Interview with Steve Bravo from Massive (Polygon)
Hello, everyone. Welcome to another episode of Line Your Own Pockets. We got a special guest here today, and it's Steve from Massive, formerly known as Polygon, which is something we're already joking in the green room. I'm sure we'll talk a little bit about. But Dave's got a little origin story, I guess, to share.
Michael:He's been talking more and more about about Massive on just things that he's doing, and things that he's playing, and things that he's building with. So why don't you take it off there, Dave?
Dave:Yeah. So I remember hearing about Polygon at the time. It was an introduction, you know, like, I remember hearing that this is a new company, market data, and I didn't think much of it. I I checked it out, and hell, okay, it's some APIs, that's pretty cool. So...
Dave:And around that time, I was using a tool called Q Collector, and this was a tool that used IQ Feed or eSignal to download market data. And I would use that into my pipeline for back testing and just downloading data and keeping data up to date. And that tool was very old, written for an older version of Windows. I had contacted the creator of it, never got a response about updating it. Like, there hadn't been updates in a long time.
Dave:So I was starting to worry about this. A couple of the traders I was working with were starting to worry about this. And around this time, Claude became really good. And I was like, okay. Well, let me...
Dave:I need to recreate this queue collector just for myself. So I thought, man, this would be a really good use of claw to create something that replaces it and is better than the Q Collector thing I was using. So I created this pipeline tool. It's available free on my website now. There's a version for IQ Feed, and there's a version for Massive slash Polygon.
Dave:And I was surprised the downloads for the Massive version are, like, 40% higher than the ones for IQ Feed. So that kinda piqued my interest, and I was like, well, let me let me reach out to to these guys and see if we can get somebody on the podcast. I think it's just like we've been talking more and more about it. It's it's... And it's gotta be such an interesting business, so I'm happy that you're on here, Steve.
Dave:So why don't you introduce yourself, tell people a little bit about your background, who you are, and where you're where you're from?
Steve:For sure. Steve Bravo. I always joke that I was born and raised in market data in many respects. It all started back in London. I was finishing my graduate degree, and one of my program mentors basically sat me down and said, what do you wanna do with your life?
Steve:I'm like, that's an interesting question. Maybe, you know, manage somebody's capital, do something like that. And he, just puts a piece of paper in front of me and says, like, look at this. And I go, alright. This is a piece of paper.
Steve:Has some company information. We're looking at Apple, some company description, pricing data. I'm like, you know, company information. He's like, yeah. You can sell that, and people are gonna buy that.
Steve:And I go, okay. So I'm like, that's an interesting kinda take. So from there, I got the opportunity to join SMP, and the world of data for me was just very siloed to the incumbent market data vendors at the time. You know, your, you know, your Bloombergs, your Refinitivs, your FactSets, and and that side. I wasn't aware of this whole different world of different types of providers who do niche things and provide different types of data to the world.
Steve:And as I sort of propelled myself into that world, I saw like, wow, there's so much out there on so many different types of, you know, data points and, like, company data, fixed income data, like, equity data, and all that sorts. So from there, I, you know, grew my career out focusing more on the buy side and quant community, getting them access to just machine readable data through FTP files and any kind of quant solution that they're using to to manage their capital, and hopped around from London to New York to Chicago. And then from there, I then came into a different realization that I just knew the tip of the iceberg when it comes to, like, data in general. I have no clue what's going on in the pricing world. What are these exchanges are doing?
Steve:You know, why is it set up that way, etcetera. From there, I actually was able to grab onto an opportunity to join Opera, which is the consolidated tape for the options industry in The US. Now, my role there, I was wearing a, you know, few different hats, but was in effect to head up their licensing department, talk to different folks who are interested in consuming options data, either redistributing it or using it internally. But that opened up a totally different world on not just, like, the data itself, how it's structured, how it's delivered, but all the nuances that come with folks that want to even start to consume it and use it internally with all these, like, regulatory policies, these different fees that they have to abide by, and some things that are, you know, common knowledge to them, not less so common to to the broader audience. Right?
Steve:Now my story with Massive, formerly known Polygon, actually starts, you know, in its early days, more or less. I got introduced to their their platform when I kept on pitching different ideas that I wanted to test out in my own kind of personal trading journey. And, yeah, was asking my my wife, like, hey. Wouldn't it be interesting if we... Like, you know, we know that momentum strategies work on on, like, a ninety day basis, and you take out commissions.
Steve:They work. But what if it's, like, a state thing? And then what if it's, like, every single day is a different state? Like, can I do that systematically doing it myself? And then she she goes like, yeah, go do it.
Steve:Go. Like, I had all this stuff built out in R, and I'm like, alright, I need to migrate this to something that is more compatible with some open API that's out there. I start doing that. Then I run into a dilemma, and I ask, well, where can I get market data through an API at a reasonable price for, like, good liquidity? And I started to, you know, stumble upon different providers, like, okay, fill out this form, talk to sales, didn't hear back, or just, you know, go to this other...
