Using News in Your Trading Strategy
Okay, everyone. Welcome back to Line Your Own Pockets. We're kinda continuing to go down the social media news kinda journey we've been having for the last couple episodes, and not talk so much about the social media side anything, but talk about actual news and news events and how we can potentially incorporate, this into our trading. We know that news moves the markets, but, being systematic with it is complicated. This is actually why I'm I'm really excited to hear Dave talk about this because I know this is part of his process.
Michael:It's not part of mine yet, but maybe I'll learn something, and I'll I'll add it in. But, Dave, you have got a you've got an announcement for the people I hear.
Dave:Yeah. So I think as we record this, we're gonna top 1,000 downloads for the podcast. So mind your own pockets, 1,000 downloads, the first, I don't know, 5 or 6 weeks. So this is pretty exciting. That's awesome.
Dave:Thanks, everybody, for listening. This is, keeping us motivated for sure.
Michael:Yeah. And it's it's you love to see it. Right? Also, as data nerds, we're definitely someone who's gonna look at at the data behind it. So it's it's good to see.
Michael:When we were talking about this, me and Dave, I was thinking, hey. This would be fun. I I like talking to Dave anyway, but it would be maybe a little niche, and then we'll see how much it's picked off. And to be blunt, it's been far outside what I thought would be, so soon. Right?
Michael:The fact that we've only just started getting going with this, I think, is super, super cool.
Dave:Yeah. And I love the feedback. I mean, I can tell that some of the topics are resonating with people and helping people, and that really kinda gives me a boost. I love to help traders because, you know, I've gone through this so many times, and I just it it just pains me to see somebody going through a trading career and not knowing about something that could just make their lives so much more efficient and so much better. So yeah, that floats my boat.
Michael:I was doing just this is completely off top before we get into it, but I was doing a CMT call, yesterday morning, with the people over in India. And they have a trading competition, and they bring on CMTs to kind of mentor and teach them throughout the trading. And the quote I gave, which I I stand by as someone who's still kinda going down the systematic trading path, not fully down like Dave is, but I'm working on it. But was, if I had, you know, learned and implemented systematic trading earlier in my career, I think I'd have way more money in my pocket and way less gray in my beard. And then that got a that got a good laugh because I I do really believe that it it is true.
Michael:And, yeah. So we're out there spreading the message, I guess, far and wide across the world at this point, which is great.
Dave:Alright. Awesome. Let's get into it. You know, last time we talked about well, 2 times ago, we had the interview with doctor Syed Kazimpour and talked about Finfluencers, followed up last week with the social media. And the recurring theme there for me is I hate the news.
Dave:I do not follow the news. I do not follow politics. I think it makes people like, I think it's literally bad for you. So, and the reason is what it boils down to for me is, yeah, you can watch the news. You can see some stuff going on, but what can you actually do about it?
Dave:You can't there's very little you can do about it. But here in this episode, we're gonna talk about some stuff that you can actually do with the news, like, how you can use it to improve a strategy. And,
Michael:I I think that's the that's the biggest thing because I was talking to people about this. Again, I'm Canadian, so our politics works a little different than anything else. But, you know, I was at the pub with my buddies. Like, oh, did you see this debate or that debate? And I'm like, abs absolutely not.
Michael:And that that was the same point that I gave. I was like, I I vote in Canada when the times of vote, and how it works for us anyway is you vote for a local candidate. And then depending on what happens with enough of those votes, in in Ottawa, it decides who's the leader. And I'm like, so I can only vote for that guy. So I only care, right, about this one thing, And I'll only look at it when it's time to make a decision.
Michael:So it's kinda like what we're talking about with trading. I know people, and I used to do this, so I'm I'm kinda outing myself, who I would spend, like, 12 hours a day just staring at charts tick up and down and up and down for no there was no point at which I was gonna make a decision. This is one of the things I think we talked about in the first episode, is that setting alerts and walking away is, like, the easiest form of of systemization. Right? If you're someone who just, I wanna buy this stock when it hits 10.
Michael:You don't need to watch it from 12 down to 10. You just need to be alerted when it hits 10. And, yeah, it's the same with news and and politics and all this kind of stuff is that if I can't control it, I'm not interested in it personally.
