So I’ve seen a few of, when I scroll through the Instagram, there are people who have probably no experience in trading talking about how you can use AI to create for yourself a trading model. And I would just like to show you, if I had an ima, I know everyone’s in love with this blank space right here, which isn’t going to change anytime soon, but there’s an imaginary red flashing light when people don’t have an understanding of markets. There’s a lot of blind spots that they have when they’re trying to put these models down. It’s no doubt, and it probably doesn’t surprise you, that these people are more marketers than they are actual traders. And so what ends up happening is there’s a lot of things that they don’t get from having a posterior, a knowledge like all of the mistakes that I’ve made. So it’s very doubtful for me that they’re going to step in and all of a sudden find a cure for cancer as it relates to trading.
But it got me thinking that you can probably program a bot to go scouring the markets to help you with research, and that can effectively save you time. Because I don’t like the idea of having people sit in front of the screen for 12 to 15 hours a day. I don’t think that’s healthy. Two studies have shown, and you can look at Andy Huberman for this on hi, the Humer Huberman Lab podcast that he has, which is excellent, where he talks about if everything that you do is 12 to 18 inches acro or in front of your face, you can become myopic or nearsighted. So you want to make sure that you’re spending a lot of time outside. He says two hours a day looking out into the distance. It strengthens your eyes and the muscles in your eyes as well. So what I can say about back testing is whatever you do with ai, if you go in that direction, or whether you’re using the premium testing engines like Mechanica or Trading Locks, or if you’re using it something like, I know Trend Spider has one, but I think it only allows you to test one idea at a time, and it only goes back two years.
That’s not enough data. You want to make sure, especially if we’re coming into a time of higher interest rates and the talk of recession, that you can test other periods of time that had similar circumstances. You can’t go back test something from 95 to March of 2000 and think that that’s a surrogate for what we’re going through Now. There’s eight names that are driving the market for the love of God. And now that you have Nino hitting the futures markets, reversing trends or what have you, fear of the boogeyman. So what you want to try to do is test when there were other difficult periods of time, how did your ideas go? Because you don’t want to get into data mining or curve fitting. Cause again, it gives you a fault. It could give you a false sense of security. You see, and I like the idea of confidence, but not under the guise of a false sense of security because then just as soon as you go
Try to put the thing to work, you’re going to get very, very different results in real time that you got when you were in the safety of a practice lab, so to speak. Two, you have to remember that if you can see it on the screen, it’s a survivor. What do I mean by that? Well, if you were going to trade a model and put yourself through the difficult times of like oh 7, 0 8, for example, you’d want to see all the names that traded in and around that period of time from say, I don’t know, 2000 to 2020 and see what happened before oh 7 0 8, what happened during that window of time, and then what was the aftermath? Because if your model would’ve had, you’ve gotten long on pullbacks with names like Lehman or Bear Stearns, for example. What happened with those names? Unless you were short, and I didn’t even know. I think the government stepped in and said, you can’t borrow and sell short bank stocks anymore. So it screwed up everyone’s models, which kind of goes against the nature to me of the US Constitution, but let’s not let that get in the way of good trading, I suppose.
Make sure that you add back all the things that blew up because otherwise the data universe that you’re looking at right now are just survivors. And if you’re going to run a system, you want a minimum of 10 years of time, preferably 20, and you hope that 20 year window of time is going to be robust enough to have included periods that were very, very difficult for stocks. Because that to me would be you’re onto something, right? So I look at it this way, if you ran a back test and came up with a hypothetical model and you were looking for funding, obviously what we’re looking for is the integrity in the model. Is it robust? Is this just you trading Google since it’s I P O? That to me doesn’t really count, you see? So I’d want to look and see how did it do during the more difficult times?
You have to remember Victor Sp Brando’s first two years of trading were a bear market where the s and p got cut by 50% over a two year period of time. So you kind of had to know how to short sell or had to learn how to do it pretty damn quickly because that’s where people were and that that’s how people were making money. Also, names that got taken over like First Tennessee, F T E n or Bkb Bank Boston merged with fleet so they don’t have to blow up. But you would like to see the Enrons, the Bear Stearns as well as the takeover candidates. How did they behave before they got taken over? Maybe there’s something that you can find out. I myself don’t know. They’re usually surpri surprise attacks in black swans in that regard. Just like Carvana coming out and catching all the short sellers, why they’re still short after the thing is down, 99% is beyond me. That’s kind of stupid trading, stupid risk management on their part.