Steve:These other websites, and it doesn't really give you good bars. They're not really what I'm getting through my broker, and there's, like, discrepancies, and and start to learn more about that. So that all culminated as I was wrapping up my career at opera where Jack Bell... Shout out to Jack Bell, just, you know, reached out to me and said, like, hey. Would you be interested in joining Massive?
Steve:And I go, I'm pretty good here. This is pretty good thing going up, but for Massive, yeah. Let's do it. This is an awesome product. You guys are doing incredible things, and I was really impressed of how they're doing it with...
Steve:Which came down to the point where not a lot of folks know that all the work and effort centered around building out the infrastructure, making sure documentation is up to par, and that folks are able to consume and utilize their day... The data in a programmatic fashion was just top notch. And that's me comparing it to a lot of other solutions that are out there. I was just thoroughly impressed. I was like, yeah, I wanna get behind this this mission, these these initiatives.
Steve:This is awesome. So that's a little bit about myself. I know Michael wants to talk about some other pieces about Massive, but based in Chicago, I've done some stints all over the place, but, you know, I like to keep myself at the core here of that quant community as well.
Michael:Well, I think first we should just kind of... We're assuming, and a lot of the audience will know, you know, who you guys are and and what it is that you deliver. But why don't you just quickly go over, you know, the nuts and bolts of of what you are as a data provider, and and what is it you're trying to do maybe different from some other data providers that they're currently playing with or using, and and then we'll we'll talk about the name after that.
Steve:Yeah. Okay. So let's get into Massive. What is Massive? Right?
Steve:It's a platform that provides programmatic access to market data, whether it be pricing, fundamentals, alternative data, partner data, you know, you name it, you'll be able to provide it. The whole point of it is to open up access to different avenues, whether you... Your application or your use case requires an API or WebSocket, or you're comfortable with flat files, copying them from our s three bucket onto your own local environment, it's all about the programmatic access point of view, and access being the the keyword there. So today, we cover US equities, US options, futures from CME, currencies, that's forex, crypto Mhmm. Indices, and then we've partnered with some great institutions like Benzinga, ETF Global, and TMX to deliver their datasets through our platform as well, which we get a lot of fantastic feedback on being able to consume that all through one place, marrying that equity data with an options data, and then something from Benzinga.
Steve:Right? We we cater to that user story, where whether you, you know, you're playing solely in equities or solely in options, you can pair them together or, you know, set up the the environment as you please.
Michael:That's good. So let's let's just get this out of the way right away. What was the name change? I think both me and Dave have said this a couple times, but my bet was the whole Polygon betting market. Right?
Michael:Is that why the the name change took place, or was it a legal thing? Or... Because the old one was way was way cooler than the the new one.
Steve:It was a lot to remove the confusion, more or less, with the with the crypto space. Right? Also... Then move into a a name that matches up with sort of the lexicon terminology we use internally about finance. We find ourselves just saying, like, oh, this is a massive event.
Steve:This is a massive return. Massive data. You know, as it comes at a time where we're looking at options trading, just hitting new peaks every single day, every single month, more or less. Like, it's just troves of data. So that name sort of exemplifies what we have, where we're going, and what we're seeing in the market.
Michael:That's cool. That makes sense too. Right? You you know, when it's there's a giant exchange for, you know, that's growing in size, it's got the same name, and not the best reputation in the world too. It would...
Michael:It could kinda make sense, I guess, to distance yourself a bit. Bet you if you Google Polygon, you get more, oh, they're getting sued, and you probably just won't go through it. Like, I don't want any part of that. You want our data? We got our data.
Michael:We're we're not we're not part of that, so I'll I'll take that as an answer.
Dave:Yeah. Yeah. That makes sense. So the one thing I think about with Massive now, and the sort of the timing of all this, because I remember when it first came out, was like, okay, great. Another market data company?
Dave:I mean, there's lots of market data providers out there. But one particular thing that stuck out at me was how well documented your APIs were, how thorough, how easily accessible you didn't have to have a login to read them. And looking back now, of course, with the AIs coming along, I guess, like, you were in the very perfect place, the very perfect time for... To have this revolution occur. And now I...
Dave:So can you talk a little bit about how how has your company's vision changed? Has it changed? On one level, think, okay. Well, you... Massive came out with this plan, and the and the whole world just sort of all of a sudden has shifted, and you're in the perfect place.
Dave:Or have you changed at all because of the AIs that have come along?
Steve:One piece is the ethos. It's very much a developer first mentality. How do we create tools, documentations to cater to that community? And one of that piece is documentation. Right?
Steve:Having sold to developers in my previous roles, sometimes the documentation is held through PDFs and some, you know, person's folder and you send it over and then the developer has to read through it and parse what's what's right and what's wrong, what's dated, what's not. And that sort of transcends to what we're building here is that developer first. How do we give you tools to implement and deploy and continue to maintain over time? And then AI came in at a perfect time because we're continuing that ethos, but now we're making it so that developers are using AI tools to, you know, assist them with coding activities. How do we cater to that?
