Candid Conversations for Financial Markets

Vincent Maurin explains how exchange data pricing impacts trading costs and argues for holistic regulatory approaches.

36 min Listen to the original Episode page
Market MicrostructureExchange DataRegulatory PolicyTrading CostsMarket MakersPrice DiscoverySIP DataAlgorithmic Trading

Summary

Vincent Maurin explains that while retail investors do not directly pay for exchange data, costs are embedded in trading fees and spreads. He argues that exchanges subsidize trading fees to generate data, creating a cross-subsidy effect similar to platform businesses. Maurin notes that monopolistic data providers may withhold information to maintain buyer advantage, but banning data fees could increase data quality while raising trading fees. He suggests that improving regulated public data feeds, rather than imposing price caps, creates competitive pressure on proprietary feeds. Finally, he warns against siloed regulation, urging regulators to consider the entire trading chain including connectivity and spreads.

Key takeaways

  1. 1
    Retail investors rarely pay directly for exchange data; costs are passed through via broker fees or wider market maker spreads, meaning data pricing affects total execution costs even for passive investors.

    Unidentified speaker (speaker_1) · 2:16

Episode sections

  1. 0:02 – 2:48

    Introduction and Trading Cost Pass-Through

    Laura Paglia introduces the topic of exchange data pricing and its impact on investor costs. Vincent Maurin explains that while retail investors may not directly pay for data, the cost is embedded in trading fees or market maker spreads, ultimately increasing total trading costs.

  2. 2:48 – 5:48

    Interconnectedness of Trading and Data Businesses

    Maurin argues that exchanges operate as multi-service platforms where trading and data sales are linked. Exchanges may lower trading fees to attract volume, which improves the data product they sell, similar to how ride-hailing apps subsidize rides to collect valuable data.

  3. 5:48 – 7:27

    Impact on Total Trading Costs

    The speaker clarifies that selling data can lower exchange trading fees to attract users, which reduces the overall trading cost for investors when combined with market maker spreads, contrary to the assumption that data fees simply add to costs.

  4. 7:27 – 11:41

    Market Maker Data Purchase and Information Revelation

    Maurin explains that in their model, only market makers buy data to gain an edge over speculators. Their quoting activity inadvertently reveals information to other investors, improving liquidity and lowering costs for those who do not pay for the data.

  5. 11:41 – 15:26

    Incentives to Withhold Information

    Referencing Anand and Fiderer, Maurin notes that monopolistic data providers have an incentive to limit data transparency. If prices fully reflect data, buyers lose their advantage, so exchanges may withhold information or sell it at high prices to maintain value.

  6. 15:26 – 19:21

    Effects of Banning Data Fees

    The model predicts that banning data fees could increase data quality because exchanges lose the incentive to withhold information. However, it would also likely increase trading fees, potentially leading to a net negative impact on market welfare.

  7. 19:21 – 21:36

    Moderate Fee Caps and Regulatory Tools

    Maurin suggests that moderate caps on data fees might help, but warns that such caps can have knock-on effects on other fees like connectivity. He notes that better regulatory tools exist beyond simple fee caps.

  8. 21:36 – 27:30

    Measuring Price Discovery Contribution

    The speaker critiques using trading volume alone to measure an exchange's contribution to price discovery. He suggests that the ability of quotes to forecast future price changes is a more accurate metric than volume, which can be driven by uninformed retail traders.

  9. 27:30 – 28:23

    Evaluating New Data Products and Market Power

    The speaker suggests tracking the total cost of all data feeds provided by an exchange to determine if new product offerings are intended to increase market power or simply offer product differentiation for customers with varying willingness to pay.

  10. 28:23 – 30:10

    Regulated vs. Proprietary Data Feeds

    In the US context, there are two types of data: regulated 'SIP' data (often top-of-the-book) and proprietary exchange feeds (faster, more comprehensive). The debate focuses on the comprehensiveness of the regulated feed rather than restricting proprietary feeds directly.

  11. 30:10 – 31:34

    Competitive Pressure from Enhanced Regulated Feeds

    Improving the regulated feed with more information (e.g., depth of book) puts competitive pressure on exchanges, forcing them to disclose more data in their proprietary feeds to maintain an edge and potentially lowering prices for proprietary data.

  12. 31:34 – 32:15

    Knock-on Effects on Trading Fees

    If exchanges charge less for data due to competition from regulated feeds, they may charge more for trading services. Therefore, regulating data fees in isolation can have negative knock-on effects on other parts of the trading chain.

  13. 32:15 – 33:02

    Policy Recommendation: Public Data Feed vs. Price Caps

    The speaker argues that having a regulated public data feed that competes with exchange feeds is a better policy tool than simply imposing a cap on direct exchange data fees, as it addresses market dynamics more effectively.

  14. 33:02 – 35:31

    Avoiding Siloed Regulation

    The biggest mistake regulators can make is looking at data fees in isolation. Regulators should consider all fees in the trading chain (data, trading, spreads) because these businesses are connected and regulating one segment affects others.

Assets discussed

AssetKindStanceContext
AmazonAMZNequityneutralAnd this is very specific to financial markets. You know, it wouldn't apply maybe to platforms like Amazon or Uber.

“Stance” is how the speakers talked about the asset, not a recommendation.

Claims worth verifying

Vincent Maurin is an Associate Professor of Finance at HEC Paris.

Unidentified speaker (UNKNOWN) ·0:32

How to verify: Check HEC Paris faculty directory or professional profiles (e.g., LinkedIn, university website). For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

The paper 'The Price of Exchange Data' was co-authored by Vincent Maurin and Thierry Foucault and published in July 2026.

Unidentified speaker (UNKNOWN) ·0:32

How to verify: Search academic databases (SSRN, Google Scholar) for the paper title and authors. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Retail investors typically do not pay directly for exchange data; costs are embedded in trading fees or spreads.

Unidentified speaker (speaker_1) ·1:21

How to verify: Review brokerage fee schedules and market structure literature on cost pass-through. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Market makers are the primary buyers of exchange data in the model.

Unidentified speaker (speaker_1) ·8:52

How to verify: Review the model assumptions in Maurin and Foucault's paper. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Terry and Short's paper identifies that only some market participants buy data and that liquidity improves for non-subscribers.

Unidentified speaker (speaker_1) ·8:18

How to verify: Locate the paper by Terry and Short in financial journals. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Anand and Fiderer's 1980s paper shows monopolistic data providers have an incentive to withhold information.