Dave:Yeah. Yeah. I think that's the right way to be. But, yeah, let's talk about news because you're right. It it moves the market for sure.
Dave:But the problem with news is it's so subjective. When you go read articles, you see, press releases, you see, earnings announcements. There's a lot of discretion that goes into how the market's gonna react. Yeah. Not very, predictable.
Dave:So as a systematic trader, you're typically trading the response to the news. Mhmm. But that doesn't mean you can't use news to improve strategies that you're already trading. And that's what, you know, once you get a strategy that works, there's lots of different ways to improve it. And that's what a lot of traders don't realize is when you get an edge, there's tons of ways to improve it and make more money with it either by reducing trades, adding trades to the system, creating an adjacent system from what you've learned with your main edge.
Dave:So Mhmm. That there there's a lot of topics there that I wanna get into in future episodes, but I wanna talk specifically now about how you can use news to improve potentially improve an edge you already have. So one situation where I've used this is a particular strategy, and I noticed that, the trades were, you know, I would take trades long and short in this particular strategy. And I noticed as I started looking at some of the news for some positions I was already in, there would be, like, a downgrade or an upgrade that morning, and that was moving the market or, you know, moving the causing the move. Mhmm.
Dave:So I thought it might make sense here or if if I'm long a stock and, there was a downgrade earlier in the day, That might not make sense, right, depending on the strategy. That's sort of the opposite of what, my direction. So one thing you can do with news is search the subject line, and you could do this programmatically. There's APIs where you can use to, to interact with the news, query for news from a specific symbol.
Michael:Yep.
Dave:And, you don't have to read the entire body of the message. A lot of times you can use you you could just do a a very simple match on the subject line to determine, okay. Is this an upgrade or a downgrade? And if so, skip that trade if it's the opposite direction of what your, system is doing. So that's a, it's a little bit nerdy.
Dave:It's a little there's some work to get that done, but it's a pretty useful thing to be able to do. And once you get it working, it's sort of programmatic and can be automated. So yeah. What do you think about that?
Michael:I like that because and again, this is something I don't use at all for my my process, the swing trading or investing or even any of the day trading algorithms I'm working on. But it makes sense as a as a refinement tool. Right. So what I was not worried about what I what I thought a lot of people when they think of trading the news, I think they think of, even systematically, I'm gonna have some sort of system and LM or whatever it is. Read the headline, determine whether it's good or bad news, and then buy or sell based off of that.
Michael:And, right, I think we both know that that probably isn't gonna yield anything worthwhile in in in the long run. But I I do like it as a refinement to basically just say, okay. There may be, may not, but there may be a headwind in my particular trade based off of if the news is is yet positive or negative. The the one thing I'll I'll kinda ask you to to counter this is that I have found the only news trading that I've ever heard anyone successfully do, and I've heard a lot of people do this, is kind of the exact opposite for that. So, like, you talked about trading the reaction to the news.
Michael:So, Kenny Glick. Right? He's, hit the bid. I don't know if you've ever had the pleasure. He's very interesting fellow.
Michael:He's been day trading for a long time. One of his main things is I'm gonna look for good news, quote, unquote, to come out on a stock and the stock to fade, or I'm gonna look for bad news come out and the stock to rally. And that ends up being a trigger. And it's the same thing with, Jason Shapiro, who, is someone that I look at a lot who uses this, commitment to trader data. And his idea is that if he can say everyone is long this futures contract, I'm gonna look for a short.
Michael:And his trigger condition is when good news comes out and that market fades. So, I'm interested to hear because, obviously, you tested it. How much of a boom was it to that portfolio? And did you did you see any of that kind of programmatically where if the thing didn't respond well to the news, it ended up being a better kinda overall trade even if the win rate or the hit rate was lower?
Dave:Yeah. Well, the beauty of this approach is you can you're essentially turning a news item into a binary yes or no, and you can include that in your back test and then see what works and what doesn't work. So, you know, one thing you could do is merge that data in with your back test and see. So one of the things I'll do is, you know, did was there a downgrade, or was there an upgrade today? And I'll include that as a 0 or 1 in my trades list in my back test.
Dave:And then I'll go back and see. You know? In my tools, I'll be able to to query for the subset that has upgrades or downgrades and, you know, query that by direction. So I can I can see very clearly? Here's the equity curve with it in there.
