|Shares Out. (in M):||70||P/E||15.7||15.2|
|Market Cap (in $M):||3,724||P/FCF||13.7||14.6|
|Net Debt (in $M):||170||EBIT||328||361|
London, October 2018
Euronext (ENX-FR) is a high quality, ~60% EBITDA margin company, enjoying the strongest competitive position in its key geographies (France, Belgium, Netherlands, Portugal and Ireland). It is offering a piece of infrastructure to hold for the long-term, unloved and mispriced by the market due to a myopic focus on cash equities business and a lack of thinking on valuation (the industry trades on P/E but doesn’t adjust for leverage amongst peers). In the stock exchange space (great monopoly style industry), it is the only business with a relevant size, decent free float and a cheap valuation. At 53€ per share (€3.7bn market cap.), it currently trades on 12x 2018e consensus EBIT and 17x PE with about c.€7 per share of hidden assets on balance sheet and €5 per share in cash. For a business offering ~45% ROCE, a serious moat and highly cash generative, this is mispriced. With a management team incentivised to keep a low profile, I see 25% upside at 12 months in a base case with a TP towards €70 per share.
ENX operates the incumbent and leading securities and derivatives exchanges in each of its four main markets (France, Netherlands, Belgium, Portugal and Ireland), providing a range of services including security listings, cash and derivatives trading, post-trading services, market data aggregation, and market solutions. Keep in mind these geographies: capital markets in Continental Europe are different from US or UK markets: the investor base is much more concentrated, there are less issuers and companies tend to be larger market cap. ENX is the only operator to offer a unique pan-European model, running one platform across various geographies with a single order book. Altogether, Euronext is the leading listing venue, cash trading venue and second largest derivative trading venue in continental Europe. It also offers market data, indices and post-trade services via some minority participations in LCH.SA, EuroCCP and Euroclear. It generates about €675m of revenue, with 45% from trading activities, 20% from market data and indices, and the rest is a mix of listing fees and services. The blended EBIT margin of this business is around 53% with an impressive cash generation of more than €150m per year lately.
Company history: how ENX came to its current form
In the 1990s, national exchanges were demutualised and listed, triggering a wave of consolidation in the 2000s. M&A has always been common in the exchange industry given the fixed cost nature of the business, the possibility to leverage technology across larger liquidity pools and cost savings potential / synergies (being historically mutualised businesses, there were many inefficiencies in operations). Euronext is no different but is often seen as not as successful as his English and German peer.
Euronext was formed in 2000 by the three-way merger of the Brussels, Amsterdam and Paris stock exchange, later augmented by the Lisbon Stock Exchange, and most recently (December 2017), by the Irish Stock Exchange. In 2001, Euronext acquired the London International Financial Futures and Options Exchanges (“LIFFE”), often seen as the crown jewel of derivatives exchange in Europe.
In 2007, NYSE was on a streak to become the first global stock exchange functioning across multiple time zones. It acquired Euronext, valuing the business at close to $10bn: this was the outcome of an intense bid process with both Deutsche Boerse and NASDAQ competing as well. Winner’s curse for NYSE as the 2008 Financial Crisis would impair its investment, resulting in a c.$2bn impairment the following year. About five years later, Deutsche Boerse would again try to acquire NYSE-Euronext but the European Commission, citing fears of a near monopoly in the derivatives market, blocked the merger.
In 2013, Intercontinental Exchange (ICE) acquired NYSE-Euronext. Following the acquisition, ICE decides to spin-off the continental Europe operations, ultimately retaining the operations of LIFFE, focusing on derivatives and not interested in a mature continental EU exchange. At the time, Euronext performance in the equities market had been hit by a slowdown in trading activity, competition from alternative venues and a dearth of stock market listings. A potential financial transactions tax on equities, to which France had signed up, was a further problem. ICE, which always portrayed itself as a growth vehicle, made clear to investors that it was not interested in retaining a cash generative but steady business. On the other hand, I am very invested in these sort of businesses.
