|
Several years ago, Canadian Imperial Bank of Commerce (CIBC) decided to expand into the US securities business to meet the needs of its Canadian clients, and to grow internationally. But becoming a big player in US government securities wasn’t the first thing on its mind. Nonetheless, the bank hired an 18-member team of government securities traders from SG Warburg, received authorization to be a primary dealer in US government debt and started running huge matched books of repos and reverse repos. Although hardly a profitable business, running such a matched book is a game that many commercial banks play in the US. Absurd as it may seem, it has allowed them to participate in securities business, prohibited by the Glass-Steagall Act of 1933. The matched books exist to help the banks pass the “90-10 test”, to comply with the Federal Reserve Board’s ruling that banks’ securities subsidiaries may earn no more than 10% of their gross revenues from securities underwriting that the banks themselves are not allowed to do. The revenues from the repo books count towards the 90% target for eligible revenues to set against the 10% the subsidiaries can earn from such “ineligible” activities as underwriting commercial paper, municipal revenue bonds, corporate debt and equity. The banks’ first push into securities came in 1980, when Bankers Trust persuaded the Fed to let it underwrite commercial paper. But a giant step towards the demise of Glass-Steagall was taken in 1987 when JP Morgan discovered a loophole in the act’s Section 20, which states that banks cannot be affiliated with a business “principally engaged” in the kinds of securities underwriting that the banks are excluded from. The debate turned to the definition of “principally engaged”. The Fed agreed that banks could set up securities subsidiaries as long as such ineligible activities contributed only 5% of the subsidiary’s gross revenues. Later that figure was doubled to 10%. Since the banks had long been permitted to deal in government and municipal securities, many quickly moved these operations out into the subsidiaries. The gross interest income earned from those operations was considered eligible revenue. Today there are 30 of these “Section 20” subsidiaries in existence. As the ineligible activities grew, so did the need for eligible ones. “If you had $250 million in ineligible revenues, you would need to find $2.25 billion in eligible revenues. And that is no small feat,” says Michael Rulle, managing director and head of the US region for CIBC Wood Gundy Securities Corp, the bank’s capital markets affiliate. The simplest way is through matched books of government securities repos and reverses. Soon the game will be over, as the Fed is expected to raise the proportion of permitted ineligible revenues to 25%. No institution could be happier than CIBC. “Banks have been screaming for relief,” says Rulle. “The rules have been limiting our ability to compete.” Rulle says the Fed, which is liberalizing the rules in a way that will virtually dismantle Glass-Steagall, has finally realized it’s “just dumb” to force banks to engage in such activities as building up huge government books just to meet a legal technicality. Through a matched repo book, with interest rates at 5%, Rulle estimates, to earn $2.25 billion in eligible revenues, one would have to find a matched book of close to $50 billion. “And there’s not many $50 billion matched books out there.” Under the 25% limit, he estimates, the same $250 million in ineligible revenues would force the bank only to have a matched book of about $15 billion, which would theoretically earn $750 million. Much to the bankers’ chagrin, however, Glass-Steagall will remain on the books. The legislation, passed in reaction to the US banks’ role in the speculative fervour leading to the 1929 stock market crash, has been amazingly resistant to change. Banks have been fighting it for over two decades, but a variety of forces have kept the act alive. First it was the banks’ own lending debacles of the 1980s, in Latin America and then in real estate, compounded by the costly government bail-out of the savings and loans industry, that made legislators wary. In recent years, the strength of the US insurance lobby, which wanted to keep the commercial banks off its turf, thwarted financial reform. Yet Glass-Steagall has clearly become an anachronism in an age of global finance, where few other countries’ banks face such restrictions, and an increasingly competitive domestic environment in which many traditional bank products have become less profitable. JP Morgan, for example, has argued in a position paper that the limitations imposed on the banks essentially instruct them to be unsuccessful “a singularly un-American idea”. The act has been whittled away by product innovation, bank lawyers and the Fed’s actions. Dreaming of mergers At the end of October the Fed also knocked down three of the firewalls separating the banks’ commercial and investment banking operations. From January, US banks will be able to cross-market bank and securities products, sell securities to other bank affiliates and streamline their management, instead of being forced to staff their securities