Given that growth is a beneficial thing, how does an exchange go about doing it? Over the years, most exchanges have grown organically, that is via the creation of new products and the search for new customers for existing products. This is in contrast to growth via mergers and acquisitions, to which we will turn shortly. Let's first look at the derivatives exchanges. In 1955, there were precisely 61 different futures contracts listed for trading on U.S. futures exchanges and a lesser number elsewhere in the world, mainly in Japan. In Table 7.1 and Table 7.2, we have listed the top 20 derivatives contracts traded globally in 1955 and then 52 years later, in 2007.

A comparison of these two charts makes it clear that organic growth, via new product development and expansion of demand for existing products, has been a major feature of the global derivatives market. Specifically,

So, starting with these 61 contracts that existed in 1955, over the next 50 years another 842 futures contracts were added in the United States. That's about 17 new contracts a year. Outside the United States, the pace was even faster, since during the shorter 25-year period from 1980 to 2005, international exchanges created 887 new futures contracts, or about 35 new futures contracts per year. Over half of these contracts (both U.S. and global) died within a few years, but some of the surviving ones contributed significantly to the growth of the world's futures exchanges. In the United States, the contracts driving exchange growth were mostly launched in the 1980s (e.g., Eurodollars in 1981 and the S&P 500² in 1982, both at the CME).

In the rest of the world, where exchanges were created more recently, the blockbuster contracts that contributed significantly to the growth of these exchanges were created in the late 1990s and into the new millennium. For example, the one-day interest rate contract at the Brazilian BM&F accounts for 55% of all futures trading at the giant exchange, yet was started only in 2004 (see Table 7.3). Or at the extreme case, the dollar/peso contract traded at ROFEX in Rosario, Argentina, represents 99% of all trading volume on the exchange, yet the contract was started as recently as 2002. In Japan, at the Osaka Securities Exchange, which has been in the derivatives business since 1987, its largest contract is the Nikkei 225 Mini, which started in 2006. So the strategy of organic growth from product innovation has been quite important.

Global data on the equity exchanges is not readily available as far back as the derivatives data, but according to the records kept by the World Federation of Exchanges, the value of global share trading increased from $9.7 trillion in 1994 to $200.8 trillion in 2007. ³ This 1970% increase in trading value was attributable mainly to more trading in existing products, since the number of listed companies increased only 37% over that same period, from 30,186 listed companies in 1994 to 41,351 listed companies in 2007. So again there was a significant amount of organic growth among stock exchanges, but little of it came from a net increase in company listings. As discussed in the later chapter on product innovation, stock exchanges in some countries have benefited from the recent rapid growth in the number of, and trading in, exchange-traded funds (ETFs).

In other cases, the exchanges have grown significantly from either the entry of new users or the growth of one of those groups of users. For example, the CME's Eurodollar contract has benefited greatly from the rapid and sustained growth of the interest rate swaps market. Swaps dealers entering into agreements to swap fixed for floating interest rate payments (or vice versa) would hedge their net risk in Eurodollar futures. In the case of energy derivatives, the rise of energy merchandisers and marketers in the 1990s has significantly driven increased volumes in crude oil and natural gas. And of course the rise of managed futures and hedge funds, which often make extensive use of derivative markets, has contributed significantly to increased volumes at many different exchanges.

One other serendipitous source of volume gains, at least at those exchanges that are or have become electronic, is the rise of algorithmic trading. ⁴ Algorithmic traders are often making tens or even hundreds of trades per second and contributing significantly to volume growth at exchanges. And whether they utilize algorithmic trading systems or are staffed by somewhat old-fashioned point-and-click electronic traders, proprietary trading shops have popped up in a number of places and also have contributed to volume growth. Of course, electronic markets, especially those with direct access, make these markets accessible to anyone with a computer and money to trade and should significantly expand the customer base in the future.