Steve:Well, documentation is now in LLM. It's in markdown. Sorry. It's in markdown files. So they can point to it and say, like, hey, here is my documentation from Massive.
Steve:You know, help me. Right? So it's that sort of mentality where now we're allowing the agent to read the documentation, understand what it is, the structure, the endpoints, what the response looks like, and then it just makes for an easier experience on their end to create these applications, these strategies, or what have you. It just... It's that effort.
Steve:Like, you know, don't gatekeep in that sense. Make sure it's machine readable, machine meaning the AI agent, and that is gonna be able to create a better experience overall.
Michael:And that's... Thank you for that. So there's a little bit about this. Right? Dave comes from a programming background, and that's, you know, he's been doing the systematic stuff for for longer than me, and he has that advantage of, you know, being a developer himself.
Michael:I'm not. Right? I I came from the trading side of things, and just kind of saw the light about systematic trading. So where I've had to kind of struggle to catch up is I don't... I can't code.
Michael:I don't know anything about any of that kind of stuff. So I rely super heavily on agents. So, you know, when I was researching this a little bit, I just was able to go, hey, Claude, go figure it out, look at it, you know, could you do this? How hard would it be? Do whatever.
Michael:And then it's able to come back and and kind of give that. So I think that makes a lot of sense going forward, is that everyone should be that facing to AIs. Because I think, and I think me and Dave are in agreement on this, that these... The whole systematic trading style of things is going to continue to grow outwards. Because that barrier for entry that was so huge just even a couple years ago is essentially gone now.
Michael:Where I can, you know, I don't open... Really open up my back testing engines, RealTest or AmiBroker at all anymore. It's just I say, hey, I've got this idea, go test it, you know, write me a report, give it all back. So having that as the kind of first point of content, I think really makes a lot of sense. It's, you know, you're not really so much talking to the people anymore, you have to kind of build these things and talk to the the bots now.
Steve:Exactly. And an interesting sort of user journey from just talking to friends in the industry, from quant friends. It's it's all sort of more or less starting in either Claude, ChatGPT, or Grok for that matter, and they're, you know, maintaining these scripts inside of their local directory, and then they're using an API service like Massive and saying, alright. I have this idea. Let me test it out.
Steve:I already have these pre built scripts that Claude helped me build. Right? And now I just need to kind of suss out a new idea and then go from there. So that's sort of the the demand that we're seeing. We're actually seeing a lot of demand coming from that, the fact that it's the ability to build upon these tools with the use of these these AI agents or, you know, AI tools, LLMs, paramount.
Steve:And to that point, we're also thinking about that broadly. Right? So we recently launched a new partnership with Coinbase where we're using a protocol called x four zero two, and the whole concept of it is allowing your AI agents or empowering your AI agents to go ping our agents.massive.com through x four zero two. They go retrieve the data they need. They pay for the data they need through their crypto wallets, and then, you know, come back and call it a day.
Steve:Right? So there's... We're thinking broadly about access. Right? How do we...
Steve:It... Axe... Programmatic access is at at the core. Develop... Or first is at the core as well.
Steve:Like, how do we expand from there to continue to broaden the ability to to provide access?
Dave:That's good. So, Steve, I'm wondering this huge shift. First, you have people like me using your product. Now, you got a bunch of people like Michael using your product. And now you've also got...
Dave:Like, you said you're a developer first. Well, Michael's not a developer. The AIs are developers in a sense. Like, on once like, I'm very curious about how your customer support has changed over this time, because I would think it would be developer centric, and then all of a sudden, you've got a bunch of nondevelopers that you're having to support, and now you got really sort of AIs you're supporting. So how has your customer support changed?
Dave:Has it gotten a lot more burdensome, less burdensome? I could see both... I could see it going both ways. I'm curious.
Steve:It goes both ways to that point. A lot of the questions that they would get are just answered naturally now through Right. Our website and documentation. They can point to it. The other side is it's it's a little bit of user training.
Steve:Right? Depend... The thing where folks might need a little bit of assistance with as well, use cases. Right? And there's a story where there was a an intern at a large fund who was accounting for 10% of our traffic, and we were like, what's going on?
Steve:He was just basically downloading the historical data every single day, and then, you know, running some kind of process. But then a little bit of user education came... Went a long way. So, well, you can just download the flat files, and then call it a day. You don't have to do that.
Steve:Now, that's a testament to our ability to handle those kind of loads. Right? But it also is we're understanding their use case suggest... Suggesting better paths forward and going from there. So it's a little bit of both.
Steve:Some of these questions that would have been easily answered just from just reading documentations now are going away because folks are pointing Claude at our website and getting those questions resolved. And then other questions are getting a little bit more into weeds as well, which, yeah, we see a lot of folks getting a little bit more into weeds with the data itself.