So look at that and you can get that data. You’ll probably have to pay for it, but it gives you a better sense of confidence by the time that you’re done. Because if you had a model, like I said, that was buying things on a pullback and those names eventually went broke, you’d want to know that. But nothing, it’s not bad if you’re only risking 1% of your capital. You see, not a big deal. But what happens when those names start to trade and then they get halted? Then you can’t get out of the position. You see, because at that point, it doesn’t matter if you have puts or calls or if you are trading the equity outright. So to me, you want to be as thorough as possible. And if you cut corners, I understand it, time is of the essence. You feel the great sense of urgency, but you also want to be very, very thorough.
Thorough in your actions, right? Because then if nothing else, it’ll give you a truer number in a robust system that can work across all different types of seasons. And when you see those results and you start to practice, you see that your results come in within model. That also can build your confidence. And it’s hard to really be good at anything in life or get trading if you don’t have any confidence in what you’re doing. You don’t really have any confidence in yourself. So the point of the message today is to make sure that you don’t cut corners. In this case, you might not think to add those names back into the data set, even though they go off the board, you can’t trade pork bellies anymore. That doesn’t, doesn’t mean that you shouldn’t back test pork bellies, right? You’d want to find that data, add that data in and see how your model would would’ve done with that, right?
Maybe there would’ve been a time where many of us stopped trading pork bellies for a few years before they were delisted because there wasn’t enough volume. And then volume and liquidity of course are two different things, but you can only go by what you can see for the most part. So you have to position size accordingly. So that’s a pro move as far as I’m concerned. If you really want to know your craft, you see what I’m saying? You can cut corners and maybe even for the majority of you, what does that mean? I don’t know, 60 to a hundred percent of you can kind of get away with it, but all you need is one or two of these names to really shit the bed, so to speak, to give you a bad drawdown number. And maybe you don’t even lose money. But what you had made during the year gets taken out by a name that you otherwise wouldn’t have been in had you looked and seen at this other data, right? From the names that were delisted, either because they went out of business or they merged with other companies. You see what I’m saying? And that can mean all the difference. So the market, yes, is full of certain type of, I don’t want to call it quality, but you can say from a statistic standpoint, if you can type in the ticker and get
Numbers, those are survivors. And you don’t want your universe of names to just be survivors because I know someone’s out there right now looking at the Russell two K, the s and p 500 and some version of the nq, whether it’s the Micros or the minis, doesn’t matter to me. And I know that there’s another group of stock traders who are just trading in and around the eight names that are really driving the markets right now. The markets are not broad based. The rallies are not terribly broad based right now. So you want to be super careful and know what does that look like? What does that look like when you have an unhealthy marketplace where only a few names, because all it takes is one piece of bad news and what happens when those eight names don’t work anymore? You see what I’m saying?
What are you going to start selling options then? This is what I’m talking about. So you want to know this before you start putting risk on, and don’t worry about missing a couple of days. Don’t worry about missing the rest of the summer to do your testing. You’re not missing anything. Markets are going secondary. Trading’s going to be around long after we’re all gone. So to me, it’s better to have a sense of who you are and what it is that you’re trying to do and have an honest look at the numbers if you’re going to do that type of testing. If you’re discretionary chart reader, good luck, right? Because that’s all to me. It’s appreciating art or poetry for some reason. Anyway, I appreciate the comments. Keep sending them in. I will address them as I can if I have something to say about it.
I usually like to speak on topics where I not only know what I’m talking about, but it comes from having started on my own with no help and having made lots of mistakes in and around that space so that I can point those out for you and show you how I had to evolve. So you kind of witnessed like an open spoken diary of the things that I’ve had to learn over time and I’m still learning, right? I’m still learning. You still learn about yourself. Then you have to conjugate that with your tolerance for risk, right? If there’s always a K Y C, know your customer. When you trade for yourself, you are the C. You’re the customer. And so you really need to know who that person is. Anyway, appreciate you being here. Please consider to like and subscribe. I’ll have ganja here tomorrow, and then I’ll be back Thursday and Friday by myself. Okay, see you tomorrow with Brandon.
Click here to get your free copy of The Inner Voice of Trading audiobook.
This is an automated transcript