When Euronext N.V listed in June 2014, the French government took the opportunity to indirectly regain influence on a strategic piece in the European market infrastructure, forcing a peculiar capital structure, and a number of “anchor” reference shareholders to ensure ENX would have a decent chance had developing independently. IPO price were set at €20 per share, with a consortium of 11 reference shareholders acting as anchors: Euroclear, BNP Paribas, BNP Paribas Fortis, Société Générale, Caisse des Dépôts, BPI France, ABN Amro, ASR, Banco Espirito Santo, Banco BPI and Belgian holding public company SFPI. After the expiration of the initial lock up in June 2017, 8 of the 11 reference shareholders entered into a new agreement for 2 years, controlling together 24% of the share capital with 1/3rd of the supervisory board seats. As you can imagine, these shareholders are mostly interested in the income potential of ENX. My research talking to bankers close to this situation revealed that a number of the banks would probably look to exit at the next expiry date due to more stringent capital rules regarding on balance sheet investments. I think this will be the catalyst for a potential third party acquisition of the exchange or unlocking the balance sheet value via share buybacks.
ENX’ sources of competitive advantage, the European market infrastructure and market shares
I think Euronext has strong competitive advantage and a big moat around its business, which remains underappreciated as a smaller scale Continental EU exchange:
As mentioned, unlike in the US market where best execution rules require brokers to connect to ALL exchange to ensure best prices, European regulation does not interpret best execution narrowly in terms of prices. As a result, brokers tend to connect to one or two venues at a time, usually always including the national incumbent. This should create a stable floor in terms of trading volume for ENX. Traders often assess the level of fragmentation of a market by looking at the number of venues one should visit to achieve best execution when completing an order. In the US, that number is close to 4.5. In Continental Europe, it is barely two (source: Fidessa.com).
Historically, stock exchange has been a winner take all market. Orders concentrated on the most liquid market as the marginal cost of a trade would decrease with the volume of orders, reinforcing the network effect of an exchange. In practice however, markets are fragmented in two ways:
(i) several lit platforms compete for the order flow (not a real problem for ENX as the trading in its securities is fairly concentrated in its domestic markets); and
(ii) a portion of the order flow is internalised off-exchange by broker dealers (so called Systematic Internalisers).
Let us address market shares concerns. The current fragmentation is the direct results of regulatory policies aiming at increasing competition and decreasing costs for investors. At a national level, market shares of incumbent exchange stabilised at around 60/65%. BATS Chi-X (owned by CBOE Europe) is the clear new entrant winner with about 25% market share at national levels. Bear in mind that this is ~65% of the lit market, which makes 35/40% of the overall lit & dark market. Historically, the estimated market share for dark pools has been around 8%, with the rest of the dark market being OTC. BATS Chi-X Europe was formed by the acquisition of Chi-X Europe by BATS Trading in November 2011, subsequently acquired by CBOE in 2017. Investors are overly cautious about this development and keep extrapolating that CBOE will continue taking market share in Continental Europe. This is ignoring some fundamental facts:
A) Until 2007, many European stocks had to be traded only their local exchanges under the so called “concentration rule”. Mifid I served as a catalyst to increase competition between marketplaces, enabling the emergence of alternative platforms to national exchanges;
B) BATS/Chi-X started its operations on the back of Mifid I and was actually backed by banks and HFs looking to push price of trading down across the industry. It was not profitable and not ran to make money;
C) When BATS Ch-X listed, it turned to profitability and started losing market share. Today it is not ran for market share but for profit and my conversation with all the exchange and a number of HFs clearly confirmed that no operators is in the market competing on price, unlike a few years ago.
So why am I so confident that ENX will not lose market share going forward? The answer: Mifid II. The new regulation intends to increase transparency by limiting the volume of equities trading in dark pools, thereby encouraging shift to lit exchanges, and banning brokers crossing networks (i.e. grossly speaking, links between dark pools of banks). However, so far, banks have circumvented the rule by re-registering as systematic internalisers. As SIs, banks can only execute trades against their own books (i.e. not client money), which forced them to bring risk capital to the trade and effectively “intermediate” trades that would otherwise have been executed directly between their two client. The whole industry has become less efficient than it was before. I was explained the following real life example:
Pre Mifid II: Bank A would receive an order from client “BigHF” to sell €20m worth of shares in company X, without moving the market. Its trader/order router would go to the market, avoiding lit venues in order not to move current price, and likely find a buyer OTC. The transaction would go privately, entirely unnoticed.
Post Mifid II: Bank A is now an SI. It cannot do this transaction OTC. However, it can use its internal SI venue, effectively having to bring capital to facilitate the trade on each side, and chop it into tiny tranches to avoid falling under certain reporting requirement. The trade is more complex and number of trades is larger as Mifid II restricts the size of ticket tradeable on SIs.
