Exchanges and Market Data - How Much Money Is Being Made?

EXECUTIVE SUMMARY

The profits that exchanges generate from market data have come under intense scrutiny in recent years, with broker-dealers and asset managers complaining bitterly about what they consider to be exorbitant, monopolistic fees. Exchanges counter that their fees are reasonable and subject to competitive price pressure. This report does not seek to settle this debate, but to provide information on the level of revenues that individual exchanges are generating from market data, what the profitability of this business is, and how this has changed over the past decade or more.

What appears to be a simple task – determining market data revenues and their growth, quickly becomes quite involved. There have been countless mergers amongst exchanges; the delivery and usage of data has changed considerably, and trading volumes have shifted.

Exchanges have seen significant growth in their information services revenues, increasing from just over US$1 billion in 2005 to over US$6 billion in 2019. This is in stark contrast to major investment banks, who have seen their revenues shrink over the past decade.

The profitability of market data at exchanges is impressively high, and we estimate that exchanges are able to achieve an average operating margin of 76% in this line of business, significantly higher than the rest of their activities. More tellingly, exchanges’ profit margins are more than twice as high as market data vendors such as Bloomberg or Refinitiv, and more than three times as high as major investment banks.

In the United States and the European Union, regulators are increasingly taking note of exchanges and market data revenues and are preparing significant changes to regulation and market structure that will see exchanges’ market data revenues come under pressure.

INTRODUCTION

Few topics generate more outrage amongst traders and asset managers than the amount they pay exchanges for market data. Not uncommonly, an otherwise sober individual can become quite excitable when the conversation turns to bourses and their fees for information services. Suddenly, arms start waving, the volume increases, and protestations such as “but they are charging us for our own data!” are blurted out, convinced they have stumbled upon a uniquely evil social injustice. Soon the discussion turns to “there ought to be a law”. Increasingly, regulators are taking note, casting a more sceptical eye on exchanges’ pricing policies, profit margins, and revenue growth for data.

Information Services Revenues at Exchanges Explode

Exchanges’ data revenues have grown at a phenomenal rate (see Figure 1), having reached a level of US$6 billion in 2019, capping a decade and a half of double-digit growth. Investment banks, on the other hand, have come under intense pressure (see Figure 4), with revenues declining over the past decade.

However, simply looking at the revenue growth this way is too simplistic. There have been dozens of mergers amongst exchanges (see Figure 3), they have made substantial acquisitions in the data space—notably in pricing and analytics, as well as indices. While the argument may be made that exchanges have monopolistic pricing power over the data produced as part of their trading activity, this argument does not readily apply to exchanges’ other information services.

In this report, we examine the revenues and the profitability of the information services businesses at eight of the world’s leading exchanges, taking into account mergers with other exchanges and acquisitions of data businesses. We attempt to estimate the profit margins that exchanges generate in their core market data offering, which is the area where they exercise potentially monopolistic or oligopolistic pricing power.

FIGURE 1. EXCHANGES’ REVENUES FROM DATA EXPLODES

The Profitability of Exchange Market Data

Disentangling exchanges’ financial reports is no mean feat, as they are not designed to allow revenues and profits from market data to be readily extracted. However, our best estimates as to exchanges’ profitability in their core market data business are shown in Figure 2. Two points stand out here.

First, the level of profitability of market data at exchanges is remarkably high, hovering around 76%. This is more than twice what market data vendors, such as Bloomberg or Refinitiv earn, and more than three times major investment banks. It is interesting to note that two market data vendors have particularly high profitability: S&P Global and Moody’s, both of whom are credit rating agencies. This is no coincidence, but ratings are a story for another day.

Secondly, the profitability of market data is consistently higher than the exchanges’ overall business. To a certain extent, this is a question of where costs are recognized internally. Data is an essential by-product of the exchanges’ trading activity. The incremental cost necessary to deliver market data, to collate it, normalise it and distribute it is quite low, but recognising part of the trading engine’s expense within market data allows flexibility in reporting profitability. How much of the trading operation’s expenses are recognised within information services is somewhat arbitrary and makes a separate analysis of market data rather problematic.

The estimated profit margins in Figure 2 might seem implausibly high, reaching above 90% for some firms. However, information from the exchange IEX gives us some confidence in these figures. IEX provided a breakout of its data costs and showed that if other exchanges had similar expenses, their data business would have margins of 95%. Even if Nasdaq and NYSE had costs four times higher than IEX for providing essentially the same service, then they would still have margins of 79%.

