The structure of the equity markets is undergoing dramatic changes worldwide. In the United States this change is happening at a pace that is unparalleled since the creation of the National Market system following the Securities Acts Amendments of 1975. Ironically, from 1997 to 2000, there was immense publicity about the development of Electronic Communications Networks (ECNs) and other Alternative Trading Systems (ATSs). Now the changes that were hyped then are actually happening.
In this paper we will look at changes that are happening in the United States in the light of four scenarios for the ultimate market structure. These scenarios represent a framework that both markets and market participants can use for evaluating strategic alternatives. We will then consider parallels for Europe and possibly Asia.
Background
The market structure that evolved in the United States in the last quarter of the twentieth century involved an equity market divided roughly equally between the New York Stock Exchange (NYSE) and the Nasdaq Stock Market (Nasdaq) as the two primary markets for original listing. During this period the American Stock Exchange declined in importance as an equity listings market, except for exchange-traded funds (ETFs). Regional markets competed for NYSE order flow and market makers traded Nasdaq issues in an over-the-counter environment.
Beginning in 1997, the Securities and Exchange Commission's Limit Order Display Rule[1] promoted the creation of ECNs, and a market structure that had been stable for nearly twenty years began to change. Most observers seemed to assume that the chaos of the late 1990s was a transient phase while the market evolved to a new equilibrium. Two important questions emerge from this assumption. First, is the assumption correct -- are we evolving to a new equilibrium, or have we entered an environment where change is continuous? Second, if a new equilibrium is evolving, what will it look like -- who will dominate the markets and what will the structure be? We will address these issues in reverse order.
Scenario planning
Scenario planning is designed to help develop strategies in situations where the future is uncertain. We will develop four extreme descriptions of possible future environments. We will try to understand the conditions that could cause these extreme cases to evolve. We will then describe the economic environment that would exist once the scenario has developed and assess its impact on market participants, both positive and negative.
Realistically, it is highly unlikely that any one of these extreme scenarios will evolve. The future environment is more likely to display aspects of each. By considering the extreme cases, however, we are better able to consider the best strategies for more reasonable alternatives.
Four scenarios for market evolution
It is useful to think about the markets along two primary dimensions. First, markets may become more centralised or they may continue to fragment. Second, markets may be dominated in the future, as they were in the past, by broker-dealers. Alternatively, investors -- the buy side -- may become more dominant (as many have predicted for years). This is shown in Figure 1, with each of four extreme scenarios represented by metaphorical labels.
Figure 1: Four scenarios for market evolution
We will explain the allusions and describe each scenario below.
Scenario 1: Götterdämmerung
In Wagner's Ring Cycle the gods are locked in battle.
The New York Stock Exchange and the Nasdaq stock market could reassert their former dominance and fight to the death with only one survivor.
Situation
The NYSE dismantles the National Market System. Without market data revenues the regional exchanges cease to exist as effective alternatives for executions at any level. Nasdaq and the NYSE quickly engage in fierce direct competition and one survives.
Environment
The SEC bemoans the lack of competitive markets but is pleased to see the end of fragmentation. It insists that all orders receive time and price priority and that all trades be printed through the surviving market. A central limit order book (CLOB) evolves for small executions, 'electronic-floor' crosses handle large trades and a bulletin board handles illiquid issues. The cost of execution for CLOB trades approaches zero as competition to provide access to the market diminishes the agency fees. Brokers only provide financial guarantees for these trades and, as automation reduces the settlement cycle, there is little settlement risk. Institutions seek free direct access for small agency trades.
Scenario 2: King John at Runnymede
In 1215 the English nobles defeated King John at Runnymede forcing him to sign the Magna Carta.
The large broker-dealers could assert themselves as the dominant centres in a process called internalisation, and wrest primary market control for themselves.
Situation
With NYSE Rule 390 gone, large firms begin to develop 'internal markets'. These markets are effectively in-house ECNs. The firms bring all orders into a central 'market system' that, for the first time, permits firms to look at all orders the firm receives in a common environment. Orders from all sources are able to interact and cross if possible. Customer crosses are priced at the mean of the national best bid and offer (NBBO). Orders not able to cross are handled in accordance with customer instructions, if any. If there are no instructions, the firm handles the orders in one of two ways. Orders that are attractive may be executed in the firm's dealer operation against inventory. Other orders are routed to the markets either by rules-based order routing or using dynamic (smart) routing systems.
