Introduction to OneTick Cloud:

OneTick Cloud provides real-time, intra-day and historic data and analytics on trading activity leading to actionable insights for sales, trading, and surveillance. With hundreds of customers, you know that OneTick is a recognized platform you can trust. We will cover these many powerful advantages to OneTick, and show you how, together, we can meet your business needs.

Webinar Description:

Regulators around the world continue to evolve market data infrastructure, and SEC Rule Changes are currently addressing issues of latency, granularity, and coverage for a more comprehensive and timely view of market activity.

This is requiring firms to stay agile, innovative, and forward-thinking to leverage these changes effectively. Firms are moving towards higher-frequency data feeds, faster data dissemination, and better alignment with the needs of both institutional and retail traders.

From a quantitative perspective, this a significant opportunity to gain access to richer, more granular datasets:

• Quants Developing High-frequency Trading Strategies: This could mean more accurate real-time data for backtesting and decision-making.
• Market Structure Researchers: the increased granularity and faster data dissemination will allow for more precise modeling of market liquidity, price discovery, and execution dynamics.

However, this new landscape will also introduce challenges, particularly around latency and data integration. Faster and more frequent data feeds will require firms to adapt their systems to handle larger volumes of data in real-time, without compromising performance.

For quants, data engineers, and market participants, the key to staying ahead of the curve will be embracing these changes and adapting to the new realities of market data. By developing the right infrastructure, refining data models, and investing in the tools needed to handle higher-frequency data, firms can turn these regulatory challenges into opportunities for innovation and growth.

In the coming years, we expect to see even more transformative changes to the way market data is captured, processed, and distributed. For those who are able to adapt quickly and leverage the power of these new data streams, the potential rewards are immense.

The future of market data is here—are you ready to embrace it?

This webinar is designed for senior professionals in quantitative research, data engineering, trading technology and infrastructure, including CTOs, heads of quant research, data scientists, portfolio managers, and technology leaders responsible for market data strategy and research platform architecture.

Speakers:

• Christopher Brampton, Market Data Management, KX

• Peter Simpson, OneTick Product Owner, KX

Watch the Recording:

Webinar Transcript:

Hi, and welcome to our webinar on the impact of US SIP changes, including Odd Lots, and recent market data changes across the world.

Today, I'll introduce who we are. Then we'll go through kind of a very quick intro to OneTick Cloud. Then we'll get into the details of the US SIP.

I'll talk about the kind of increase in data volumes, regulatory change, and with a regulatory change, and then we'll cover other changes outside of the US SIP, and also how we're expanding our cloud coverage in the coming months.

So first of all, Chris Brampton's gonna be presenting. He's in charge of all of our market data onboarding within our OneTick cloud market data as a service platform. Chris has been with OneTick for the last eleven years, and prior to OneTick, he was with Interactive Data, now ICE Data Services.

And there's me, Peter Simpson. I'm responsible for the OneTick cloud platform, and I've been with OneTick for the last seven years, and prior to that, kind of history and kind of financial markets.

So OneTick cloud, we provide market data on demand across global equities, futures, spreads, and options.

And that's both real time data, where we're providing access to real time intraday last value caches, and then historic data. And the historic data can go back to 1993. As well as the tick data that we're collecting and storing, we also provide access to AI feature sets, whether that's kind of one minute trade and quote bars, daily bars, kind of volume splits, or market share daily bars.

And for regions of the world where there's fragmented liquidity, we also provide composites. So that would also include creation of custom NBBOs for the markets where there isn't a regulatory NBBO. Now overall, we cover around six million symbols, where a big chunk of that, kind of roundabout four and a half million are options, but there's still that's nearly two hundred and twenty five thousand equities and ninety six thousand ETFs and over two hundred thousand futures and then two hundred and fifty thousand future spreads.

That all equates to around two terabytes of data we're collecting and storing every day, And overall across our platform, we have a history of around four petabytes of data.

Now people can access the data either as files, as Parquet over our s three bucket, or through our REST API. They can write SQL. They can write Python, and they can use their MCP server to write natural language and get queries back and get data back.

