Line Your Own Pockets

Michael: 00:00

All right, everyone, welcome to another episode of Line Your Own Pockets. In this one, we're going to talk about our column library today and some ways that you can make additions to it, a way that I kind of discovered a new indicator that I want to make additions to it. And we're going to give some columns that we think potentially people could look at. And these are ways to help filter and optimize what you're doing currently to hopefully make a strategy just a little bit better. So how are doing today?

Dave: 00:28

I'm doing great. Yeah. This is good. I love this topic because even more than the strategies themselves, the column library is if there's any secret sauce, this is it. And just to go over what the column library is and why I call it that, I come from I'm I'm a software developer.

Dave: 00:47

Been developing software for a long time. And in software development, the way to scale yourself is to create libraries. So you got a bunch of different software projects that you might be working on. When you first start developing, you you do with the common thing, which is copy and paste some of the common things and functions that are used throughout the different projects. Mhmm.

Dave: 01:09

But you quickly realize that when you find a bug, all of a sudden, you have to make changes in all of those different projects. The the thing to do there is create a library that's shared amongst all the product other projects. So then if there's a bug, you change it in one place, all your projects update automatically. It's just a way to leverage what you're doing, what you're coding, and it just makes you a lot better. And it helps you think about it helps you think in, like, more strategically about your code.

Dave: 01:41

What sort of functions would make sense to leverage it and and have it work across all projects? And so with the column library, it's the same thing, but for strata trading strategies.

Michael: 01:54

Yeah. That's what I wanted to bring it back to. So what what we're talking about is that if you have indicators or call them whatever you want, right, indicators or candlestick setups or whatever, and you're telling your back tester to record that on entry. So what is the value of, you know, the MACD or what how far are you from the two hundred day moving average or whatever? And as soon as you record all of these data points from the moment that you would have entered the trade, now you have this kind of big, robust data set that you can go back and you can say, you know, start asking questions of the data and, hey, is it better if this thing's occurring or that thing's occurring?

Michael: 02:31

And we do have, you know, a whole episode about this, so you can go back and look, and we will remember to put it in the show notes here so you guys can can go take a look at it. But today, we want to talk about kind of adding to that. Right? So, you know, this is something that I've picked up from Dave, and I went through, and I did the work, and I put everything that I thought I would need in there. But I think just like having a process for increasing and getting new strategies and adding those new strategies to your trading.

Michael: 03:02

It's just as important as having a process to add or to try to open yourself up to new indicators or columns that you could add into this library for kind of what we're gonna talk about, which is, you know, the ability to expand out what you're doing quite a bit.

Dave: 03:18

Yeah. And every time you add a new column to your library, just think of the power it has. I mean, you can when you find something that's predictive in one strategy, the first thing you should do is go back and add that column to all your strategies that you created before that. Oftentimes, you'll find it's predictive there also. And when you find that, that's just extremely powerful because you can go back and improve that strategy.

Dave: 03:46

You could in a lot of different ways, you could use it to exclude the worst trades. But really the more powerful way to do it is you could could start with a different starting point potentially and start with something that has more trades with more average profit per trade and optimize from there. You end up with something that is just more profitable. So it's never ending process, and you're constantly improving your comm library, and it just gets more and more and more powerful over time.

Michael: 04:16

Yeah. So where we're where we're starting from this is that, you know, one really nerdy thing that I do is the CMT every year puts out the Charles Dow award. Right? It's just an award that they run. Anyone can apply for it.

Michael: 04:29

And the idea is that people are coming up with new trading or investing concepts or indicators or relationships or whatever it is, they write papers about them, they submit them, and they pick one. And one of them was a way to kind of normalize the MACD. And this is what spurred on this episode is because right now, putting like an absolute value of the MACD, for those who don't know, you can moving average convergence divergence, you can Google it, there's a whole bunch of resources there. But right now, it's kind of useless for this because the way it's set up, it doesn't compare across securities or through time. Right?

Michael: 05:10

You take two securities, they're gonna have completely separate values, and then they're pretty much irrelevant for it. It's one of those indicators that's useful only as a visual tool for people who believe in it kind of only. But this one article that this person did where he calls it the Mac DV, and it's just a normalized version of it, all of a sudden, he normalized it for ATR. So he just made it so that across time and across securities, all of the values are normalized. And then that kind of got me thinking, okay, well, now that this very old school, like built in like the sixties or traditional indicator is now been changed just a little bit, Well, now it's interesting for a column library because it has that comparable aspect.

Michael: 05:51

And just by, you know, doing some nerd stuff and reading these kind of papers from time to time, that light bulb moment went off, and then you go back to work. And then you gotta be like we talked about, put it into the column or put it into the library and just start testing strategies that you currently already have in this kind of new light that you wouldn't have seen unless you put yourself in a place to expose yourself to an opportunity of, oh, this is a new way to look at something that's actually quite old.

Dave: 06:21

Yeah. I love that idea of, normalizing the data. You're gonna end up with something that's way more useful across your strategies and across different securities. I think it's a great idea. So it reminds me of one of the columns that I added that is useful.