Dave:Here's the equity curve with it with it not in there. So you're really you're sort of making an indicator out of an you know, a custom indicator out of your news items in a way that you that matches the theory you have about your strategy. So, you know, I'm not here to say, hey. This is gonna work or not work in your strategy, but what I'm trying to tell you is, hey. This is a process that you can use to get your answer yourself.
Dave:And once you find that answer, you know, maybe it's maybe it's difficult to, imagine doing that in real time programmatically with the system you're creating. But, you know, you have all the time in the world to back test. You could pull that data at your leisure and see what it would be like. And then you could use that to figure out and motivate yourself. You know, if the data looks good and it's worth doing, I mean, it's a simple cross benefit analysis at that point.
Dave:Like, here's how much my strategy would improve if I was able to get this data in real time. Then you can cross the bridge, but, okay. What do I need to do to get it in real time? How do I do that? Yeah.
Dave:So it's obviously worth something to me.
Michael:So that's, again, interesting. So I'm starting to think of of, again, some of the, some of the strategies I'm running now with it. But, that's for very easily binary news. Right? Upgrade, downgrade.
Michael:It's very easy to go, okay. This one should be good, and this one should be bad. Do you then ignore news that's kinda muddier? Or do do you do you move on from that, or do you just say, hey. These are the things I'm looking at.
Michael:Because I could see, like, a news event of, you know, you lay off some staff, and it seems like it sometimes that's a great thing, and the stock's gonna do super well, and sometimes that's that's garbage. But that's gonna be something that's a little bit more, in the weeds and nuance. Is that a I'm gonna try to figure out a way around it, or is it I'm gonna ignore it because I can't I can't make heads or tails?
Dave:Yeah. I think I would ignore that type of stuff. This is for stuff but the best way for this kind of stuff to work is to, you know, watch this the trades that your system makes, and then go look at the news and see if there's anything that makes sense that you can match on that was that's gonna happen a meaningful enough, time to be able to show up enough in your back test, and you see that it's common enough and it makes sense, then it might be worth, putting it together. The and, you know, the more common things that are gonna resonate with a lot of traders in a certain way that that that have a direction are typically more likely to be, you know, correlative with your, with your strategy. So stuff like, like you mentioned, probably not, but stuff like upgrades, downgrades, offerings, that would be another one that, you might wanna look at.
Michael:Yep.
Dave:So but I like what you're saying. You know? It is a fundamental thing. Like, the first people thing people think of is, okay. Yeah.
Dave:Let me trade this news. You know, have a system based on looking for this type of news item and do something.
Michael:Yeah.
Dave:That's probably not the best approach. And it's well, it's I'm not saying it can't be done, but this you know, using it to you know, using specific news events to exclude trades from your already working system, that's something that is worth exploring.
Michael:Yeah. And I could think of a a couple new sets that could be binary. Right? If you just did earnings meet. Right?
Michael:Did they did they, beat or did they miss? You could do, you know, a couple things that you could turn into a binary event. But I I think that's where kind of my brain stopped caring about news. I never thought about that. What because I always just looked at it and said, okay.
Michael:Well, I don't know how to interpret, right, a layoff in a company because it right? Is that, a layoff because they're gonna get leaner and meaner and make more money, and that's something we should be interested in. Or is this layoff because, right, they're, you know, they're struggling and they don't have rights. So there's a lot of layers behind that. But I guess what you're just saying is if you just stick pretty surface level, find things that you can turn into, some sort of binary event, and then it becomes easy then to run that against an already existing set of of trades that you have and say, does this make my life, does it make my life better or worse?
Dave:Yeah. So there's another related technique that I've used in the past, and I use this in a couple of my systems now. So in addition to you know, you have the ability to look at the subject line. You have the ability to look at the body of a news item. I don't do any of that.
Dave:I only look at the subject lines.
Michael:Mhmm.
Dave:There's another way you could look at this, and what I've done is, you could count the number of news items that appear on a given day. K. So Yep. This would this is good in gap strategies. Like, sometimes the stock is gapping, and, you know, you you you know, that's a pretty common strategy that people look at.
Michael:Yep.
Dave:In the event that there's several news items for that, that might be a proxy for something that might be predictive to your model. So that's another way to look at it. And you could even get fancier with, you know, looking at the average number of news items that appears for a stock.