Michael:That's what I would have assumed as well. Right? That's the... All of the kind of level one questions would be easier to sort out because people are asking, you know, agents as opposed to you. But that also means that anything that reaches you has probably gone through a couple agents and failed.
Michael:So then by the time it hits you, it's gonna be it's gonna be more complex. So that would be probably a problem with staffing and everything as well, making sure everything's kinda trained up and and and good to go there because you're not just getting the how do I get started questions anymore because that would just be taken care of by the bots more or less.
Steve:Yeah. And that's precisely the... So port team is incredible. They're all engineers, developer engineers. They understand, like, the nuances of the technology, and and we pride ourselves on that.
Steve:And that's sort of the core piece there. Getting more interesting questions as they as they pose it up to me.
Dave:So I mentioned... So I had this group called the Mabe Git Mastermind. Bunch of really good serious traders, several 7 figure traders that... In this collaborative group. And I mentioned I was gonna do this interview.
Dave:They submitted... I was sorta overwhelmed, like, submitted a bunch of really good questions. So I'd love to go through some of these to get them some answers.
Steve:For sure.
Dave:So the first one, and this comes from a professional trader. He's asking about the difference between pro and non pro usage in Massive, and he asked specifically, like, any sort of advice for private traders who wanna get a really fully accurate picture with the best information and data possible from US... From your US equities feeds relative to what you can get. Like, so he sounds like he's just... He's willing to pay whatever to get as accurate a data as possible.
Dave:What's... What are his options, or are there options for him?
Steve:I see. I see where the question's coming from. So a couple of things there. Let me start there. The...
Steve:Let's start with the definition. Right? The pro versus non pro. That is a distinction that's made on the exchange level rather than on the Massive level. So each exchange and consolidated tape and so on and so forth has their own sort of definition of what a non pro versus a pro is.
Steve:Right? Some of them can go as far as to say, like, you know, anyone that works as a financial institution is a pro, or some others are anyone that is registered with FINRA is a pro. Right? Now that aside, the data that you're gonna receive through Massive, it would be the same whether you're a pro versus non pro. It's just based on where do you fall within that categories in...
Steve:With respects to the exchange. Right? So if you're a pro in the eyes of the exchange, then you're... You would have to go through the pro onboarding process at Massive. If you're a private trader, no affiliation, you'd fit the the the category of a non pro, then you would be on the non pro plan.
Steve:But at the end of the day, it's the same... We're connected directly to the exchanges, delivering the same data, same tech. The fact of the matter is how the exchange is treating you.
Michael:Yeah. And we we have some experience from that. Well, Dave probably wouldn't. We both worked with trade ideas for some time, and we found that the exchanges were very mad if if you did that classification wrong. So, you know, it's definitely something you have to do.
Michael:And and, you know, looking at the questions that you guys submitted, Dave, I I'm gonna go back a couple because you skipped the one that I care about. Any plans for historical fundamental data, market cap float? I'm not so interested in the, you know, like sharp ratio or the, you know, quick ratios or any of that kind of stuff. But I, for one, like to play. So so, know, Dave is...
Michael:He will only ever day trade. I'm a little bit more on the swing trader or long term time frame. So I'm more interested in in things, institutional ownership and short float and even insider buys and sells, different things like this that are... You can call them alternative data, think is the the snappy version, but whatever you wanna call them. Any plans for those?
Michael:Because I'm always interested to see if I can grab something. The harder the data is to get, the more I think that there might be an edge hiding there that people people haven't discovered yet.
Steve:Yeah. I think that's... I know there was, like, some questions, like, what Quanta asked, and it's like, the harder it is to get, the more interesting it is. Like, I remember getting a phone call at one of my previous roles, and it's like, do you guys have data on how many, you know, doughnuts Krispy Kreme sold in 2024? I go, no.
Steve:But let me see. Like, there's there's some kind of alpha there, I guess. But with respect to the question itself, yes, you know, like, market cap and and flow, there's different ways to understand that. We have a research team that is dedicating time to understanding the nuances of that, and collect it, and tag it, and normalize it as well. Fundamental data as well is gonna be an initiative of how do we tag and categorize these filings and distribute that onward to to consumers.
Steve:I was talking to some folks. There are a lot of nuances here as well with filings themselves. You have main table, and then you have footnotes. Being able to categorize everything and timestamp everything is part of what the team is focused on as well. Now, when you ask a quant, there's alpha on every single step of that collection process.
Steve:From reading the the PDF or the main tables, to understanding and adjusting for footnotes, and even to the point where you upload that data to your database, that then gets a database date. You know, they... Versus, you know, when does the market get... Become aware? When they...
Steve:Would I... My trading algorithm would have received it? There is, like, alpha and all those steps. And those are all components that the team is kinda considering, and there's some stuff that they're looking at, of course, not gonna divulge on this piece that goes really into the weeds on how we can make sure that folks are maximizing the value that they would get from these type of data sets that we're working on.