Both ESMA (EU Regulator) and the AfM (French) have publicly stated that increased transparency has not succeeded so far given the increase use of SI and periodic auctions (https://uk.reuters.com/article/uk-eu-markets-mifid-analysis/light-or-dark-six-months-on-mifid-2-rules-divide-equity-traders-idUKKBN1JP0L7).
Three CFOs of European exchange have told me that the regulators are currently reviewing the regulation to clamp down on SIs and the surge of periodic auctions effectively being used to circumvent the rules around OTCs. The timetable is unclear but this will be a net positive for Euronext as it will be harder and harder for SI volume to be traded. It is realistic that the OTC/SI volume will eventually finds its way to lit exchange.
I also take comfort from the fact that market share relative to MTFs has been relatively stable and that incumbent exchanges still operate with the most liquidity in each security than new entrants.
Euronext business model and financial review
Euronext makes about €600m in revenue. About 45% of total revenue is trading related and 80% of that is cash equities trading. ENX is the platform with most exposure to direct trading revenue, which partly explains the valuation discount to large peers LSE and Deutsche Boerse.
Cash trading (this includes equities, ETFs, Fixed Income, Warrants etc) – 36% of revenue (€190m)
o Generally good momentum recently (with the exception of 2016 due to low volatility in H1 with the UK referendum and broader impacts across capital markets in Europe – ADTV was down 15% that year). This business is a price and volume game. Volume is dictated by external, macro drivers (generally market momentum, overall valuation, volatility, macro news).
o Price has been under pressure in recent years with the development of non-national exchange venue (e.g. MTFs and dark pools), but overall, ENX has maintained a decent yield. ENX is more expensive to trade on, but it differentiates itself through market quality, depth and protection. For background, when speaking with a few Dutch and French PMs, I understood they tend to appreciate the platform and are not very focused on the cost of trading as alternative offerings simply do not match the liquidity and pricing of security. Keep in mind the geography as recent regulation in Europe makes it harder for MTFs/dark pool to take on a more significant market share like in the US as the regulator is not purely focused on costs but also alternative quality metrics (e.g. Best Bid and Offer, relative spread, market depth etc).
§ Pricing plans include a fixed fee per trade as well as a volume dependent variable element. The price structure varies greatly according to the package selected by market participants, with typically a sliding scale based on the value traded going from 0.95bps for the most expensive package and lowest volumes to less than 0.15bps for certain ETFs.
§ ENX cash trading yield (i.e. the charge per value traded vs. the average daily traded value) has been relatively flat around ~0.5bps and ADVs of ~€8bn / day (in 2017, total cash volume was ~€2,000bn).
o Certain analysts I spoke with told me they were concerned that prices might continue to erode over time. I do not think this is the case and recent years have shown that ENX is defending well its cash yield. In addition, ENX covers markets with lower liquidity than the UK, US or Germany, making it harder for new entrants to come in and compete. I think that market share and prices are more stable because of (i) a hard floor on market share due to the auction (i.e. morning and closing auctions have to take place on the exchange) and domestic retail flow and (ii) generally less price aggressiveness from BATS/Chi-X since their acquisitions by CBOE (they have to make money now!). In Europe, best execution is not solely based on prices and Euronext is always top of class on these metrics. Very different from the US market where best execution is purely on price, forcing all market participants to connect to all exchanges.
Market Data – 20% of revenue (€105m)
o This is a very good business, high-margin, capital-light and insulated from macro and changes in trading volumes. There are three broad products in market data:
§ Real time data: price, trade and order book data on all instruments sold to data providers and investment banks
§ Historical and reference data
§ Indices: subscription fees for data and analytics for indices such as AEX, CAC, Bel 20, PSI
o In recent years, ENX improved revenue yoy by providing new products and services as well as price increases.
Today, almost no growth is expected by the sell-side here and I think it is misleading: ENX is still significantly cheaper than LSE for data and LSE has been able to raise prices almost double-digit yoy with little push back so far (clearly, it will change). I expect it to perform well going as many passive investors with appetite for new index concepts and benchmarks increasingly use it.