Michael:That was gonna be my question. Yeah. Because if you took something like NVIDIA, right, if you took something like NVIDIA, it's it's in the news a 1000000 times a day every day regardless of what happens. So so you're thinking almost like, like an oscillator in some respects. You're comparing, is this a spike, dramatically enough on on top of a normal baseline?
Dave:Yeah. And, you know, a lot of times for you know, you you mentioned a video. Of course, there's a lot of news for that. But most symbols, there's not a lot of news for on a given day. So the fact that there is a news item at all for any symbols
Michael:Mhmm.
Dave:Is a notable thing. So, of of course, it's gonna matter a little bit less when you get to the bigger names where there is more, there are more news items each day. But for the vast majority of stocks that you might be trading in, there's, you know, systems that you can create for. You're gonna you know, news item is a notable event that and just regardless of what the news item says, you get a programmatic view of the status of the news items just by counting them and and doing a simple count and determining, you know And that's a numeric number you can include in your back test very easily.
Michael:Yep.
Dave:And, you know, it it's even better than a 1 or a 0. Right? You're gonna have a a a an integer with a a number there. So the more news items, perhaps the the stronger the news event that was that day, and you can easily see that from backtest. That's
Michael:a that's a really interesting one as well. Now you'd have to, I guess are you looking at a single news source? Because there's a lot of this thing, there's a lot of news copying that happens out there. So one article comes out, and then, you know, maybe it gets copied 30 times across 30 different. So you'd need either to focus on some primary news sources or have some sort of way to say, no.
Michael:This is just the same news article, right, copy and pasted 7 times. Yeah.
Dave:It's possible. You could do some filtering to to detect if it's the same thing. But, you know, then I would argue, well, if it's being copied that often, then they're copying it for a reason, and somebody thinks it's important enough to copy. So there's probably some validity in just you know, I I think there is validity in just purely a naive look at the the total number of news items.
Michael:That's a very interesting one. Now that gets a little bit more, you know, programmatic than I think some people could do. But if you have those kind of chops and you can go and and pull that data and, you know, we're not gonna get into kinda how you do it personally, but there's probably a lot of, what do you think, public sources out there that you can you can grab from that
Dave:Yeah. There's, there's one called Tingo. I think it's t I I n g o. There's they have an API for the news. There's several.
Dave:There's there's a wide variety of quality of these APIs and the and the news content. So, so so it is notoriously kinda difficult to find a a great news source, and none of them are really perfect. But anything you can get to to try to improve your system is well worth looking at. And the more, you know, the more trades you have in a system, the more potential this has to improve it. Yeah.
Michael:Well, and then that would be that would be similar to any any indicator. Right? If I wanna refine a system down to see if even just a generic RSI or something like that is gonna affect it. Yeah. If I've got 20,000 trades, it's gonna be way easier to go through and and look at that, a little bit tighter.
Michael:Now, kinda lastly, and this is something that I'm kind of interested in, and I know you want to leave to last, but, have with I'm a much bigger AI guy than you. That's kinda why I'm checking myself. But, any work done that you know of, or do you see any merit in the ability for some of these LMs to be able to read and come up with some sort of sentiment based off of, those more nuanced article articles that I was talking about sooner where you could potentially feed it a bunch of news articles and say, hey. Do you think this is positive or negative? Then using that as a a binary input and then and then testing that as well.
Dave:Yeah. I think you could totally do that. Yeah. There's, there's ways you could come up with your own sentiment analysis of news. It's actually I mean, there's plenty of tools to do that.
Dave:That there many of them are free. You can use Python or R to do sentiment analysis on news items. The beauty of that is there's you know, we've talked about putting, you know, basically reducing news items to an indicator. And the first one we talked about was 0 or 1, does it appear in the in the subject line? The other one was, you know, the number of news items.
Dave:Mhmm. Instead of an analysis, you could actually come up with a score. So it could be basically a double or a float. It would be a score that would, that'd be, you know, that'd be pretty granular, what you could come up with there, and that would be even more powerful to include. And there are plenty of companies that do offer this.