Unidentified speaker (speaker_1) ·12:22

How to verify: Search for the paper by Anand and Fiderer from the 1980s. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

The Canadian Stock Exchange offers a free data feed specifically for ETFs.

Unidentified speaker (speaker_1) ·27:10

How to verify: Check the Canadian Stock Exchange website or regulatory filings for data fee schedules. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Primary source

The speaker references a paper titled 'The Price of Exchange Data' by Vincent Maurin and Thierry Foucault.

Unidentified speaker (UNKNOWN) ·27:30

How to verify: Search for the publication 'The Price of Exchange Data' by Vincent Maurin and Thierry Foucault in academic databases like SSRN or Google Scholar. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

In the US, there is a regulated data feed called 'SIP' data, which traditionally contained top-of-the-book quotes.

Unidentified speaker (speaker_1) ·29:10

How to verify: Check SEC filings or Nasdaq/CBOE websites describing the Consolidated Tape Association (CTA) and UTP Plan, which define SIP data requirements. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Primary source

The US SIP feed has progressively included more information about the depths of the book beyond just top-of-the-book quotes.

Unidentified speaker (speaker_1) ·29:53

How to verify: Review SEC regulations or exchange filings regarding the implementation of Level 2 or depth-of-book data in the SIP. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Primary source

The speaker's research finds that more information in the regulated feed leads exchanges to disclose more data in their proprietary feeds.

Unidentified speaker (UNKNOWN) ·30:34

How to verify: Verify by reading the full text of the paper 'The Price of Exchange Data' by Maurin and Foucault. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

In Canada, regulated fees are currently only for display fees for professional users.

Unidentified speaker (UNKNOWN) ·34:32

How to verify: Check Canadian Securities Administrators (CSA) notices or Canadian exchange websites (e.g., TMX Group) regarding market data fee regulations. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

More trading in Canada is being done by algorithms using non-display data.

Unidentified speaker (UNKNOWN) ·34:40

How to verify: Review reports from the Canadian Securities Institute or market structure studies from Canadian exchanges regarding algorithmic trading volumes and data usage. For recent events, verify 2026-08-16 through 2026-11-16. Use another period only when the speaker explicitly names it.

