Questions for the future

Tee Williams

I have recently been thinking about the future of the securities industry, and particularly exchanges and markets, in the context of the changes that began about forty years ago. If we think about the next forty years, what are the questions the industry leaders must address to make reasonable decisions that will ensure their firms and exchanges survive and prosper? I contend that the following eight questions are critical:

1. Will trading in individual securities or contracts coalesce into a single or dominant market in each trading area, or will fragmentation continue or increase?
2. Will markets remain largely local/national, or will trading become truly global?
3. Will ‘best execution’ concerns be defined so that individual transactions can be measured against an objective standard?
4. What underlying investment decision-making strategy – active stock selection, passive investing, or quantitative/model-based investing – will generate the dominant share of orders?
5. What will be the primary pricing model for securities and futures trading?
6. Which of the primary securities/futures business models – exchanges, broker-dealers, institutional investors, and/or vendors – offer the best investment and employment opportunities in the future?
7. How will the tension between market transparency, which is necessary for broad market participation, and the concern of those trying to move large positions with minimal market impact be resolved?
8. What are the prospects for market regulation over the next forty years?

These same eight questions might have been asked in 1967. Firms that thought about these questions, or at least had the right answers, are still here. They are, in spite of humble origins in many cases, the dominant financial institutions today. Mighty firms with no answers are gone. I will look at the changes of the last forty years as they relate to these primary questions.

In the late 1960s, the markets were in many important respects unchanged from the beginnings of trading four centuries before. Traders on the floors of exchanges in Amsterdam, Frankfurt, London, New York, Paris, Tokyo or Toronto in 1967 would find trading was largely similar to that which first occurred at the Amsterdam Stock Exchange in the early 1600s. Moreover, trading in over-the-counter markets, but for the telephone, was essentially similar to trading in the Bourse family tavern in Bruges, Belgium in the 14th century or the coffee houses of Amsterdam and London thereafter. Trading was person-to-person, deals were recorded by hand, paper certificates were sent out for manual transfer of ownership, and exchanged for bank drafts once the new owner was recorded.

One important factor makes the current moment very much like the period of the late 1960s. In the late 1960s most firms and markets were only just beginning to use computers in their businesses. The changes that have happened since could not have taken place without those computers. Now we are just beginning to come to terms with a networked marketplace. In spite of a dramatic number of customer web sites, most firms still operate as if their operations are isolated islands of technology. As the isolation ends, the structural changes are likely to be as profound as the introduction of automation forty years ago.

Question one: Dominant market or fragmentation?

In 1967, the dominant markets were in Amsterdam, Frankfurt, London, New York, Paris and Toronto, and Chicago for futures. Tokyo was growing rapidly. In each of these financial centres there was one dominant exchange, and in some cases a secondary exchange. In the last forty years, these dominant marketplaces have remained remarkably stable with a few significant changes. Nasdaq, the CBOE, Liffe and Eurex were created from scratch and have become global market centres. In the rest of the world, exchanges were created and have become important regional centres in the former Soviet block and in Asia.

For the first thirty of the last forty years, most attempts to create competitive trading venues – and there were many attempts – were miserable failures. However, beginning about ten years ago, a number of alternative trading systems (ATSs) were created that have been successful.

What changed to make recent competitive markets successful? In all of the cases that I can find the following characteristics apply to trading venues that have sought successfully to become competitive with established markets:

• Each was automated, although automation alone was not a guarantee of success;
• Each targeted a segment of order flow in a market with diverse trading motivations;
• Each priced transactions to appeal to the firms that controlled the order flow for the targeted segment;
• Each used some mechanism to provide a guarantee of liquidity for those situations when no natural match was present in the trading venue; and
• Each left the control of the execution in the hands of their trader/members.

Looking at the next forty years, can we expect the same level of stability in markets? Or, should we expect that, having found the key to competing with existing markets, new market venues will continue to be created to replace existing venues gobbled up in combinations such as ArcaEx-Euronext-NYSE and BRUT-Inet-Nasdaq?

Question two: National markets or global?

