Michael Spencer is fond of boats, planes and automobiles, especially Aston Martins. He also fancies Château Cheval Blanc and horseback riding. The founder and proprietor of Icap, the world's biggest interdealer broker, can afford such indulgences: The London Sunday Times, in its 2004 Rich List published last month, proclaimed the 49-year-old Spencer the richest person in the City, London's financial district, with a fortune estimated at £330 million ($606 million). Spencer's annual take-home pay is nothing to sneer at, either -- £4.93 million in 2003, according to Icap's annual report for the fiscal year ended March 31.
But then, the Icap CEO has been having a rather good run of it lately. In the ferociously competitive interdealer-broker world, where single-digit profit margins tend to be the norm, Icap's are 20 percent. The firm's operating profit in the fiscal year ended March 31 surged 38 percent, to £170 million, on a 21 percent spurt in revenues, which reached £801 million. The company's London-listed shares rose nearly 50 percent in the 12 months through March and have soared more than 500 percent since July 1998, when Spencer began buying up rivals in a concerted consolidation drive. Icap's £1.6 billion market cap has it knocking on the door of the FTSE 100 index.
The No. 1 interdealer broker handles about $600 billion worth of securities a day, giving it a 27 percent share of the $4.3 billion in interdealer-brokerage commissions generated annually worldwide, according to Spencer. (Rivals say that his reckoning is about right.) The 2,900-person firm maintains 23 offices in 21 countries but derives 45 percent of its revenues from Europe and the U.K. and another 45 percent from the Americas; Asia accounts for the remainder.
Icap's £800 billion in revenues is roughly double that of its closest rival, the Tullett Liberty unit of London-based Collins Stewart Tullett. Collins Stewart, however, is poised to acquire another interdealer broker, Prebon Group, a move that in one swoop will boost Collins Stewart's revenues to about £750 million. In any case, both firms are well ahead of the next-ranked interdealer brokers: the U.K.'s Tradition Financial, with £320 million in revenues, and New Yorkbased GFI Group, with £200 million. A major competitor in electronic trading, New Yorkbased interdealer broker Cantor Fitzgerald, is privately held and so doesn't disclose revenues. But in traditional voice brokerage -- a business dominated by Icap and Collins Stewart -- Cantor is a minor player.
In a fitting symbol of Icap's ascendancy -- in addition to Spencer's coronation as the latest Croesus of the City -- the firm in May left the ramshackle Edwardian town house in Finsbury Circus that had served as its headquarters for most of its 18 years and moved into an expensive-looking steel-and-glass edifice in the City's landmark Broadgate development. There Icap rubs shoulders with the European Bank for Reconstruction and Development, Société Générale and UBS.
"Icap has been an incredible growth story, and Spencer is a very big part of that success," says Jamie Hooper, director of U.K. equities at Isis Asset Management, which owns some 3.6 million Icap shares, or 0.6 percent.
Yet Icap's lovely results should not lull investors into a false sense of security. Competition among interdealer brokers is as fierce as ever. "Commissions have been under pressure for 20 years," points out Icap's chief operating officer, David Gelber, who used to be head of interest rate derivatives at HSBC Bank. "I see no reason for that to change. The best we can hope for is a slowdown in the rate of decline."
Indeed, interdealer brokerage may be the most cutthroat business in finance, which is saying something. A handful of leading firms shuffle vast quantities of bonds, derivatives and currencies around the globe between big securities firms, all of them demanding ever-more-rapid (and discreet) execution of trades, in exchange for commissions that are doled out in basis points.
So bitter is the rivalry among these specialized over-the-counter market makers that it can erupt into very public animosity. Icap and Cantor Fitzgerald are notoriously antagonistic, often confronting one another in court. In an infamous episode last summer, Cantor Fitzgerald International's president, Lee Amaitis, was asked in London's Royal Courts of Justice whether he had once called Michael Spencer "a fucking fat, greasy fuck." Amaitis, now Cantor vice chairman, pondered a moment, then pointed to Spencer, sitting in the spectators' gallery. "I do not recall calling him that," the Cantor executive replied. "Obviously, as you can see -- Mr. Spencer is sitting over there -- he is not that type of person. He is not fat."