FIGURE 2. EXCHANGES’ DATA BUSINESSES ARE HIGHLY PROFITABLE

FIGURE 3. DECADES OF MERGERS AMONGST EXCHANGES

FIGURE 4. WHILE EXCHANGES EXPAND, INVESTMENT BANKS SHRINK

Consolidated Tapes

This report is global in scope, looking at exchanges in Europe, Asia, and North America. While in most of these regions, exchanges are responsible for managing their own data distribution, the US market is rather different, requiring the use of consolidated tape for equities market data. In Europe, there have been ongoing discussions around the topic. We consider the situation in both of these regions.

UNITED STATES

In the US, the Securities Exchange Commission (SEC) has long mandated an exclusive, consolidated tape as part of the National Market System (NMS). Under NMS, exchanges are required to provide their best quote and last trade, which is then consolidated and distributed to the market under programmes whose economics are determined by the SEC. The profits from this activity are then shared back with the exchanges providing the data, based on their trade volume and the percent of the time they post the inside quote.

FIGURE 5. US CONSOLIDATED TAPE – A DECADE OF NO GROWTH

The revenues from the US consolidated tapes are shown in Figure 5. The approximately US$400 million generated by the consolidated tape plans has remained fairly constant over time.

In addition to being required to make their data available and through the consolidated tapes, US exchanges are free to sell their market data directly to their users – in this report we refer to these data as proprietary market data. Sourcing data directly from the exchange affords market participants a speed advantage, since the consolidation of data from different exchanges necessarily adds some latency. Also, exchanges can provide additional data through their proprietary feeds, such as depth-of-book, which are not available on the consolidated tape.

Some US market participants have complained bitterly that this system essentially forces them to pay for the consolidated tape and to also buy exchanges’ proprietary data. Not only that, the Order Protection Rule that is part of Reg NMS means that any exchange, no matter how small or how limited in liquidity, displaying the best price must have order routed to it. As a result, market participants need to have access to data from all exchanges.

Very recently, the SEC announced that it would start to consider some fundamental changes to the distribution of market data under NMS. The SEC has recognized changes in technology and how exchanges’ proprietary data feeds provide a superior product in terms of both latency and information by including depth-of-book. The SEC is now proposing to include depth-of-book in the NMS data plans, but more importantly to introduce competition into the system by allowing multiple, competing data consolidators, in place of the current, exclusive approach. Also, the SEC is proposing to introduce a new type of entity—self-aggregators—which will be broker-dealers who choose to combine market data from different exchanges for their own use. Currently, the SEC sets the pricing for users of consolidated data, and the creation of this new entity type “self-aggregator” strongly suggests that the SEC will determine prices for exchanges selling their proprietary market data feeds to broker-dealers as well. This understandably has US exchanges very concerned, but details about the SEC’s proposal have not yet emerged.

EUROPEAN UNION

In most of the rest of the world, exchanges are not obligated to participate in a consolidated tape. However, in the European Union (EU), there are increasingly discussions about putting a consolidated tape into place, modelled on the US approach. MiFID II had made provisions for a market-provided consolidated tape, but none of the market data vendors expressed interest in providing this. The regulatory burden simply appeared too high and demand too low. 

The European Securities and Markets Authority (ESMA), the EU securities markets regulator, has expressed frustration that market data prices in the EU did not drop with the implementation of MiFID II and that no consolidated tape has emerged in Europe. ESMA is now proposing to create a consolidated tape and take a more active role in monitoring exchanges’ prices for data.

However, looking at the US market should strike a note of caution. Having a consolidated tape has done little to reduce the costs of market data there. Those hoping it will do so in Europe are likely to be disappointed. Aside from the economics of a consolidated tape, the geographic dispersion of the EU also poses a problem. Wherever the EU consolidated tape is physically located, trading firms will be forced to place their trading servers in close proximity to avoid any delays in receiving information. Essentially, EU regulators will be picking which city in Europe will become dominant in trading, a point likely to be bitterly contested.

EXCHANGE BY EXCHANGE

In this chapter, we analyse market data revenues and profitability at eight of the world’s leading exchanges. We provide information regarding the growth, breakdown, and profitability of data businesses at these exchanges. While doing this, we take into account mergers and acquisitions, changes in the number of users, and transaction volumes. Where information regarding profitability is available, we provide it; where profitability information is not available, we provide an intelligent estimate.

Between exchanges, we see a substantial degree of variability. Some have seen transaction volumes grow, others fall. Some have been actively engaged in acquisitions in their core exchange business, others have been more circumspect. Most have seen the number of users of displayed data decline, while non-display usage has increased. Despite all these variations, one constant emerges—revenues from market data have grown considerably, and this is true of exchanges across the world.