Now all of this data infrastructure is hosted in AWS in US East, and it's supported by a set of teams. They cover a market data onboarding team led by Chris who can process the feeds, process the PCAP files, and then map data, producing data bases and tables and fields. And there's our reference data team. They cover kind of corporate actions, adjustment factors, symbology mapping, market calendars, and, continuous contract definitions.

Then we have a data loading team that can manage the operation of loading data every day from all of the markets globally.

Then we have a DevOps team that manages the whole networking compute storage across our AWS environment. And then finally, we're creating these derived AI feature sets. So there's a team that are kind of writing the analysis to produce those.

And finally, we have a custom analysis team specific to particular customer requests where we're building new examples and building custom analysis for a particular customer's requirements.

Now we're going to talk about some of the changes that are happening to market data and, in particular, to market data for US equities sent over the US SIP feed. So the SIP, the Securities Information Processors, the CTA and the UTP, Basically, they're the organizations whose job it is to collect trading quote information from all of the US stock exchanges and consolidate all of that into a single day feed, well, two feeds. A CTA feed and UTP feed. UTP covers everything listed on NASDAQ.

CTA covers stocks listed everywhere other than NASDAQ. So by taking these two feeds, get a full view of all of the trade and quote activity on US equity exchanges. Now, it's a very convenient feature that has, over the years, undoubtedly driven the higher liquidity and interest in the US equity markets. Historically, there has been some limitation in the data provided by these feeds, leading to an information imbalance between your average user and, for example, sophisticated users who might take direct feeds from the individual exchanges.

So to address that information imbalance, the SEC has started changing regulations as to what's included in the slip feeds. And so we're going look at some of those that have been happening over the last year and that are still ongoing now and will continue through 2026 and on as far as 2028.

So in November 2025, started addressing one of the key areas where there was an information imbalance between the SIP feeds and direct exchange feeds, which was the way that quote sizes, bid and offer sizes are represented.

So historically, when the SIP sent a bid size or an offer size, what they actually sent was the number of round lots. Now, the lot size for most stocks in the US is one hundred shares. So they would send, for example, a bid size of two and that could mean anything between two hundred and two ninety nine shares.

You wouldn't know the actual number of shares, only that it was two round lots.

And of course, this also means that if there was less than one round lot at the best available price, you wouldn't actually be able to see the best available price at all because that simply wouldn't be included in the sip feed. What you would actually see would be a worse price that had more than one round lot.

So there's clearly some important missing information there if your average user wasn't able to see the best prices. So the first stage of addressing that was last year in November when the SIPs started sending quote sizes in shares rather than lots. So instead of two, they'd send two hundred. They still didn't send the actual size and they still didn't send the odd lots at all if there was less than one hundred shares, But at least they started reporting the sizes in shares in a more natural way.

In February, the SIPS also started addressing another area where there was some missing information, this time in trade data. One of the things that has grown in recent years and now accounts for around five percent or slightly more of trading activity is the ability primarily on retail platforms to trade fractional shares. So instead of having to trade whole shares one, two, three, four, five shares you can actually trade half a share or a hundredth of a share or some other number. Now for a lot of stocks, it doesn't really make that much difference whether a trade size is, for example, one or one point one.

But when you look at stocks that are trading at a very high price, like if we use an extreme example like BRKA, Berkshire Hathaway A, trading up around seven hundred thousand dollars or whatever it is now, most of the people who might actually want to trade this just simply can't. I mean they have seven hundred thousand dollars So most of the activity is actually fractional as a result. And that wasn't going out at all in the SIP feed. Or rather, you could see the trades, but all you'd see was one share, whereas the actual size being traded could be, for example, a hundredth of a share.

So in February, the SIP started publishing that fractional share information. So instead of the sizes saying one, they actually had the full precision or at least six decimal places precision so that you could see all the fractional trading.

Then in April came the second phase of the changes to quote sizes, where the SIPs actually started reporting the odd lots. So this was by far the biggest change.

So this means that those prices before that the average user couldn't see because there was less than one round lot at the best price now those became visible.