A festering source of resentment between the two firms is a dispute that goes back to 1993, when Cantor allegedly poached an Icap broker in New York; a court eventually ruled in Icap's favor but awarded it only modest compensation. At the moment, Cantor and its electronic brokering affiliate, ESpeed, are suing Icap and its electronic subsidiary, BrokerTec, for $64 million for allegedly infringing on ESpeed's patent for its trading system. Icap dismisses the June 2003 complaint as "without merit," and ESpeed won't comment on it.
In this hypercompetitive and hypercharged atmosphere, Spencer, a brash and blustery former trader, is following his gaming instincts -- he once supplemented his income playing backgammon -- by wagering the future of Icap on the rise of electronic trading, which has long been anathema in the human-broker-dominated world of interdealer brokering. Nonetheless, he is hedging his bet by turning Icap into something of a hybrid, straddling electronic trading and traditional human, or voice, trading. The firm's influence is such that this new arrangement is bound to affect the way in which the trading of OTC securities evolves.
"I never shared the view of the techie trendies that all interdealer brokers were going to be replaced by some great dot-com in the sky, swept away by the tide of history," says Spencer. "But it is clear to me that technology will change our business over time, and to survive and prosper, we had to embrace electronic trading."
The trend toward electronic trading in some of the biggest and most-liquid interdealer OTC markets, including benchmark U.S. Treasuries, U.S. and European repurchase agreements and European government bonds, is unmistakable. Serge Marston, the director of fixed-income trading and e-commerce at Deutsche Bank in London, expects the $3 trillion credit-default-swap market to be largely electronic in the not-too-distant future. Boston-based financial technology research firm Celent Communications predicts that 60 percent of the U.S. fixed-income market will be trading electronically by 2008. Commenting on Icap, Celent analyst Sang Lee says, "If you have a presence in voice and offer electronic trading on the other side of your business, you can take a coordinated approach."
Not all interdealer brokers embrace this electronic future. Terry Smith, CEO of Collins Stewart, argues that "not all instruments are amenable to electronic trading." Benchmark, or on-the-run, U.S. Treasuries have been a huge success, particularly for ESpeed, Smith contends, because "they are pretty simple. You choose an issue and say whether you are buying or selling. What makes for a product that can be successfully traded electronically is a very large, deep market with lots of liquidity, lots of bargains and small ticket sizes."
But, he continues, "there are very large parts of the business that do not fit electronic trading," such as nonstandardized securities with a credit element, like swaps. Moreover, he adds, "there are other parts of the business that, while you could do them electronically, there are many reasons not to. Why do interest rate swaps electronically? I would guess even the biggest players, like Icap, write only 50 tickets a day." Icap's closest competitors -- Collins Stewart (counting Prebon), GFI and Tradition -- together control 40 percent of the market, and none operates an electronic trading platform of any consequence.
That leaves Icap's BrokerTec and Cantor's ESpeed to fight it out for interdealer brokerage's online future. Cantor recast its strategy after the September 11, 2001, terrorist attacks killed 658 of the firm's employees, including many brokers, at its World Trade Center headquarters. In the wake of the tragedy, Howard Lutnick, the CEO of both Cantor and ESpeed, deemphasized voice trading and placed his marker on technology.
He believes that as more and more products -- fixed income and derivatives, foreign exchange and equities -- go electronic, ESpeed's sophisticated trading tools and low-cost execution systems will come to the fore. "We have a full complement of products to roll out . . . on a platform that is already built and paid for," Lutnick told analysts in a May conference call. "The marginal cost of extending this platform is very low, [which] keeps us very much excited about our future growth."
But as Collins Stewart's Smith sees it, an all-electronic broker can't compete with a voice-and-electronic hybrid like Icap. "Their best option," he says of ESpeed, "is to use the money they generate to buy back into voice brokering to create liquidity that can be encouraged to move onto ESpeed."
Rupak Ghose, a specialty-finance analyst at Credit Suisse First Boston in London, agrees that with its two-pronged strategy, Icap can maintain its current leadership in such products as U.S. government and agency securities, European repurchase agreements and Japanese government bonds, while relying on its voice brokers to steer the most-liquid and routinely traded thin-margin instruments toward BrokerTec. "ESpeed likes to talk about the number of products it offers, but the fact is, BrokerTec is doing a better job of actually getting liquidity and market share," contends Ghose. Nevertheless, he adds that "Icap will have to become less like a broker and more like an exchange. It is well positioned to do that, but there remains significant execution risk."