While some may seize upon this as evidence of the monopolistic pricing power of exchanges, such a discussion is beyond the scope of this report. In part at least, the shift at exchanges from transaction fees to market data revenues is a natural and deliberate progression. Revenues from transactions are volatile and highly cyclical, while revenues from market data are far more stable and predictable. Naturally, exchanges have tried to shift their revenues from transaction-based sources to more stable subscriptions.

Australian Securities Exchange

The Australian Securities Exchange (ASX) stands out on several counts. First, it is remarkably efficient, having some of the lowest operating costs per trade in the world, despite the exchange’s relatively small size. This has given ASX the highest operating margin of any of the exchanges covered in this report at 75%.

On the market data front, revenues have grown at a relatively modest pace of 8.2% annually since 2005 (see Figure 6), while transaction volumes have grown far more quickly (see Figure 7). At the same time, the number of users relying on displayed data from the ASX has declined markedly, losing more than half their users over the past five years (see Figure 8).

FIGURE 6. ASX’ DATA REVENUES INCREASE

FIGURE 7. BUT ASX TRADING VOLUMES GROW EVEN FASTER

FIGURE 8. ASX WORKSTATIONS DECLINE

Chicago Mercantile Exchange

While the Chicago Mercantile Exchange (CME) has been singled out by some market participants for what they consider to be particularly egregious price increases in terms of market data, these increases are not readily visible in Figure 9. Annual growth since 2006, taking into account acquisitions, has been a relatively tame 6.4%. However, since 2013, data revenues at CME have been growing closer to 10% annually, allowing the exchange to increase market data revenues by over 70% in the six years between 2013 and 2019.

FIGURE 9. MODEST INCREASES IN MARKET DATA AT CME

FIGURE 10. SINGLE-DIGIT TRANSACTION VOLUME GROWTH AT CME

With CME we see the importance of taking mergers into account. Without factoring in the acquisitions of NYMEX and CBOT in 2007 and 2008, CME’s growth rate would look far larger, creating a misleading image, but it is interesting to note that the growth in market data revenue has been far greater than growth in transaction volumes over the same period (see Figure 10).

Deutsche B?rse

A few interesting points can be seen in Deutsche B?rse’s market data business. First, revenues have increased almost every single year since 2005 (the one exception being a decline in 2011). Also, revenues from indices account for about half of Deutsche B?rse’s information services revenues, with indices growing far more quickly at 9.4% than the exchange’s core data service, which only grew at 3.7%. Nevertheless, the 3.7% growth in core exchange data is well in excess of the growth in transactional volume at either Deutsche B?rse’s derivatives exchange Eurex, or in its equities trading volumes (see Figure 13).

FIGURE 11. ALMOST CONSTANT GROWTH IN MARKET DATA AT DEUTSCHE B?RSE

Secondly, the change in the profitability of the data business is remarkable. From a unit that had an operating margin of only 35% in 2005, profitability more than doubled by 2009, when the exchange’s data business enjoyed a profit margin of 80%. Since then the profit margin has come down somewhat but is still hovering around 70%. It appears that Deutsche B?rse treated market data as a kind of low-margin, utility business through 2005. A change in strategy is evident, with Deutsche B?rse starting to maximise the profitability of data around 2009.

FIGURE 12. A DOUBLING OF PROFITABILITY IN DATA BUSINESS

FIGURE 13. EUREX VOLUMES FLAT, EQUITIES MODEST GROWTH

Hong Kong Exchanges

Hong Kong Exchanges’ (HKEX) trajectory in terms of market data followed an interesting path, as can be seen in Figure 14. The jump seen in 2013, after several years of declines, coincided with the acquisition of the London Metal Exchange (LME) in late 2012. We believe that the vast majority of the HK$170 million increase in market data revenues in 2013 was due to the LME acquisition. LME did not break out its revenues in 2013, but in 2018 the LME generated just over ￡25 million – about HK$250 million - from market data. 

FIGURE 14. MARKET DATA REVENUES AT HKEX INCREASE MARKEDLY SINCE 2012

The decline seen through 2012 turned around with the acquisition of LME, and since then HKEX’s market data revenues have increased by about 25% or 4% annually, a far lower rate than the increases in trading volumes that HKEX witnessed over the same period (see Figure 15 and Figure 16).