In the last forty years there has been much talk of global markets. In the early 1980s, the large universal banks were worried about ‘global limits’ for customers accessing both credit and securities services in multiple locations. In the middle 1980s, investment banks talked about ‘moving the book’, meaning twenty-four hour trading in which a global inventory would be managed in successive markets as the markets in one time zone closed and the markets to the west opened. However, in times of stress, such as the market crash in 1987, talk of global markets faded, and most firms reverted to trading in local markets.

Now we see the beginnings of a truly global market for a few large companies such as AstraZeneca, Microsoft and Sony. Traders actively arbitrage securities for the same company in multiple global market-centres as if the securities were perfect substitutes for one another.

More interesting, small US companies are reported to have made their initial public offering (IPO) in London because of the restrictive reporting requirements of the Sarbanes/Oxley laws, and non-US companies are electing not to list in the US for the same reason.

Since the beginnings of investing, exchanges and regulatory bodies have acted with impunity to impose reporting requirements, trading restrictions, taxes such as stamp duties, and any other restrictions that they have seen fit. Likewise investors, particularly institutions, have treated investments out of their home countries as ‘foreign investments’ for the purpose of diversification. Few investing institutions actually manage their portfolios globally; few broker-dealers actually manage their inventories and capital positions globally, even if they actively trade globally.

In 2047, is it reasonable to expect the same level of national independence to exist? If not, what happens to those financial centres that continue to act as if they have complete control over all who would wish to use their markets? What happens to countries that ignore, or are slow to recognise, the alternatives that are open to their investors, listing companies and intermediaries?

Question three: An objective measure of best execution?

In both Europe and the US, the concept of best execution has become a dominant issue for regulators. One of the primary problems with the idea of ‘best execution’ is that it is currently defined on a transaction-by-transaction basis. If you consider the success of market making that evolved from the idea of a National Market System in the US and Big Bang in London, it rests on the economic realisation that, for a market making inventory, groups of trades can be more effectively managed, with lower risk, for the entire inventory or portfolio than for individual transactions. Before May Day (May 1, 1975) when commissions ceased to be fixed by the New York Stock Exchange, most broker/dealers could not conceive of how they would make money if commissions were reduced and could go below the ‘standard cost’ for a transaction. They failed to take into account the opportunity to reduce the costs of clearing and to profit from growing order flow, and to profit from relationships among multiple customers rather than individual transactions.

Regulators and investors, however, want an objective measure for each transaction. Ironically, this can have the unintended effect of creating a de facto fixed charge for transactions if intermediaries have to ensure that each transaction meets an objective criterion.

What is likely to be the impact on the total cost of trading if we continue to press for a transaction-by-transaction measure of quality?

Question four: The dominant investment model?

Since the beginnings of institutional investment, substantially all decision making has been based on a portfolio manager who decides which securities will do well and which securities will do poorly and issues trading decisions based on his (predominantly men back then) judgment. About forty years ago, academic researchers began to develop a series of hypotheses that together have come to be called Modern Portfolio Theory (MPT). MPT called into question many of the methods of managing money that had existed for centuries. Nevertheless, many portfolios continue to be managed by active security selection.

Three important new investment alternatives began to develop at about the same time based at least in part on MPT. Passive or index trading tries to replicate the performance of major market indexes rather than trying to ‘beat’ the index. Passive trading is based on the theory that markets are efficient and it is difficult to consistently earn higher returns than the returns on the market as reflected in an index. ‘Quantitative investment’ is my umbrella term for all types of trading where a computer model rather than the judgment of a portfolio manager is used to make decisions to change the components of a portfolio. (Obviously, judgments are built into the quantitative models.) Finally, hedge funds (another umbrella term for funds that are permitted to take short positions) began to become significant in the 1960s. Each of these newer styles – there is much overlap among the three – has the impact of shortening the time horizon for investment re-evaluations. Moreover, many quantitative investors and hedge funds attempt to profit from transient market imbalances, ignoring the ‘worth’ of the securities traded.

More active trading and less dependence on judgments of ‘intrinsic value’ generally result in more volatile markets and higher turnover rates. In 1968 Wall Street nearly collapsed because NYSE trading volume had reached 20 million shares per day. We now have daily trading volumes that exceed four billion shares on the NYSE. Moreover, the widespread growth in wealth and the demand for investment services that occurred in Western Europe and North America during the last forty years are currently spreading to Eastern Europe and Asia. This growing wealth may well spread to much of the rest of the world in the next forty years.