Icap is already acquainted with some of the pitfalls of electronic trading. In 1999 it launched a fixed-income e-trading platform, Electronic Trading Community, that allowed clients to make trades by "hitting" prices on a screen or by calling an Icap broker. ETC attracted little interest and less liquidity, initially losing money at the rate of $30 million a year. Not easily discouraged, Spencer spent £181 million in May of last year to buy BrokerTec Global, a Jersey City, New Jersey, e-brokerage platform, from a consortium of 14 contentious banks and Wall Street firms. Negotiations took months; the remnants of ETC were folded into BrokerTec.
Barely one year later, thanks to BrokerTec, Icap is processing some $305 billion of securities each day electronically -- roughly half of all its trades. Most are Treasury bonds, repos and other straightforward, highly liquid securities. BrokerTec, brags Spencer, "has transformed Icap overnight from a modest operator in electronic brokering of fixed-income securities into the clear global leader." Spencer is convinced that adopting electronic trading will propel Icap's share of the global interdealer-broker revenue pie from today's 27 percent to 35 percent within the next few years.
Nonetheless, BrokerTec's transaction volume is far more impressive than electronic trading's contribution to Icap's bottom line. The interdealer broker's Electronic Broking Group, which BrokerTec overwhelmingly dominates, reported a pretax operating profit of £4.7 million on £62 million in revenue for the fiscal year ended March 31. (BrokerTec's share is for the ten months following its acquisition in May of last year.) In other words, the electronic group chipped in about 8 percent of Icap's sales and less than 3 percent of its profits. ESpeed, meanwhile, earned $56 million pretax last year on $157 million in revenues.
BrokerTec's impact on Icap should significantly, perhaps dramatically, improve. The unit's margins have suffered from the cost of, and the lag in, finishing infrastructure initiatives, such as cross-wiring BrokerTec and the old ETC. Once development costs have played out and systems are humming, Icap's costs for incremental electronic trades should be minimal, making for plumper margins.
It's true that commoditized liquid markets that have migrated to electronic trading, such as benchmark U.S. Treasuries, command far lower commissions than do voice markets for more esoteric securities. For instance, Icap reaps margins of 40 percent or more on voice-traded swaps. Yet the ability to trade electronically gives clients more chances to trade. One European repo client of BrokerTec does 450 trades a day, which would be physically impossible in a voice-brokered market -- no one could write all those tickets. And as markets in more and more OTC products go electronic, trading volumes should grow, expanding margins ever wider because of the minimal expense of carrying out the next trade. Thus, even though overall revenues for markets that switch largely to electronic trading may well be lower, their volumes and margins ought to be a good deal higher.
ESpeed, as an established e-brokerage platform, is a case in point -- and therefore a source of both envy and encouragement for Icap. The electronic broker's operating margin last year was 36 percent on a 21 percent surge in volume, to $31.7 trillion. Until the World Trade Center disaster, Cantor, too, was pursuing a hybrid strategy. CEO Lutnick had referred to his voice brokers as a "built-in sales force" for ESpeed. (Cantor executives declined to be interviewed for this article.)
BrokerTec and ESpeed "have different strengths," says Celent analyst Lee. "ESpeed is ahead of BrokerTec in more-liquid U.S. Treasury benchmarks, whereas BrokerTec dominates in off-the-runs. If you remember, before 9/11, Cantor's strategy was exactly what Icap and BrokerTec are doing -- voice and electronic trading through a hybrid model. That may give BrokerTec the edge in new products."
Icap's New Yorkbased U.S. chief, Steve McDermott, contends that electronic trading is fundamentally changing the whole interdealer-broker industry. He likens the change to that of a trucking business converting to a pipeline business. Trucking has many competitors and comparatively low barriers to entry, McDermott points out, while pipelines have few competitors but steep entry barriers. Or, as Spencer puts it, "If someone says they can launch a de novo electronic platform to compete with BrokerTec and ESpeed, I would say that that person is surely smoking pharmaceuticals."