FIGURE 15. STRONG GROWTH IN EQUITIES TRADE VOLUMES AT HKEX

FIGURE 16. STRONG GROWTH IN DERIVATIVES VOLUMES AT HKEX

Intercontinental Exchange

The Intercontinental Exchange (ICE) has pursued one of the most aggressive acquisition strategies of any exchange. ICE has snapped up several other exchanges, most notably the New York Stock Exchange (NYSE), then part of NYSE Euronext. ICE’s acquisitiveness has not been limited to other exchanges, with several acquisitions made in data services. These have included:

• Interactive Data Corporation (IDC) for an eye-popping US$5.2 billion in 2015. IDC’s main business was providing evaluated prices for bonds, where the firm competes with Bloomberg and IHS Markit.
• Standard & Poor’s Securities Evaluations (SPSE) in 2016, a further acquisition in the evaluated pricing area, consolidating IDC’s position.
• BofA Merrill Lynch’s Global Research Index Platform in 2017.

As can be seen in Figure 17, most of ICE’s growth in data services is due to the acquisition of IDC. However, even if we remove the IDC revenues in Figure 18, market data revenues have still grown at a very healthy rate. ICE’s market data revenues have closely following the growth in derivatives trading on the exchange (see Figure 19). Unfortunately, ICE’s financial reports do not allow us to break out data revenues from derivatives, where volumes have grown rapidly, from the exchange’s equities trading business on NYSE, where volumes have declined.

ICE does provide information regarding the profitability by line of business and for its Data and Listings segment, the operating margin is 53%, a figure which is unusually lower than the exchange’s overall profitability of 59%. However, Data and Listings includes a disparate number of businesses, including evaluated prices, exchange data, and listings. When IDC was acquired, its operating margin was about 40%. If we assume that the Pricing and Analytics business has continued at this operating margin and that Listings have a similar margin—which is in line with Nasdaq, who does provide profitability for listings—we can obtain an estimated operating margin of 70% for the exchange data segment, which encompasses data feeds, connectivity, and desktops.

FIGURE 17. ICE’S EXPLOSIVE GROWTH IN DATA

FIGURE 18. ICE GROWTH STILL STRONG WITHOUT IDC ACQUISITION

FIGURE 19. ICE DERIVATIVES VOLUMES EXPAND DRAMATICALLY

FIGURE 20. THE SHRINKING NEW YORK STOCK EXCHANGE

Japan Exchange Group

The Japan Exchange Group (JPX) is the result of the merger of the Tokyo Stock Exchange (TSE) and the Osaka Securities Exchange (OSE). As can be seen in Figure 21, JPX’ revenues from market data have increased at a fairly modest rate, but have increased steadily nevertheless. However, JPX’ volumes have grown more quickly than market data revenues.

The growth in the number of equities transactions has on TSE has been particularly pronounced (see Figure 22) and can to tied to the introduction of a new, high-speed trading system called Arrowhead in 2011, which allowed TSE to attract a number of high-frequency trading firms, particularly outside of Japan.

FIGURE 21. JPX’ DATA REVENUES EXPAND MODESTLY

FIGURE 22. JPX TRADING VOLUMES INCREASE

London Stock Exchange

The London Stock Exchange (LSE) has made a concerted effort to push into the area of indices and made two major acquisitions in this space:

• The remaining 50% of FTSE for ￡450 million in 2011. FTSE’s primary business was providing indices.
• Frank Russell Company in 2014. Frank Russell consisted of an investment management group, as well as Russell Indexes, focused on selling index data.

As can be seen in Figure 24, most of LSE’s growth in its Information Services segment is due to these two acquisitions. If we remove the index business in Figure 25, an interesting trend emerges. The exchange’s real-time market data revenues have actually modestly declined over the past decade, decreasing by about 18%. However, this has been more than compensated for by Other Data, which includes end-of-day prices, historical prices, and reference data such as the exchange’s securities master file SEDOL. Here, revenues have more than doubled over the past decade.

LSE provides overall profitability numbers for its Information Services group, but not for exchange data and indices separately. To estimate the profitability of the exchange data segment we look at the overall profitability of Information Services – 56%. This is a respectable number, but below the margins at most other exchanges. We know that the FTSE index business had an operating margin of 45% when acquired in 2011, and Russell Indexes was at 49% in 2014. These figures are in line with margins seen at other index data providers. If we assume that the index business has remained at the same level of profitability, then we can determine the returns on just the exchange data piece, as shown in Figure 23. Our estimate indicates that the profitability of LSE’s exchange data business more than doubled between 2013 and 2016, reaching 76% by 2018. On the surface, this seems difficult to achieve, but we have seen other exchanges, such as Deutsche B?rse, similarly double their profit margins for data in equally short timeframes.

Trading volumes on LSE have only shown modest growth over the past decade (see Figure 27) and the number of users of the exchange’s displayed data products is gradually declining (see Figure 26).

Finally, the LSE’s acquisition of Refinitiv, due to close shortly, will see LSE’s revenues from data multiply. Already, LSE generates more in revenue from data than trading. With the addition of Refinitiv, the transformation of the stock exchange into a data business will be complete.