subsidiaries with non-bank employees. Last month, Comptroller of the Currency Eugene Ludwig also announced a new rule to allow banks to offer more financial products, including securities. This would affect those US banks not structured in the bank holding companies that fall under the Fed’s jurisdiction. Though Ludwig was short on specifics, there has been some speculation that he intends to go much further than the Fed, raising the limit as high as 49%. Bankers say the changes will free them of most of the hoops they have had to jump through to compete in the securities business, allow their equities, high-yield and corporate debt business to grow substantially, and eventually result in the long dreamed-of bank-broker mergers. “What it might mean is that banks will no longer be content to just grow the business, but will want to acquire it,” says Donald Layton, vice chairman of Chase Manhattan Bank. Although many banks, like Chase, have opted for an internal growth strategy “for integration and culture reasons,” he says. “This now puts the alternative back on the radar screen for us to think about.” For the immediate future, says Layton, Chase will be building up its equities business substantially. According to senior bankers, the 10% cap has meant that banks have had to focus on two of the three most attractive “ineligible” areas high-yield bonds, investment-grade corporates and equities. For example, JP Morgan has concentrated on the latter two, Bankers Trust has targeted high-yield and equity, and Chase has opted for the two debt categories. In the near future, these banks are likely to be scurrying into whichever area they’ve so far ignored. Although the New York City-based banks have been the leaders in the effort to get rid of Glass-Steagall, many bankers and analysts argue that regional and foreign banks could benefit most from the changes. The restrictions on banks and their subsidiaries sharing directors have prevented some regional banks with smaller staffs from having Section 20 subsidiaries. Moreover, it is the regionals and foreign institutions that haven’t generated sufficient growth internally that are widely viewed as most likely to start the acquisition bandwagon rolling. For example Hugh McColl, chairman of acquisition-minded Charlotte-based NationsBank, told a recent analysts’ meeting that the bank hopes to buy a non-New York City-based brokerage by 1998. Earlier this year, Swiss Bank Corporation (SBC) chief executive officer Marcel Ospel said the Swiss bank was also on the prowl in the US, looking for “smaller-scale acquisitions aimed at filling niches”. The 10% limit has been a problem for virtually all the banks. “It’s something we have to manage,” says Rulle. “People are struggling to maintain their 10% cap,” adds Robert Slaymaker, chairman of BA Securities, the securities subsidiary of Bank of America. “It’s an inefficient use of the balance sheet and therefore of capital.” These low-earning assets also diminish the banks’ return on assets and their net interest margins, both of which can depress stock prices. Slaymaker contends that those in the business longest have the biggest problem as both their eligible income and their matched books have grown the most. The Fed tracks the ineligible revenues on an eight-quarter moving average, based on banks’ quarterly reports. But the banks monitor the percentages daily. “You look at trading profits, and the deal pipeline, because if a couple of big deals hit, all of a sudden the ratios go way up,” says Michael Capetides, CIBC Wood Gundy’s general counsel. The success of the high-yield team CIBC hired last year forced it to beef up its matched books. “You look at it every day and make a decision every two weeks or every month to take the matched book to the next level. It’s not an exact science, but it’s crucial.” Once a bank gets behind, he notes, “you’re in a lot of trouble, and running very large books gets to be expensive. You not only have to handle your current ineligible revenue totals, but you’re going back and making up for past time.” The problem is compounded by the fact that the ineligible businesses earn more than the eligible ones, and it is revenues, not assets, that are being calculated. “If you have $10 worth of government securities on the books earning 5%, and $1 worth of high-yield assets earning 12%, the 90-10 test doesn’t work,” explains Capetides. The Fed’s firewalls also prevent the securities subsidiaries funding their activities through the bank. This is why the matched books are so popular. Essentially, they are a self-funding mechanism for buying government securities. Here’s how it works. When a bank buys a security, it funds the purchase by selling the security with an obligation to buy it back the next day a repurchase agreement or repo. Then the bank buys the security back (a reverse repo) and sells it again. Capetides explains that true matched books start with a bank “reversing them in and sending them right out again”. Capetides estimates that hundreds of billions of dollars in matched books exist solely to meet the 90-10 test. Wall Street firms have increased their trading desks to accommodate the banks. But as banks’ need for the matched