Dave: 06:42

Like, sometimes you don't even realize that something's not normalized. Like, for example, a percent gap, like today's gap in a stock, that seems like, okay. That's pretty normalized. But when you after you think about it for a bit, you could actually use the gap in terms of ATR, which is a way more accurate and more normalized thing. So something you think might be normalized, once you mull over it and think about it some more, you realize, oh, actually, it could be even more normalized.

Dave: 07:18

So I really love that.

Michael: 07:20

Well, and that's I think that's a perfect example. We we can hit on that a little bit. Like, a lot of people, yeah, they just look at it and they say, okay. Well, you know, I I wanna look at the biggest percent gainers or the biggest, you know, percent gappers or something like that. And that's just standard.

Michael: 07:34

You you pull, like, 20 traders, and they'll all say the same thing. But the example I always use to talk about why you need to normalize data is a $1 stock that gaps 10 percent, 10¢. Right? It could be a big deal. It could just be a misprint.

Michael: 07:50

It could be any number of things. If Apple ever gaps up or down 10% in a day, they're writing about it in the papers and, you know, it's all over the news and because Apple's a multi $100 stock, and it's much kind of slower. So that's why I like you take these things and you normalize them where you say, okay, how much does this stock kind of normally move in a day? Well, if it's gapped way more than it normally moves, well, that's something that's potentially interesting, and you can go and just by looking at the you're you're essentially comparing the stock to itself as opposed to comparing the stock to something that's a little bit more arbitrary. And you could do the same thing as, oh, I want stocks that gap up a dollar.

Michael: 08:29

Like, again, on that dollar stock, that's a big deal. It's a 100% gap. On Apple, that's a normal day. You know, comparing each stock, and I think this is where a lot of traders start to gravitate to things like futures and and things like that because it it helps them they don't have to worry about this process because they're looking at the same thing every day, but they miss a lot of edge and opportunity because they can't normalize these things.

Dave: 08:53

That's a great point, Michael. You don't even really have to normalize if you're just looking at futures. So so that that's you know, I've never really thought about that. That's totally true. Somebody's coming from futures.

Dave: 09:04

They're not used to having to do this. Like, why would they be? So it's so interesting.

Michael: 09:09

It's funny. You hear from old school and discretionary traders, and they say things like, you know, each stock has, like, its own personality. And you you'll hear things like that, and it's it's not untrue. But what they're just saying is that some stocks are more volatile, some stocks are less volatile. Right?

Michael: 09:25

Some stocks, you know, probably more in the news, so they gap more and whatever. And this is why you'll even see individual traders float towards you'll hear the guy say, oh, you know, I trade Apple and Nvidia, and that, you know, he's got, like, five things that he likes to trade. And I would say it's not that he cares about the stocks at all. He's just coincidentally found stocks that fit like an ATR universe and have fit like an average volume and liquidity universe that he likes. And, yeah, I think that's a huge reason why people trade things like futures.

Michael: 09:54

Because if I'm looking at this one thing every day, it's not that I learn how it or I I feel the personality or the vibes of whatever the the instrument is. It's just that I just know roughly how much it moves based off of how much it normally moves.

Dave: 10:12

Yeah. Totally true. Alright. So I'm guessing that there's listeners now who, in the back of their mind, they have this question. So I'm gonna be the devil's advocate to you and let you

Michael: 10:22

Sure.

Dave: 10:22

Knock this out the park. You ready? Alright.

Michael: 10:25

Make it easy on me.

Dave: 10:26

It's Alright. So, Michael, I know that all these indicators are kind of bullcrap. Right? Like, I know that you're supposed to have as few indicators as possible on the chart. I know the beginning traders add a whole bunch of indicators.

Dave: 10:40

It's all they're all worthless. You're talking about adding these to backtest if they're worthless on the chart. Why in the world like, why are you even thinking about this? Why are they valuable at all?

Michael: 10:50

The one that always gets me is they always say, well, indicators, they're useless because they're lagging. And I actually saw my buddy JC, who's the CEO of All Star Charts. He actually answered this on stage in front of, like, 50 people. And it was about technical analysis or indicator and technical analysis. These are all crap because they're all just lagging based off price.

Michael: 11:11

He just looked the guy dead in the eyes. Like, listen. If you have any future information, you let me know. He's like, everything we're doing is based off of past information. I hear that argument all the time.

Michael: 11:20

It's like, oh, you should just use price action and not indicators. And I think the simplest answer to that is that indicators are just ways to approximate price action, and there are ways to be able to crunch large amounts of data in price action. Someone says, oh, I only buy stocks and uptrends. Great. What do you mean?

Michael: 11:37

Like, what what is that specifically that criteria? And without having to go and and talk about higher highs and higher lows on a on a chart or anything like that, it's much simpler to say, okay, above moving average. Right? If because it's mathematically, if it's above moving average, it's been moving up for some period of time. So, you know, is there one indicator that's amazing and you you buy Lambos and and yachts with it?

