Line Your Own Pockets

Michael: 00:00

Alright, everyone. Welcome back to another episode of Line Your Own Pockets. We're gonna get into the tools of the trade in this one, which I'm super excited for. We talked a lot about, you know, theory, and we've been going through different ideas for backtesting and using social media and news and all that cool stuff, and we'll continue to go down that road, of course. But we're gonna get a little into the nitty gritty and just talk about what tools you actually use.

Michael: 00:23

And it's it's gonna be interesting because I know, we me and Dave are gonna approach us from a little bit of a different angle where Dave's a little bit well, a lot better when it comes to coding and and that type of thing. And I'm looking for more user friendly and and different names like that. So I think it'll be interesting. So, Dave, I know you were mentioning that you write an email that's pretty popular that talks about this. So why don't we get started there?

Dave: 00:49

Yeah. So I have a free email backtesting course. So, yeah, it's a back backtesting course where all is free, and it comes over email, and there's several emails that come out once a day. It's pretty popular. One of the emails in there is on choosing the right backtesting platform.

Michael: 01:09

Mhmm.

Dave: 01:11

It's I think it's a critical choice because backtesting, I think, is a superpower. It helps you gain confidence in your system much quicker. It's like a cheat code to getting confidence. And the better you are at it, the better you'll be as a trader, and that lasts forever. I mean, this is a skill that you'll develop over the course of your entire trading career.

Dave: 01:38

So it's important to have to make the right choices when you I I think this is probably the most important tool that you use in your arsenal.

Michael: 01:48

Yeah. Especially if you're if you're planning on automating it. Right? It's it's it's all that matters. You know?

Michael: 01:54

It's you have to develop a system that tests well and that you're making sure is tested robustly, and and there's no funny business going on in the back test. And then you let the system run, so it's that. And then reviewing what actually happened are are kind of the 2 major parts of being systematic with your trading, especially if, again, it's gonna be automated in some way. So you've you've got to make sure that you dial in that back testing side of things and that you have to make sure, obviously, that you have the right tool that's giving you an honest account as as best it can of what will happen in the real world and all that. And as more and more technologies pop up, there's way more of a field to go into.

Michael: 02:33

You know, we I I know we sound like old men. We talk about this, but when we started trading, there was the back testing actually meant. And for those who don't know, paper trading actually meant writing on a piece of paper what you would do. So people would just scroll back on their charts and just, you know, move arrow by arrow, bar by bar and say, okay, what I bought here and write down on a piece of paper and then, you know, tick forward a couple of candles and I would have sold here and then seeing what those results would be. So it's for us, I think it's probably one of the best inventions that has happened in trading since charts.

Michael: 03:07

Like, you can you can just go through and ask a system questions and build a system and go back. You know, for me personally, with being more longer term in nature, I go back to the year 2000, and I start all my back tests there, and I can go back even further if I wanted to to 1950 for some of these indices. So, yeah, you definitely gotta make sure you got this taken care of.

Dave: 03:29

Yeah. For sure. It's, like I said, it's a superpower. And there's definitely there's definitely things that I look for in a backtesting platform, and not all backtesting platforms measure up. And so why don't we just go through the things that I look for?

Dave: 03:48

Sure. And these are, like, hard requirements for me whether I'll even consider using a backtesting software. So the first one is it has to backtest the entire US equities market. If it can't do that, I am not interested. I just think there's so much there's such an easier path to creating an edge when you're looking at the entire US equity markets.

Dave: 04:15

So, yeah, that immediately rules out several pop very popular back testing softwares. Right? TradeStation. TradeStation does not back test the entire market. Yeah.

Dave: 04:27

I think TradingView has a back tester in there. Yeah. Can't back test the entire market. Transpider? I think Transpider can't back test the entire market.

Dave: 04:37

So And I just think it's such a critical feature. I would not consider any software that doesn't do that.

Michael: 04:46

Well and we should again, like everything, you should talk about what it is that you're looking to do because this might not be for everyone. Right. So we're going to talk about this as well. If you're someone who just wants to trade a single symbol like an S and P 500 or futures, I'm thinking Bitcoin, something like that. It might be more okay to use something like this.

Michael: 05:09

But for us who are who are looking at equities, you're a 100% right. And now that the reason I think there's so many that do a single stock is it's just way easier. Right? You bring up the stock and it just runs through the candle math. And this is why quite often, if you're looking at a back testing engine, that will be a different beast than a charting and alerting type platform.