Do your own research

  • Verify the specific trading fee structures and data subscription costs for your broker and exchange of interest.
  • Review the latest SEC filings or exchange rulebooks to understand the current scope of SIP (Securities Information Processor) data versus proprietary feeds.
  • Check recent academic papers or industry reports on 'Terry and Short' studies to validate claims about liquidity improvements for non-subscribers.
  • Analyze whether your trading strategy relies on 'display' or 'non-display' data, particularly in Canadian or other international markets, to assess regulatory exposure.
  • Compare the latency and depth of book information available in regulated feeds versus proprietary feeds for your asset class.
  • Investigate if your exchange has recently introduced new data products or changed fee structures that might indicate increased market power.
  • Look for regulatory comments or proposals regarding 'knock-on effects' of data fee caps on trading fees and connectivity charges.
  • Assess the total cost of ownership for data feeds, including connectivity fees, to determine if lower data prices are offset by higher trading or access costs.
Full transcriptexpand
0:02
Laura Paglia: Welcome to Candid Conversations for Financial Markets Podcast Series. I'm your host, Laura Paglia, President and CEO. Our podcasts are designed to provide insights, transparency, and provocative yet achievable proposals for matters that move financial markets in Canada.
0:20
Laura Paglia: Stock exchanges earn money, both from bringing buyers and sellers together and from selling information generated. by trading. Regulators worry that the price of that information is too high,
0:32
Unidentified speaker (UNKNOWN): but
0:32
Laura Paglia: a lower data bill is not necessarily a lower cost for investors. Our guest, Vincent Moran, Associate Professor of Finance at HEC Paris, co authored a paper in July of this year, The Price of Exchange Data. The research challenges both the claim that selling data necessarily produces better information. information and the claim
0:52
Laura Paglia: that restricting data fees necessarily benefits markets. Welcome, Professor Morat.
0:59
Unidentified speaker (speaker_1): Thank you very much, Laura. It's great to be here and thank you to the Canadian Forum for Financial Markets for hosting me today. It's great to talk about this issue.
1:07
Laura Paglia: We are delighted to have you. So Let's start with why the price of the market data should matter.
1:16
Laura Paglia: Who ultimately pays for it and how?
1:21
Unidentified speaker (speaker_1): Yeah. Well, so that that's that's a good question that retail investors might ask themselves. Uh, because not always obvious, you know, uh who pays uh for for exchange data. So I would say, you know, the simple answer is that uh the price of exchange data adds to other trading costs in a way. So so let's take a simple example. like you're a retail investor, you want to
1:41
Unidentified speaker (UNKNOWN): trade
1:41
Unidentified speaker (speaker_1): a mining stock mining stock on a Canadian stock exchange. So you will you would basically place an order with your broker. And so maybe that broker is you know paying for market data from the different exchanges
1:56
Unidentified speaker (speaker_1): with which it's connected. Right. So so uh and even if your broker uh uh himself or herself doesn't pay for that data, maybe maybe the market makers that are quoting uh on these exchanges would pay for this data. And so that will maybe impact uh the spreads that they are quoting. So the kind of the the the idea
2:16
Unidentified speaker (speaker_1): is that you know, even if it's not directly the retail investors, somebody along this trading chain, let's say, uh would would pay, you know, for this, for this, for this data, and that will increase the total trading cost. uh between the ultimate buyer of a security and the ultimate seller of that of that security. So I would say that's that would be my my first answer uh to this
2:35
Unidentified speaker (speaker_1): question.
2:37
Laura Paglia: Okay. And in in that context, the exchange earns money from the from the trades and from selling information about the trades.
2:48
Unidentified speaker (UNKNOWN): So
2:48
Laura Paglia: that's two sets of businesses. What changes when we look at those businesses together instead of separately?
2:58
Unidentified speaker (UNKNOWN): Yeah,
2:58
Unidentified speaker (speaker_1): no, that's that's a good question. In a sense, that's that's at the heart of our of our research. So I'm I'm glad that you asked this this question. Uh so you know, to be fair, in a sense, exchanges you can view them as multi-services Services platforms. And you know, for a long time, the only thing that exchanges were doing were basically to bring people together to trade. and they were charging trading fees for that's for that service.
3:18
Unidentified speaker (speaker_1): Uh now they provide a host of ancillary services and data is is one of them. But you can also think about you know collocation uh services and so on. And so basically what what our what our you know papers say is that you kind of need to consider uh you know the s the sale of trading services and the sale of data together because these businesses kind of
3:38
Unidentified speaker (speaker_1): interact. And in a way, I think it would be surprising if I told you that there's no relationship between this these two businesses. So let me try to be a bit more specific.
3:47
Unidentified speaker (UNKNOWN): So
3:48
Unidentified speaker (speaker_1): one obvious way in which they interact is that basically the input for the data that the exchanges are selling. So these are quotes and trades basically you can think that the exchange will think well the more quotes costs and trades I have on my platform, the better the data that I'm going to supply.
4:02
Unidentified speaker (UNKNOWN): And
4:02
Unidentified speaker (speaker_1): so that can give exchanges incentives to reduce the cost of trading, so to lower the trading fees, you know, in order to get you know better data that they can they can they can sell. So that's that's one reason. Uh and here you can really think about you know a ride hailing apps, or think about Uber or La Lyft or these kind of companies.
4:22
Unidentified speaker (speaker_1): You're probably, as a customer, you're probably not paying marginal costs for your ride, uh, maybe because it's these platforms are collecting data about you and that's valuable for them. So that's that would be that I would say is the kind of the closest analogy you can you can think
4:34
Unidentified speaker (UNKNOWN): of. Now
4:34
Unidentified speaker (speaker_1): for for exchanges, there's something more which is that uh you know market participants who buy this data to you know inform their trades or to place their quotes, they use the data in a sense on the exchange who sells the data. you know. And so that has further implications uh because, for example,
4:54
Unidentified speaker (speaker_1): you're not going to use data on an exchange that charges super high trading fees. So basically maybe the exchange will lower the trading fees for you know to increase the use value of that of that data.
5:07
Unidentified speaker (speaker_1): And so finally like one reason in which interact is that you know because the data you set you sell is used on on your platform as an exchange uh now you know the data that you sell and the quality of that data will will maybe impact trading volume so there will be no there will not be an effect from your data business back to your traditional business of charging
5:27
Unidentified speaker (speaker_1): trading fees uh for providing trading services so so in a sense the the point of our research is try
5:33
Unidentified speaker (UNKNOWN): to
5:33
Unidentified speaker (speaker_1): make sense of this these complex interactions between the two businesses
5:37
Unidentified speaker (UNKNOWN): i
5:37
