base of profitable strategy. Right? That think that's the biggest, like, no fun hurdle, right? You go from a losing trader to, okay, I've got one or two strategies.

Michael: 41:37

They're doing okay. I'm making a few bucks. And that's to me that like the mind opening moment because you're saying, okay, there's two things to do is to try to optimize those strategies a little bit and then to bring in more and more. But at least you're starting at a nice positive base that you're trying to build upon as opposed to going from absolutely nothing to something. Now it's from something to potentially something better.

Michael: 41:59

And that to me is where, again, this this, like you said, just gets a lot more fun, because I'm constantly thinking about, you know, you're making money and you're saying, Okay, well, do I I look at it as kind of how do you give yourself a raise? And what are the things that you can do to give you know, you're working at a job, there's nothing you can do, either your boss likes you or he doesn't. But when you're working here, you're saying, Okay, well, I'm making this and you know, that's my base amount, and that's everything's fine. What if I just tweak this? Does that increase my income by 5%, ten %?

Michael: 42:30

And like you mentioned, we're pushing on the edges here, which is, again, totally fine. But yeah, we're pushing on the edges to try to raise things a little bit, but you know, you compound those out over time, you know, little tiny edges here, there over time become huge, hugely different things.

Dave: 42:49

Yeah, mean, is how you get better as a trader. You've heard me say it before, the best way to come up with an awesome strategy is to start with an okay one and start improving it. So that's that's the way to do it.

Michael: 43:03

So, yeah, I think that was a that was a great one. It was a long one because I think we're we're passionate about the topic.

Dave: 43:08

And, you know, I could talk for four more hours about this. I mean, I've already I'm thinking of, three or four more episodes we could do about this.

Michael: 43:17

I know. That's I love it. And the thing I like about it, and if this is you, you know, post a comment or whatever where you haven't even thought of this problem yet, because I think there's a lot of people that are kind of on the journey, and they haven't gotten to that point, or they might not even realize that just by a different sorting mechanism, a different filtering for the best trades, they couldn't improve results. So I think the takeaway for the audience is just that, right? Make sure that you're you're going through your trading and seeing are you if there's 20 trades to take, how are you deciding which of those those 20 trades to take if you can't take them all?

Michael: 43:53

Because that could be the difference between an amazing winning strategy or just an okay strategy.

Dave: 43:59

For sure. There's lots of I know we'll do an episode on buying power, running out of buying power because that is a very fun problem to have.

Michael: 44:09

And yes. Yeah.

Dave: 44:10

That's there's a lot to talk about there.

Michael: 44:14

Yeah. We'll we'll we'll save that to the next one just just to make sure. But that was great. Again, I I I thank you, Dave, for your help because this was kind of a I have a question for Dave and let's just do it on air. So if you like that kind of, you know, really in the weeds nerd talk, certainly let us know because like you said, we could we could be going for hours here if if we didn't, you know, if we didn't have other things to do, we could probably hear till tomorrow talking about different ways to do it.

Michael: 44:40

So if you like that kind of stuff, definitely, you know, hit us up. We love the comments, we love the feedback, we love the questions that you guys give. And this has been great. Now I have a whole bunch of work to do on the weekend because I've got all these ideas. But until next time, I'm Michael Noss.

Dave: 44:58

And I'm Dave Mabe. Thanks for joining this online, your own pockets.