If we consider the markets in another forty years, how will institutions manage the huge inflows of investment funds and how will the markets handle dramatically more active and volatile trading patterns?

Question five: The primary pricing model?

Forty years ago, the majority of revenue to intermediaries came from commissions. With the end of fixed rates, there was a strong move toward spreads as a means of pricing. This was particularly true in the US because the spread on equities was guaranteed to be USD0.125, or one eighth of a dollar, the minimum price variation (MPV). Other countries did not have this artificial limitation, but some, such as Canada fixed the MPV at a CAD0.05. Whatever the causes of comparatively wide spreads, transaction-based pricing was a source of immense revenues. Firms became extremely profitable.

Individual dealers, traders or investment bankers over the last forty years who were good at selling securities and who had a good list of contacts could count on income that was dramatically higher than a counterpart with comparable skills in other industries such as manufacturing or retail. For an individual, the key was knowing which institutions might be willing buyers or sellers of securities, or which corporations might be available for sale or interested in mergers. For good sales people this knowledge produced immense revenues for their firms, which in turn permitted the firms to pay those critical individuals with the right contacts, i.e., a good ‘Rolodex’, commissions that were staggering.

During the same period, exchanges and vendors of market data were extremely successful in charging based on individual users of their information. Users and/or display devices were the primary method of charging for market data. Since everyone who invested or traded needed to have a display device, the profits were extremely high.

Looking forward, volumes seem attractive for firms because, as we have seen, any reasonable estimate suggests that transaction volumes are likely to continue to grow. Even if the price for a single transaction becomes extremely low, the aggregate revenues should be tremendous.

With the Internet and other means of selling securities, and with more trades happening based on models and computers directing the trades, the value of the individual diminishes in the trading process. Individuals do not go away, but they become less critical. If the value of a salesman’s contacts shrinks, his or her commissions can be expected to shrink as well. Likewise, if the importance of individuals in the trading process diminishes, then pricing based on display devices or users will not provide optimum revenues to vendors and exchanges.

Looking at the next forty years, should we expect individual incomes to remain abnormally high in the securities industry, or will they fall more into line with other industries? Also, can exchanges and vendors continue to charge based on users or display devices as algorithms replace traders? Or is there a need for a new unit of pricing that more accurately reflects the changing value and usage of the information that is being sold?

Question six: Exchanges, broker-dealers, or institutional investors?

Forty years ago, exchanges were localised entities controlled by the local brokerage community. Broker-dealers in the US and Japan, or brokers and dealers as separate entities among most of the British Commonwealth countries, were small and poorly capitalised. Independent investment companies were relatively new and very small. Banks and insurance companies controlled most investment funds and were huge by comparison to securities intermediaries. In Europe there were universal banks, but commercial banking dominated their activities. I remember making a comparison in a report in the late 1970s, and I found an SEC document that showed the total net capital of the US securities industry was about USD4bn, while Citibank alone had assets of about USD80bn. It seemed that commercial banks could easily dominate the smaller securities industry if they chose.

In reality, however, the last forty years has been the period of broker-dealer dominance. First, broker-dealers have been more aggressive and creative. Banks accepted the limitations that kept them from engaging in brokerage activities, while brokers actively sought ways to circumvent the rules and offer banking services. Second, and this may be related to the first, the last forty years have been dominated by the process of securitisation – the transformation of bespoke financial transactions into standardised securities with liquid secondary markets. Forward markets became futures markets, short-term loans became commercial paper, syndicated government loans became government bonds, many corporate loans became high-yield bonds, and mortgages and other loans were packaged as asset-backed securities. Also, investment services such as mutual funds, private pensions and unit trusts have become dominating forces in the trading markets. In short, those organisations in the best position at the beginning of the period to dominate the last forty years – the banks – did not do so.

Now many exchanges are public companies. Also, many execution venues have elected to organise as brokers rather than as exchanges. Many banks and broker-dealers have combined, often including investment management arms as well.