Laura Paglia: i i see thank you So you've described various scenarios where the exchange would be incentivized to lower the cost of the data. Um
5:48
Laura Paglia: does that necessarily always mean that the investor's total cost of trading falls?
5:56
Unidentified speaker (UNKNOWN): Yeah,
5:56
Unidentified speaker (speaker_1): so so just to make sure I understood your question. So what we're saying is that the fact that you're selling data might might lower the trading fees that are charged by the exchange. So not necessarily the price of the data itself. But uh but you know, so if if I understand that correctly, um yes, so you're right that in a sense the trading fee that you know that is paid to the
6:16
Unidentified speaker (speaker_1): exchange is just one component of the total trading cost for for an investor. Uh to that, you want to add at least uh the spread that is you know charged by the market maker between the best bid and the best, the best, the best, the best, the best offer.
6:31
Unidentified speaker (speaker_1): And so, you know, what our model says is that if the exchange sells data, they will lower the trading fees to attract more volume. Right. So that's that's one obvious reason why they do it. Um and also to make people willing to buy their data, because again, I'm I'm not gonna buy data and use it on an exchange if it's you know prohibitively costly to trade on that exchange.
6:51
Unidentified speaker (speaker_1): Um and so that's about the exchange trading fee. Now, in our model, it turns out that when the exchange lowers their trading fee for that reason, the total trading cost also goes down, right? So if you also include uh the market maker spreads, which is what you know an investor would do to try to compute their the cost, the cost of the cost of trading, right? So that's kind of the of
7:11
Unidentified speaker (speaker_1): the leading effect in our models. So trading fee goes down. Uh because the exchange is incentivized to lower this trading fee to get better data and to to sell this data at at a higher price. And that leads to a decrease in the overall trading cost for investors. That that's correct.
7:26
Unidentified speaker (UNKNOWN): Okay.
7:27
Laura Paglia: Um thank you. You you mentioned the market maker, the role of the market maker. In your research model, the market maker buys the data
7:36
Laura Paglia: from, and this is a model from traders who never pay for it and can benefit. So
7:43
Unidentified speaker (UNKNOWN): could
7:44
Laura Paglia: you walk us through that? Um
7:47
Laura Paglia: the result
7:49
Laura Paglia: is the result that the market maker is revealing its market knowledge through quoted prices?
7:59
Unidentified speaker (speaker_1): Yeah, so so yeah, that that's an interesting point. So basically. I understand your question right. So, and as you say, it's it's a model, so we we do not pretend to capture every aspect of that of that situation. So uh let me start by by by a bit of empirical research, which is not from from us, right? But a paper by Terry and the short and
8:18
Unidentified speaker (speaker_1): quarters. It's really hard to know in practice This who buys you know the data and which data do they buy, because as we as guess we're gonna going to talk about later, there are there are you
8:26
Unidentified speaker (UNKNOWN): know
8:26
Unidentified speaker (speaker_1): tons of data products out there with different speed, different you know, degrees of granularity and so
8:32
Unidentified speaker (UNKNOWN): on.
8:33
Unidentified speaker (speaker_1): So in in the case of like the the New York stock exchange, Terry and and and his quarters, uh they they were able to identify who buys the data. So they can tell you, okay, this market maker buys the data, this trading participant doesn't buy the data, right? And so I'm saying this because in our model it's true. that in equilibrium, the
8:52
Unidentified speaker (speaker_1): only players that buy the data are the market makers, right? And so what what happens then is that
8:59
Unidentified speaker (UNKNOWN): the
9:00
Unidentified speaker (speaker_1): market maker buys the data, other investors don't. And it in a sense, in equilibrium, via the quotes, you know, the market makers will kind of reveal part of their of of their of their data. And just to be clear, they're not doing it out of the goodness of their heart, you know, it's just an it's just an equilibrium outcome that the quotes that they post will
9:19
Unidentified speaker (speaker_1): kind of help investor figure out what's the informational content of the data in the first place.
9:24
Unidentified speaker (UNKNOWN): And
9:24
Unidentified speaker (speaker_1): so then you might ask, okay, so why would the market maker buy that data if ultimately its quotes will reveal that data and so that will lose their information advantage? And so that's that's where like you know the model is kind of useful. because there we say
9:40
Unidentified speaker (UNKNOWN): the
9:40
Unidentified speaker (speaker_1): market maker basically, you know, getting the data is kind of a way to get to get ahead of other investors, right? So you might have other investors in the market that are speculating on other types of information. You can think of you know credit card information, everything that is not market data.
9:56
Unidentified speaker (UNKNOWN): And
9:56
Unidentified speaker (speaker_1): so basically, for the market maker, you've you they're thinking: if I buy the data, I get ahead of this. these people, I reduce my losses uh with these speculators, let's say, and then I can make more money on like you know, the typical retail investors uh that trade not on information.
10:15
Unidentified speaker (UNKNOWN): So,
10:15
Unidentified speaker (speaker_1): in a sense, what we get out of this is that it's profitable for the market maker to buy the data. That's why that's why they do it in the first place, that that allows them to increase their profit.
10:23
Unidentified speaker (UNKNOWN): Uh,
10:23
Unidentified speaker (speaker_1): but at the same time, you know, because the market makers are quoting prices on this exchange. these exchanges, their quoting activity kind of reveals part of the data
10:32
Unidentified speaker (speaker_1): to
10:33
Unidentified speaker (speaker_1): investors. And where I want to, if you, if you don't mind, where I want to finish with this, is basically what we get out of this is that even if a subset of investors, let's say market makers in our case, buys the data,
10:46
Unidentified speaker (UNKNOWN): it
10:46
Unidentified speaker (speaker_1): doesn't mean that only them benefit. Other investors can benefit from that.
10:51
Unidentified speaker (UNKNOWN): And
10:51
Unidentified speaker (speaker_1): here I'm going back to the to this paper that I mentioned by Terry and the Shot and Quoters. What they show is that
10:58
Unidentified speaker (speaker_1): only some market participants buy the it's it's called the nice integrated fit data in their in their case. Only some market participants buy the data. And of course, for them, like kind of trading cost goes down, profit goes up, so that's what you would expect.
11:10
Unidentified speaker (UNKNOWN): But
11:10
Unidentified speaker (speaker_1): they also show that liquidity and trading liquidity improves and trading cost costs goes down. Also for people who don't subscribe to the data. So there seems to be this positive spillover
11:20
Unidentified speaker (speaker_1): from people who buy the data and maybe reveal it implicitly via their quotes and people who don't buy the data. So that's that's kind of what we're capturing in the in the model.
11:29
Unidentified speaker (UNKNOWN): Well,
11:30
Laura Paglia: thank you. And in in looking at these incentives that you're researching and the effect of those incentives. Does
11:41
Laura Paglia: the concept of selling data can that give an exchange any reason to make the data less informative? So, for example, is there an incentive to hold back useful information?
11:54
