On 4 April 2007 NYSE Euronext begins trading in Paris and New York as a fully merged company, the world’s first truly global stock exchange group, under the ticker symbol NYX.
News announcement by NYSE and Euronext
Towards Market Central
In the first edition of the Handbook, published almost 20 years ago, I contributed a story of the future where established exchanges had shut up shop. The story went that securities trading was to be conducted by automated proprietary trading systems wired to Market Central, a global electronic order matching network operated on a commercial basis.
In an article in the 2003 edition of the Handbook I mused whether the established exchanges, in spite of heroic attempts to modernise, commercialise and consolidate, might be bypassed by the very financial intermediaries that historically had sponsored them.
If you venture into London’s Square Mile and look for the London Stock Exchange (LSE) in Old Broad Street you will see nothing but a construction site. The vaulting tower and cathedral-like trading floor proudly erected by the Council of the Stock Exchange in the early 70’s is now in the hands of property developers. The action now takes place somewhere else, on rows of computers located in anonymous grey buildings, and the administrative office is now near St Paul’s. Whilst the old market floor provided a time and place for trading activity, the new arrangements are free from time and location, and open the door to round-the-clock global trading. The London Stock Exchange is of course not alone, and this pattern is replicated the world over and sets the stage for consolidation and exchange linkages to form Market Central.
The idea of Market Central was planted in my mind when Joseph Hardiman, erstwhile head of Nasdaq, visited the LSE in 1985 when I was engaged there on the preparations for the Big Bang automation programme. He shared a vision of connecting both quote-driven markets to form the beginnings of a 24-hour global market. In turned out at that time, and of course more recently, that the overture was rebuffed. Actually the Stock Exchange Council at the time was more interested in doing something with NYSE, but a working model could not be achieved, in part because of the incompatible systems of trading.
Most exchanges are now public companies, often listed on themselves, and are subject to the competitive forces, rigours, opportunities and threats of the capital markets game. Essentially they are providing automated order matching services.This is rapidly becoming a commodity business, given the rapid fall in technology costs and overcapacity in service provision. Many established exchanges have a high cost base and rely on selling market information, a by-product of the order matching process, to make ends meet.
Against this backdrop rapid consolidation of exchanges is now underway. In fact there is a veritable frenzy of exchange merger and acquisitions activity in the race to build a bigger and better exchange, to achieve scale economies and to build barriers to entry.
Many exchanges have become the speculative playthings of hedge funds that can, and do, dictate exchange strategy.
In the meantime, having sold out their membership of traditional exchanges, the sell-side players are free to provide exchange-like services either individually or in cooperation using new models. This is against a backdrop of enabling regulations, such as the US Regulation NMS and the EU Markets in Financial Instruments Directive (MiFID) that are designed to increase competition in exchange service provision, and to give investors a better deal.
Last year we saw the first moves by Tier 1 investment banks to provide alternatives for trading and market data consolidation in the guise of Projects Turquoise and Boat respectively.
Where is all this heading?
Merger mania
NYSE triggered the latest round of consolidation in a staggering USD9bn transaction with the tiny Archipelago Electronic Communications Network (ECN). Archipelago provides order matching services via a simple public limit order book on low cost commodity computers, in stark contrast to the arcane specialist trading methods at NYSE and the old hat and expensive systems employed by their facility provider, SIAC.
Pioneered by Instinet in the early 80’s, the ECN model was to gain favour with the buy side as alternative trading venues to the traditional floor- and specialist-based models of the established US exchanges and the Nasdaq quote-driven competing market maker system. The ECN model, in particular, was to suit hedge funds and their need for fast automated order execution. ECNs blossomed in the wake of burgeoning hedge fund growth and enabling regulation, principally SEC Reg ATS. Later Reg NMS was to endorse fast execution, and in fact came to threaten traditional trading service models by linking fast execution to the trade through rule[1].
Nasdaq and subsequently NYSE were to gain no favours with investors when members were found to be front running their orders, causing SEC intervention and sanctions.
Nasdaq came close to being choked off from order flow by ECNs in the early 1990’s. It was only heroic efforts under the leadership of Frank Zarb, leading to the introduction of a public limit order book into the quote-driven system of trading, that saved the day. Called Super Montage, the idea was resisted by the Nasdaq market makers, but in the end it was a question of reform or die. The measure was also decried vigorously by the ECNs with Instinet in the vanguard. When it was clear that Frank Zarb’s efforts would prevail Reuters, who owned Instinet, put it up for sale and it has now been absorbed by Nasdaq along with other ECNs. Of the pack, Archipelago had chosen to hold out and reversed into the Pacific Exchange to gain exchange status, which in retrospect appears to have been a good move.