There are limits to what organic growth can do for an exchange. We have established that exchanges worldwide have added many new futures contracts over the past few decades. We also noted that most of these contracts are simply not successful. In countries where derivatives markets are newer, there have been some recent contributors to significant volume growth, as pointed out earlier. But in the more mature markets, it is much more difficult to do this. First, in mature markets, the best fruit has been picked long back. The blockbuster contracts were those that were created in the 1970s or 1980s (thus most of the trading volume at the CME comes from contracts created in 1981 and 1982). So it is difficult today to find brand-new ideas that translate into futures contracts with high volumes. Attempts to create clones of successful contracts at other exchanges generally do not work either, because of the liquidity-driven monopoly problem. What this means is that once an exchange has established a reasonable degree of liquidity, typically with a relatively large number of buyers and sellers who are always ready to deal and therefore create relatively tight bid-ask spreads, no one will trade the same product at a new exchange that decides to list it, because there are few buyers and sellers to trade with. So attempts to establish clones of successful contracts usually fail. This results in a monopoly in that product for the exchange that lists first and develops a reasonable degree of liquidity.

There may be more room at newer international exchanges to grow organically. As shown in Table 7.3, the top products at these exchanges were typically started much more recently; the average age of the blockbuster contracts is about 10 years. In China, the top product is soybeans because the government has not, as of late 2008, given the go-ahead for the new financial exchange to start trading. Once it does, there will be a huge opportunity for growth in financial products, and it will be only a few years before the top product is a stock index or government bond rate rather than soybeans. Note also that the top product on average accounts for over half of total exchange volume in these markets. But this is not much different than the United States. Before the mergers with the CBOT and NYMEX, the CME's top product, Eurodollars, accounted for about half of total volume, and the top two, Eurodollars and the E-Mini S&P 500, accounted for about 75% of total CME volume. One does worry a bit, however, about MexDer (Mexico) and ROFEX (Argentina), where a single product accounts for 93–99% of total volume.

We started this chapter by saying that there were two ways for an exchange to grow. The first was to grow organically by either adding new products or adding customers to existing contracts. The second way to grow is to add products to an exchange's portfolio by simply buying those products indirectly through the purchase of another exchange that has contracts that are different from those at the acquiring exchange. And different exchanges almost always have different products because the liquidity-driven monopoly tendency leaves most exchanges with a unique set of products. For example, when the CME and CBOT merged in the summer of 2007, there was not a single duplicate active contract between the two exchanges. Not a single one of the CME's 88 contracts was actively traded at the CBOT. So the CME acquired a solid set of new product lines in this acquisition, including futures and options on U.S. Treasuries and grains. Likewise, when the CME bought NYMEX, it acquired new product lines in derivatives on energy and metals.

Though there have been mergers throughout the history of futures and stock exchanges, they were relatively rare. The CME, for example, experienced spectacular growth until 2007 based only on creating new futures and options contracts and developing their customer base, without ever participating in a merger. The CBOT, during its almost 150-year history, has absorbed only two other exchanges, both incredibly tiny. ⁵ The Kansas City Board of Trade and the Minneapolis Grain Exchange have had long histories without any merger or acquisition. On the stock side, the New York Stock Exchange and the London Stock Exchange trace their histories back over 200 and 300 years, respectively, but neither participated in a merger till 2005 (NYSE with Archipelago) and 2007 (LSE with Borsa Italiana). There have, of course, been mergers of exchanges, especially when one of the two exchanges is beginning to have trouble surviving on its own and needs to be folded into another exchange to consolidate the administration of the two.