The key to profitability in electronic trading, with its prohibitive initial costs, is to achieve a critical (and growing) mass of transactions. Therefore, during the protracted negotiations over BrokerTec, Icap fought for a guarantee of continued order flow from the electronic broker's owners, including Goldman, Sachs & Co. and Merrill Lynch & Co. It was "the most difficult 18 months of my life," recalls McDermott, who was also battling cancer.
Voice brokering and electronic trading can also work synergistically to provide liquidity. "There are only two ways to get liquidity onto an electronic platform," reasons McDermott. "One is to have a consortium of dealers commit to provide liquidity, and the other is to have liquidity in voice and have it move electronic." Laughs Spencer, "Every time we try to move a product electronic, all the traditional voice brokers can do is pray that it doesn't get critical mass."
CSFB's Ghose reckons that Icap is winning the liquidity war. He notes that the firm announced on March 16 that it had a 47 percent share of all trading of U.S. Treasuries among interdealer brokers, outstripping ESpeed's 42 percent. Ghose says that six months earlier, those figures would have been reversed and that three years ago ESpeed controlled 65 percent of the interdealer-broker Treasury market. Icap's May trading statement asserted that its share of Treasury trading -- both electronic and voice -- was up to 51 percent. Observers warn, however, that market share figures are inherently suspect because so much of the trading is conducted in strict secrecy.
Will Spencer's straddle strategy succeed? The roulette wheel is still spinning on that. Some of the firm's customers, though, say that systematically combining electronic and voice brokering is a smart move, because it gives both Icap and its clients more options. "Icap is a very good voice broker, and it makes nice margins in that business," points out one customer, Michael Davie, co-head of European interest rate trading at J.P. Morgan Securities in London. In a market like swaps that is "inherently more complex than Treasuries," he says, "an Icap voice broker is offering a premium service. There is a degree of value added that it is hard to see being extracted from a screen." Yet Davie also says that "if more markets do go electronic, Icap is well positioned to fight for a top seat at the table, even at the risk of cannibalizing its voice business."
Deutsche's Marston foresees a lesser but lasting role for voice trading: "There will always be a place for interdealer brokers to help facilitate access to liquidity in the market and to facilitate the movement of liquidity among the dealers. Though I am a great believer in markets moving electronic over time, there will always be parts of the market that will be voice-broked."
Collins Stewart is counting on that. The firm still conducts only a fraction of its trades electronically, and CEO Smith downplays Icap's BrokerTec strategy. "I'm sure it's been said that Tulletts is a dinosaur," he says, "but the technology is not the most important part of the equation. Icap is not massively advantaged in this regard. Certainly, we and perhaps others have the capability to move electronic."
The rivalry between the Tullett and Icap bosses -- both hard-charging, larger-than-life figures -- is the stuff of City legend, but Smith can be charitable toward Spencer: "I know Michael very well and regard him as a friend."
Spencer's attitude is typically acerbic. "If you told me Tulletts was launching a competitor to BrokerTec, I would laugh," he says. "Actually, I would be pleased, because they would piss loads of money away with no chance of getting it back."
SPENCER HAS ENJOYED THIS SORT OF ROUGH AND tumble all his life. The son of the financial secretary of the last British colonial administration of what was then called Malaya, he was born in Kuala Lumpur in 1955 and grew up there and in the Sudan. At nine he was packed off to boarding school at Worth Abbey in Sussex, England. He studied physics at University of Oxford's Corpus Christi College before launching his City career as a stockbroker at Simon & Coates (now part of J.P. Morgan Chase & Co.) in 1976. Three years later he joined Drexel Burnham Lambert but was fired in 1982 for concealing a $110,000 loss on a foreign exchange trade. From there he went to futures brokerage Charles Fulton, which he departed in 1986 to found Intercapital with £40,000 of his own and three now-long-gone partners' money. The firm was one of the first brokers in the fledgling interest-rate-swaps market, and it quickly established a reputation for adroit trading as well as for market swagger, risqué advertising and lavish parties. Intercapital once hired St. James's Square for a Wild West party, complete with a staged barroom brawl. Spencer, in his trademark red braces, began commanding more column inches in British tabloids than in the financial press.