Whenever there was an odd lot quote that's at or better than the NBBO started being sent in the SIP feed so everyone could have access to their information. Now of course with the new information being published in the feed and there are a lot of odd lot quotes out there, it's a large portion of the activity. With that extra information came a significant increase in data volume. So also issues in relation to the data volume that everyone had to process in order to handle all of this information.

US equities already quite a large data set and by adding all of this odd lot information, the overall quote volume didn't quite double but it was pretty close. So it had a significant impact as well on the data volume that everyone had to consume in order to get a full view of the information in the SIP feed. Now these changes are continuing through this year. So in November 2026, there's going to be another potentially fairly impactful change, which is that the exchanges will start allowing stocks to be traded with half penny tick sizes.

So right now, anything trading at one dollar or higher has a minimum tick size of a cent. So it doesn't matter whether a stock is trading at a dollar or at a thousand dollars, the minimum, the smallest price increment is a cent. So obviously for anything trading around the dollar, that's proportionately a very large movement. If a single tick is one percent of the share price, then it's the large bid offer spread.

So to enable slightly more granular bid offer spreads in circumstances like that in November, anything that has a time weighted spread of a cent and a half or less will have its tick size reduced to half a cent. Now this has the potential to have a significant impact on data volumes because if you reduce the tick size then obviously you have algos that respond very quickly to changes in the market that are going to take advantage of that. And so they're going to change their price entering orders out much more quickly if there are more granular tick sizes that they can use. So there's potential for a significant rise in data volumes again.

However, what we've found is that with the way that stock prices have risen recently and the very high valuations that we're seeing and the high prices that stocks are trading at, as well as above average volatility, what we're actually seeing is that because of the requirement for a time weighted spread of a cent and a half or less, most of the liquid names actually don't have that. And so when we look at some of the most heavily traded names out there like Sky or QQQ's, we're seeing that they have average spreads that are slightly above that. So if the market stays where it is right now, and if market conditions stay as they are, we actually expect that this will probably have a small impact, but it won't be as large as it could be.

If conditions were a bit quieter and valuations weren't as high, then this time weighted spread criterion could potentially lead to a lot more stocks trading in half penny increments. Then in December, right now of course, US equity trading on the main exchanges isn't an all day thing.

So the pre market begins at four am and the post market finishes at eight pm Eastern. So again, access to the markets and in particular interests to, for example, the Asian market, looking to trade on US equity exchanges. The exchanges in December twenty twenty six, December of this year, will start moving to a near twenty four hour trading model. There'll be a bunch of exchanges like NASDAQ, NYSE ARCA and CBOZ EdgeX, which will move to a twenty three hour trading day. So the market will still close at eight pm, but instead of there being an overnight break until four am the next morning, these exchanges will actually open an hour later at nine pm and that will be the start of the next trading day.

Finally, looking ahead to May twenty twenty eight, this is the final part of the changes that the SEC wanted to introduce: the reporting of Odd Lot, which was rather than reporting the best odd lot prices if they're available. So right now, if there's an odd lot that's at or better than the NBBO, you can see the best available odd lot price.

But in May 2028 what's actually going to change is that SIPs will have to publish not only the best odd lot prices provided there at or better than the NBBO, But if there are other odd lots behind that, they actually have to publish all of them. Anything better than the NBBO basically, they have to publish every level of odd lots. Now that hasn't been implemented yet because the SIPs pushed back on this essentially. Their feeds were not set up to report as what is essentially market depth and there was relatively limited interest in the market.

And even now, there still appears to be limited interest in this. So it's unknown what kind of impact this is going to have when it arrives. But it appears not to be a focus kind of in terms of what the market considers to be valuable information. But anyway, it's still due to happen in May 2028.

So yeah, Peter?

Okay. So we've seen changes in liquidity. As Chris was mentioning, it's a very active time in the market. We store the data every day in a compressed form, and we can see back in 2024, getting into 2025, we were kind of averaging around thirty gigabytes of data a day.