Unidentified speaker (speaker_1): Yeah. No, so so so I see where your question is coming from, and and this is uh, you know, a point that uh we as academics are familiar with, and when you talk to practitioners, they will say, Okay, what do you mean? Like, what does it mean to hold back on information?
12:06
Laura Paglia: Yeah.
12:07
Unidentified speaker (UNKNOWN): So
12:07
Unidentified speaker (speaker_1): so first let me say that uh it's kind of a classic result in in economics, and then I'm gonna tell you, you know, the way it works
12:15
Unidentified speaker (UNKNOWN): in
12:15
Unidentified speaker (speaker_1): uh in a model, and and then maybe I'll I can map this out to to uh
12:19
Unidentified speaker (UNKNOWN): to
12:19
Unidentified speaker (speaker_1): what happens in practice.
12:21
Unidentified speaker (UNKNOWN): Uh
12:22
Unidentified speaker (speaker_1): so the basic idea, you know, there's a there's an academic paper from the 80s actually by Anmaty and Fiderer that shows the following.
12:30
Unidentified speaker (UNKNOWN): If
12:30
Unidentified speaker (speaker_1): you're the monopolistic provider of some data, and exchanges are in a sense, you know, the Toronto Stock Exchange is the monopolistic provider of data coming from the Toronto Stock Exchange, right? I mean that's that's a tautology, but but but that's true. Then you know you don't want to disclose,
12:46
Unidentified speaker (UNKNOWN): you
12:47
Unidentified speaker (speaker_1): don't want to sell too much data. And this is very specific to financial markets. You know, it wouldn't apply maybe to platforms like Amazon or Uber. And the reason is that you know, at a high level, when you sell data to market participants, right, well, they're gonna trade on that information. And that information is gonna be reflected via quotes and prices, right?
13:07
Unidentified speaker (UNKNOWN): So
13:07
Unidentified speaker (speaker_1): if at some point you know
13:09
Unidentified speaker (UNKNOWN): too
13:09
Unidentified speaker (speaker_1): much information is revealed by quotes and prices, uh, then it doesn't make sense for marketing. marketplace plans to buy the data because they see too much of it in the prices already. Right. And so basically that gives you an incentive on exchange not to be too transparent, right? So to withhold some of the data that you're selling, because otherwise, all the data is reflected
13:29
Unidentified speaker (speaker_1): in into prices. Nobody has incentives to you know to buy to buy that to buy that data. So, you know, you know, in a our paper, the way it works is that. I mentioned this these market makers and the speculators. So again you can think of speculators as let's say you know uh we don't view it as a negative term but it could be hedge fund or like any type of
13:49
Unidentified speaker (speaker_1): trader trading on like non market information. Again think about them trading on credit card information about companies and so on and so forth.
13:55
Unidentified speaker (UNKNOWN): So
13:56
Unidentified speaker (speaker_1): as I said look in the model the market makers buy buys the data
13:59
Unidentified speaker (UNKNOWN): you
13:59
Unidentified speaker (speaker_1): know basically to reduce their information disadvantage vis a vis-a-vis vision the speculators, right? They want to get ahead of them and reduce their loss vis-a-vis these these traders.
14:08
Unidentified speaker (UNKNOWN): And
14:09
Unidentified speaker (speaker_1): so, in a sense, if the market maker has data that is too good, right, there's no room to to make profit for speculators. But if there's no room for speculators to make profit, market makers do not have incentive to buy the exchange data, the market data in the first place. So this is kind of a similar logic to the one I explained before.
14:25
Unidentified speaker (UNKNOWN): Uh
14:25
Unidentified speaker (speaker_1): and that's what we
14:27
Unidentified speaker (UNKNOWN): that's
14:27
Unidentified speaker (speaker_1): why we get this result. So you might ask me, okay, so in practice, what does it mean? What does it mean to resort information and so on and so forth?
14:35
Unidentified speaker (UNKNOWN): So
14:36
Unidentified speaker (speaker_1): first, you know, it's hard to have access, you know, to to know, sorry, how granular the data is that exchanges have access to, right? So they might have like very good data, but they might only sell part of it. It will be kind of hard to tell.
14:49
Unidentified speaker (UNKNOWN): But
14:49
Unidentified speaker (speaker_1): one way you can tell. that they're doing kind of part of this is that there's a lot of product differentiation. And for example, like the best data would be sold at a very high price, right? And you might say, well, that's maybe because only some customers are willing to buy it, but that's also a way for the exchange to limit
15:08
Unidentified speaker (UNKNOWN): the
15:08
Unidentified speaker (speaker_1): set of participants Participant that have access to this, to this very good, this very good data. So it's not a proof of what we we're saying in the model, but that's somewhat symptomatic of maybe this behavior of you don't want the base data that you have to be available to the broad broadest set of market participants.
15:26
Unidentified speaker (UNKNOWN): Okay.
15:26
Laura Paglia: So let's let's say the regulators theoretically say to the exchanges, you can't charge for data. Nothing.
15:35
Unidentified speaker (speaker_1): Right.
15:35
Laura Paglia: Um what does uh your model makes
15:39
Laura Paglia: predictions about
15:42
Laura Paglia: um better information but worse overall outcomes can you uh walk us through what your research shows in that type of a scenario
15:53
Unidentified speaker (speaker_1): yes so that's that's that's a great question so basically the what we're looking at is is a bit of an extreme form of regulation so you know like in in in Canada as I understand it, uh, you know, exchanges cannot do anything they want in terms of of data fees, right? So there's some guidance, there's some interval that uh there the the data
16:13
Unidentified speaker (speaker_1): fees have to lie within this interval. Um, and of course, you know, we would like to capture uh this kind of limits on your ability to charge fees. What we do, you know, because it's simpler in the model, is to say, well, suppose you cannot charge for data alone. at all, right?
16:28
Unidentified speaker (UNKNOWN): And
16:28
Unidentified speaker (speaker_1): so here we find two results that are kind of interesting, I think, which are the results that you referred to. Uh, first, you know, marketplaces will tell you, well, if you cannot charge for data, then you know the quality of our data will decrease, right? And we find the opposite. So, so so why is that? First, in a sense, you know, data is a
16:48
Unidentified speaker (speaker_1): byproduct of the exchange activity, right? So they they get some of that of this data for free, quote unquote, because you know, market makers, brokers, investors are trading on their exchange, right?
16:59
Unidentified speaker (UNKNOWN): So
16:59
Unidentified speaker (speaker_1): if you you know, in a sense, you could think that the marginal cost of producing the data for for exchanges is probably close to close to zero. Um, so and the second thing is that
17:09
Unidentified speaker (UNKNOWN): if
17:10
Unidentified speaker (speaker_1): you cannot monetize data, you will only care about data to the extent that selling better data, having a more transparent marketplace, would increase trading volumes. Because as a retail investor, you're thinking, well, you know, if that market is very transparent, uh, I'm not afraid of you know my quotes or my trade being picked up by speculators, I will trade more,
17:30
Unidentified speaker (speaker_1): right? And so basically, what you see is that because of these complementarities between the businesses, just because you're as an exchange or you're making money. out of trading fees, you have a you have an incentive to kind of sell very good, very high quality, very high quality data. Right. And so when I take
17:49