Michael: 12:01

Absolutely not. But these are tools in a tool belt, and they they approximate certain things in price action. So it's important that you, a, use them, and, b, you understand what they're telling you. Right? And that's where I think people fall apart is they're looking for the magic this line crosses this line, and and I make a million dollars.

Dave: 12:19

Yeah. Well, you know, even price is lagging. Right? Mhmm. Even in real time, as soon as the price happens, that's by definition, it's in the past.

Michael: 12:29

I Well, let's I've And let's hit on that. That's not only so that's just us us technical guys. Imagine the fundamental dudes that are getting they're caring about last quarter's earnings. Yeah. So a company will come out and say, this is what we made three months ago.

Michael: 12:45

Here you go. It's you wanna talk about lagging or, like, here is the unemployment numbers that occurred six months ago that's gonna get revised seven times between now and then. Yeah. It goes back to that that quote JC is like, if you got any any future data, let me know. But everything we're doing is relying on the past.

Michael: 13:02

We're just looking for things that tell us certain things about the data that we can we can play with. So if you're gonna harp on on technical people, I don't ever want to see you quote anything about fundamental analysis ever.

Dave: 13:16

So true. Yeah. So so you're you wanna tackle this Mac DV, you wanna put that in your com library.

Michael: 13:22

Mhmm.

Dave: 13:23

What's the plan? What are you how are you gonna do that?

Michael: 13:26

Well, the beauty of of the way I discovered a lot of these or the way I'm doing these with the Charles Dow award, because I'm on, like, a bit of personal mission where they've been doing this for, like, fifty years. I'm gonna I'm gonna read them all eventually. Mostly, I skim them. I put them into a shout out notebook LLM. I don't know if you've tried this one, Dave, but you can put any anything into this, and it turns it into a podcast like you're listening to right now with with a guy and a girl just just talking about it.

Michael: 13:54

And I do this all the time with stuff that I would fall asleep reading. I skim it myself as quickly as I can. You upload the paper to this, and then I take the dogs out in the woods. And I just listen to this completely made up. Definitely can tell it's AI.

Michael: 14:07

No personality. No no soul behind it at all. But they're explaining all of the the aspects of this. And because it's open, because he just is easily explaining what he's doing, it's very easy to go into a real test or an amnibroker or whatever and just code up the differences to these indicators. And then I can add them to the column library and then start my tests after that.

Dave: 14:28

So do you have a sense of this from the strategies you're trading now and that you've back tested? Do you have a sense of where this one is likely to be valuable or where it's gonna have an impact?

Michael: 14:40

Yeah, I'm gonna look at all of them, but just from, you know, reading the paper and then and then listening to it, I can very easily understand what the indicator is trying to do. So it's like we're talking about with the the price approximation. This one is now normalized, so it can do things like mean reversion and overbought and oversold. And I have some mean reversion strategies that are using similar indicators currently. So there's kind of two jobs I wanna do.

Michael: 15:06

One is I just wanna replace the indicator I'm using with this new one and just test. In theory, it should be similar, maybe a little better, maybe a little worse. But like you talked about, if it brings in more trades and it's somewhat similar results, then that's good. And then, yeah, put it into the library to use in the cruncher or some of the own kinda analytics that I do myself.

Dave: 15:26

Yeah. So you'll so you'll run it through the cruncher, and I and some of them, I've I'm assuming MACD is sort of important so you can go back and look and and just get an apples to apples comparison. You could see which one you can see literally if it improves it and and by how much just by looking at the report and the cruncher. Right?

Michael: 15:44

Well, yeah, and then whether or not it shows up at all. So what I'll be interested to see is, you know, maybe because I wasn't able to put MACD in before, but I was putting relationships with two moving averages because it's more relevant. If that never showed up when I run my tests and then I put this new one in and it shows up, then that's information of itself because it means, you know, the MACD, which is inherently a trend following indicator as a measure of of strength of a trend, I can I can take a look at that and say, okay? Well, this measure of strength that I tried to approximate before didn't work, but now it's in here. And like we talk about with everything, that opens up a whole bunch of doors to why.

Michael: 16:22

Right? And then what else potentially could I build in to to confirm whether or not this trend is important or not?

Dave: 16:28

Yeah. So the so the traders that use MADE Kit and the Strategy Cruncher, it's it's kind of a typical path. So when they start using it, they'll, you know, run a back test. They've got all the columns that are included in MADE Kit, and they run it. They put it through the cruncher, and all of sudden, they see 10 ways that they never would have thought about of ways to improve their strategy.

Dave: 16:53

Maybe they lop off 20% of the trades to really and then the remaining 80% of the trades are just way better than the entire set. So that's how you the the typical improvement. But, really, the most powerful way to get use out of it is strategy two, three, or four based on what you've learned there by what it's told you. Because now all of a sudden you could say, okay. I know that this improved the strategy.