As we move to the end of 2025 and 2026, now we're on average around sixty and we're spiking up to one hundred. We can see here April twenty twenty five that was a big spike around kind of the US kind of tariff announcements, then it calmed down a bit. But from October onwards it's been very active and effectively data volumes are twice what they were a year ago. Now if you look at just message rates as we consume real time SIP data, we're seeing a similar pattern as you would kind of expect, where we used to, the second half of last year, receive around two billion messages a day for US equities, and now that's kind of ramping up to nearly six billion messages a day. Now we don't just provide access to US equities in real time, we also cover a whole set of derivative markets, So the the CME markets and the ICE markets and EUREX, and also kind of European equities. And we can add markets in real time based on customer interest.

When we're collecting this data, we do it twice because we need to have a highly available environment so we can ensure no data loss and ensure very low latency access to this data. So when we combine all of the real time data that we've been seeing, we've seen this kind of growth in liquidity, just for US equities, but also across all of the other markets we're consuming, where back the second half of twenty twenty five we were collecting around four billion messages a day and now that's kind of spiked up nearly to fourteen billion and in the last few weeks it's been around ten billion.

Now we're just going to go over how we've implemented these changes to the US SIP pages that have been introduced particularly the Odd Lot release which the SIPs introduced at the end of April and we're actually going to be moving forward with at the end of this month on the twenty ninth of June. Now SIPS of course added the odd lot quote information and we are going to provide that. However, because essentially all regulatory requirements in relation to, for example, best execution are still based on the NBBO, which still only includes round lots.

So the NBBO hasn't changed as a result of any of the odd lots changes. It's still based on the best round lot prices.

So because all of the regulatory requirements still relate to the NBBO and they completely disregard the odd lots. So while the odd lots are there, you won't necessarily get filled at them or at least you're not guaranteed to be filled at them. We'll actually be leaving the existing quote tables the quote and NBBO tables as they are. So the exchange quotes in the quote table will continue to represent the round lot quotes and the NBBO and the NBBO table will continue to represent the round lot NBBO, the regulatory NBBO. No changes will be made to those and like that users can just continue using that information as is.

But separately, to provide all this additional detail that's become available, we will be adding two new tables, two separate tick types, which we named quotecomp and nbdocomp. What those tables will do is that they'll behave much like the quote and NBVO tables that we currently provide. One contains exchange quotes, the other contains the NBVO. They will provide the best overall quote, including both round lots and odd lots.

So if there is an odd lot available, then that will appear.

If the best quote is a round lot, then that'll appear. And if the best quote is a combination of the two, then the sum of the two will actually appear in the quote.

And we've done this because well, that way it just makes it simpler to see the overall liquidity that's available without having to look at round lot quotes and odd lot quotes separately. As a result of this, well, anyone who chooses to take the new tables, quote comp and NBBO comp, will likely see a doubling in data volume. That's what it looks like at the moment. But of course, as I mentioned before, anyone who just wants to see the round loss information can continue to take that and there will be no impact at all.

And we'll be backfilling this information to first of June twenty twenty six as of the release date on the twenty ninth. So looking forward to these changes that I mentioned earlier in relation to twenty four hour trading and half penny tick sizes. In terms of data volume impacts, we actually expect that the move to twenty four hour trading, even though it adds another seven hours or so to the trading session the effect on data volume should actually be fairly small. Right now there are ATSs in the US that offer overnight trading.

So for example, Blue Ocean ATS and OTC Markets' Moon ATS.

They offer trading in US equities while all the exchanges are closed. That already exists.

However, while there's decent volume on there, it's nothing even close to what you see on the main exchanges during the day. So when this twenty four hour trading change or twenty three live trading change goes live, there might be a slightly bigger impact than that because more people just trade on the main exchanges.

We don't expect it to be more than about five or ten percent in terms of overall data volumes. The tick size change, as I mentioned, has the potential to be quite a bit bigger.

But that does depend on how many of the most liquid names are captured in the tick size criterion in that threshold of time weighted spread of one and a half cents.

And right now, yeah, because of where the market is, the most liquid stocks have spreads that are larger than that in general. So it's equally likely that it's going to have quite a small impact. But it remains to be seen exactly what impact that will have on data volumes.

The fewer other changes that have been occurring outside US equities here. Was quite an interesting or unusual one on the National Stock Exchange of India to their order and trade data that they provide, where essentially the Indian regulator noticed that order and trade information they were publishing included a bunch of non public information.