Unidentified speaker (speaker_1): away the profit you're making from the data, because I tell you you cannot charge for them, what's happening is that I also kill this incentive that we just discussed to limit the quality of the data, right? Because you want to limit the quality of the data to increase the market value of that data.
18:03
Unidentified speaker (speaker_1): If you don't earn the market value of that data, that incentive is gone. So ultimately,
18:09
Unidentified speaker (UNKNOWN): our
18:09
Unidentified speaker (speaker_1): model, what our model says is that if you cannot charge for data, data quality might actually increase instead of decrease.
18:16
Unidentified speaker (UNKNOWN): And
18:17
Unidentified speaker (speaker_1): then what's the over effect? So for the over effect, that's where you care about the trading fees. Because essentially, what our model says is that if exchanges cannot charge for data anymore.
18:27
Unidentified speaker (speaker_1): then they will tend to increase their trading fees.
18:31
Unidentified speaker (UNKNOWN): And
18:31
Unidentified speaker (speaker_1): the reason is kind of simple, is that
18:33
Unidentified speaker (UNKNOWN): the
18:34
Unidentified speaker (speaker_1): fact that I can sell data was one reason for me to keep my trading fees low in the first place, right? Because I want people to buy my data and to use it on the on the exchange.
18:43
Unidentified speaker (UNKNOWN): And
18:43
Unidentified speaker (speaker_1): so if I cannot charge for data, you kill my incentives to keep my trading fees low, right? So trading fees might increase.
18:51
Unidentified speaker (UNKNOWN): um
18:51
Unidentified speaker (speaker_1): and as a result the overall impact on uh on uh on uh investors and welfare as we as we say can be can be negative so on the one end you increase data quality on the other end trading fees might increase and we find that often the net effect actually on market quality can can actually be can actually be negative
19:09
Laura Paglia: well thank you and and as you mentioned this was an a this was an extreme scenario to prove a
19:16
Laura Paglia: point. What if we went in the middle and
19:21
Laura Paglia: a moderate fee restriction? Would your research does your research point to having the same essential effect?
19:29
Unidentified speaker (UNKNOWN): So
19:29
Unidentified speaker (speaker_1): that that's a very good question. And in a sense, I wish we had done that before the podcast. So I can I can answer the question more precisely. So So what can what I can tell you is that yes, you're right. So we look at this extreme ban on on you know like data fees so to make a point. Um you know I suspect that basically we looked at the two cases where there's no limit
19:49
Unidentified speaker (speaker_1): and a case where you cannot charge. So I suspect that you know the optimal should be somewhere somewhere in the middle. So I I I I I would say that yes I think our research would support the fact that a moderate
20:02
Unidentified speaker (UNKNOWN): uh
20:02
Unidentified speaker (speaker_1): you know uh cap on data fees can can help.
20:06
Unidentified speaker (UNKNOWN): What
20:06
Unidentified speaker (speaker_1): I do want to say though is that, you know, in a revised version of the paper that is not out yet, uh
20:12
Unidentified speaker (UNKNOWN): what
20:12
Unidentified speaker (speaker_1): we find is in in a sense there are there are kind of better tools, you know, if you want to achieve the objective of improving market quality. And the reason is that
20:20
Unidentified speaker (UNKNOWN): if
20:20
Unidentified speaker (speaker_1): you set a cap on the on the on the data fees, as I said, this this is bound to have a knock-on effect on other. fees that the that the exchange is charging. So we we're focusing on trading fees, but you know, market participants are often complaining that oh you know, data fees are not the only thing that we're paying. We're also paying for connectivity fees and so on and so forth. Um and so basically what we're
20:40
Unidentified speaker (speaker_1): saying is
20:41
Unidentified speaker (UNKNOWN): that even
20:42
Unidentified speaker (speaker_1): if you put a moderate cap on data fees, you know, it's bound to have another effect on other fees that the exchange is charging.
20:49
Unidentified speaker (UNKNOWN): Uh
20:50
Unidentified speaker (speaker_1): and so that that's why the kind of the welfare consequences are are not are not as clear. But the short answer to the question is that if the only tool that you have at your disposal uh is you know a limit or a cap on the fees that exchanges charge for data, it's likely that the moderate one would would help. Then of course, you know, empiricists will tell me okay, what's the right level?
21:10
Unidentified speaker (speaker_1): And that's where it gets really, really messy and I don't want to overclaim about what what our model can can do.
21:15
Unidentified speaker (UNKNOWN): Okay.
21:16
Laura Paglia: Very helpful. Thank you. And um as um as you've mentioned um Canada is currently revisiting how we assess professional real time market data fees. And there is a current consultation that asks uh for input on how to measure a marketplace's contribution to price
21:36
Laura Paglia: discovery, including odd lot trades, intentional crosses, opening and closing auctions.
21:44
Laura Paglia: Why is the amount that you trade not necessarily the same as the information produced or that you get?
21:54
Laura Paglia: And
21:55
Laura Paglia: what would you measure if that's the case? What would you measure instead of just volume?
22:02
Unidentified speaker (UNKNOWN): No.
22:02
Unidentified speaker (speaker_1): Yeah, that's that that's a good point. So so um, you know, as as I as I said before we started the podcast, you know, like uh this this gave me an opportunity to read a lot about the the Canadian context, which I realize to some extent is close to the US context, to some other extent closer to the European context. Uh so so let me tell you what I think based on my my understanding
22:22
Unidentified speaker (speaker_1): uh of this of this you know data fee methodology that is that is being used to uh to set like bounds uh for for data fees that exchange can can charge.
22:32
Unidentified speaker (speaker_1): So so so you know my understanding of the Canadian context is that in a sense um even when the data is is kind of sold via the information processor so the the that is that is run by the the the TMX uh it's basically investors that buy the data that have to pay the fee that is charged by the by the exchange that is provided. providing the data. So it's
22:52
Unidentified speaker (speaker_1): uh but then there's there's uh the regulator comes in and says, look, you cannot charge too much for the data that you that that you're selling.
23:00
Laura Paglia: Yeah.
23:01
Unidentified speaker (UNKNOWN): So
23:01
Unidentified speaker (speaker_1): I think my my understanding of you know the the current way it works that you know you're gonna rank exchanges according to their contribution to let's say price discovery uh in a broad sense. And and of course it's kind of hard to it's kind of hard to measure. uh this in this price discovery contribution. And so I guess where I'm a bit surprised when I see the the
23:21
Unidentified speaker (speaker_1): details that trading volume in itself uh is not necessarily telling you something about the quality of the data that the exchange is is providing. It really depends what what you mean by quality of the data.
23:33
Unidentified speaker (UNKNOWN): If
23:33
Unidentified speaker (speaker_1): I take a narrow view that you know
23:35
Unidentified speaker (UNKNOWN): how
23:36
Unidentified speaker (speaker_1): much an exchange contributes to price discovery is basically how much the quotes on the exchange kind of forecast future price moves for that for that security or for that asset,
23:46
Unidentified speaker (UNKNOWN): you
23:46