Dave: 17:20

So let me step back, think of another idea that where I can use this and start from a completely different starting point that is has more trades in it. And you you that's the real power is you find these relationships in the data, and then you can find a second or third strategy that is just significantly better with more trades than your original strategy. It's it's a pretty common path that and it's I love seeing this because that's exactly what happened to me, you know, ten or fifteen years ago when I created these tools for myself. It's just it kinda blew my mind then. And, so it's really exciting to see when people go through the same steps that I was going through.

Michael: 18:07

Well and and what's interesting, I I think, is, you know, it's just gonna sound like really cheesy and motivational or whatever, is that it's not that it so much always answers the question, but it actually gives you more questions. And that's what I found that's interesting about, you know, the Cruncher, but also just the column library in general, however you want to do your analysis on it, is that you see that, okay, five things are all kind of trend following. And you go, okay, well, that's interesting. I didn't think the trend would have affected this strategy. Not, you know, doesn't mean you just necessarily plug something in and away you go, but now it's it's giving you a question of why.

Michael: 18:46

And I find sometimes the most illuminating things with the crunchers is when they're not at all what I expected. You know, for example, I'm so we talked about this with Ambien Broker, one of the strategies I'm currently building is just an opening range break. This is a second day play, know, the SMB guys talk about this all the time is day one, you have a huge gap up and a whole bunch of noise. I'm getting kind of old, so the strategy I want to is gonna avoid some of that. And then it's looking for day two and three breakdown, right?

Michael: 19:15

So all of the degens have forgotten about this, and they're on to the next crazy, you know, Discord room pumped whatever, and I'm shorting it when everyone else has kind of forgotten about it. And I expected, okay, the more volatile it was on day one, maybe that kind of has some effect, or the more relative volume or something like that. And if I run that, and I have these pre these assumptions in my head, and they're completely wrong, to me that's way more interesting. Where it says, no, no, no, you actually want like no volume today, not a lot of volume, you want nothing. Then it makes me think, so for this particular strategy, it makes sense.

Michael: 19:49

It's like if everyone's forgotten about it, the volume should just dry out of them, and I'm gonna be trading these more liquid securities. But just the fact that when I brought it in there, I had assumption a, and it told me the exact opposite of that, that's really interesting because then it's just like, okay, now it's time to go back out in the woods and think about it for a little bit with the dogs.

Dave: 20:08

Yeah. And it's it can be humbling. Right? Because you think you know these things. You think your intuition and experience in the markets have taught you these things that and then when you're confronted with the fact that you might be wrong, like, wow.

Dave: 20:23

Okay. So I have been humbled so many times with data like that that it's just too many to count. But, yeah, I mean, that's how you grow, and that's how you learn, and that's how you sort of noodle with your ideas to figure out, okay. How can I use what I've just learned here? And if you just learn something, then it's not it's probably not commonly known in the trader community, and you and and because of that, you could use that to to find some edge.

Michael: 20:52

Yeah. It's the same thing as, you know, we talk about the building a a trader group or a trader community as well. I kind of equate it to someone who just challenges your beliefs. You say, okay, you know, I'm gonna take this trade because this thing happens. And and discretionary traders do it all the time.

Michael: 21:09

And if you have a group that is constantly, you trying to crap on your idea, and the whole point is that you want that because if it can withstand all of that, then it's probably a pretty good idea. And if it can't, it probably sucks. And the ability to have kind of a machine do that, because it doesn't know what your idea is, it doesn't right, you're just giving it raw data, it has no idea what it is that you're trying to accomplish with the trade, and it's just spitting back out some data, and you can go, okay, well this is either good or bad. And yeah, I just find, just like anything, know, if I had someone who was trading with me and every time I said, oh, I'm look gonna look to buy this stock if you if you're discretionary. He said, oh, yeah.

Michael: 21:46

It looks great. If if that's the Yeah. If that's the feedback every time, that's that's a bit useless. Right?

Dave: 21:51

Yeah. Like chat GPT being so so complimentary every time you ask a question. Yeah.

Michael: 21:58

Yeah.

Dave: 21:58

So so it's funny. So I'm one of the guys I'm coaching now, he said to me the other day, thank you for bursting my bubble. Like, I he had just asked me this question. I think, you know, his hopes were really up. Like, he saw this he saw something in the cruncher, and he was like, man, this looks really good.

Dave: 22:18

What do you think, Dave? So I looked at it and I could recognize I've I've had this exact same feeling before. Saw the same relationship. And I was like, actually, this is sort of illusory and you should move on to something else. Here's why.

Dave: 22:35

So I'm going to explain that to him and his I burst his bubble a bit, you know, because you've got such high hopes for it. And, and it's, it's really when you have high hopes like that. And then somebody says, well, actually it's not right. You shouldn't take that approach. That can be a little bit demoralizing, but you want to learn that way earlier in the process.

Dave: 23:01

Mhmm. You wanna be confronted with that early. I mean, the worst time to be confronted with that is to have the market confront you with that. And when you're trading money and you realize, okay. I'm losing here.