Dave:They'll give you they'll they'll do sentiment analysis for you and and give you a feed of it so you can retrieve it easily. And, I I think there's you know, I don't use that in any of my models, but I've looked at it and it's I'm imagining that over the next couple years, I'll be including it in some model. I I don't know what it will be, but my guess is the way things are going and the cost of those models, and I think they'll probably continue to come down and be more prevalent with more competition. I think I'm guessing that I'll be using one with, the sentiment in there that would make sense.
Michael:Yeah. And that's the one I think that interests me as a longer term trader the most because just from experience and from my discretionary trading, sentiment generally follows price and catches up to price. So the ability to find something that, maybe is setting up technically but is universally hated. I think there would be a little bit of more of an edge to that than there is to to to something else. So, for example, if, you know, if a stock is rallying and everyone loves that stock, that may not be as much of, something that I'd wanna participate in because the news is already disseminated.
Michael:But what I would what I would be interested in to see is is a stock that is universally hated, and all the news is coming out is is negative and negative negative, but then the price is not then reflecting that. I think there would be a very interesting, divergence setup because and I'm just using an example that we're seeing right now with, with all the macro study that I do in China. Right? Everybody hated China. They were taking it out of ETFs that were world ETFs.
Michael:It was the worst thing in the world. And then over the last month or so, we've had this in insane rally. That kind of thing is interesting to me where I'm I'd be interested in building indicator and say, okay. Is all the news for this thing terrible? And then it's moving higher because that would be again, just means that, there's probably some people caught offside.
Michael:So I I like that idea very much when it comes to longer term type moves.
Dave:Yeah. And, you know, even what you're describing there is still something that would improve a current strategy. I don't I find it hard to believe that you could find something where sentiment analysis alone would be the signal.
Michael:Mhmm.
Dave:That that I don't feel like has legs, or I haven't seen anything to where that would make sense. Just because, you know, if you think about the news I mean, we're we're talking about the news here. Yeah. It's not it very often is brought up out of thin air. There's no hard data.
Dave:It's not we're not looking at price and time and volume here. We're we're talking about the news, and often that, you know, is is opinionated. It's can come out of thin air for any reason that anybody, you know, any PR firm comes up with. So I think just by the very nature of it, it's, it's gonna be a little bit vague, a little bit nebulous, and a little bit less, predictable and, you know, numbers based than what we're really accustomed to and what we're really comfortable with.
Michael:Yeah. It's, so, like, one of my I think, a perfect example for that for one of the systems that I run, is short float. So it if I look at it, and this is basically a simple trend following system, the higher the short float, the better. But if you remove short float entirely, it's still a profitable system. So the idea is that the, how much a security is shorted isn't indicative that it's going to move higher, of course.
Michael:Right? Because generally speaking, if something's very short, it's probably a junk company, and it's probably on its way down over the long run. But those that divergence is when things becomes interesting. When you take something that's, like, a stock that's hitting new all time highs and it's heavily shorted, and then you combine those two things together because you know that, yes, this is the the sentiment of the market. I think short float is just a numeric sentiment kind of indicator, and then there's the price action that that's contract to it.
Michael:So, yeah, definitely agree that anything that you do when it comes to news, it has to be first, you have some sort of technical system that makes sense and and is profitable, and then you just look to see how this is additive or or whether it's additive to to that overall system. I do think the trap is when people go, okay. This is good news. I wanna buy it, because there's just so many layers on yes. It's good news, but did the market expect better news?
Michael:And and and what is, like, all the the layers that could occur there?
Dave:Yeah. I mean, most people most people have trouble thinking in ranges. They think in binary. So, you know, they're looking for hard and fast rules. Okay.
Dave:When the news says this, do this. But, really, I think the beauty of trading and what keeps me excited about it is that there's so much edge out there and opportunity for edge. And once you find an edge, the the the ways to improve it and the bar for improving it is so low. I mean, it it that's it just it's constantly there there's constantly something you can be looking at to make a significant difference in your strategy that's meaningful, and it's it it just it's never not it's it's never boring to me. It's Yeah.
Dave:And that's because you're so close to real meaningful edge and improving your edge at any point. I mean, I see traders all the time that they're so close. They're trading something that that's working well, but can be so much better with just a just some tiny tweaks that they're not aware of. It's, it's just an exciting industry, and that's that's really the reason I love it.