Nonetheless, Intercapital prospered. The firm took on more and more products -- swaps became a major generator of profits -- and relentlessly grabbed market share from other interdealer brokers. But in the late 1990s, Spencer confronted a dilemma. The impending unification of Europe's major currencies into the euro would dramatically shrink the spot foreign exchange market, which accounted for about 35 percent of the firm's trading revenues. So, like the savvy trader he was, Spencer positioned Intercapital ahead of the market.
"We either had to get a lot bigger, or we had to accept niche status," he explains. "We took the gamble to go global." As Spencer saw it, "The days of a dozen subscale, undercapitalized brokers operating as partnerships were over," and those firms would be replaced by "three or four very big global players succeeding by driving volume through their business."
His consolidation play began in July 1998, when Intercapital bought London's Exco Moneybrokers for £25 million. Exco was losing money -- the purchase price was one-tenth its market value when it floated on the London Stock Exchange in 1994 -- but it had a significant U.S. presence as well as a cash securities business that complemented Intercapital's derivatives operation. "The business was shot to pieces and was absolutely ripe to be taken over," says Icap's Gelber, "but no one had attempted consolidation in our industry, so we were doing something new." As a by-product of the deal, Intercapital became a public company through the back door: It kept Exco's London Stock Exchange listing. Today Spencer owns 22 percent of Icap, and the remainder is widely held.
Spencer next bought Garban Securities, another British interdealer broker. With a market cap of £135 million, the firm was almost identical in size to Intercapital, and it was a major player in two markets in which Icap was weak -- Eurobonds and U.S. Treasuries. "We always viewed Garban as our best-run competitor," says Gelber. "Look at our business and you see Steve McDermott, who runs the U.S.; Ron Purpora, who runs U.S. securities; Gary Smith, who runs U.K. securities; Brian Massie, who runs Japan; our finance director, Jim Pettigrew -- they and many more are Garban people." He adds that this merger of equals worked only because "Michael was utterly ruthless about making sure that the right people were running the business, even if that meant parting with old Intercapital friends." Spencer sacked 800 brokers and cut costs by £75 million, or 15 percent.
"What everyone forgets today is that back then the doomsayers thought voice-based interdealer brokers would go to the wall," says Roger Barton, head of European e-commerce at Goldman Sachs in London. In fact, a number of Intercapital's rivals from the late 1990s have merged or disappeared, including Liberty Brokerage Investment Corp., Marshalls Moneybrokers and Tullett & Tokyo.
Spencer's deal making maneuvered Garban Intercapital (as the firm was known until the name was shortened to Icap in 2001) into a position to benefit from a bull market for OTC products that outlasted the bear market in stocks. Over the past four years, Icap's daily trading volume has more than tripled. Today the firm controls 40 percent of all brokered swaps.
Icap's derivatives and money market group boosted its revenues 17 percent, to £309.2 million, in the 12 months through March. Pretax profit climbed 31 percent, to £68.5 million. Revenue from securities markets, the other sizable contributor to Icap's earnings, rose 5 percent, to £365 million, and pretax profit increased 17 percent, to £66.5 million. More-modest profits came from the firm's energy, electronic brokering and information divisions.
Icap remains a creature of the markets in which it acts as a broker. Spencer began his acquisition spree when equities reigned supreme, bond issuance was in the doldrums (imagine -- government budget surpluses!) and the OTC markets were out of favor. Thus he was able to pick up bargains among hard-pressed interdealer brokers. But now fixed-income and OTC markets are flourishing.
Indeed, in the week after Icap disclosed on March 31 that its quarterly trading volume was moderately lower than it had been in the record-setting fourth quarter of 2003, its stock dropped 12 percent, to £2.70. Some observers, however, see that as an overreaction. "The secular trend is in place," says London money manager William Claxton-Smith of Insight Investment. As of June 24 the stock had rebounded slightly, to £2.76, and was 17 percent below its 52-week high of £3.31.
For how long will the markets bless Icap? Arguably, with government surpluses seemingly a thing of the past, there is no fundamental reason why the firm's business engines should go into reverse. And even if they did, Spencer, displaying his usual bravado, believes he can compensate for lower overall volume by commandeering greater market share. "The small, the inefficient, the weak would perish, and consolidation, which would happen anyway, will happen faster," he says. "If I was a competitor of Icap, I would be very, very afraid."