So for example, non disclosed order volumes that appeared in their order data. The regulator noticed this and essentially just told them to stop immediately publishing this information.

And as a result, within the space of a few days, the exchange had discontinued their order and trade data services at that point and replaced them with new services that omitted this additional non public information, such as non disclosed order sizes and kind of routing and algorithmic properties of all the order flow that went through their platform.

Of course, we had to develop to this new service and as a result, this information also disappeared from our feed.

CME also introduced quite a significant change. Historically, CME had always had an overnight break where Globex CME Globex CME's electronic trading platform had always had this maintenance window after starting at four pm central time, five pm eastern, where essentially everything just stopped for forty five minutes before the pre open restarted for the next day.

Due to the way the crypto market works, if you look at crypto exchanges, of course, one of their main features is that you can trade on them 24/7.

It doesn't matter whether it's the weekend or any other time. You can always trade on most crypto exchanges.

To address the discrepancy in availability between CME's crypto contracts, which just like everything else on CME, closed every day for an hour and closed over the whole weekend. At the end of May, CME changed this so that at four pm central, the break was reduced to rather than being an hour, was just two minutes. So now the market closes at four pm. All of the usual end of day stuff happens in the space of a minute or two, and then the market reopens at four zero two.

Once it's reopened at four zero two on Friday, then they do actually shut down from two to four am on Saturday. But otherwise, CME crypto product trade for the entire weekend. Aside from that, there have been other smaller changes that we've implemented as well. So for example, Athens replaced their data feed.

We developed a new one that's available now. There have been enhancements to order book information for Canadian venues and trade information for Dublin, Euronext, Korean venues and others. So there have been lots of other little changes that have also been taking place in recent months.

Okay, so over the last year, as well as changes to how we've been receiving the data, we've also been expanding our coverage. If you just go over slightly a year ago, the Montreal data set went from just including futures to expanding to also including options. Then at the start of this year, NSC India, which we've just been discussing, we expanded the coverage from just futures to also include options and go from a level one to a full level three book depth, given that we're consuming the PCAPS both for the equity market and the derivatives market. That equally applies to Bombay, where it went from equities only to also including derivatives. Then in terms of how we support futures, we added in support for continuous contracts, specifically the first twelve months rolling, plus able to look at the front month with the contract with the maximum volume or the contract with the maximum open interest rather than just rolling expiry.

And people can query those with bar symbology, so the exchange product code and then the month, or they can specify the Bloomberg symbology for continuous contracts. Then still in February, we added the MaiEx futures exchange.

Then getting into March from Taiwan, so another futures exchange.

We also started adding our fixed income data set in terms of the FINRA TRACE data set. Slightly more esoteric New Zealand dairy features, which are hosted by the Singapore Exchange.

Then Bitnomial, so another kind of a crypto venue.

Then at the start of June, we added historical US earnings events. So if people are querying the US equity data they can now combine it with the time when earnings were released and the times when the earnings conference call was released.

We also added US market share statistics.

So if you're looking at stats around how individual exchanges, their quotes compared to the NBBO, or their trading compared to the NBBO, they're also additional data sets. Now we're continuing to add data sets across the rest of the year.

In late June, we're adding key data metrics for US equities.

We're also adding US book depth, and we'll have a separate webinar talking about that, where we're adding the level three book depth from PCAPS for the CBO markets, the kind of Nasdaq markets, and the NICEE markets. So four CBO, three Nasdaq, and four NICEE. And that will allow people to query book depth for each of these markets individually and also as a combined composite across nearly all US liquidity. Now as we go into Q3 and Q4 we'll be continuing to add more venues, especially more book depth across other regions around the world. Thank you for watching. If you have any questions, please let us know.

Fantastic. Thank you, Chris. Thank you, Peter. We're now going to answer some of the questions that have been coming in.

Are there any changes to open and close auction with respect to twenty three point five trading in the US?

No, those will still happen at the same time. There's no plan changes. The reopen at nine pm Eastern, so that's just continuous trading. They're not going have an opening auction then.