Michael: 05:29

There there's some that do both, but it it gets very, very hard to process that amount of data. Right? Looking at one symbol and, you know, I bring up a daily chart of Apple and have it saying, okay, every time the Mac d crosses, this happens. Again, someone who doesn't code, I know that'd be fairly simple to do. I could do that in Excel spreadsheet, right, with with one dataset and one symbol.

Michael: 05:51

It gets very, very hard and it gets more complicated when it comes to things like buying power, like in which order do the trades come through if you're testing the entire market on on which ones that, that you have to make sure that you have all of those things. And again, that's why I think a lot of these will talk about when it comes to actual backtesting engines. They have to have that data in some other fashion because that's the hardest part. Right? You're you're talking, you know, 6,000 ish, 10,000 roughly symbols that are are existing and trading every day with any kind of liquidity.

Michael: 06:23

And you're right. If you're not testing all of those, there's a whole bunch of biases that we can talk to. I don't know how many times. The reason I'm ranting a little bit, how many times people send me a back test for, like, the S and P 500 and say, oh, look. You know, it's one thing that just triggers me is that, hey.

Michael: 06:39

75% of the time, when this happens in the S and P 500, the mark it's up. And I'm like, well, 75% of the time when anything happens in the S and P 500, it's up. So you're already creating a bias there, as well that you then need to you need to deal with. So looking at a wide breadth of symbols, I think, is is the easiest way to make sure that you have some sort of robustness in your testing. If it works on Apple but it doesn't work on Tesla, then there's probably something going on.

Dave: 07:06

Yeah. I like what you said about it. It's, it's easy to look at a single stock chart, and that's one of the side of kind of the core of this whole issue. Yeah. It's really easy.

Dave: 07:17

I mean, just think how many people look at the chart of the s and p five 100. I mean, probably every trader in the world. Mhmm. And even every investor in the world probably looks at that chart on a somewhat regular basis and thinks about, oh, maybe there's a strategy I can trade with it. So that's a lot of eyeballs.

Dave: 07:35

That's a lot of competition. That that's gonna be a very efficient market, which is difficult to trade.

Michael: 07:43

Yeah. And, you know, just again from the it's kind of trying to find an example, but a charting platform will just put out, we have backtesting, and then they put in that single symbol backtesting because, again, it's glorified Excel spreadsheet. Right? This, then that, and this. So it becomes from a developer's point of view, just an easy thing that you can just tack on to the end of all the things that you're charting platform does and say, okay.

Michael: 08:07

We can also back test. And it's like, well, yes. In theory, you can also back test.

Dave: 08:11

Yeah.

Michael: 08:11

But you're not doing the the best thing or term that I've heard to search for is portfolio level back testing. So if you're googling right? Because that's, again, assuming that you have a starting portfolio and what is happening to that portfolio day after day across a wide breadth. But, yeah, a lot of the most popular names get knocked out pretty immediately just because, yeah, they they can only do a single symbol at a time.

Dave: 08:34

Yeah. So let's talk about which ones do do this, backtest the entire market. So, I've got a list here that I think is comprehensive, I believe, as as of this moment. K. So trade ideas, the odds maker, is one that backtests the entire market.

Dave: 08:54

And as far as I know, it's way faster than any of the other ones. The only downside with it is it you can only backtest 3 months' worth of data, which for some strategies, that's gonna be sufficient for. But, you know, it's the longer time frame your your strategy is, the less sufficient that's gonna be.

Michael: 09:15

Well and and this is probably a good way to do it. We'll go through 1 by 1 and just do kind of a pros and cons list because, yes, so trade ideas. Right? It back test the whole market. It does it incredibly quickly.

Michael: 09:27

Again, only 64 ish days of of data. So what I like about it, because it is so quick, is that to me is a good starting point, when it comes to this. Now if you're swing trading, if you're long term investing, that kinda knocks that out entirely because if your hold time is 10 to 20 days and you only have 64 days of data, that becomes really, really hard because you're only getting 1 or 2 cycles of having trades go on that could be a problem. But it can tell you really quickly if it's something to look to look at further. If you do a back test, especially go back to intraday trading.

Michael: 10:03

If you do a back test and there's absolutely no edge in anything you did over the last 3 months, it's probably not worthwhile digging too much deeper on that. You can probably just scratch that one off. Now if you back test and there is a bit of edge for that period of time, it may not mean that your strategy is complete. You may wanna go look at other back testers or start paper trading for a period of time or do something else to increase that data set, but it very, very quickly gives you a good and quick feel of, am I on to something? Is this a waste of time?