Listing – 16% of revenue (€84m)
o Admission fees charged to issuers on IPOs, follow-ons and rights offering as part of broader corporate finance transactions. Annual fees for securities traded on ENX’ markets
o Whilst most issuers will get a view of fees before choosing a listing venue, in my experience it is clearly not a decision factor in their choice of listing location
o ECM activity tends to be cyclical and linked to macroeconomic development / GDP and overall valuation cycle: more activity at the top of the cycle. Recent development has been very positive with double digit growth in most years… however 2016 was a bad year (down 3%) as the broader European capital markets were quiet for an extended period of time with the Brexit referendum and general market fears around the euro.
o Clear monopoly here, as ENX is market of reference (e.g. required to be listed on ENX to be included in leading CAC, AEX, BEL indices etc.). It will be cyclical but will continue to perform well over time.
Post-trade – 13% of revenue (€72m)
o 2/3rd of post-trade revenue come from ENX’s 11% stake in LCH S.A. LCH is the main clearing house for the London market. Through its 11% stake and a private agreement with LCH, ENX receives a share of clearing income in exchange for a fixed fee plus a variable feed based on derivatives volume. Margin is essentially fixed trough that contract at around 45%. The agreement is for 10 years and was renewed in November 2017.
o 1/3rd of post-trade revenue is from custody and settlement from Interbolsa Portugal, the Portugese CSD.
Derivatives trading – 8% of revenue (€40m)
o Similar to cash trading, albeit for a different product, derivatives operates under a similar model of volume and pricing.
o The principal types of derivative contracts traded are equity and index products as well as commodities. The main competitors are Eurex (DB1) and ICE Future Europe.
o The derivative market is a lot smaller (ADV of ~€14m; total 2017 value of volume traded of €3.7bn), with a price charged per lot (contracts traded). Current price is around €0.29 per contract. I understand that this is high compared to Eurex (which goes as low as €0.1 per contract – according to a Eurex employee, this is as low as it gets for it to make money). Euronext is able to keep high prices as the market structure favours the incumbent: first, markets are national / regional and given the low liquidity of these products, it is very hard to move market shares. Liquidity is sticky and comes in bulk because for traders / PM want to manage capital efficiently and not spread out their margin calls on too may platforms. Market share in France it is roughly 50 / 50, in Netherlands it’s around 80% for ENX, Belgium is ~60% for ENX and Portugal has no derivatives. They have been very stable over time.
o A former competitor, TOM, (for The Order Machine, a derivative platform founded in 2010 by a group of banks) shut down in June 2017 after a failed strategic sale. This shows the difficulty of establishing a foothold in this business, which is all about scale and liquidity. Competition is national and by products (e.g. Dutch equity options, French equity options etc.) with most markets setting as a duopoly between Eurex and Euronext. I note that ENX used to be the grand champion in most of these markets but excessive fees led it to lose some share to competitors over time.
o Not a core part of the story and I would not be surprised if this part of the business might be sold going forward.
Market solutions – 6% of revenue (€34m)
o Another high margin low capital business: ENX markets IT and technology solutions, such as its Universal Trading Platform, and more recently Optiq, to smaller exchanges.
o Most technology provided to clients include software for reference data management, price calculation and market control functions. Exchanges and venue operators can take advantage of fully hosted trading and clearing platforms without investing in data centre or network infrastructure. A simple example of the type of product is the MAR compliance software for MTFs: MAR compliance service monitors trading activity for suspicious behaviour. With a web-based user interface, it can be offered to market regulated markets, MTFs or Sis without the need for IT development or deployment.
o Growth has been fairly stable over the years at around 6%. Again, I think it is totally underappreciated by the market and has a nice runway in front of itself. ENX has just finished the upgrade to its latest trading platform, Optiq, and is now marketing it to smaller exchange this year. When speaking with technologists at smaller exchanges using a different platform there seem to be a recurring theme that DB offers a good product but not flexible at all (no development on top of it possible), and LSE is difficult to deal with, especially when technical incidents occur. Time will tell but Optiq looks really promising at the latest partnerships with Luxembourg SE is an example of ENX winning shares over DB in that space.
Revenue: looking holistically at revenue drivers, where do we stand? Management has an absurdly low target of 2% CAGR over ’15-‘19e: bear in mind that they have a clear incentive to budget low as one of the key metrics for their performance scheme is on actual delivery vs. budgeted results. Clearly not interested in aiming high. I believe they will do much better in 2018e and 19e because of the various points I raised above. On top of that, volatility has remained extremely low, volumes have held well, and a surge in volatility would increase volume. The market is very doubtful about ENX growth strategy entitled “Agility for growth”. ENX suffers from expectations mismanagement and a poor PR with investors:
- In 2016, Euronext launched its 3 year strategic plan called “Agility for Growth”. Euronext announced its intention to deliver up to €70m in additional revenue by 2019 on six “strategic growth initiatives” such as new products in MTFs, indices and services to issuers. These were supposed to be developed in-house in certain cases, or through M&A, as ENX has always been very opened about inorganic growth opportunities.