books grows, their already marginal profitability declines, and can even turn negative. “At a certain size, it becomes difficult to find counterparties, and it becomes more expensive,” he notes. Probably around the $10 billion range, he estimates, “you no longer make money on them”. At such a point, for every $100 in US securities a bank repos out, he calculates, it might only get back $98. But net profit is not important for the purposes of meeting the 90-10 test. The calculations are based on gross revenues in this case the interest earned on the securities that, for accounting purposes, are still owned by the bank. Banks have also tried to move fee-generating businesses into their Section 20 subsidiaries to meet the 90-10 test, although they are reluctant to divulge what they’ve done. “It’s a little proprietary,” says a senior banker at a major New York bank. Competitors speculate that Chase, a leader in loan syndications, has booked that activity, with its substantial fee income, in the Section 20. Other businesses deemed eligible include project finance, securities brokerage, private placements and M&A advisory. But some banks have got into trouble with the Fed. Bankers say the Fed would not allow one institution to put its substantial mortgage banking activities in the subsidiary. And earlier this year, SBC became the first bank to be fined by the Fed for going above the limit. Bankers believe the Fed deemed SBC was including inappropriate businesses in the eligible category. While bankers agree that the liberalization of the limit is clearly the most important of the pending changes, the firewall relaxations will also make life easier and less costly. Of the three, the one many bankers say will have most impact is allowing bank employees to market the services of the securities units. “We rely on our relationship managers, our commercial bankers in the field, as the primary contact point with our clients,” says Sandy Curlett, president of First Union Capital Corp, the Section 20 subsidiary of First Union, a regional bank headquartered in Charlotte. “Having had a firewall which prevented those bankers from actively making clients aware of the capabilities of our broker-dealer is like being told we can use our hands, but we’re not allowed to move our arms.” In practice, Curlett says the restriction involved “extraordinary amounts of training”. A banker had to be taught he could “talk to a client about the private placement market, but he was only able to say ‘We have a broker-dealer that can do that’. He couldn’t say ‘Here’s what they’ve done’.” Ending prohibition Curlett believes this restriction was probably more onerous for a regional bank like First Union, spread across 13 states and focused on middle-market companies, than for the New York-based banks, like JP Morgan, which focus on Fortune 500 companies and long ago turned many of their bankers into investment bankers in the Section 20 subsidiary. But Rachel Robbins, general counsel for JP Morgan, says the bank was also hurt by this firewall. “It impeded the natural flow of business… You had to have different people covering the same client, or else change the structure and move people out of the bank so they were in the holding company or some other subsidiary.” The latter wasn’t an option in its international offices, she says. “There we often had two people talking to the same clients. It was clearly not efficient.” Moreover, she notes, even in the US the bankers in the Section 20 subsidiary couldn’t make credit decisions, “so we had to involve more people”. The prohibition against interlocking directors is also being eased. In the past, directors, officers and employees of a bank could not work in its securities affiliate, and vice versa. Under the new rule, employees can work in both organizations and some directors can serve jointly as long as they don’t comprise a majority of either organization’s board. A shared chief executive officer is still prohibited. Some bankers are relieved by this relaxation, saying they wanted to put more senior people from the bank on the boards of the Section 20s. Staffing the Section 20s with bank holding company employees, or just moving individuals out of the bank and into the securities subsidiary, created other problems. “Given the convergence of products, it often makes sense to make the trader who trades loans for the banks also trade for the Section 20 subsidiary,” says Robbins. Morgan will not need to make changes at the top levels of the subsidiary. But traders in emerging market loans will be able to trade emerging market securities and be officers in both units. But the benefits of this relaxation, according to Bank of America’s Slaymaker, will be constrained by legal factors. “A bank employee cannot pitch some securities business without being registered by the National Association of Securities Dealers, and you cannot be registered by the NASD as an employee of the bank. The advantage of being able to cross-pollinate the two is limited by securities law.” One banker said the main problem for his institution has been that all bank holding company employees were also employed by the bank. “When we set up the Section 20 subsidiary, the board