Michael: 10:34

Should I move on with my life? And then I think from there, it's a good expansion. The beauty of if we're just going over the whole software, the beauty of that is that when you pay for that and trade ideas, you get all the other stuff. You get real time scanning and alerting, and you have the ability that whatever you're back testing, you can generally pretty quickly then auto trade it as well. So it it does that.

Michael: 10:55

I just that one limitation of not being able to go back too far, I think you then mitigate in some other way either through forward testing. I know they call it, which just means training in real time. Or you go and, you you use another back tester or another method to make sure that you have that that data that's robust enough that you're comfortable.

Dave: 11:15

Yeah. So, basically, the best traders that do a lot of backtesting, they'll use trade ideas as a prototype. Well, just like you said, like, they're prototyping their stuff before like, it it it lets you see very quickly before you devote a whole lot of time and effort into coding out a real back test that works over a long period of time. So why don't we go over the other ones that on on my list here? And then trying to decide whether we should go back and talk about the other requirements I have for backtesting, which I haven't gone over yet.

Dave: 11:52

Before we go over this list, what do you think?

Michael: 11:54

Let's go through the list, and then let's, as you come across, and I bet you there'll be ones that either have requirements or don't, we can kinda talk about it as we go. But, yeah, let's let's go through it because each one is gonna have some massive benefit and then probably some sort of flaw that we have to deal with. So let's go through that. And then if there's anything that's still on your list of of requirements, we can go back and and make sure we get.

Dave: 12:17

Okay. The next one is Amibroker. So amibroker, all one word. This is a very popular software for developers. So it it has its own coding language.

Dave: 12:34

It's it has its own charting package, and it's a bar by bar backtester. It can back test intraday. The it's a very extensible platform. I've used it and, you know, created many strategies that I trade using AmaBrokers. I I love it.

Dave: 12:53

I've been using it for over a decade. There are some downsides to it. The one is, you know, you are learning a completely new language that is only applicable to Amibroker. Y'all there's also some data, there's a bit of a learning curve in determining, you know, how to get your data into AMI broker. Does it automatically flow into it?

Dave: 13:20

So there's some data management aspects of it too that to to learn.

Michael: 13:25

So, yeah, as and and I think this will be interesting. As someone myself who is not a coder, and and is getting kinda just more into the the very basic, I don't ever see myself building apps with a very basic, you know, scripting and and things like that to do, when it comes to testing. I tried Amni Broker. I still have a a subscription to it, and I wanna play with it more. Pros, I would say is because it is so popular, I would say probably the most popular when it comes to just raw, backtesting.

Michael: 13:56

The benefit to that is there's a lot of things that you can find online to help you with that and, AIs. Right? So I you know, there's a couple custom GPTs you can have that will help you code and, you know, you can give a code and ask for feedback and do certain things like that which is really really interesting. The massive downside, the reason that I actually moved on to another back tester is the forms I found absolutely like, atrocious and, like, toxic and just downright mean sometimes. And I think that part of the problem of it existing for a long period of time.

Michael: 14:34

Right? People are sick of of the newbie coming in and trying to just ask questions and and figure out things, But as that newbie who is trying to do that, it it just wouldn't work. And I found even the the CEO and the owner just fighting with people in their in their forms. So there was a very big lack of community. So if you're someone who is very technically savvy, I would imagine any broker can do a whole lot of things.

Michael: 15:00

But, yeah, I was just completely turned off as a learner. So I ended up moving on to another software, which will I'm sure we'll talk about near to the end. But, yeah, it was it was it was gross. It was mean. I didn't like it.

Dave: 15:11

Yeah. Good vibes. You're not alone. Yeah. I've interacted with Tomas a a good bit.

Dave: 15:16

And, yeah, he can be he he doesn't hold back. He's not, he's not holding anybody's hands, and it is interesting. I mean, that's one way to do support. It's not a way that I would use to do it. And I wonder I mean, I've thought many times as as harsh and terse as he is in those forums, I wonder if he'd had a different approach, if the software would be even more popular.

Dave: 15:45

I'm I've heard what you the the what you just described. I've heard from many traders. Many traders say the same thing. Kind of kind of vibe.

Michael: 15:53

Yeah. And, again, it would be different, I think, if if someone was, you know, just more savvy in those things, like like someone that you are who who has the ability to look at something that would be completely foreign from a code point of view, but you have all this backing on on how to approach these problems and how to solve these problems and how to do this, that, and the other thing. Whereas someone who kinda comes in like myself and just gets completely stuck, and it's one of those I could spend 8 hours trying to get myself unstuck and it just wouldn't happen because I don't have that background. It just became quite quite the turn off. But so what I ended up doing, that's why I end up going to the the chat gbt store and there was because it's so popular, there are many gbt's that were trained on the entire code base.