The premium paid by NYSE for Archipelago appears to have been justified by two key factors:
- As a means to trigger demutualisation and to free the exchange from the protectionist tendencies of the Floor Brokers and Specialists
- As a hedge against the hugely complex hybrid market system, NYSE’s answer to NMS, not delivering the desired results
Now we have NYSE Euronext, the appendage of course a result of exchange consolidation in Europe.
Having failed in its aggressive USD6bn bid for LSE, Nasdaq is now reported to be putting its ruler over OMX, itself the result of Scandinavian exchange consolidation.
OMX and then Deutsche Börse have also had a run at the London Exchange and been rebuffed.
LSE in the meantime is reported to be seeking to hook up with the Bombay and Tokyo Stock Exchanges.
Last year the Chicago Mercantile Exchange (CME) and Chicago Board of Trade (CBOT) announced a proposed merger. Migration of automated operations to CME Globex and trading floor consolidation is planned. Globex is also used by the New York Mercantile Exchange, perhaps paving the way for further consolidation. CME Executive Chairman Terry Duffy has described the goal of the CME/CBOT merger as seeking to deliver ‘significant efficiencies along with new products and technologies as quickly as possible to our global customer base’.
So all in all there is every prospect of Market Central emerging as a consequence of exchange automation, consolidation, and standardised electronic access using the now widely adopted Financial Information Exchange (FIX) protocol and internet technologies.
But are things as simple as this, and why are the big investment banks intent on getting back into the exchange services business?
Back to the Future
I used to think that automated trading systems threatened to give broker-dealers a hard time. In fact with colleagues I developed the Tradepoint Investment Exchange in the mid 90s in anticipation of disintermediation of the sell-side by institutional investors driven by the desire to reduce trading costs. Tradepoint provided direct electronic market access for institutional investors via FIX. They could trade anonymously via a public limit order book and clear and settle through a central counterparty facility at the London Clearing House.
It turned out that we were somewhat ahead of the game.
Whilst they have been the subject of some attrition and considerable consolidation, sell-side players have proved to be adept at adopting the Internet and FIX connectivity, direct market access and algorithmic trading technologies to provide highly successful and attractive trading services both to institutional and retail investors.
Orders continue to be routed principally via the sell-side rather than directly to trading venues.
Retail brokers add value by providing low cost immediate execution via the web, and provide the necessary settlement guarantees and credit. Execution may be achieved by routing orders to a third party venue, or by matching orders internally, or by executing an order on a principal basis at, or better than, the best price in the market.
For institutional investors, a package of Prime Broker services offered by the Tier 1 banks includes credit on fine terms for hedge funds, FIX connectivity and fast execution of large orders across multiple trading venues with minimal market impact costs. Market share for burgeoning hedge fund order flow is hotly contested. The prospect is that this style of trading will be increasingly adopted by the traditional long fund managers in an attempt to increase returns.
Trading Services Supply Chain
So the status quo is represented by the above picture.
Retail and institutional order flow is routed via the sell-side to available trading venues, or is executed internally, conditions permitting. Trading venues match orders typically through some sort of continuous auction based on the familiar public limit order book, although other techniques such as single price auctions, hybrid quote/order-driven mechanisms and crossing at a reference price are deployed. (It is now fashionable to refer to crossing systems, somewhat mysteriously, as dark liquidity pools).
Trading venues, as a by-product of order matching, provide pre- and post-market information which they sell to financial intermediaries and investors, and this can form a significant part of income.
Clearing and settlement operators offer central counterparty services which deal with settlement risk.
So what could upset this happy state of affairs?
Well, the surging growth of hedge funds is attracting top-end trading and technology skills and it is quite within their grasp to engineer their own direct market access and algorithmic trading techniques. Access to clearing facilities can be done by arrangements with existing clearing members. This could have the effect of unravelling the neat prime services bundle. The motivation of the hedge funds for this would be to obtain a better deal.
So the big sell-side players could be motivated to return to a position where trading facilities are owned and controlled by them as a defensive measure.
There are other drivers too.