There was an earlier wave of exchange consolidation driven by the first round of technological change in communications—the telegraph, ticker tape, and telephone. ⁶ Before the development of these early communications devices, exchanges were established in multiple commercial centers throughout a number of countries. For example, the United States once had 250 different stock exchanges. India had 23. England had 20. Information had to be delivered by foot or horseback, so every significant urban center needed its own market. With the development of the telegraph, the telephone, and the ticker tape, it became much easier to both know the almost real-time prices at distant exchanges and to trade on those exchanges. So distance began to matter much less, and traders began to seek out the exchanges with the greatest liquidity for the stocks in which they were interested. This resulted in a tendency toward consolidation and the elimination of duplicate products at multiple exchanges spread all over the country. Of course, the regional exchanges that went into decline would sometimes fight the trend by getting the state to intervene. For example, there was a rule in India that required each company to list itself on the exchange closest to its headquarters, even if it was listed on the country's major exchange at the time, the Bombay Stock Exchange. So firms based in Calcutta (today called Kolkata) would generally list at the BSE but would be required to also list and pay listing fees to the Calcutta Stock Exchange. After the screen-based National Stock Exchange (NSE) was created in 1994 and the BSE shifted from floor to screen to survive, virtually all trading shifted to the NSE and BSE, and the 20 or so regional exchanges were left with no trading activity, though they still earned listing fees because of this rule. The rule was finally dropped early in the new millennium, and then the regional exchanges really did die. ⁷

Though the Indian consolidation took place with very little merger activity, the case has been just the opposite in Japan. In fact, we know of no country in which there was such a rapid period of consolidation as there was in Japan during the 1990s and early 2000s (see Figure 7.1). Japan has long had a large number of physical commodity exchanges in commercial centers of various sizes spread throughout the country. ⁸ In 1991 there were 17 Japanese derivatives exchanges based on physical commodities. By 2007 those 17 exchanges had been consolidated into four, a 75% reduction in the number of physical commodity exchanges. For example, the Tokyo Grain Exchange had absorbed three other exchanges: the Tokyo Sugar Exchange, the Hokkaido Grain Exchange, and the Yokohama Commodity Exchange. The Kansai Commodities Exchange had absorbed five exchanges: the Kobe Grain Exchange, the Kobe Raw Silk Exchange, the Osaka Grain Exchange, the Osaka Sugar Exchange, and the Fukuoka Futures Exchange. And in the largest consolidation, the Central Japan Commodity Exchange had absorbed six other exchanges: the Nagoya Grain and Sugar Exchange, the Nagoya Textile Exchange, the Toyahashi Dried Cocoon Exchange, the Kobe Rubber Exchange, the Osaka Sanpin Exchange, and the Osaka Rayon Exchange. ⁹ In fact, by the end of these 21 mergers, only one old exchange name survived from the 1980s and that is the Tokyo Grain Exchange.

What drove the consolidation was the fact that some of the more regional exchanges simply could not compete with the larger, more liquid exchanges in Tokyo and Osaka and thus were not able to continue to operate a viable business. These Japanese mergers often involved government coaxing and were a mix of government guidance and exchange strategy. Remember, these exchanges are still member owned, so mergers are a bit more complicated. So it was the Ministry of Agriculture, Forestry, and Fisheries (MAFF) that recommended that the Tokyo Sugar Exchange merge into the Tokyo Grain Exchange. However, the Yokahama Commodity Exchange found itself in a very difficult financial situation and requested that the TGE take them over. ¹⁰

The consolidation in Japan was driven by changes in technology—specifically, falling communication costs, as mentioned earlier—and it involved weaker exchanges that were losing volume being taken over by stronger exchanges that were gaining volume. And this was the way things went worldwide. For example, in 1979 the Cocoa Exchange in New York had been for a decade shrinking about 3% per year, while the Coffee and Sugar Exchange had been expanding 8% per year. In a very logical move, the much stronger and more rapidly growing Coffee and Sugar Exchange took over the Cocoa Exchange to become the Coffee, Sugar, and Cocoa Exchange. ¹¹