Unidentified speaker (speaker_1): could have an exchange where most of the volume is totally uninformed, right? So it's basically like retail traders like maybe uh like me trade on the exchange. So there's a lot of volume, right? But you wouldn't say that it contributes a lot to price discovery, right? Because you know this volume is not is totally totally uninformed.
24:03
Unidentified speaker (speaker_1): So again, um I want to be clear that this is not an aspect that we're looking at specifically in the model, but from the lens of our model, I would say that if you want to measure contribution to price discovery, maybe you would look at the quotes of an exchange and what's their ability of these quotes to forecast future price changes. You know, like you take an
24:23
Unidentified speaker (speaker_1): exchange and say, if I look at the quotes from this exchange. you know how well can I forecast future price moves over and above the information I get from from from other from other exchanges right so
24:35
Unidentified speaker (UNKNOWN): I
24:36
Unidentified speaker (speaker_1): think it's there's a bit of a
24:37
Unidentified speaker (UNKNOWN): of
24:38
Unidentified speaker (speaker_1): a question of what does it mean to have good data what does it mean for an exchange to contribute to the you know to to price discovery
24:45
Unidentified speaker (UNKNOWN): I
24:45
Unidentified speaker (speaker_1): think it's a very broad question and I'm sure that you know regulators are doing you know uh uh as as well as they can. I f I think the one dimension that I felt was a bit missing is that volume in itself,
24:56
Unidentified speaker (UNKNOWN): you
24:56
Unidentified speaker (speaker_1): know, is not really indicative of whether this is volume that tells you about you know uh future asset values or future securities values. So that that's that's the kind of one thing that
25:07
Unidentified speaker (UNKNOWN): uh
25:07
Unidentified speaker (speaker_1): that I would that I would that I would add maybe to that to that framework.
25:11
Laura Paglia: Thank you. But The the consultation also looks at narrower product specific feeds alongside the broader data offerings. And
25:23
Laura Paglia: in your view, when is charging separately for a different feed legitimate? Okay.
25:31
Laura Paglia: And when is when should a regulator worry about the market power?
25:39
Unidentified speaker (speaker_1): Yeah, so that's that's that's a good question. So again, you know, uh and maybe listeners would be disappointed with what I say. I I would say that product differentiation is is very important in practice. So you know you see a lot of data feeds, uh different data feeds with different level of granularity, maybe different level of speeds that are being sold.
25:57
Unidentified speaker (UNKNOWN): Uh
25:57
Unidentified speaker (speaker_1): it's not an aspect that we look at specifically uh in the in our research, but you know, again I'm gonna try to talk about this through the laps of our of our for research. Um
26:07
Unidentified speaker (UNKNOWN): so
26:07
Unidentified speaker (speaker_1): I think product differentiation in a sense is not is not a bad thing. So if I take the the US context, for example, uh you have this uh SIP, so called SIP data,
26:18
Unidentified speaker (UNKNOWN): uh
26:19
Unidentified speaker (speaker_1): which is provided relatively cheaply uh to investors,
26:23
Unidentified speaker (UNKNOWN): and
26:24
Unidentified speaker (speaker_1): it's not as comprehensive as the exchange direct fit. fits, maybe. But everybody understands that most investors do not need to have like full depth of the book, maybe uh, you know, data from exchanges.
26:36
Unidentified speaker (UNKNOWN): And so,
26:36
Unidentified speaker (speaker_1): in a sense, it it kind of makes sense to
26:39
Unidentified speaker (UNKNOWN): uh
26:39
Unidentified speaker (speaker_1): to have like two products because you're catering really to two different populations.
26:45
Unidentified speaker (speaker_1): For the question that you're you were asking about this you know, specific product offerings from the candidate Canadian context, I would say that
26:50
Unidentified speaker (UNKNOWN): it
26:50
Unidentified speaker (speaker_1): depends a lot on what you observe. So, in a sense, if if you introduce a new product uh as a way to increase you know the total price that uh investors are paying overall for the for the for the for the for the main product and for the specific product, that's probably you know raised raised the flag, right? But so when I looked a bit at
27:10
Unidentified speaker (speaker_1): what these specific product offerings were, so I found an example of the Canadian Stock Exchange. um offering like, you know, uh data feed specifically for ETFs and I think other products.
27:21
Unidentified speaker (UNKNOWN): And
27:21
Unidentified speaker (speaker_1): in that case, this data feed was for free.
27:23
Unidentified speaker (UNKNOWN): Why?
27:24
Unidentified speaker (speaker_1): So it would be it's hard to make a case that, you know, isolating a group of securities, let's say ETFs,
27:30
Unidentified speaker (UNKNOWN): and
27:30
Unidentified speaker (speaker_1): providing a data feed on this security that is free, it's hard to argue that, you know, this goes against, you know, investors' welfare and investors um,
27:38
Unidentified speaker (UNKNOWN): you
27:38
Unidentified speaker (speaker_1): know. uh against investors investors' benefit. So I think overall what I would I would say is that okay, suppose an exchange says, okay, I'm gonna introduce a new product-specific offering in terms of data feed, I will try to keep track of the total you know, uh cost of all the data feeds that they're that they are providing, right? And so
27:58
Unidentified speaker (speaker_1): that might give you an indication of whether they're using that uh, you know. to basically increase their market power,
28:05
Unidentified speaker (UNKNOWN): or
28:06
Unidentified speaker (speaker_1): maybe whether it's just you know standard product differentiation uh that can benefit customers with different different willingness to pay for data.
28:16
Laura Paglia: Is um
28:18
Laura Paglia: is restricting data fees good policy?
28:23
Unidentified speaker (UNKNOWN): So
28:23
Unidentified speaker (speaker_1): so what why what do you mean exactly by restricting data feeds?
28:27
Laura Paglia: Our discussions have been and um our discussions have been looking at um what regulators should look at, what the market incentives are.
28:37
Laura Paglia: Is it good policy for a regulator to be
28:42
Laura Paglia: examining these things and looking to actually restrict what data fees should be?
28:51
Unidentified speaker (speaker_1): Okay, so I I think it's an interesting question. So usually I we we look at it via a different angle, which is we kind of you know basically think that um at least in the US context, so maybe it doesn't apply fully to the Canadian context, uh you have two sets of products. So you have this kind of
29:10
Unidentified speaker (speaker_1): regulated uh data fit called often called seed SIP data. I mean it's a bit more complicated. There are three plans for different types of securities, but essentially think about it as I want to know the top of the book for some securities. I don't care about getting it super fast. Uh uh, and then then I then I will get it from the
29:30
Unidentified speaker (speaker_1): from the regulated uh so-called SIP data. Uh and then and then you know you have exchanges that are selling proprietary. data feeds. And of course, they would tell you that it's both faster and more comprehensive, right? So because they want to keep they want to keep an edge, right? And so the way I think because of the US debate, the
29:50
Unidentified speaker (speaker_1): way it has been debated recently in the US
29:52
Unidentified speaker (UNKNOWN): is
29:53
Unidentified speaker (speaker_1): how comprehensive should let's say the regulated feed be, right? So for a long time it only had top-of-the-book top of the book quotes. Now they're they're including progressive. progressively uh you know uh more information about the depths of the of of the book. And so what what