Michael:Yeah. Because I think that last step is one of the hardest, because when you're looking at this kind of alternative data or something like short float, for example, reported every two weeks, but on a lag. So if you're not very careful to line up when you would have known that data, you can make something that looks way better than it than it actually does. Because it's like, oh, well, short float on this date was really high, so maybe I should have bought it then. But if, you know, you wouldn't have known that until three or four days later, it it could really mess things up.
Michael:Just for the listeners to know why this data is so hard. Because unlike price data that just updates every tick, and it just is what it is, this kind of data, it exists at a time, but then it's reported another time, and then sometimes it's revised later down the road, and and all kinds of things that make it, I think, way more way more messy than that pure price data.
Steve:Right. Exactly. And for context for everyone, I spent a lot of time selling point in time data in previous roles. So getting to the minutiae of those deltas is extremely important for folks to understand and things we're considering here at Massive.
Dave:Alright. So here's a question from Oliver, and I'm not gonna read the whole thing because there's... I can sense a lot of frustration in this question. So... But basically, it boils down to when you look at data from different data providers, a lot of times you'll see differences.
Dave:Sometimes sometimes pretty substantial differences. And when you're doing backtesting in in the kind of people that I work with, they're looking at the details. So why is it that you have data discrepancies across providers? I mean, aren't we getting all this from the same provider? Like, why why would you see different candle data, different...
Dave:Like, why why does that happen?
Steve:A couple of different pieces there. It it starts with, of course, making sure everyone's directly connected to these providers. The second piece is... The question that you have to ask yourself is, are these vendors normalizing the data? Are they dropping packets?
Steve:Are they removing quotes? You know, we're going down to nanosecond time stamps if their systems are handling it, and especially if it's like bar data. Right? Like, if... Are you taking into consideration every single tick or not?
Steve:Then goes into trade conditions. Are you agnostic to trade conditions? Are you only taking into consider certain... Consideration certain trade conditions as opposed to others? Are you, you know, looking at the whole entire picture?
Steve:So there's a lot of nuances to that, and depending on how they... What sort of standard they adopt, you might see slight, you know, deltas between one provider and the other. So boiling down or drilling down into exactly how something is calculated is is important. Now, on our end, we'll calculate the bars using a sort of a standard we can share publicly after that as a follow-up. But we also provide you the tick data so that you can then discern and say, hey.
Steve:These are the trades or or these are the trades I wanna, you know, take into consideration within my bar as it's streaming into my application. And because you have... That might be your secret sauce. That might be your alpha. Right?
Steve:And so on and so forth. So to answer the question, differences in methodologies, if everyone is connected directly. To further it on is if there is something that you want to consider in trade conditions, you can always take a look at the tick by tick level data, stream it to yourself, and calculate it yourself.
Dave:Yeah. That's that's a good point. So you're not getting bar data from the exchange, or nobody's getting bar data from the exchange. Every provider is building bars and then transmitting those to subscribers. So...
Dave:Yeah. And if you look... I I was sorta got into this world at trade ideas when I saw some bad tics in the morning, predictable bad tics. You start looking at trade conditions, there's a massive, so to speak, list of them. I mean, there's so many trade conditions, and if you're not handling them properly, you will get sometimes subtle, sometimes really big differences in bar data.
Dave:So, yeah, I could see how that specifically, you could end up with with data that just looks very different across providers. That's so interesting.
Michael:Yeah. Well, yeah. And I I think it's, you know, you you can look at it through charting platforms. I think it's the easiest way for the audience to understand is that sometimes there's that bad tick in the morning, and and sometimes there's not, depending on the charts. And they go, well, did the price actually trade there or not?
Michael:Those are egregious, but if you think for every egregious example, there's probably five or six, like, little minute examples as well.
Steve:Exactly. And if there are two traits, this is something I've found that look... That came in at the same exact time stamp, identical in the fact of every single component of it is identical outside of, you know, a decimal off? Are they considering both? Which one takes priority if they if they arrive at the same time?
Steve:Because the one thing is since the market is fragmented, there's trading going on across all these exchanges. So, theoretically, from, like, a, like, a time and, like, physics perspective, same trade might happen at the same time at, you know, at different places. So which one do you prioritize at the end of the... At the close of that bar?
Michael:And this is the... I think the the same question when we were talking with IQ fee that I asked. And are you shocked about how this all works as well as it does? Every time I hear about the more of, like, the back end, there's, like, all these different exchanges, and things come in at different orders and things. It just seems so wrong.
Michael:It just seems like there'd be just such a much simpler and better way to to to do this, to have some sort of like universal source of truth, like single flow, whatever. Is it just purely it's already this messed up and it's existed this way for so long? It's... I think the example I gave last time is it almost feels like you... Someone built a house, and then other people just kept building things on top of the old house, and no one bothered to say, you know, at some point, you just gotta rip it down and and build a nice new building as opposed to trying to build more modern buildings on top.