- The delivery of “Agility for growth” has been mixed so far. In 2017 and 2018, ENX spent about €34m on acquisitions for what they call “Corporate Services”; Corporate Services include the acquisitions of Company Webcast (online webcast services), iBabs (online board portal services for issuers) and Insider log (online insiders list and PDMR list management tool for MAR compliance). Together, these acquisitions would have represented about €12m in FY revenue in 2017 – far off the €70m mark. Other initiatives were also launched for investors with for example a partnership with Morningstar for some new indices or a new MTF for bonds (nothing ground breaking in my view).
- In the meantime, in February 2018 at the FY results, management announced the postponement or cancellation of three of the six area for growth, resulting in a revision of guidance for the “agility for growth” programme from €70m to €55m. Whilst core results of the business were good (double digit revenue in growth, good margin expansion, good EPS expansion etc.) share price went down 3% on the announcement and analysts turned very bearish…
- FY2018 will be the first year of full contribution from Corporate Services, which could create some surprises, but at H1 2018, it only represents €8m in revenue contribution. When asked the question on earnings call, management sounded rather embarrassed and offered no explanation as to how they intend to reach the €55m. CFO said “On the 55m, I believe that I cannot give you more details. […] there are other initiatives that rely on critical mass and therefore you should not expect a linear growth.”, soon completed by the CEO: “it’s a very diverse level of advancement […] the developments ought to come in the coming months. So I think the right pace for posting and looking at what the real numbers and the real delivery is with the annual results as we have this year when announcing the revision from €70m to €55m.” Clearly, it does not help to build confidence.
I do not think all is bad with ENX and to me, it looks like management might have been a bit more opportunistic recently and focused on acquisitions outside of the “agility for growth” programme, which do make strategic sense in my view. The acquisition of Irish Stock Exchange (EUR33m in revenue in 17a) was completed in Q1 2018 and provides ENX with the largest listing venue for Fixed Income globally. FastMatch (EUR19m revenue in 17a) is probably a bit more controversial as it evolves in the highly competitive world of FX trading, was completed in FY2017. Together, these two represent about €260m in capital deployed and I am more excited by their potential (notably margin improvement for Euronext Dublin) than the agility for growth programme. Brokers are not giving any credit to margin improvement and assume disappointing revenues at FastMatch: last quarter was disappointing but it ignores this was due to a dispute with the previous CEO of FM, who is now gone.
On the expenses side, Euronext is a high fixed cost business with relatively stable and high margins in recent years, around 55% at EBIT level. In recent years, Euronext did a good job at keeping its employee base under check, having reduced its FTE base from 880 at the time of the IPO to about 690 today. Recent M&A activity is the primary reason for the increase in headcount. The rest of operating expenses is composed of system & communications, professional services, accommodation and some other expenses deemed to be for marketing, professional membership fees and corporate management. Together they make about 20% of the revenue and are stable over time.
The clearing business has a fixed margin at around 45%, entirely variable. The recently completed acquisitions will dilute margins in the short term but offer optionality longer term. The Irish Stock Exchange has a number of quick-wins (migration to ENX existing Optiq trading platform, head office cost – e.g. I was told there are more printers at ENX Dublin than in the Paris HQ with less than a 1/3rd of employee base).
In terms of balance sheet, Euronext is a very capital light business. In fact, when I visited the control room for all European markets in Paris, I was surprised to see that only 2 operators were on duty and responsible for market surveillance that day. The total asset base is around €1.4bn, of which goodwill and intangible assets are the largest account with €662m at Q2 2018. Goodwill arose mostly through the complex M&A / spin off history of ENX. The next biggest assets are ENX’ stakes in LCH Clearnet, Euroclear and the few ventures it invested in last year (c. €250m in total).
Overall, Euronext generates a very decent return on capital employed, which I estimate historically at around 55% (important to bear in mind that 2013/14 were littered with restructuring costs and various balance sheet movements related to the parent and 2017 is impacted by the acquisitions which are on balance sheet but not yet reflected in the accounts).