30:10
Unidentified speaker (UNKNOWN): in
30:10
Unidentified speaker (speaker_1): in that context basically what we're looking at is not necessarily
30:13
Unidentified speaker (UNKNOWN): uh
30:14
Unidentified speaker (speaker_1): restricting the proprietary data fits that exchanges can can sell,
30:19
Unidentified speaker (UNKNOWN): but
30:19
Unidentified speaker (speaker_1): it's more like putting putting competitive pressure on the exchanges by improving the kind of the regulated uh feed that that is available. at a at a lower cost to investors and so that's what's interesting is that we in in the most recent version of the paper we do look at that
30:34
Unidentified speaker (UNKNOWN): and
30:34
Unidentified speaker (speaker_1): there we find we find kind of kind of two things. So basically uh the more information is contained in the regulated feed the more information uh exchanges will disclose in their proprietary feeds. Why? Because they want to keep an edge, right? So you know basically before the
30:54
Unidentified speaker (speaker_1): regulated feed only had you know top of the book data. So basically, by selling depth of the book data in your proprietary feed, you already had an advantage as an exchange. Now, if you know the the the the seed data already already contains that, then you will have probably to to to to to show even more
31:14
Unidentified speaker (speaker_1): data. in your proprietary direct feeds right so so basically that that's a force that that push uh pushes exchanges to to uh sell more and better quality data at the same time what what what we what we show is that essentially you're putting competitive pressure on the exchange right so they can charge less for their
31:34
Unidentified speaker (speaker_1): data why because there is this kind of a similar product
31:37
Unidentified speaker (UNKNOWN): that's
31:38
Unidentified speaker (speaker_1): being provided
31:39
Unidentified speaker (speaker_1): for relatively low cost, right?
31:42
Laura Paglia: Yes.
31:42
Unidentified speaker (speaker_1): And so that has kind of a similar effect as you know restricting the fees that exchanges can charge for their data. You know, basically, before they were the monopolistic provider of the data,
31:52
Unidentified speaker (UNKNOWN): now
31:52
Unidentified speaker (speaker_1): you have some SIP data that contains almost the same thing as the direct feeds. And so then you can charge less for your for your for your proprietary uh data feeds. And so that you know, going back to our previous discussion, that we show can have a negative knock on effects on trading fees because again, exchanges can charge less for data, they might charge more for
32:12
Unidentified speaker (speaker_1): trading services.
32:13
Unidentified speaker (UNKNOWN): But
32:14
Unidentified speaker (speaker_1): to conclude on that, what we find is
32:15
Unidentified speaker (UNKNOWN): that in
32:16
Unidentified speaker (speaker_1): a sense, like having some kind of
32:19
Unidentified speaker (speaker_1): we call it public data feed. So it's it's not really public, right? But what we mean is
32:23
Unidentified speaker (UNKNOWN): that
32:24
Unidentified speaker (speaker_1): it's more regulated data feed that competes with the exchange feed,
32:28
Unidentified speaker (UNKNOWN): uh,
32:28
Unidentified speaker (speaker_1): can be a better policy actually than just imposing a cap on the on the on the direct on the feeds of the of the exchanges, right? So you you kind of you kind of want to have this this this tool as well, as opposed to just a simple uh cap on the fees that the exchange can charge. That's that's kind of what our our our research will will show.
32:48
Laura Paglia: And so that's that's very helpful. Yeah Is that
32:53
Laura Paglia: the single biggest mistake regulators should avoid when setting market data policy the cap? Or is it something?
33:02
Unidentified speaker (speaker_1): Well, you know, I I I I I wouldn't I wouldn't want to say that because so
33:07
Unidentified speaker (UNKNOWN): in
33:07
Unidentified speaker (speaker_1): a in a way, I would say the biggest mistake and it's easy for us as academics to claim that, you know, it's a mistake because we know that uh regulators and market participants operate under constraints that you know that we don't face because we're writing like nice, beautiful models where everything is relatively simple, uh the rate is not simple. Uh, I would say
33:27
Unidentified speaker (speaker_1): um I think what you want to avoid is to is to have a is to have a silo where you just look at some kind of data fees in isolation. And so in our claim we make is that.
33:39
Unidentified speaker (UNKNOWN): you
33:39
Unidentified speaker (speaker_1): kind of want to have a look at all the fees that are being paid, you know, uh, let's say along this trading chain. And that could be data fees, that could be trading fees, that could be market maker spreads, you know, like uh so on and so forth. Because to the extent that these businesses are connected, and we're not the only one, of course, to say that you know the business of selling data and
33:59
Unidentified speaker (speaker_1): the business of providing trading services are connected,
34:02
Unidentified speaker (UNKNOWN): to
34:02
Unidentified speaker (speaker_1): the extent that they are connected,
34:04
Unidentified speaker (UNKNOWN): uh,
34:04
Unidentified speaker (speaker_1): you're gonna have this knock-on effect. So if you try to regulate one segment in isolation, let's say data fees,
34:10
Unidentified speaker (UNKNOWN): you're
34:10
Unidentified speaker (speaker_1): probably gonna have an effect on other types of fees. And so that's that's something that that we should look at. And if you want to take an example that you know I think it's relevant for the Canadian context, even without looking at you know trading fees, you could say, okay, maybe some you know for some Canadian securities. exchanges are very competitive, so trading fees are really not important.
34:31
Unidentified speaker (speaker_1): Uh
34:32
Unidentified speaker (UNKNOWN): what
34:32
Unidentified speaker (speaker_1): my understanding is that the the fees that are regulated are only uh display fees for professional users, right?
34:40
Unidentified speaker (UNKNOWN): But
34:40
Unidentified speaker (speaker_1): you know, there's more and more trading that is being done by uh algorithm that use non-display display uh data.
34:46
Unidentified speaker (UNKNOWN): And
34:47
Unidentified speaker (speaker_1): what I understand is that the non-display data is not part of the of the of the discussion at least not yet from my understanding
34:53
Unidentified speaker (UNKNOWN): and
34:53
Unidentified speaker (speaker_1): so it's a simple example in which you
34:55
Unidentified speaker (UNKNOWN): know
34:56
Unidentified speaker (speaker_1): in in equilibrium you know there is people using uh display data people using non-display data and so if you really try to narrow narrow it down too much at one specific issue you probably understand that issue better but you might miss this this knock-on effects on on on other segments either of the data market or uh for
35:16
Unidentified speaker (speaker_1): other services that exchanges are providing.
35:19
Unidentified speaker (UNKNOWN): Well,
35:19
Laura Paglia: thank you very much. Uh that was uh very, very helpful. Uh Professor Moreth. Thank you very much for joining us today.
35:27
Unidentified speaker (speaker_1): Okay. Well, thank you very much, Laura.
35:29
Unidentified speaker (UNKNOWN): It
35:29
Unidentified speaker (speaker_1): was great to be here.
This page was produced automatically: the audio was transcribed with nvidia/parakeet-tdt-0.6b-v3+sortformer-v2.1 and interpreted by Qwen3.6-35B-A3B-UD-IQ3_S.gguf. Transcription and interpretation can contain errors; always confirm against the original episode. Nothing here is investment advice.

More from Candid Conversations for Financial Markets

All insights