The question then for the next forty years is: which functional area within the securities markets is most likely to be dominant? Where should a young person seeking a career in the industry look for employment, and how should an investor evaluating the industry rank the prospects of differing industry segments?

Question seven: Transparency versus market impact?

For the last forty years, regulators and markets have struggled to increase transparency. Forty years ago, the US SEC forced the over-the-counter market first under the self regulation of the National Association of Securities Dealers, previously a trade association, and then to create the NASD Automated Quote system or Nasdaq. Dealers accepted a quote advertisement system, but fought last sale reporting, fearing that by reporting his trades in illiquid markets a dealer reporting a trade would alert others to his position increasing his risk. When the SEC forced last-sale reporting, to the amazement of dealers, transparency caused the markets to grow, and the increased liquidity masked positions much better than secrecy.

With the exception of Big Bang, transparency has been a success ever since and one of the primary goals of regulators. In the case of Big Bang, the success of the 1975 Amendments to the Securities Acts in the US created unrealistic expectations for Big Bang. Too much capital flooded the market and firms blamed the losses that resulted on the initial mandate that every trade be reported. Now we are on the brink of the Markets in Financial Instruments Directive (MiFID) mandating full disclosure in all of the European Union countries.

In spite of the success of transparency, we now have a rapid growth of so-called ‘dark pools’ in which trades are negotiated between anonymous principals with no quotes reported. This phenomenon reflects the fact that as securities markets have increasingly adopted limit order books as the primary tool for executing securities, trade sizes have dropped dramatically, making large orders more expensive and cumbersome to execute. Also, it has long been the case that very large orders seek to maintain anonymity by using inter-dealer brokers (IDB) and ‘beards’ (brokers who obscure the identity of the principal.) In fact, the dark pools are actually the electronic equivalent of an IDB. Moreover, order management systems, very fast order routing systems, and execution automation permit principals to withhold orders from market centres until they become marketable. This creates a ‘virtual book’ of orders that overhang the market but are not displayed to other traders in the market.

A more curious issue also arises. As we get more information about securities trading on multiple, fragmented markets it seems that, like Heisenberg's Uncertainty Principle, our increased knowledge of the true prices of securities may not be commensurate with the cost and effort expended in trying to understand prices more precisely.

Looking out forty years, where will we strike the balance between transparency and the need to be anonymous? What new types of information will be required, from what sources, and what means will we employ to get it?

Question eight: Market regulation?

Finally, from the late 1960s to the present, markets in most countries have become more regulated. The SEC in the US has become more powerful and forceful. The Commodities Futures Trading Commission, the Financial Services Authority, and regulatory bodies in a number of countries with new or expanding trading markets have all been created to regulate and promote national markets. Each of these bodies has authority that ends at its national boundaries even though, as we have seen, markets are finally becoming global and technology makes it possible for entities in one country to access and participate in markets in other countries on an even playing field with locals.

Market participants avoid regulation even when the regulation is in their own best interest. In the case of the United States, for substantially every securities regulation on the books, someone went to prison and others were bankrupted or injured financially. Nevertheless, investors and traders generally seek to avoid regulation. It is difficult to look at the next forty years and not expect that traders and investors will actively seek to use those markets with little regulation and avoid those markets with much. It is also hard to believe that some countries will not seek to become ‘trading havens’ with limited market regulation as countries have sought to be tax havens.

Against this backdrop, you have to know that someone will do something stupid. With modern technology, stupid can become very big, very fast and very ugly.

So the question for the next forty years is how soon will national regulation be eroded? What will be the nature of the disaster that causes us to realise that we have to develop some new mechanism for trans-national market regulation? And what form will that regulation take?

‘You've got to ask yourself one question: “Do I feel lucky?” Well, do you, punk?*’

If you have your future or your capital invested in a broker-dealer, an investment company, an exchange, or a vendor that serves the markets, you have to consider the eight questions above. One thing about the future: it is a good bet that there will be one – a future that is. It is less certain that every current participant will have a part in it. Certainly the firms that did not ask the right questions and respond with the right answers forty years ago are not here now.

* Harry Callahan (Clint Eastwood); Dirty Harry; Warner Home Video; 1971