Firms find that market-making is decreasingly profitable and focus primarily on proprietary trading using the information from the orders passing through the internal market as the basis for in-house trading decisions. The internal markets provide a real-time view of supply and demand. Match rates for the firm's order flow are initially low because factors such as investment research tend to bias customer orders in an individual security. Firms with diverse order flow (i.e. a mix of retail, institutional and correspondent orders) find the match rates to be acceptable. Firms find they can increase their match rates if they actively seek 'out-of-phase' orders -- orders from traders that tend to be counter (i.e. are on the opposite side) to the majority of firmorders -- to increase the potential for crosses.
The NYSE initially holds the line against in-house crossing, believing that it 'destroys price discovery and kills the chance for price improvement'. When Primex and Nasdaq InterMarket exceed 10% of the total listed market, the Exchange changes rules to permit firms to 'print' internal market trades on the NYSE. The NYSE soon develops a system that permits orders traded in-house to be exposed to the floor.
The SEC is troubled by the potential for conflicts of interest but is unable to craft a reasonable alternative. One major issue develops when the number of orders executing within firms becomes a large portion of total executions. There is a perceived loss of transparency. Connecting major firms' internal markets directly to CTA/CQOC and ITS[2] solves the problem
The SEC makes certain that internal markets protect retail customers. Institutions initially object to the concept of internal markets but their unified front quickly crumbles when many institutions discover that they can cut very attractive deals in which the institution is paid for its order flow deals by major broker-dealers.
Environment
For retail investors there is little perceptible change in their executions. Super-low-cost trades become rare unless the characteristics of the trade are attractive to broker-dealers. Institutions find internalised markets attractive or not depending on the types of orders they typically place. Institutions with orders that are out of phase with typical orders find that they are in high demand. Out-of-phase orders are coveted by broker-dealers and brokers bid for the right to execute the orders. Institutions that have more typical orders find that execution costs rise.
Broker-dealers without diverse order flow have a hard time competing. There is little interest in the services of dealers except on large orders and less liquid securities. There is much consolidation among those firms that historically focused on trading. Pure agency firms do well, but wholesale firms and other primary market makers have a difficult time surviving independently.
Exchanges become less important. An exchange becomes a place to layoff unattractive orders and to rebalance inventory. Most of the regional exchanges do not survive. The NYSE becomes less of a general marketplace and focuses on providing special services to its members. Nasdaq finds that having become for-profit puts it in direct competition with its members. Nasdaq becomes a price-reporting facility and is dependent on systems efficiency to survive. ECNs choose among three strategic roles. An ECN can survive as an adjunct to an agency brokerage operation. Alternatively, an ECN can become a facility in support of major internalising broker-dealers. Finally, the ECN can give up its dream of becoming a self-sustaining market and become a portal to other markets.
Scenario 3: Chêng Ho
During the Ming Dynasty the civil servants became more powerful than the emperor and attempted world exploration and conquest.
For thirty years it has been predicted that institutional investors, whose order flow dominates the equities markets, would assert their economic muscle and seize control of the trading process.
Situation
Institutions become increasingly unhappy with the perceived broker-dealers'conflict of interest in their role as intermediaries and seek an intermediary of their own. Several alternatives are tried, but one workable solution is the creation of an independent, mutually owned intermediary. This is very similar in concept to the creation of Knight Securities by broker-dealers, but the new broker/dealer(s) is(are) mutualised among the institutions instead.
Brokers counter institutionally-owned intermediaries by offering to run proprietary 'internal markets' for institutions. These systems are similar to State Street's Lattice system and permit institutional fund complexes to cross orders internally. More aggressive sites begin to seek added order flow from other customer groups and other institutions. The additional order flow increases the match rates.
Regional exchanges, under strong competitive pressure, begin to offer membership to institutions. Clearing broker-dealers are created to clear and guarantee trades.
The effect of both systems is to reduce institutional commissions dramatically. Institutions with attractive order flow (high probability of natural crosses) are paid for order flow. Other institutions pay rates near zero for simple orders and lower commissions for more complex trades.
Environment
While institutions (e.g. Fidelity) were concerned about sending to their own broker-dealers (e.g. Fidelity Capital Markets), they are very happy to use a mutually owned broker/dealer. Institutions participating in the ownership of intermediaries receive both lower commissions and dividends from their ownership.
Retail investors with smaller orders see little impact but those with larger orders (roughly several thousand shares) that now might match with an institutional order have a harder time executing.
Broker-dealers see their role diminish. They have a smaller cushion from simple trades to subsidise harder trades and the cost of complex trades becomes higher. Institutions, however, are unwilling to pay the true cost of complex trades. There is significant contraction in the business. The primary source of broker/dealer revenue becomes proprietary trading. The overall size and importance of the broker/dealer community diminishes.