Michael:And I just... I wonder now with, you know, the recent announcement, Robinhood's gonna allow $2,473.65 over the weekend trading and all this. It's like, are we just... Is there gonna be a a like a tipping point here we should all be worried about with all of these fragmented? Or or am I just being paranoid and everything's running fine back there?
Steve:You brought up, like, a plethora of pieces that I can dive into and speak like hours on, because I... The first piece, say the the architecture, we'll boil it down at the piece, is that exchanges do serve a purpose at the end of the day. They serve as a venue to express your opinion on the market. Right? Each exchange has some kind of feature that's interesting, that folk...
Steve:Why folks trade there. Right? Is it because they prioritize certain types of blocks? Is it because there are certain types of traders present there? Is this the certain type of orders that they provide?
Steve:So there is sort of value in that sense. Now, some exchanges, you can only trade a per specific contract exclusively on that exchange. That's why you trade there. Right? Now stepping away from there, it goes well, then let's talk about the data points point of view.
Steve:Why is it... Why can't it just be consolidated into one piece? It's it is... And the way they set it up is through these very sophisticated networks, more or less concentrated in Chicago and New York, New Jersey. But the piece is since, I would say for the most part, all the exchanges have different protocols, you then have to adhere to different sort of standards as you consume the data.
Steve:So to visually map this out, we'll say all these exchanges are disseminating data into a centralized hub where you calculate the NVVO. That then needs to be reading different protocols. You pull it in. You calculate the NVVO. You almost talk talk about it as like a bus.
Steve:Once the bus is full, you send it out, and then that goes through a different channels onward. That's easy for, like, equities. When you're talking about, like, options, that gets more complicated just because options is more of like a, you know, a diner. It's like, go in, you go to diner, open up the menu, you have, like, hundreds of options that you can, you know, select from, and then every one of those options is a separate market, more or less, and the amount of data that goes through there is enormous, massive. Right?
Steve:So rebuilding it from the ground up, you have to, we also have to consider, like, the the nuances of the different protocols, like, if there's, like, a stand... Centralized standard, is that... Do they wanna do it? At the end of the day, you know, once the exchanges went private, they have say on how they wanna run their business and and how they wanna provide that service to the industry, and we sort of come to a head there. Now...
Michael:Well, I guess that makes the most sense too, is that at the end of the day, I'm always thinking of, you know, exchanges like there's, you know, some sort of government body, but at the end of the day, they're all for profit companies. So having them all sit down at a table and agree to something and play nice is gonna be almost impossible unless there is a regulatory body that says, no, you you have to do it this way. Everyone's gonna do the way they think is gonna work the best for them and and make the most money. So, probably near impossible, I'd say, at this stage.
Steve:It's possible. Right? You have the SEC, you have the CFTC, Who are those bodies who do, you know, hold these committees and talk about pressing issues in the industry and how do you address those sort of innovations in the industry? Twenty three five trading, same thing happened. How do you get all the exchanges in one place and say, know, this is coming.
Steve:How is everyone gonna get prepared for it? Some folks are gonna introduce a new protocol. Some folks are gonna introduce a new channel. They more or less have to conform to some level of standardization. So now that also brings up to a separate point where Massive does engage Massive does engage in sending letters out to the SEC on what our opinion on is on certain topics, on what they're doing, how the exchanges are doing it, etcetera, and that helps.
Steve:Right? So it it puts... It gives sort of a voice of the market, puts them in a perspective on a topic from folks that are disseminating data to folks who are using the data and making trade decisions on that data.
Dave:Alright. So here's a question just as you were talking there that came to my mind. Twenty four seven trading, or twenty three five, whatever we're going towards soon, surely that's keeping you up at night a bit. I mean, I'm thinking about specifically those flat files that get generated each day. If you've only got a certain amount of time, I mean, you got plenty of time now, but you're gonna have a lot less time, a lot less room for error before the market opens back up again, and people have to have that data to make decisions for the next day.
Dave:How are you thinking about that? Is that is that keeping you guys up at night?
Steve:Is it keeping us up at night? I would say perhaps some folks. We have a clear plan of, like, you know, this is what we're gonna do. This is how we're gonna prepare for everything. It's...
Steve:We sort of frame it up in a way of it's exciting. Right? This is a new innovations to market that has been in other markets like futures for some time, and now it's being introduced to equities and then soon options. Right? So those are all questions we're looking to solve.
Steve:Like, how do we get our flat files ready? How do how do we calculate bars? Right? Because the next trading day technically starts at 8PM. And then you might hear on the...
Steve:In the industry, it's like, it's 9PM, Steve. It's technically 8PM because that hour buffer is just a buffer that is not acquired, but suggested that the exchanges take. So there might be an exchange that comes back online at 08:15, and now we need to be ready for that.