had to be non-employees of the bank. We didn’t have anyone like that.” And, although in theory the subsidiary was operated separately, “in reality functional controls exist. Policies are levied down from the holding company level, but it made for very impractical working relationships between the subsidiary and the bank.” “The interlock firewall,” adds First Union’s Curlett, “has been a problem in the sense that we’ve got managerial talent who know the institution well who really couldn’t manage in the bank and in the 20 at the same time.” The third firewall change will let banks purchase a wider range of securities from their Section 20 affiliates for investment or to accommodate clients. Bankers concur this is the least important of the changes. But banks with significant asset management operations will no longer have to place trades externally, keeping commissions in-house and helping to preserve confidentiality on some trades. First Union, whose recent purchase of Keystone Investments made it the third largest US bank manager of mutual funds, believes this change will be of most use in expanding its retail business. Now most of its Section 20’s distribution is institutional. “And anything that can be done to broaden the distribution” will be helpful, says Curlett. In the past, even customers who wanted to do something as simple as buy a mortgage-backed security could not get their bank to buy it from the Section 20. “Even though we were making markets in that, they’d have to buy it from somebody else,” adds Bill Maxwell, chairman of NationsBank’s Section 20 subsidiary. A NationsBank employee would have to call an independent broker, who would then call the Section 20 desk. “It made no sense at all.” But the greatest impact of the impending changes will be the bank-broker mergers that most analysts, bankers and M&A specialists feel are inevitable, if not imminent. “For every 90 pounds of gossip, there’s an ounce of reality,” cautions Chase’s Layton. Although it might have been possible for big US banks to buy boutiques prior to the limit change, no one really knew for sure. Now, he says, “you’re more credible if you’re talking to someone. They can see running room.” A 25% limit will make boutiques fairly easy to swallow. That immediately puts the hot technology firms, such as Hambrecht & Quist, Robertson Stephens, Cowen and Montgomery Securities, on the alert. Analysts and bankers also reckon that even some of the largest securities firms could be purchased, at least technically. “Now bigger banks have the opportunity to buy bigger firms. It ratchets up the whole thing,” says CIBC’s Rulle. Since neither securities firms nor banks publicly break down their revenues into eligible and ineligible categories, there’s still some guesstimation involved. But most brokerages have huge eligible revenues from asset management, government securities trading and M&A. Independent brokerage analyst Perrin Long notes that interest income is a vast majority of many investment banks’ revenues. At Morgan Stanley, for example, 64% of revenues is interest income. Of the rest investment banking may total 25%. “Then you take out M&A and merchant banking revenues that they have, one can’t help but to think that all your major firms could be acquired, even Merrill Lynch,” he says. To make a hypothetical case for acquisition, bank analyst Raphael Soifer, of Brown Brothers Harriman, takes a more conservative approach. He assumes that 40% of a securities firm’s gross revenues derive from ineligible activities. “If the firm’s gross revenues are, say, $10 billion on total assets of $100 billion, a bank acquirer would have to supply an additional $6 billion in bank eligible revenues (plus the $6 billion in such revenues which the firm already has) to create $16 billion in gross revenues to meet the 25% test.” Under the 25% rule, the acquirer would need $100 billion in matched-book assets, which he claims is “not out of the question”. Under the 10% limitation, matched-book assets would have to total more than $500 billion. Even with the limit, he argues that the largest securities firms still pose a “significant challenge” to the banks. Merrill Lynch’s annualized gross revenues for the first half of 1996 were $24.4 billion. However, at Morgan Stanley they were only $13.4 billion, $13.7 billion at Lehman Brothers, $9.6 billion for Salomon Brothers and a mere $5.5 billion at Donaldson Lufkin Jenrette (DLJ). European dimension Some small firms with little eligible revenues might prove even more difficult. Analyst Long points to Alex Brown, which has been eyed by many bank suitors in the past and might be a good target for a regional bank because its focus on initiation public offerings (IPOs) dovetails neatly into a retail strategy. Alex Brown has very little government securities business, and a huge corporate bond trading and underwriting operation. Moreover, since it is an IPO house, its stock is already priced at about twice book value, leaving little room for a deal in today’s market. Soifer believes that German, Swiss and Dutch banks, which can live with a 10% return on their investment, are in the best position to buy a top player. US and UK banks, whose stockholders