And the closing auction will still be at four pm with the post market period from four to eight pm as it is now.

Question,  Are US options markets also expanding their hours to mirror US equities?

In general, the answer is basically no.

CBOE is currently planning to extend their trading hours slightly for options on some of the more liquid equities out there. But in general, just because there's a variety of additional considerations when it comes to the options markets, such as option expiration, exercise, assignment, all of this extra stuff.

There's strong incentives for processing to be more or less synchronized.

So we haven't seen any drive from the options markets really to, well, the equity options markets at least to change their trading hours so that they're different to other venues.

So right now, yeah, I mean, you have some exceptions like SPX options on C1 have an overnight session.

And as I just mentioned, you will have options on some of the most liquid equities out there. So like the Nvidia, the Amazons and so on are going to have a pre opening session. I think it starts at like seven thirty Eastern. That's the plan.

So a short pre opening session, and they're going to trade for another fifteen minutes end of day. But apart from that, that's what we're seeing in at least the US options markets.

Right. Thank you.

Another question here. Are there any non US markets that have announced they'll be expanding their hours?

Yeah, I mean, again, there isn't any great push towards overnight trading in global equity markets. I mean, there's been a few equity markets that have been expanding their trading hours over time, but it's not a huge coordinated thing. Like the German markets now, those trade from around seven thirty am, eight am to ten pm Central European Time. So that's been expanded fairly recently.

In Korea as well, you have a new Nextrade ATS, which now offers I think it's eight am to eight pm trading, something like that.

And of course, yeah, Japan Next in Japan has offered overnight trading for years. But the pace of progress there is pretty slow. We don't see any rush towards moving towards twenty three point five trading as in the US.

Peter, if you have the answers on how do you actually get notified of upcoming changes?

Okay. Anyone that is registered for OneTick Cloud and permission to a particular dataset, we will notify them via email of a change. Also, if they go into OneTick Cloud, in their profile page, there's a list of past announcements irrespective if they're permission to the dataset or not.

Oh, fantastic. Great. Another question here, maybe newer to OneTick Cloud, is how do you go about retrieving the data?

Okay. Well, there are a few ways. If you're permissioned, you can pull down Parquet from our s three bucket for each table and each database, or you can use our APIs, which are kind of REST, SQL, or Python to pull down data. And we provide hundreds of examples of how to pull down data and how to run increasingly complex analytics on that data.

Fantastic. We'll be sure to include the link. I'll make sure that's available for people to check out. Another question here, will changes always be backloaded?

Yes. And they'll be backloaded to the point in time when they start. So for example, for the US Odd Lot data, we will be backloading the history to the point when the regulations changed.

The same goes, not really regulatory changes, but when we're adding in AI feature sets, as well as kind of taking that feature set forward, we then go back through the history. So you can look at the kind of summary of the information for key metrics going back in time.

Thank you, Peter. Let's see. Another question here we have.

For US book depth, what about other US venues?

MEMEX, IEX. They're on our list to add in Q3.

first point is getting all the major venues out and available for querying, and then we'll add the minor ones.

So that will include as dated, Memex and IEX plus others.

Great. Couple more questions here.

Do you have book depth for non US markets?

K. We provide book depth for most European primary markets, most derivative markets globally, and we're continuing to add book depth for secondary markets and markets in APAC.

So for example, if you wanted to look at full kind of l three book depth from PCAPS for CME, you can do that today.

Great. One last question here. Can you use MCP and/or AI against OneTick Cloud data?

Yes. So you can either write queries in SQL or Python against all of the data that we're providing on demand, or you can use our MCP server to just write in natural language what you want, and that will build the queries for you, which you can then execute.

And you can choose whether you're gonna be writing SQL queries or Python queries, depending on whether you're more of a Pandas person or more of a SQL person.

And that MCP server really covers the the combination of understanding syntax. So for example, how to return a return the TUAP at spread, and then coverage so that if you're looking for US data, this is where that data is available.

Thank you all for your time today. Thank you, Chris.

Thank you, Peter.

Do reach out either on our website, www.onetick.com or www.kx.com. And you can always email info@onetick.com to learn more.

Thank you both so much. Thank you to all.