Scenario 4: Gulliver bound by Lilliputians
Gulliver awoke in Lilliput to find he was completely bound by hundreds of strings. Individually the strings could be easily broken, but together they were too strong for Gulliver to break.
With some relatively straightforward operational changes, trading could devolve to the point where eBay-like electronic markets would permit investors to trade with little or no need for conventional intermediaries.
Situation
A number of execution alternatives begin to emerge that are individually quite small, but together represent a substantial fraction of the market:
- Listed companies create internal markets that permit existing shareholders to sell their shares to other investors. This creates a proprietary secondary market for the companies shares. A clearing broker/dealer is hired or created to guarantee trades.
- Web portals create trading environments where investors can post 'interests to trade.' These portals register as broker-dealers to facilitate the transactions.
- Institutions, wanting to participate in these emerging markets, create their own sites where they are prepared to act as 'market-makers of the moment' for securities in which they have an interest. While these sites seldom offer two-sided markets, there are in aggregate enough such sites to provide general execution immediacy.
Environment
At first these markets are highly fragmented. The cost of searching for liquidity is high even though the execution costs are low. Technology quickly solves this problem. The sites offer both counterparty search engines and smart orders.
Counterparty search engines -- 'stock bots' -- create a list of potential trades that a prospective customer can access. Traders place orders with potential counterparties using financial guarantees from intermediaries or posted credit lines. Bilateral agreements between the portals facilitate clearing.
Smart orders allow a would-be trader to send an order into the Internet armed with specific instructions on what to trade and at what prices. The order can be active or passive (the order either executes or simply reports back its findings). The search process creates latency -- a lag between the time when the order is entered and when its executed -- in which prices can change, altering the economics of the trade. Creating directories throughout the Internet that keep updated lists of bids and offers collected from multiple sites solves the latency problem. Smart orders can check these directories while in transit to update their search lists of potential targets.
Summary
These four distinct market structures are based on who controls (or exerts the primary influence on) the markets and whether the markets are fragmented or centralised. While none of these environments are likely to evolve as described, they present a framework for evaluating less extreme alternatives.
The possibility of stable equilibrium
One outcome that seems unlikely is that a highly integrated, stable environment will evolve. First, each of the existing trading environments satisfies the needs of one or more important constituencies. Moreover, innovation and economics are creating new alternatives. And even established competitors are changing at a frenetic pace.
Implications for Europe and Asia
It is common for spokespeople from European markets to react with alarm and disdain to the chaos in US market structure. It is true that most national markets in Europe trade in a single market, or in linked markets as in Germany. However, if one looks at Europe as a single entity rather than a collection of economically independent countries, the parallels are strong. For cross-border trading in the most liquid securities there is great fragmentation and it is hard to construct a scenario for the creation of a single pan-national market. Past attempts to create a pan-European market have failed. So long as marketplaces can court user groups with diverse needs, unity will be hard to achieve.
Asia has not yet begun to consolidate and so the traditional, national market remains strong in most countries. The presence of large companies operating in many different countries suggests an environment where each market centre can begin to poach significant amounts of order flow in companies from other countries that are well known to local investors. Inevitably, local markets will begin to compete both within the region and globally.
Conclusion
The inter-market competition and fragmentation in the US is more likely the norm than a transient condition that will fade in time. Technology creates an environment where location matters little, tradition matters less and national pride matters not at all. Perhaps the best commentary on the future of the markets is a portion of a poem entitled The Second Coming by William Butler Yeats, written in 1919 about the 'troubles in Ireland:
Turning and turning in the widening gyre
The falcon cannot hear the falconer;
Things fall apart; the centre cannot hold;
Mere anarchy is loosened upon the world,
The blood-dimmed tide is loosed, and everywhere
The ceremony of innocence is drowned;
The best lack conviction, while the worst
Are full of passionate intensity.
The idea of a 'centre' to the market has been lost. It is difficult to predict the future. Alas, we do not lack for those, 'full of passionate intensity', who are only too pleased to preach their vision of the future.
[1] In 1997 the SEC began to require Nasdaq market makers with non-discretionary customer limit orders to display the order in the market maker’s quote if the limit price was better than the market maker’s existing quote. Electronic limit order books registered as brokers (termed ECNs by the SEC) became a means for displaying customer limits in Nasdaq.
[2] The Consolidated Tape Association and Consolidated Quote Operating Committee (CTA/CQOC) are the entities in the US that administer the collection of last sale reports and quotes (respectively) for NYSE and AMEX-listed securities. The entities are currently owned by the exchanges.
CTA/CQOC also collects and allocates the revenues from exchange fees among the participating exchanges. The Intermarket Trading System (ITS) routes orders among the exchanges.