It has always been a bone of contention with the sell-side that they produce the order flow which keeps exchanges alive, and yet have to pay for the information that is generated as a by-product of order matching. Market data is the life blood of a big investment bank, is vital to core trading and risk applications and can represent a substantial cost.
By taking control of the execution process and capitalising on their captive order flow the big banks can possibly generate the best reference price in a range of exchange-traded securities, and then become net producers rather than consumers of market data.
Also by internalising execution of order flow the big banks can save on exchange execution and clearing house fees.
Boat and Turquoise
In August 2006 nine investment banks launched Project Boat. The consortium comprised ABN Amro, Citigroup, Credit Suisse, Deutsche Bank, Goldman Sachs, HSBC, Merrill Lynch, Morgan Stanley and UBS.
Project Boat will pool and disseminate market information on a real-time platform which will allow equity dealers to comply with MiFID pre- and post-trade transparency regulations.
Soon after, a similar consortium announced Project Turquoise. The plan is to build an alternative trading platform that will operate as multi-trading facility (MTF) under the EU MiFID regime.
At this time, Boat appears to be merely an attempt to achieve MiFID pre- and post-trade transparency compliance for those banks choosing to operate as Systematic Internalisers at minimal cost, and also to put the squeeze on exchanges regarding market data costs. The extent to which the pre-trade quotes will be actionable for anything above retail market size is questionable, and may well limit the impact of this development at least in the short term.
Project Turquoise is more interesting as the consortium can potentially capitalise on their captive order flow to achieve critical mass in the most liquid European equities and thus become the preferred venue for trading these. They could publish their reference prices through Boat. This could hit the established exchanges in Europe hard as a large part of their income comes from transaction charges and information sales arising from activity in the leading European stock. In competitive strategy terms this could be a classical creaming operation by the big banks.
It will be interesting to see what trading model is adopted for Turquoise. It could be simply a jointly owned public limit order book, or a hybrid approach which capitalises on the marker making capacity of the consortium members. Another more sophisticated approach, which may have merit, is for each bank to operate their own order books or hybrid structure and to execute orders at the best consolidated price across consortium members and third party systems. In this scheme, if a consortium member cannot match the best price they could route their orders to the best venue available through a shared order routing hub. This in effect mirrors the way the National Market System operates in the USA and could have benefits. For example the consortium members could:
- reap the cost benefits of internalisation
- compete amongst themselves in the development of better trading mechanisms
- provide for easy admission of other banks to the scheme of things
- keep the shared facilities to basic order routing and market data consolidation hubs.
The US National Market System
NMS: a model for Project Turquoise?
Conclusions
I concluded my 2003 article as follows:
I now think the new technologies, combined with the power of the big consolidated houses, pose a significant threat to the established securities exchanges. The usefulness of purpose-built centralised execution venues may have served their time.
The drivers for this rewiring of established connections are the potentially significant cost savings accruing to the big houses, but perhaps more importantly their desire to influence more closely the shape of securities markets to come. Their key cards are:
- They are the conduit for the lion's share of order flow to the established exchanges – the capital markets business is highly concentrated with 20 globally-positioned full service investment banks accounting for over 96% of the business.
- They are the originators and the primary distributors of the products traded on established exchanges.
- They have access to more skills and resources and have more commercial clout than established exchanges.
- They are more agile.
- They apply significant risk capital to market making and are generally active in this capacity in wholesale and retail market and across the gamut of capital markets instruments.
- They are globally positioned.
- They have years of commercial experience at the sharp edge.
Established exchanges are now effectively in play and they are mere minnows compared with the big investment banks that use them. The rate of consolidation and globalisation of exchanges has been eclipsed by similar developments amongst their big users. There is over-capacity in exchange services, and little discernable differentiation amongst many of them.
I stand my ground.
Notes
1 The trade through requirement for NYSE listed securities was instigated by SEC as an investor protection measure in 1975. It required an order to be routed to a better-priced market regardless of how long it took to be executed. Under the traditional floor-based Specialist system there was no guarantee that the investor would get the price advertised when the order completed, and in fact the risk existed that the order would not be filled at all. Rather than protecting investors, the trade through rule had the effect of providing monopoly protection for the outdated and rapidly aging manual market and its system of Specialists. Reg NMS sought to cure this by obligating US exchanges to either provide fast guaranteed execution at advertised prices, or be cut off from National Market System order flow.