Dave:Interesting. That's be hard for... I mean, you can you can maybe provide the data for that, but man, there's gonna be a lot of traders where all of a sudden, they're gonna have to rearchitecture their whole overnight process. Bought a bought a provider, so they're gonna have to reoptimize and and reorganize their overnight process. Yeah.
Dave:That's... It's gonna be a massive shift.
Michael:Is it... The only thing that's on purpose? Are you doing that on purpose at this point? Is he doing more I that why the
Dave:name change makes sense now. Right?
Michael:Listen, is he slipping a 20 under the table every time you say Massive? Because I wasn't brought into that that conversation, but... No, you're you're right too. And then also, it just seems like everything's changing. The the market seems to do this, where there's nothing changed, like, fundamentally, and then everything changes at once.
Michael:Like, the biggest one since this twenty four hour trading, I think, was that there was like a week period where all the brokers went to zero commission trading. I just... It was just like every day you woke up and boom and boom, and it's everyone... And I was like, jeez, and that was a huge shift in a very short time. And then nothing happened for years, and then it seemed like, okay, Robinhood did twenty four hour trading.
Michael:And then everyone else did, and just over... And now you... It was, what, two days ago as of recording this that they came back out and they said, screw it. Casino's open on the weekend. Have fun.
Michael:And it's... You know, the... What's gonna happen is it's the same as everything else. You're gonna have everyone else starting to go through and and update that as well. So that to me is a a pretty decent shift.
Michael:And that, you know, to Dave's point is, like, things like, I don't know, server maintenance and all of that kind of stuff. If it's literally like crypto, which was gonna lead me to my point, you guys already have to have some experience with that. So do you guys look at it and say, okay. We're doing this with crypto. It's the same thing, just way bigger when the equity markets do it too?
Steve:Perhaps not as a reference point, but more from, like, a, you know, how's futures handling it. Right? Because we we provide futures on the platform, so we we sort of, like, have that sort of as a learning experience as well, and then understanding, like, how failovers work, how, you know, the... That kind of, like, backup plans work as well has been a massive learning experience. I'm gonna stop to to get ready for that.
Steve:And definitely, there are a lot of learn... Learning lessons in crypto that were taken away to that. Now the the piece with crypto versus sort of your traditional asset classes is the regulation and the documentation and, like, sort of rules and and guardrails that are existing in these traditional asset classes that do help us sort of conceptualize how we should be treating certain things. So there's a lot of, you know, discussion. There was a roundtable in DC, I think, a couple of weeks ago to talk about the topic.
Steve:Folks post some interesting questions, and there's a lot more deliberation there. Whereas, you know, other markets had been just like, just open it up and we'll fix things as it goes, because like the volume is just like... It's gonna grow over time. Whereas here, we could see volume just instantly shift.
Dave:So I know we're running up on time a bit, and I want... There's several other questions I wanna get. So maybe we can go through these in more rapid file... Fire fashion. I saw that I can group a couple of these into one.
Dave:So couple of people are asking, alright. What about level two data? And, like, I don't think you provide that, but maybe you do. Do you have plans for that? The other one is what are...
Dave:What he calls calculated indicators, like the tick, the trend, advanced decline, these kind of aggregated statistics. Are you thinking about more of those that you could provide, or how are those provided now with your service?
Steve:Let me start the second question, go back to the first. So the second question, we hired a few researchers to embark on what we call our market insights and proprietary datasets who are gonna be working on something like this. That being said, since we provide you the raw data, you can calculate the metrics yourselves. There are dozens of different ways that folks calculated Black Scholes and Binomial Tree models to calculate their Greeks. So we give you the data so that you can sort of put your spin if you want something to work with.
Steve:We're sort of exploring those avenues as well to see what we can provide you in that event. Going back to your first question about level two, we do have plans to onboard that, hopefully, at some point next year. Twenty three five has consumed a lot of our resources to get ready for that for this year, and then from there, we're gonna expand and add additional exchanges, more depth, more regions. So, you know, geographically expand our side, and then add proprietary datasets and market insights, which answers the second question.
Dave:Okay. Great. So here's a here's a really good question that I wanna ask you. So this one's from Nicholas. He says, what do sophisticated quant clients ask you about that retail traders don't usually even think about?
Steve:I've been thinking about this. I always think about this because thinking back to my time, you know, working with quantamental quants, right, they they think about data cleanliness, you know, how do how do I find some differentiated data on your platform. There was even a quant that I was talking to who knew that... When I was at the firm at the time, they knew that we had superior data. So what their alpha was, I know that there's a larger market data vendor that provides the same data who most people use, and he would not call out the incorrect data point at that larger vendor because he knew there was alpha in that delta.
Steve:So he would know that we had better data, and he would not intentionally not correct the data at that larger vendor because, like, you know, that's how he would play it. So then the question that he would pose is like, alright. Like, can I get this through some kind of programmatic access to compare it in real time so that I can then trade based on, you know, when that data gets collected and published at the other vendor, and then here as well? So quants go into the weeds. I call them artists.