This principle of a dynamic, growing exchange absorbing an exchange in decline was played out in an important U.S. merger that took place before the exchanges began to switch to screens and become public companies, and it illustrates the difficulty of executing a merger between two member-owned exchanges. Until its merger with the CME in 2008, which is part of the new age of exchange mergers, NYMEX had experienced a single merger, the one with COMEX in 1994. The NYMEX and COMEX merger is a great example of how difficult it can be to execute a merger between two not-for-profit, member-owned exchanges. For years, the two exchanges had talked about the possibility of merging. In the autumn of 1987, consulting company Arthur D. Little, Inc., recommended that the two exchanges merge their boards, their administrative staffs, and their clearing mechanisms. COMEX officials had supported the merger for some time, but it now appeared that NYMEX officials were also ready to buy into the Arthur D. Little recommendations. So it was not surprising that in the summer of 1988 industry publication headlines trumpeted, “New York exchanges edge towards merger” and reported that NYMEX officials had finally reached a decision to go ahead with a merger plan. ¹² Two weeks later, the merger was off, supposedly because of pushback from NYMEX members. Feelings and rivalries between exchange memberships run deep, and the summer 1988 merger of COMEX and NYMEX proved to be just another false alarm. Then in early 1990, the two exchanges announced that they were exploring a merger and that they would form a joint governing board that would oversee any merged exchange. ¹³ After a year of discussion, on February 19, 1991, COMEX announced that it had rejected a NYMEX proposal and NYMEX announced, “Having reconciled all differences over material terms, including a generous offer to allow trading privileges, we are saddened by the failure of our colleagues to accept the proposal.”¹⁴

Finally, three years later, in the spring of 1994, came the real deal. The two exchanges announced that they were, in fact, going to finally merge. On the evening of April 25, 1994, the members of the two exchanges had an opportunity to cast their votes. The NYMEX required a simple majority vote to pass the resolution and 83% of the members voted in favor of the merger. COMEX had a stricter standard, requiring a two thirds majority. They barely made it, with 72% of the COMEX members voting yes.

Why were so many COMEX members still opposed? What is important to understand is that COMEX had once been the big exchange based largely on gold trading and NYMEX the much smaller one. As gold fell out of favor and investors turned to other assets (such as U.S. Treasury securities) as a safe haven in uncertain times, and as the importance of crude oil increased, not only were there the normal jealousies over losing rank, but COMEX members began to worry about their future. Seat prices are a good measure of the health of an exchange. The COMEX seat price, which had been $325,000 in 1974, had fallen by two thirds, to $100,000, shortly before the 1994 vote. This was the result of falling volume and revenues, including a 1992 volume drop of 27% and a net revenue decline of 80%.

But in the months leading up to the vote, volume and revenue began to increase. So, the Exchange's improving fortunes made many COMEX members feel that they no longer needed to merge with the larger NYMEX to have a viable exchange. Right before the vote, COMEX members began circulating letters among themselves arguing that the proposed NYMEX payment was too small and the loss of control was too great. In fact, one COMEX director wrote that “the merger will transform COMEX into an essentially colonial backwater of NYMEX.”¹⁵ On March 1, one group of COMEX members took the legal route and sued both NYMEX and COMEX in a class action suit to block the proposed merger. But when the vote finally came on April 25, 1994, 72% of the members, 5 percentage points above the required 67%, was enough to approve the merger. On July 22, 1994, the CFTC gave its approval, and the merger became a reality. The deal itself made COMEX a wholly owned subsidiary of NYMEX and gave COMEX members $62 million—$42 million at the close of the deal and $20 million spread over the next four years. It should be noted that in contrast to the mergers of the new millennium, this was not a full merger, since each market retained its own memberships and trading rights.

The only U.S. exchange that is the result of multiple mergers involving a number of exchanges is the New York Board of Trade (NYBOT), which in 1997 was purchased by ICE and recently renamed ICE Futures US. The final merger to form NYBOT involved combining the New York Cotton Exchange (NYCE) with the Coffee, Sugar, and Cocoa Exchange (CSCE) in 1998. But the NYCE had already acquired the almost dormant New York Futures Exchange from the NYSE in 1995. And the CSCE was itself a 1979 merger of the Coffee and Sugar Exchange and the Cocoa exchange. The roots of these components go back to 1870 for the NY Cotton Exchange, 1882 for the Coffee Exchange (renamed the Coffee and Sugar Exchange in 1916), and 1925 for the NY Cocoa Exchange.