Michael: 16:41

Some actually pulled a lot of the form down and trained on them and those helped substantially. So and it was much nicer while it was helping. So these are things I can definitely help if you're someone like me who has I have a lot of knowledge in trading. I've been doing it for a very long time. I don't have any knowledge when it comes to coding.

Michael: 17:00

Like, I'm really good Excel and, you know, basic, visualizations of of data. But so when I approach these things, it's okay. I need to be sure that it's something that I'm going to use for a long time because the amount of time it's going to take me to really get proficient at it is gonna be way longer than maybe someone else who's just having more of a more of an issue when it comes to comes to those things.

Dave: 17:23

Yeah. And and the ability to code in this period in history, it it's never been easier to start coding because of chat gpt. So

Michael: 17:36

Yep. You

Dave: 17:36

know, I have some of the traders that I work with, 2 years ago, I would not have recommended that they used Amber broker for their back testing. Mhmm. But today, I totally do recommend, depending on their skills, Amber broker, and it's worked out really well for couple people in particular. One is, you know, going full time with their trading, and they have used Ambit broker. But it's not somebody that I would have recommended 2 years ago.

Dave: 18:05

So it's it's it's really an exciting time for sure.

Michael: 18:08

Yeah. I love I know some people get some people are very anti the AI. Some people are very pro. I am very pro because, again, for me, it was one of those and I I I know we've had chats probably 2 years ago of of me talking to Dave and saying, okay. I need to learn to get better at my analysis job, which I do for, you know, CMT and things like that, analyzing market, and to get better in my own trading.

Michael: 18:32

I I need to learn to code. So I was talking to Dave about what kind of languages and everything to learn, and then the AI came out, and that that is completely gone for me because I just see a world where I don't need to really learn Python or or even the Amni broker code too extensively. I need to know enough to make sure that the AI is not doing anything funny, but it it it there's much more of a guided thing. And then as I'm scripting and as I'm building these things, if I have a block, I can just copy the and I can paste it in there, and I say, hey. What's wrong with my code?

Michael: 19:02

Give me, you know, 5 suggestions or something. And that usually helps. So it is like if you it's something you're trying to get into, it is like having your own teacher or tutor or something that will help kinda nudge you in the right direction as you as you learn and kinda get you around a lot of the, you know, the funny business for for us layman's.

Dave: 19:20

Yeah. And even you know, it's great for beginners, but even for experts that have been doing this a long time, it's incredibly helpful. I mean, so I just ported one of my trading applications to DOS. So, I've got a robot that trades through DOS from a trade ideas, alert window. So I use chat gpt as part of that project.

Dave: 19:52

And so it's a it's a pretty big extensive project and one that, you know, like, really the first one I've embarked on since chat gpt was widely available. It's it probably saved me 60% of the time that it would have taken me without it. It's really I mean, that's not earth shattering, but that is significant. Right? That is and I think the developers that really learn how to work with it and use it are gonna be the ones that do really well.

Dave: 20:22

If you're a developer and you're not using it in your workflow or you're you find yourself complaining, you know, about nits and gnats about what it does or doesn't do and and are finding ways not to use it, you're gonna be left behind.

Michael: 20:39

I, I know we're going on a tangent, but I think it's a good one, so we'll keep going. But, I remember I showed it to my dad. Right? And my dad actually owned an accounting firm, and and he was always bragging because they were the first accounting firm ever to get a computer. That's how how far back we're going.

Michael: 20:55

Right? Everything was done on paper. He's the first one to get, like, the old school, like Commodore 64 things that just cost an arm and a leg back then. And I remember showing him this because he's still using Excel and, you know, and all of the things you can do. And he's like, this reminded me of when they invented Excel and, you know, because I talked about all my existential fears of I'm gonna be completely replaced and there's not gonna be jobs left and, you know, what's my son gonna do and all this and he said, no, no, no.

Michael: 21:21

He said, Excel didn't replace accountants. Excel, there's more accountants than there's ever been in the world. He said Excel replaced accountants that didn't learn Excel. And that kind of was an moment where he's which exactly what you're saying that, you know, all of all of a sudden, it's not that there's way less accountants. It's that they're just way more efficient, so they're doing way more.