Michael:It's a, yeah, it's a giant puzzle. And, right, that's that's the whole point of it is that you spend, especially and this is, you know, go back to what I was talking about in the beginning. Especially when things are systematized, the the puzzle becomes that much more interesting. The I think the frustrating part from what most traders is going in every day and actually battling the market themselves and then not having any time or energy or mental energy to then improve things. But as soon as you systematize the entire trading part of things, your brain really does open up to, okay.
Michael:Now I can think more of of what this next step is. So this this news chat has been absolutely great for that because, again, it's not something I've just ignored entirely, and now I've got this whole thing to walk around the woods and and ponder.
Dave:Yeah. I think, you know, a lot of people think of systematizing something. They think that's robotic. Creating is really very much a creative endeavor, and you're very much like an artist trying to come up with ideas, being creative, trying to come up with things that that people have never thought of before, building on what you've done in the past, developing your skills. There's really there's a a surprising my my wife is an artist.
Dave:Actually, did this thing here. But there there's so much commonality that artists and other creatives have with traders that most people just don't realize. And it's it's an important part of having a long term edge as being, you know, how creative can you be in working on being more creative.
Michael:Yeah. And and, you know, the in this chat I had with the CMT and and the students there, I the conversation came to one one of the questions was basically, you know, how much time do you spend throughout your day as a more systematic person trading? And the example that I always like to give is, like, a professional athlete where, you know, if you're a boxer, you're hopefully only spending 1% of your day actually getting punched in the face. Right? The rest of the time is is working on all these other other things, and that's the yeah.
Michael:It's the same with this where if you have a, a methodology that you do, and hopefully that doesn't take a lot of your time, then, yeah, the the bandwidth that you freed up to do those more creative things becomes way bigger. And, that's the kind of trap that I think a lot of traders are stuck in that I think we we do a good job and and will do a good job of pulling them out of. And there's no value sitting there watching every tick. Right? And, again, if I could have all those hours back, I'd be I don't know.
Michael:I'd probably be, like, super jacked and really rich. Yeah.
Dave:So, yeah, I guess just to sum up here, I wanted to touch on 2 thing 2 companies that do sentiment analysis that, I just wanna mention that people for people to look at. Mhmm. One is called Likefolio. So that's a company that, a friend of ours, Andy Swan, started with his brother, I think, and is still involved with. They just sent him an analysis.
Dave:I've looked at their data, and it's it doesn't cover the entire market, so it only has the top names. That may have changed, and I'll, I'll reach out to him just to see by the time this episode airs. But Mhmm. That's not one one firm to look at. The other one is a company called Financial Modeling Prep, FMP.
Dave:Okay. It's a source for a lot of API data. Fundamentals, news, sentiment analysis they have. That's another one to look at. That there's a lot it's a surprising amount of data that they have available there.
Dave:And if you're a, you know, a data nerd like us, that would be one to take a look at. We should put that in the show notes.
Michael:Yeah. Yeah. I'll copy those. I think that's great. And and I think this was a great chat.
Michael:I I love these that are completely outside of of what I do. Right? And then that's a good rabbit hole that we've been going down with the social media and the news and all that, as someone who's been a technician and then working to systematize what I've been doing from a technical standpoint, it's just something that I kinda left on the cutting room floor 20 years ago. Right? And and it's kinda cool to see that, there's some uses to it that I should I should start thinking about.
Dave:Yeah. I think, you mentioned cutting room floor. That's there's a lot of stuff that a lot of traders leave on the cutting room floor that has a lot of value. And there's at least a couple episodes we could talk, about cutting room floor stuff that traders leave. So I'm I'm glad you mentioned that.
Michael:That's awesome. Well, as always, more ideas for more episodes, but we should wrap that one up for now. I again, I appreciate we have hey. Check now. Have we crossed a 1,000 just while we were doing this?
Michael:Or Yeah.
Dave:I was just gonna check.
Michael:There one more time.
Dave:Yes. We have. A a thousand one now. So pretty awesome.
Michael:Well, congratulations, Dave. So, yeah.
Dave:Congrats. This is great.
Michael:And we appreciate we appreciate all the feedback we're getting and everything. And I look forward, every day that we go to record this. So, I'm having fun, so I don't see any anything that can stop us now. So, thanks for tuning in, and see you guys next time.
Dave:Yep. See you next week.
Creators and Guests