Dave: 23:12

I wonder why. And it takes you half a day of digging through log files and comparing back to us to figure out that, okay. Well, this isn't quite what I thought it was. That's the worst time to have it. So the the earlier in your process you could be confronted with that and find out that your idea wasn't quite as great as you thought it was is is always a good thing.

Michael: 23:35

Well, it's good. And I think we've all done that early in the career. So I think it's good to do it. One of those things earlier rather than later is, you know, you're looking to I think we've all done it where we run a back test, and the backtest looks amazing. It's like just straight up into the left, there's no pullbacks, there's no nothing.

Michael: 23:52

Go, okay, well, I've figured out, I hacked the market.

Dave: 23:54

Yeah, yeah.

Michael: 23:55

Good for me. And then you start trading and you go, okay, there's a slippage thing, or, you know, it's getting filled at a weird time or whatever it is. It's just glad that I've learned that early enough that Yeah. You know, and it's the same, you know, this is a I think the importance of bringing in these new columns is that you're always in a state where you don't know what you don't know. Right?

Michael: 24:17

You don't know that if you only had two or three things that you looked at for every trade, for example, like I know, you know, the common one, is it above or below a 200 period moving average or or something simple like that? Well, if you're not bringing in columns that make sense, but maybe you don't think they have any effect at all, then you're kind of you don't know what you're missing is I think the biggest point I'm trying to make is that I'm, you know, the way I'm doing is I'm looking at these things and I'm saying, well, I don't this was a very interesting write up about this this new Mac D and Mac DV. I don't know. Like, I'm gonna put it into the the Cruncher, and I just may never see it. And that's the important of, you know, ranking these things.

Michael: 24:56

But it's one of those, if you're not, if you don't have a process and a plan to exposing yourself to new filter ideas, you just really never will. And whether it's kind of the way that I'm doing it, where I'm just absorbing content, or, you know, someone else observing the market or however you plan on doing it, it's it's super important that you're just letting in the fresh thing because it's, I think, even more important than doing a new strategy. Because I can go back and it it could do nothing. I could run, I could rerun, I could retest all these strategies and have absolutely no no benefit whatsoever. And that's okay.

Michael: 25:35

I just wasted a bunch of time. You don't know till you get in there. Or it could make some or all of my strategies significantly better just with this one thing, this one paper that I I read and that I listened to out in out in the middle of the woods.

Dave: 25:49

Well, even when a column that you think of and you think might have some effect and then you go and put it through the cruncher and it doesn't show up in the list, even that's valuable because you you had this idea, you had sort of high hopes for it. Now you can move on. You can put that. You know that that's not valuable, at least for this strategy, but it's in your library. For other ideas in the future, it might be important.

Dave: 26:10

So you've you've got this process. And even if it didn't pan out, it's still that's still valuable information even right then, not to mention the fact that you're gonna, you know, potentially use this in the future anyway. So so I wanna go back to something you mentioned that we just think it was really good where, and I know a lot of people a lot of traders think this way is, you know, how does this strategy work when price is above the fifty day moving average or below some moving average?

Michael: 26:42

Mhmm.

Dave: 26:43

A very common thing people want to do is add a column that's either a one or a zero. Is it above the fifty day moving average or is it below? Mhmm. And you can do that, but a way more valuable thing to do is to to reduce that or or make a comment that says distance from the fifty day moving average. It's every bit as good as the one or zero and way better because when you reduce something to a one or a zero, this binary thing, for you know, if it's just barely above, that's one.

Dave: 27:23

But if it's, like, way, way, way above, that's still a one if you're reducing it to a binary. Where if you have this continuous variable that's, like, distance from fifty day, you're it it's just way more precise. There's you're you're allowing the cruncher or whatever tool you're using to regress across that entire continuous variable, and there's just a lot more information there that you're you're losing that precision on if you reduce it to a one or zero.

Michael: 27:52

Yeah. And I think that's important too. We talked about before is you you need to know what these indicators mean, and you need to know kind of roughly how they're calculated because and this is funny because they make you do this in the CMT, and then I did that five years ago. Then I come to this stuff and I just forgot it all, so I had to, relearn. But it's important because in some aspects, the indicator just does it on its own.

Michael: 28:18

For example, I don't care what the absolute value of the 200 period moving average is. I don't care if it because it's just gonna be a price. It's it's a useless useless thing to me. But a RSI, for example, the absolute value of an RSI, that's important. So there's some of them that you should compare price to, and there's some that you should leave as the indicator is itself and, you know, all kinds of of things in between.

Michael: 28:41

But you're right. I don't want to you know, for your example, above below moving average, you know, you could look at that and saying, well, I don't want why would I just set up a is the RSI oversold? Which generally speaking is under 30. If I just set that up as a binary option, you know, like you said, it would it would do something. But instead, if I just put the absolute value in there and then I see it pop up as under 30, then I know it's great.

Michael: 29:05

But, yeah, as soon as you do a binary option, well, what if it's 31? Like, do I just ignore the trade entirely? Does not and and only the data will tell me that where it's, you know, it's this 30 area that they made for oversold is just an imaginary line that someone put in and said, oh, they usually bounce when they're under this thing.