Michael: 21:41

So as opposed to a whole team of useless bookkeepers to change one number on on Excel thing now, it's it's that. And then when you went to go get a job, it wasn't, do you know how to, you know, do a debit and credit ledger? It's how good were you at Excel and formulas and pivot tables and things like that. So, yeah, I look at that as that next logical leap that, yeah, if you're a coder, if you're really anyone and you're not using this tool, then eventually it'll come a day where someone else is using that tool and they're able to do 10 times what you're able to do, and, you know, they'll end up replacing as well. So that's my optimistic view, I guess, of AI is that it just makes us all really super efficient to things, and there's always more things to do.

Dave: 22:25

Yeah. I concur, and, yeah, I'm sure we'll have other episodes where we'll bring up Jet GPTs. I'm sure this is not the last time we're gonna talk about it.

Michael: 22:33

Absolutely. But, just bring it back to Ami Brokers. So, again, that was my way around the biggest downside, which was the the the the complexity and the lack of help and all of that. There was a whole bunch also of, like Udemy Coursera, like a lot of these kind of self made courses, a lot on YouTube and all of that. So that's the beauty of being a very well known and a very well used platform is there's a lot more other people who are teaching it.

Michael: 23:00

So you can go and find a teacher that you like. These Udemy course I took was, you know, pennies. So you can go and I think it's $10 for a whole like introduction course. So you can go and you can find a course you like, and and I think that will certainly help you and get you around this. But, yes, it is it is very complex to use for me anyway.

Dave: 23:20

Yeah. So let's go to the next one on my list here, and that's RealTest. So this is the, I almost call it the anti Ambivroker. Right? We we both know Marston Parker, the guy that created this.

Dave: 23:37

Mhmm. Almost completely opposite of Tomas Remembroker, very friendly guy. We may, hear from him on this note pretty soon as a matter of fact. Oh, dropping hints. But so you you had more experience with real test.

Dave: 23:55

I've used it, but you use it on a regular basis. Why don't you talk about it a bit?

Michael: 24:01

Yeah. So Realtest is the one that I switched to, after after Amni Broker. And part of the reason so, Morrison is a market wizard. So if you haven't read those book series, I I highly recommend, you know, it's Jack Schwager. He goes and he interviews traders.

Michael: 24:17

And to become a market wizard, you know, it it's a long process. It's not just you put that you're a millionaire on Twitter and and he lets you in. Right? So you know the people there have legitimacy, and he has a chapter in I think it's called the unknown market wizards, the most recent one. All about automated trading.

Michael: 24:33

So as I was going through and as I was looking at this, it was interesting. And then there's also a gentleman that I follow on Twitter just called the Chartist from Australia, and he had switched from Amni Broker to RealTest. And the reason he did that is he runs multiple portfolios and multiple systems, and that's kind of the crux of RealTest. So the downside will get out of the way right away. There's no intraday.

Michael: 24:56

You could load in your own data if you want it, but it's a bit convoluted. It's focused primarily on daily, weekly, and monthly candles for its testing. But the it's built from the ground up for multisystem testing. And I just believe that that's we'll hear more soon now that they've let the cat out of the bag, but I believe that was on purpose for Morrison because he believes in in I do as well that the key to kinda consistent returns is running multiple systems. So he wanted to build something up that was more user friendly, from the ground up that was designed to work with multiple systems.

Michael: 25:30

So you can test a portfolio across 5 different systems, and it will take your money and it will allocate it based off how you want it to and kinda split it up. And that's from the ground up. So now what I like about it is it's, again, the exact opposite of Amibroker. It's more niche. It's it's smaller.

Michael: 25:48

It's definitely not nearly as popular. So I had to build and train my own GPT, which actually will put link in the show notes because I think I can just share that link and people can go. Where I just fed it all the help docs and everything. So it's been a little bit more complicated like that. I found from a non coders point of view, just the scripting behind it way easier on on how to import data and how to how to run tests is way, way easier.

Michael: 26:14

And the form has been super nice. You you see, like, John Boillinger, the creator of Boillinger Bands is in there. Larry Williams, who's done Williams R and a lot of other indicators. They're in there as well. And, yeah, it's just a a lot more kinda welcoming community because I think it's a newer software, so everybody who's using it is a little bit newer.

Michael: 26:32

And that's been really, really helpful. But, yeah, so those are the pros. Again, I think the anti AMI broker is a 100% the right way to describe it.