FIGURE 23. ESTIMATING LSE’S PROFIT MARGIN REVEALS HIGHLY PROFITABLE DATA BUSINESS

FIGURE 24. LSE – STRONG GROWTH DRIVEN BY ACQUISITIONS IN INDICES

FIGURE 25. REAL-TIME DATA DECLINES, BUT OTHER DATA MORE THAN DOUBLES

FIGURE 26. LSE DATA TERMINALS IN DECLINE

FIGURE 27. LSE’S TRADING VOLUMES GROWING SLOWLY

FIGURE 28. REFINITIV ACQUISITION WILL LEAD TO FOURFOLD INCREASE IN DATA REVENUES AT LSE

Nasdaq

Nasdaq’s revenues from information services have grown considerably, as can be seen in Figure 29. However, much of this growth has come from Nasdaq’s expansion in the index business, rather than the exchange’s core market data offering. With Nasdaq, we can compare the exchange’s revenues from the US tape, that is through the SEC-mandated consolidated tape, and its own proprietary US data products. It is interesting that in 2005 the vast majority of Nasdaq’s data revenues came from the consolidated tape, with proprietary data sales being almost an afterthought.

Over the years, this has changed significantly. Revenues from the US tape have stagnated and even fallen, while proprietary data revenues have taken off (see Figure 30). It should be pointed out that Nasdaq only provided a breakout of its US proprietary products through 2015. From 2016 to 2019, we have had to make estimates. However, since the US tape plans do reveal the payments made to exchanges, we can estimate Nasdaq’s proprietary data revenues for 2016–2019 with some confidence.

FIGURE 29. NASDAQ’S FOURFOLD INCREASE IN DATA REVENUES SINCE 2005

FIGURE 30. NASDAQ’S PROPRIETARY US DATA PRODUCT GROWS FASTER THAN DATA OVERALL

FIGURE 31. A CURIOUS FALL IN PROFIT MARGINS FOR DATA

In terms of profitability, Nasdaq has shown a steady decline in its Information Services business from 2012 to 2019. The decline seen in Figure 31 in 2018 can be explained by Nasdaq’s acquisition of eVestment. However, the steady ongoing decline from 2012 is best explained by the increased revenues in index-related data, which typically has a lower profit margin than the exchange data business. Backing out the index business allows us to estimate that the profit margin for Nasdaq’s exchange data business is about 79%, about the same as in 2012, before index data started to grow rapidly.

While Nasdaq did acquire the Philadelphia Stock Exchange (PSX) in 2008, PSX’s overall revenues were only a small fraction of Nasdaq’s and only had a limited impact.

FIGURE 32. NASDAQ EQUITIES TRADE VOLUMES BOUNCE AROUND – BUT WELL OFF-PEAK IN 2008

Singapore Exchange

The Singapore Exchange (SGX) has the dubious distinction of having the least profitable market data business of the exchanges considered in this report (see Figure 34). Nevertheless, SGX has been able to increase its revenues from market data and connectivity every single year for the past fourteen years (see Figure 33).

The trading volumes of SGX’s equities and derivatives trading volumes are shown in Figure 35. While equities trading has been in decline, derivatives trading has more than compensated for this.

FIGURE 33. SGX DATA REVENUES CONSISTENTLY GROW

FIGURE 34. PROFIT MARGINS FOR SGX DATA BUSINESS IS SURPRISINGLY LOW

FIGURE 35. SGX’ EQUITIES VOLUMES DECLINE WHILE DERIVATIVES GROW

LOOKING FORWARD

For better or for worse, exchanges’ market data businesses have come under increasing scrutiny, particularly in the EU and the US. In the US, the SEC has been more reluctant about approving new exchange data pricing plans. Until 2018, NYSE’s and Nasdaq’s changes in fees for market data went largely unchallenged. However, in a ruling in late 2018, the SEC rejected these two exchanges’ price increases for depth-of-book data. More recently, in February 2020, the SEC announced it would deliver proposals to revamp market data distribution, in what looks like the SEC taking a far more active role in setting the prices exchanges can charge for data.

As we have seen, ESMA in the EU is also aiming to regulate market data and plans to create a European consolidated tape.

US and EU exchanges will have to contend with far greater regulatory oversight in regards to their pricing for market data, although the form that this will take is not clear yet. At the moment, the broker-dealers and asset managers who have complained so loudly about the cost of market data appear to be gaining the upper hand, with exchanges fighting a rear-guard battle to stave off regulators, who are intent on limiting how much exchanges earn with market data.