Dave: 29:24

Yeah.

Michael: 29:25

Yeah. So you you gotta keep that value on there. Right?

Dave: 29:28

Yeah. It's it's better to take, like, a naive approach and let it tell you what oversold is. That's really what you're doing. You're if you remove the assumption, then you can let the data just tell you what the optimal value is. I mean, that's a way better approach.

Dave: 29:41

Alright. So let me go over so there's a couple more here that I have added over the years that were sort of light bulb moments for me.

Michael: 29:49

Okay.

Dave: 29:51

One is position and range. So instead of, like, above yesterday's high or below yesterday's low, position in yesterday's range is, like, the same sort of concept, but it but it can be applied not only the yesterday yesterday's range, but all sorts of different situations. And when you want when you watch your trades come through, you're gonna develop theories about how things work or how you know, when things are better, when trades work out better in certain situations. And, really, probably one of the best ways to brainstorm for new columns like this is to when you feel like you wanna override the system, like, feel like you wanna exit early because of why, then, like, figure out what the why is and how you can reduce that or or come up with a column for that, add it to your backtest, and and test your theory. That's a great way to do it, and that's how you that's how you improve and get better over time.

Michael: 30:54

Well, yeah. And just to explain and define the position in in range, if you just visualize yesterday's candle and then a 100 would be the top end of the range, right, zero would be the low end and 50 in the middle, and then you could go right above a 100 below zero. And just like we talked about with this that this MACD that's spurted on is that's normalized because you're not saying some people might say, oh, you know, if I'm gonna do a mean revision thing, maybe I wanna short stocks that are are really oversold, and they're getting up to yesterday's range. That's that's the idea. Well, you think about programming that in, you could do something like plus or minus, like, 10¢ or something.

Michael: 31:34

But that doesn't right? It doesn't make sense because on some stocks, that'd be the whole range on some stocks. So this was just a way of saying, okay, well, you know, maybe it's 80% to a 120%. I'm gonna use that band of yesterday's highs, my area to short. Well, that makes more sense because you can apply that across instruments, and you can apply it throughout time, and and you can kind of build models and systems like that.

Michael: 31:56

And that's one of those things that, yeah, I immediately added that to mine as well because it's just it's it is the simplest way to approximate, you know, price action and things like that.

Dave: 32:07

Yeah. So another approach that I've taken, which is which it didn't dawn on me at the beginning, which it should have. So, like, when I was mentioning earlier, distance from the moving average or something, like, whatever that is, it's you could use that for all sorts of different things. But what is that distance? What should that distance be?

Dave: 32:30

Like, you should make it just like we're saying before, gap percentage versus gap in terms of ATR. It's the same thing. Use use ATR for that those distance columns, and it's, like, automatically normalized.

Michael: 32:46

ATR, your most moved used, like, indicator across everything for because for me, it's a 100%. Like, everything I do is normalized. And again, if you wanna read the paper, it's it was the 2025 Charles Dow or 2024 Charles Dow award winner. But that's what he did, is he took the normal MACD and he divided by ATR. Right?

Michael: 33:08

So it's just one of those things that it's hilarious to see the simplicity of it. However, no one's really done it and and kind of published anything about it for like forty years. But, yeah, I just wonder because, yeah, for me, ATR, I use it so much across everything. It's just it's I think it's like a quant's best friend.

Dave: 33:26

Yeah. I think it's I'm sure it's used across most of minors. I've it contributes to a lot of them. So, yeah, I think that's totally true. Alright.

Dave: 33:37

So here's another one that dawned on me after the ATR. So there's some strategies I have where the stop distance is not determined by ATR. So the distance to the stop, if you're calculating in terms of r and use expectancy for your performance, which you should know about, look that up if you if you're not familiar with that. You can do a lot of these measurements for that specific strategy in terms of r. So distance to the stop, which may or may not be correlated to ATR, may be exactly ATR for some strategies, but this particular one did not include ATR in that calculation.

Dave: 34:21

So now distance in terms of our two different things was highly I mean, very predictive for this particular strategy. So that's another way to and that one's pretty perhaps signal specific or system specific, but there are some probably some ways you could use to apply that across different systems that you have. That's a it's a really good one that was predictive in a lot of different ways.

Michael: 34:47

So, yeah, so what you're saying is, you know, say you're you're shorting a stop and you're putting the stop stop at the high of the day, or just just for example, you're taking whatever that distance is as the risk value, and then you're feeding that back into into a way to normalize a lot of these things and saying, okay, you know, whatever this risk is, I I want to use that as kind of a threshold for a way to normalize across this because in some trades, for that example, you might be risking 1% and some might be 10% and some might be 50%. And, you know, by doing that distance, it's it's interesting. And that's that kind of makes me think of, I do a lot with today's range as well, right? So you're looking at ATR, which is just a moving average of the normal range, but sometimes you want to take that moving average out entirely, you want to say, okay, today the stocks moved a buck. And I wanna then normalize everything to that, and you could use, I guess, a one period r or ATR for that.