Dave: 26:40

Yeah. And, you know, RealTest, I'd like to mention here the second requirement I have for any backtesting software, and this is a big one that I think is not widely known and not widely considered. And that's the the ability to add your own custom columns that flow through to the trades CSV file that you end up with when you run a back test. That is that is the number one thing that has saved me a bunch of time, and it that is the one reason I've been able to come up with so many strategies is that part of my workflow. So in any backtesting software, when you first start using it, the tendency is to code up something, run the back test, look at the equity curve, change something in your code, make it better, run the back test, look at the equity curve, and do this over and over and over and over.

Dave: 27:46

That is very, very time consuming, and it's gets you you can make progress doing that, but a way, way, way better way to do this is to run one back test and include all the custom columns you'd need. Just create any rules you might want to create, and then do your analysis in the CSV file. That that alone has saved me, I don't know, probably it's I'm probably 20 times faster, 25 times faster as far as my, back testing workflows. So it's just huge. And I and because of that, I would not consider any platform that doesn't allow that.

Dave: 28:31

And I can tell and you can kinda tell when platforms don't allow that or don't have that as a feature. It's sort of a red flag for me because I'm like, okay. Well, you you can't have done this for a long time and not and not have that feature in the back test. So that's sort of that's a big red flag for me.

Michael: 28:52

Well and, you know, I've been talking about this a bit, and we can talk about it a bit more as well where I've been working to get, a little bit more involved in intraday trading and this is something that we've figured out, ourselves when dealing with real test is that I've been you can in real test. I've been using just, again, the daily candles doesn't mean you can't do any intraday trading. The systems I'm I'm working on for intraday trading all use daily candles anyway. And, yeah, I I didn't even know this until Dave was like, hey, can you try this? And I went in and I was able to, to take care of that as well.

Michael: 29:30

So anything that you can, kind of, think of, you can add as a column. So, then give it to Dave, and he's doing all his wizardry over there with it. But it makes perfect sense and not something that I really thought about too much, but you're right. Even just from an Excel point of view, being able just to take, you know, all of sort by all of your losing trades and just start charting out all of the the different data points. Yeah.

Michael: 29:53

You may notice here's your losers, here's your winner. And and, you know, this is what the MACD or whatever you decide to to put in ends up outputting. And it makes a lot of sense that you don't really think about how important that is, until you find somewhere that doesn't do it. So, I know Amibroker, you said, definitely does, and I know Realtest does. So the other ones, I'm not too sure.

Michael: 30:17

Well, I know, Transpider and TradingView wouldn't, but they're disqualified already because they're they're a single instrument. So any of the other ones on the list that you know can or cannot?

Dave: 30:30

Great Ideas definitely does, and that this was and it it did not used to. But, you know, when I was there, I made sure that was included because just I I just knew how important it was. So that's included. So there's 2 other platforms I wanna mention, that one is very popular, and that's Python, just a Python programming language. You know, a lot of developers are comfortable with Python.

Dave: 31:02

Python has sort of won the scripting wars, I would say.

Michael: 31:07

Mhmm.

Dave: 31:07

And it's incredibly popular in finance. In fact, I was just up in, New York City, just got back a couple days ago where I participated on a panel at an event put on there by the CFA Society of New York. And the topic was Python and finance and how, how valuable it is, how many things you could use it for. It's an incredibly valuable skill to have to program in Python, and it's really widespread. There's so many things you can do with it.

Dave: 31:39

It's very powerful. But you one of the things you could do is back test. You can create your own back tester. There's some libraries you can use to, back test your own data. So you can, you know, you can download data.

Dave: 31:55

You can parse through it with your own Python code. And that's the ultimate in extensibility, being able to do that. I mean, there's no limit to what you can do if you have your own you know, that's a low level thing to do, to be able to, create your own back tester from scratch. And, you know, it's not it's actually not that hard. In fact, my daughter, who is quite interested in trading, she just got her data analytics master's degree from, NC State.

Dave: 32:24

And

Michael: 32:24

Congratulations.

Dave: 32:25

She's quite interest yeah. She's quite interested in it. And what and and she knows Python. I said, you know, I challenged her to create your own back tester in Python and start back testing some data. So I sent her some data to look at, and it's not that hard to create sort of a basic back test with Python.

Dave: 32:47

It's it's easier than you might think.

Michael: 32:49

So I I guess the the big question from the the non trading or the non programming guy like me is going back to the AI, as someone who's just interested in learning more about it and it would be interested in getting into something like that. I am sure the answer will be that, because there is a lot of, a lot of use for it. A lot of people using Python is, again, probably the most popular, if not one of the most popular on Earth. Probably pretty good one to grab one of those AI assistants and and and help you along, you know, check your code or or write some code snippets for you or things like that.