Beginning in the new millennium, M&A activity no longer was something that occurred occasionally because there was a weak exchange that needed to be acquired by a stronger exchange; it occurred out of necessity, because of the fundamental transformation of exchanges away from floors and onto screens. When exchanges became electronic, economic cost curves shifted, and there were very, very strong economies of scale to be gained from processing an increasingly greater number of trades on an exchange. In a floor-based world, adding new products involved incurring significant costs to earn the new revenues that would accrue from increased trading. New floor space had to be found to build the pits for the new contracts, and new bodies had to be put into those pits to make markets and trade. In a screen-based environment, all one needed was more server space to match orders and disseminate prices for the new contracts.

So, exchanges that could expand trading in their existing contracts or could add product lines (whether by launching new products or acquiring them via an acquisition) would significantly lower their average costs per contract traded and therefore increase their profits. Electronic exchanges left behind with a small number of products and a small level of trading activity would experience much higher average costs per contract traded and much poorer profits. To survive and prosper, exchanges had to get bigger by increasing the number of contracts traded on their electronic platforms. This need to get larger resulted in a frantic quest to merge with other exchanges.

This significant ramping up of the economies of scale factor created by the shift to screens drove exchanges to want to merge, but it was another transformation we’ve discussed that made it easier for them to merge. That of course was the shift from private clubs to public companies. To create a merger between two public companies, you have to convince the shareholders that it's a good idea. For many public companies, a large number of shareholders hold small ownership positions in the company, and these shareholders rarely get in the way of a merger proposed by management and the Board. Even the large shareholders generally view their stock as a financial asset, and they are looking for a good return on their investment. So when a merger makes sense, shareholders rarely get in the way. Not-for-profit member-owned exchanges are a different animal altogether. Exchange members often have a strong emotional view of their exchange, and their exchange membership generally represents a much larger part of their net worth than does a stockholder's shares in any given company. The exchange is the place they go to work every day. They have a bond with other exchange members, not dissimilar to the bonds shared among classmates at the same high school or college. For exchanges in the same city (like the CME and CBOT, or the NYMEX and COMEX), there are often rivalries, as there would be between two football teams. So when members are making a decision regarding whether to merge with another exchange, especially a cross-town rival, it is not a cold, steely-eyed business decision but instead one infused with a significant degree of emotion.

In addition, though there are clear rules for managing mergers and acquisitions among public companies, this is much less the case for nonprofit, member-owned institutions. So it should have been no surprise that the merger between the two not-for-profit exchanges, COMEX and NYMEX, took over six years of on and off and on again negotiations, whereas the CME/CBOT merger of two publicly traded companies took less than two years. ¹⁶

Once an exchange becomes screen-based, it has a strong need to merge to lower average costs and pump up profits, and if it also becomes a publicly traded company it becomes much easier to merge and it almost has to engage in the merger game. Let's take a look at a few mergers that have been part of this new age of exchange M&A. The first is the story of two mergers entered into by the New York Stock Exchange. The first merger, with Archipelago, actually brought the NYSE into the modern era of electronic, publicly traded exchanges. The second merger, with Euronext, was a new age merger, with the special characteristic that it was the first trans-Atlantic exchange merger.

Whether the exchange lists derivatives or stocks, screen-based trading by exchanges that had become public companies created a situation where every exchange had to get bigger to take advantage of the new screen-driven economies of scale. Exchanges had to absorb or be absorbed, and this has led to a frenzied period of M&A among exchanges. Most exchanges do only one or two mergers, but they discussed potential mergers with many more exchanges than those with whom they ultimately tied up. The period of active M&A is not over yet. There are hypothetically economies of scale to exploit until there is only one exchange left in the world. We don't believe that the entire world will converge to a single exchange, but we do believe that these strong economic forces will continue to propel mergers for some time to come.