Michael: 35:47

But the I wanna just look at specifically today's action and then start to normalize all my indicator for that particular action for today.

Dave: 35:56

Yeah. I think that's great. And I've got one more concept here that I think people will get a lot from. So all that we've talked about so far are what I call entry columns, like columns that you can know about at the time of entry Mhmm. And therefore, you could create a rule based on.

Dave: 36:20

These columns, at least the way in Amibroker, these are calculated and collected, they're actually assigned to the trade at exit time. So you keep track of them at entry time, but they actually get assigned to the trade at the exit. So it's not looking into the future. There's you're still adding the values that you could have known about at the entry, but they're assigned at the exit. What that allows you to do is you could add some intra trade columns based on things that have happened since the entry.

Dave: 36:54

Now you can't, of course, optimize on these for the purposes of, like, making an entry. But what you could do is run it it allows you to create one backtest and and test perhaps three, four, five, many different exits. So you could have a column for, alright, what was the profit at noon? You could have another one for, okay, what was the profit at 02:00PM? What was the profit at just before the close?

Dave: 37:31

Mhmm. Then instead of having to run, like, four or five different back tests just to give you this one answer, like, what's the when's the best time to exit? You've got one back test, and you can just look at each of these columns and see, okay, how do they compare? And you can see just from that, just from a single back test, you can get your answer.

Michael: 37:51

That's why I like doing this podcast. I never thought of that. That makes perfect sense. Because I always, you know, immediately thought of forward looking things are faux pas and you want to avoid them at all costs. But right?

Michael: 38:03

If you think about it, right, it's a 100% right, is that that's a way to save a lot of time. Alright. Well, we got one of the I hope that this is what the audience goes through every now and then. Right? They hear something, us rambling on about something like, oh, yeah, that's something.

Michael: 38:17

But now I have homework. Right? I'm actually writing this down.

Dave: 38:21

You

Michael: 38:21

guys may have saw me pull out the Apple Pencil because I have got I've always got this thing sitting here I'm writing on. I'm gonna add that into because some of the like, guys know I'm going through this Ameren broker journey with this opening range breakout, and that was gonna be one of the questions. Right? I've noticed for the stuff that I'm doing anyway, profit targets don't I don't wanna use them. Right?

Michael: 38:41

I wanna let the thing go till the end of the day because I'm okay with a lower win rate because what I notice every now and then, you just blow one out of the water. And because, you know, this is in addition is the way I'm looking at this trading for me to other stuff that I do when I don't need the smoothest equity curve in the world, I'm okay with sacrificing smoothness for a potential overall profit, and it's the simplest way to put it. But that was gonna be the next question. And my plan was literally to run four back tests, one entering at the close, one entering at noon, one entering maybe after the first hour, and then go from there. So now I've got some homework to do myself.

Dave: 39:16

Yeah. And I'll just point out that, you know, if you were doing this in a back test that took two seconds to run, you would have never thought this would never have dawned on me. Like, I would never Yeah.

Michael: 39:28

You just would have done it.

Dave: 39:29

Never thought about it. Like, why why would you this is you know, as we've said in a couple episodes ago, I think, a really quick back tester kind of creates some bad habits. And having a longer back tester, you know, it's maybe it's a little bit polyan, a little bit, you know, silver lining, glass half full mindset of me. But it makes you think about, okay. What's really the best way to do this?

Dave: 39:55

Like, there's probably a better way. And and the fact that a backtest takes way longer in Amibroker kind of creates this situation where, okay. Well, I'm gonna try to think about a way to get all my answers with just one backtest.

Michael: 40:08

Yeah. And for the way I look at it now, and and just so you know, I've explained my setup. I've got a Mac that I do everything on. I've got a PC that's locked away in the other room because I hate Windows that much. And it just runs backtests all day.

Michael: 40:20

That's that's the whole the only thing I use it for is for real tests and for for Amibroker. So that kind of stuff actually works perfectly where I can run one monster backtest that has all these exit times and everything, and then I copy and paste the data onto into Excel on the Mac, and then I just set it to do something else. And it's okay that that thing's, you know, working its its butt off for however long, because now I've got more time more data in my hand that I can do to keep myself occupied while the computer goes and does something else. And I think that's a good thing to look for and optimize as well is that if I'm running the thing, and then I'm just sitting there not doing anything for a long period of time, Well, that sucks. But if I'm doing something like this, where I'm getting all this column data, and I'm getting all the the exit time data now, and all of this, then I've got a whole bunch of work to do.

Michael: 41:10

And it won't be as painful because I'm not just sitting watching this thing kinda tick up and say, hurry up and finish. Instead, I've got other work that I can do to move that along. But, it was a good one. I never thought of that one.

Dave: 41:23

Yeah. I, you know, I provide a calm library with Makekit, and it's so I I love seeing traders use it because I know, like, almost every trader that I've ever worked with, I I can see clearly before they do that they're taking probably 20% or 25% too many trades and that one simple rule could improve their life dramatically. So it's really fun to see that when somebody goes through that and they I get an email from like, oh my gosh. I never I never would have known that this had the effect on the strategy. That's so I I love seeing that.