Dave: 33:27

For sure. Yeah. There's, you know, I I remember you know, I've been in the industry a long time, and I used to use a scripting language that was sort of in a bitter rivalry with Python. So it was called Perl. I'm sort of showing my age here.

Dave: 33:45

But I sort of grew up and, you know, at the early part of my career using Perl for just about everything. And, like I said, there was a bitter rivalry at the time between these two languages, and it wasn't clear at all which one would win. And, eventually, Python just really took off because of a I think because of a particular library called Pandas, which allows you to basically, it's like an interface for data and CSV files that's just really easy to use and a lot of support for it. And then I would say probably 10 or 15 years ago, it really gained steam, and and more APIs that you would see that would come out would have support for Python natively, and that was where but all the sample code is being used was that it would always have some Python there. So it really became very popular, and, you know, the whole industry sort of, rallied around it, I think.

Michael: 34:49

So that's interesting. And and, yeah, it's one of those how how deep into the guts do you want to go, I guess. Right? So, you know, if we're talking about, you know, levels to things, when it comes to complexity, something like trade ideas that's completely gooey user interface, it's gonna be great, but you're gonna be locked in a little bit of a box when it comes to the things you can do. And then Python would kind of be the exact opposite of that where you have to get really, really technical and really, really into it, but you you can do whatever it is that you can feasibly think of and then and then be able to program to do.

Michael: 35:27

So kind of where you exist on that on that curve. So I I would say it's, you know, going back to with Realtest and Amibroker, they're on that curve as well where Realtest is probably a little bit more simpler, but a little bit more locked and then Amibroker is a little more complex, a little bit more open. So that might be a good field for which people, look through it as well. Right? Is is saying at what level of complexity do I I want to get involved based off where you are?

Michael: 35:54

Now I would say you should and something that I wanna do as well. Eventually, you should be able to to move between them with some relative ease. You should be able to use multiple with some relative ease, and and you probably wanna always be going to the most complex and most free thing for some bit at some point. But, you know, look at yourself where you are now and where you're coming into trading from. If you're someone who has a whole lot of trading knowledge but absolutely no programming or coding knowledge like myself, you may take a different route than someone who's the complete opposite of that who's, you know, maybe an engineer or a coder by trade, and you're just getting new to the markets, you might enter, right, this whole curve from a different area.

Michael: 36:34

But, yeah, eventually, I think, you know, the answer is you kinda wanna move a little bit, as well, like, a little bit from from everything, because, again, there's pros and cons. So you may go to something like we talked about trade ideas, do some prototyping prototype testing, and then you may, you know, go to an ambi broker or a Python or something after that, because there's none none of them that are perfect might make sense to kinda utilize a few of them.

Dave: 36:58

Yeah. I I think you're totally right, and a lot of it does depend on your skill set. It's and your background, how technical you are, how comfortable you are in Excel, how much you like Excel. That's one of the questions I always ask people. I have this road map product that I offer people where I give you a custom road map of, you know, how what you need to do now to improve and get better as a trader.

Dave: 37:25

And one of the things I ask in there is, how much do you like Excel, and what sort of technical skills do you have? So and, you know, as part of that process, you know, there's there's not one single answer that's gonna be perfect for everybody. There's a depending on your background and your where you are and what sort of strategies you're trading, your time frame, the best back tester for you is probably gonna be different. I mean, I've got traders that I work with that are great at Python and are very comfortable with it, and that's what they use. Others that, actually have started using Python and have migrated to Amibroker just because of the, there's a lot of functions in Amibroker that are specific to trading that you don't have to create.

Dave: 38:24

They're just they're ready for you. So it's it's it all depends on you you and your background and, your skill

Michael: 38:33

set. Yeah. So even, I guess, if you have the capabilities, it might just be sometimes you say, listen. I just want something that's got some some built in, ease of use or ease of functionality and and then just kinda go from there. That makes absolutely perfect sense that, you know, even if you have the ability to do something, you just might not want to.

Michael: 38:52

Alright, Dave. So does that wrap up your wrap up your list there?

Dave: 38:55

Got I've got one other one I wanna mention. So I mentioned the the event in New York that I was on the panel for. I met the founder of QuantConnect, which is a software that it it's actually you can use Python, and it's an event based backtesting software. How is that different? Well, it doesn't you you can use bar by bar data, but you can also, write code to handle events, which is more complicated.