Steve:They're very creative. They have different ways. They're Picassos of this industry. They have different ways of, like, painting the picture of the market that they see, and it all boils down into the types of data that they are interested in or what their opinion, more or less, requires from the data industry. Right?
Steve:To to Michael's point, they're thinking about date stamps, they're thinking about different pricing points, etcetera. When it comes to pricing data, they go into the weeds. Right? They go into sort of tick level time stamps, deltas between one tick to the to the next, just the general order flow, etcetera. So they're going into, like, sort of like in sequence, in time, does this make sense?
Steve:Am I receiving the cleanest data overall? It almost brings me back to a book by the founder of Rentec. I'm not sure if you guys have read it. I think it's called The Man Who Solved the Markets, something like that, where there's like yeah. He he...
Steve:There's a chapter in that where one of his colleagues literally sat there and went through tick by tick by tick by tick, looking at the data to make sure it was clean. And then at the end of it, he had the cleanest data out there. So the quants there are looking at that sort of level of granularity and questioning, you know, does it make sense from one to take to the next?
Michael:Yeah. That was that was the Jim Simons piece. Right?
Steve:That's Yes. Yes.
Michael:You know, from our, you know, one of the the GOATs, I guess, from from our group. Right? The greatest hedge fund manager, some people say the greatest investor of all time who refused to hire anyone from Wall Street ever. It was all purely quantitative kind of multi system development. Now, we're running a little bit over, but Dave's itching there.
Michael:Do wanna hit him with one more before we wrap up? Do you have an itchy one?
Dave:Let's get... Let's get one more in.
Michael:Keep it for the whole day, but one more. Okay.
Dave:Alright. So do you have any guidance for dealing with corporate actions, splits, mergers? This is sort of it seems like it should be clean. It seems like there should be an algorithmic solution, but once you get into the weeds, you realize there's actually some discretionary decisions you have to make about which, you know, for for some of these. Do you have any guidance on that?
Steve:The guidance is that we're working on it in the same avail of the fund... Fundamental and, like, the market day... Market cap. It's it's... We are...
Steve:That's an active project that we're we're looking at on how to tag it, and how do we roll it up. Right? If there's a split, you know, does that create a new entity? Or, like, sorry. If that's a spin off, like, does that create a new entity?
Steve:Do we tag it intentionally just so you can see that sort of flow? And at the end of the day, it's like, how do... How can someone map that entity from one end to the other, understanding that there was some kind of corporate action in the middle? So that's something that we're actively working on at the moment.
Michael:Well, good. Well, you know, before we... Again, we don't wanna take a a massive amount of time. See, I... Everyone else everyone else got their swing.
Michael:Right? So I I had to do one. But yeah, we don't wanna take all day, but why don't you just leave the the people with, you know, what... So say you're someone never heard of this before, you're interested in taking a look, what would you suggest them go to watch first or to read first or to to look into first, to see if this is a good solution for
Steve:them? Certainly. Well, I'm always of the, you know, reader of good content, like that book from the man who saw the market understanding that level of of ticks. There are a plethora of YouTube videos out there that can get you interested on how to get into systematic trading. But the question you have to ask yourself is, if there is a solution or a GUI based application that you're using today, which you can't seem to really conform to what you're thinking about, then getting access to a programmatic platform like Massive would be a great fit.
Steve:Whereas, like, there is, like, a specific formula, there's a specific way of calculating something that you can't really get what your current solution to to do, then getting the just sort of fire holes of data and calculating yourself in real time would be a a fantastic play. Great starting point is you're a local LLM. Right? Talk talk to Claude. Talk to ChatGPT.
Steve:Point it to our website. Express your opinion, and and see... We'll start building from there. We have a team of builders here, and myself included, where I have a directory in Claude just called Alpha, where I sort of test out ideas and and go from there. So that's that's another great starting point.
Steve:Then last but not least, and this is a plug for Massive, we have great blog posts and new case studies that are gonna be published on our website, massive.com, where you can start reading about these different datasets and how they're being implemented with some sample code, and you sort of start piquing your curiosity on how to handle some of these datasets, and start sort of sparking some creativity on your end on what is the possibilities with here. Because at the end of the day, with programmatic access, it's, you know, how creative can you get? Because you're getting access to the lowest level so that you can build from there.
Michael:Awesome. Well, let's see. Thanks for coming by. As always, I I always love these look into kind of the the underbelly of how everything works, because I do think a lot of people just take some for granted, especially discretionary traders that are just looking like a trading view or something. They they kinda take for granted all of the things that go in to get you that that tick on your screen, and, you know, something us nerds spend a lot more time worrying about.
Michael:But, yeah, thank you for coming by. And as always, I'm Michael Nauss.
Dave:And I'm Dave Mabe. Talk to you next week online in your own pockets.
Steve:Thanks, y'all.
Creators and Guests