Dave: 42:07

It's awesome.

Michael: 42:08

Even, you know, even more so. Right? So let's say in this hypothetical universe, I'm currently testing and I'm optimizing the system for clothes at the end of the day. Let's say I add this column, and it turns out that a clothes that a a sell at noon or lunchtime does almost as well, even if it underperforms ever so slightly. Well, that's huge because now that's put me on my next path, which is okay, now I need something that is probably scanning for, like, unusual outlandish moves that happen at lunch or into the end of the day.

Michael: 42:42

So not only, right, we're talking about, you know, indicators that could filter out bad trades and things like that, but that, this kind of one tidbit may open up a whole bunch of buying power that, right, if I exit this one strategy, if it's using up a lot of buying power and I exit it all at noon, well, now I've I've got the freedom in the room now for strategy two that I know is gonna be completely different because it's gonna trade during a period of time in which in which strategy one is kinda completely closed.

Dave: 43:10

Yeah. I think that's great. I mean, that's you're exactly right. It doesn't have to be like, you don't have to find something that's way better. You could find something that's almost as good.

Dave: 43:19

I love that.

Michael: 43:20

Yeah. If

Dave: 43:20

it looking back at my results from the February, and my best year during that period, I had a lower expectancy than the other years, but I had, like, 10 times the number of trades that than when I started. Oh, it's it's not that and I remember my mindset during that period was, hey. I just done automation, just implemented that. It's like, wow. Okay.

Dave: 43:50

Like, I can actually relax my threshold for what a profitable strategy is because I can all of a sudden, can take as many trades as I want to. I did. I took I took literally 10 times the number of trades that I had just a couple years before. Each one of them weren't wasn't great, but you add up enough for those trades and there's a whole lot of profit there.

Michael: 44:14

So yeah. So I think I think you kinda guys saw in real time what we what we wanted to do or what we would the point we wanted to make in this podcast was that you need to have a plan to open yourself up to thinking about new columns or thinking about new ways to optimize existing strategies. Because, again, as someone who's a prolific consumer of content, right, I picked up the Mac DV by by reading this paper, and then by sitting down with Dave, I picked up another one. You guys don't have that, I think that's the main takeaway is not, you know, we gave you some some good ones, I I definitely think, but it's the process of how am I going to reach a little bit outside my bubble. We talk about it all the time, especially when things are going well, traders kind of silo themselves and they keep doing the same thing that's working because it feels when things are going well, it feels like magic.

Michael: 45:02

Right? I woke up in the morning, I turned on the robot, and, you know, I I made a thousand bucks. It feels like like magic. But you have to have that process to opening yourself up to what's the new thing because you never know if there is something out there. So now I've got two things.

Michael: 45:19

I've got the MACD, and then I need to go and add these these time columns in as well and a whole bunch more work to do. So hopefully, you know, the the bot continues to tick on like magic for a little bit longer, and then I'm not working from a an area of of detriment. I'm working from an area of strength where now I'm going, okay, well, now let's go and explore these new things while this guy over here just just makes me some money, and then I'll come back to him and hopefully it makes more money.

Dave: 45:48

Yeah. I mean, you're it it's really what you're really doing is just becoming more efficient. And when you do that, you increase your ceiling dramatically. I mean, that's what you're doing as you, you know, figure out how to make a strategy more efficient, how to use automation to take advantage of it, do what you said, which was brilliant. Can can I exit these at noon and create a whole new window of potential for a strategy that's profitable?

Dave: 46:22

I mean, that's that's what you can tell I love this stuff. I mean, this is so much fun for me to solve this, and, I just I just love trading and everything about it.

Michael: 46:33

Well, good. Well, I'm glad you do a podcast. But it would be really weird if you did a podcast about trading. You just hated every part of it. Yeah.

Michael: 46:41

But yeah. So I again, I we gave you some stuff. I think we gave you a process or or a way to think about how to add more columns, you know, observe the market, consume content. There's a lot of good stuff out there, a lot of crap. But, know, I'm sure people who probably stumbled upon this process or this podcast are probably decent at sniffing that out anyway, hopefully.

Michael: 47:03

And, yeah, just make sure you're always open. You're always open to kind of figuring out what's next.

Dave: 47:08

Yeah. I think that's a important part of being able to adapt is not thinking you found the answer. I mean, I see a lot of traders that get stuck in that when they

Michael: 47:17

Mhmm.

Dave: 47:17

Like I said, the the only thing harder than creating your first straight trading strategy is creating your second because you get stuck and you think you figured it out, and you always need to be open to the fact that you might be wrong or it might be, you know, your your strategy might fade. You might need to have a completely different idea. So that's the hard part.

Michael: 47:39

But as always, I'm Michael Noss.

Dave: 47:41

And I'm Dave Mabe. Talk to you next week on line your own pockets.