Dave: 39:30

It's more powerful. So I I was I'm pretty impressed with the platform. It's the one thing that it doesn't do that I told Jared is the the guy that was on the panel that I, talked to about it, it doesn't have the ability to add custom columns, which is very curious. I'm very surprised that it didn't. So I and I told him you know, explained why that's so important and, why it's such a a missing hole in in this in the software there, and, I think he's gonna add it.

Dave: 40:04

So maybe by the time this gets released, he he will have added that. But, yeah, it's just such an important piece that, I I I can't use a back tester without that. But so hopefully, you'll have that. That's another one to keep keep your eye on.

Michael: 40:18

Well, and that's also another, I think, large benefit to some back testers over others that really can't be quantified. Right? I mentioned how my experience with the CEO of, Amni Broker, and you're saying that you're able to talk to the person who made QuantConnect and give him some suggestions about maybe add this, maybe remove this. So I I think that's something as important as well as how flexible, that person is and how flexible it is to kinda add or remove things, from their technology. Right?

Dave: 40:50

For sure. Yeah. Like, how long they've been around, how flexible they are, how pleasant it is in dealing with support. I mean, you're right. That's, you don't wanna have friction, and you want you you wanna feel good about the software you're using.

Dave: 41:05

And

Michael: 41:07

yeah, it's

Dave: 41:07

it's an important thing.

Michael: 41:08

It's where is your barriers and where does the software do the best to help with your barriers. Right? Like, I fully know my barrier is not, you know, technical analysis or or looking at the markets. My barriers, I don't really know how to code. So the thing that I will look for will be different for others.

Michael: 41:27

If you're someone who doesn't think you'll need much support, then then that am new broker barrier may not affect you. Right? So, yeah, definitely make sure you're you're diving in, you're going through that. QuantConnect is something that I played with a little bit, but then not knowing how to code, I just completely said no shot. But now with that was cup before JAT GBT, so now that it's there, that's something that when they get that feature that you're interested in, I may actually go back to because I did like a lot of the things that could be done with it.

Michael: 41:55

It did seem to help streamline a lot of that that Python related, approach. So so I'll be someone interested in it. Maybe someone will talk to you in the future.

Dave: 42:06

Yep. For sure. The the the point of this is choose something and get started. That's the most important thing here. Because like I said, to open this, backtesting is a superpower for traders, and there's no question about that.

Michael: 42:22

Well and it also for me, it goes to a lot of the, the problem that people have that they end up trading too much because they feel like they have to work more. And we talk about this a lot. Our trading is not about working more. It's about, you know, you've gotta build your system, then you gotta go for a walk and and and let it work. And that's hard for a lot of people.

Michael: 42:40

A lot of people, I think, that get into trading. Right? Because type a, we wanna work hard. We wanna succeed. I think backtesting is a great outlet for that.

Michael: 42:48

Right? Anything, you know, a, you should do it for all the reasons we talked about at the beginning of this that it's that superpower, but, b, if you're someone you're like, well, I need more things to do, well, great. Right? You've got all the you've got all the work in the world that you can do. You got all this free runway.

Michael: 43:04

Let your systems work. Leave those alone, and then work on refining or building new ones or testing ideas or whatever. You and this is something that you can do anytime. You know, if you're someone who has a job, you can, you know, run your back tests during the weekend or overnight or whatever. But, yeah, definitely grab something, get started.

Michael: 43:22

You know, you won't know what you won't know what it's missing, and you won't know maybe what you need until you get in there and you start playing. And then maybe you'll you'll move on to something else. But, yeah, definitely get in there and definitely get going.

Dave: 43:34

Yeah. I mean, this is really it's really a path to confidence. And like I've said many times, if you it's easy to trade with small size, But if you're gonna trade with meaningful size, you need to be confident in what you're doing and think a lot about how do I get more confident over time with this. And backtesting is a way to do that.

Michael: 43:57

Well, awesome, Dave. That was another good one. We ran a little bit long, but, I think it's worth worth doing, and and I'm not gonna cut off a good conversation. So, as Dave teased, soon we're we're gonna have another interview on with one of the the CEOs of 1 of the back testers. But I think this is, an interesting route that we'll we'll go down for a little bit about talking a little bit with the tools of trade and and and what you need to get into it.

Michael: 44:20

So I had a lot of fun. If you guys had a lot of fun as well. Right? We appreciate all the feedback and the comments. It's it's just, amazing, as, the more I hear.

Michael: 44:32

Right? It's just been glowing. People are are are super happy, and we're always super happy to do it. I get look forward I look forward when we get to report these. So talk to you guys next week.

Dave: 44:42

See you next week.