require closer to a 17% return, would find it much more difficult. “A continental European bank could afford to pay two times book if the target firm has a return on equity of 20%.” Today, he says, Lehman Brothers is trading at 0.9 times book, but only earns around 10%, while DLJ is earning a little over 20% and trading around 1.5 times book value. And, although small, DLJ is a top firm in equity, merchant banking and high-yield securities. DLJ is the best buy, he argues, for an institution that wants to make significant inroads as a global investment bank. “DLJ is the most attractive candidate. You’re getting a number four firm for less than a market capitalization of $2 billion.” The caveat is that the bank would have to factor in the cost of employee guarantees, which might leave little for investors. DLJ, previously owned by the Equitable Companies, has been on the block before, with both US and foreign banks looking at it. Through its 60% interest in Equitable, French insurance company Axa tried but failed to sell DLJ. Instead, notes Soifer, it sold a portion of the shares to employees in an IPO. But Axa still owns 80% of DLJ. Moreover, employees may be more interested in selling now that they are stockholders. (DLJ did not return calls for comment.) But there is still a big hitch to a bank purchase of DLJ and many other investment banks that has nothing to do with Glass-Steagall. Banks organized in a holding company structure are prohibited by the Bank Holding Company Act from owning more than a 5% stake in industrial companies. Any buyer of DLJ would have to divest or vastly restructure its substantial, and very profitable, merchant banking holdings. Such institutions as Bankers Trust and JP Morgan already structure their merchant banking portfolios with convertible or other non-voting securities to comply with the law. Banks most often mentioned as potential buyers include such continental Europeans as SBC, Dresdner Bank and ABN Amro – institutions with global ambitions whose US presence trails Deutsche Bank and Union Bank of Switzerland (though neither of the latter two are ruled out). Another candidate is Bayerische Vereinsbank, whose attempt to buy Oppenheimer was scuttled by Glass-Steagall late last year. Of the big US banks, Soifer thinks Bank of America and Chase are the most likely acquirors. Reportedly, Chase has considered Salomon, DLJ, Merrill and Lehman. Bank of America is said to have looked at both DLJ and Robertson Stephens, but declined to comment. Citibank is considered less likely, since its unhappy purchase of Vickers da Costa has left a bad taste. JP Morgan is viewed as culturally opposed to buying, although chairman Sandy Warner has told analysts that, just because it hasn’t bought anything “doesn’t mean we’re not looking”. Bankers Trust, which could use a brokerage acquisition to boost its business and reputation, is probably too small to buy a big firm. Regional banks with acquisitive tastes, apart from NationsBank and First Union, include PNC, Norwest and Banc One. But they would be more likely to buy smaller, regional retail-oriented brokers like AG Edwards in St Louis or Raymond James in Florida. Numerous obstacles remain. The first, and most often stated, is that bankers seem reluctant to buy the highly cyclical stocks at the heights they appear to be near in today’s bull market. Moreover, argues Soifer, a successful deal would have to be friendly to keep employees and clients from bolting. And none of the brokers is much interested in selling, given their current health, say bankers. “We’re in a point in the cycle,” says NationsBank’s Maxwell, “where investment houses are doing very well, which makes it a difficult time to acquire someone. I think people will be very patient.” NationsBank’s McColl told analysts that even though he wanted to buy a non-New York-based brokerage, “no one is for sale”. Banks might be best advised to wait for a market turn. “If we saw the Dow Jones Industrial Average at 4,000 … that would create a good buying opportunity,” speculates Soifer. That would be more than a 33% drop in the Dow, similar to the crashes of 1987 and 1974. The caveat to his scenario is an “irrational buyer who comes in and says I don’t care what the price is,” says Soifer. “Once a big deal happens, and a big firm gets bought, that sets in motion a dynamic of scarcity value.” This is the argument used by M&A specialists to try to convince bankers to be first, which is fraught with its own complexities. Primarily, bankers worry about what would happen in Washington if a big bank-broker merger was proposed. Such a purchase may be technically possible under the new rules, but Glass-Steagall does still exist. “It’s an issue for bank CEOs to be the first out of the box,” says a senior executive whose bank has looked at several brokers. “Most [of them] realize that if a significant bank attempted to buy a significant securities house, the deal might end up in the political domain in Washington.” The headaches would only be starting. What would get the feeding frenzy under way? He says that for a big US bank, a regional bank buying a regional broker would not cause a stir. “A foreign bank with nothing in the US buying Lehman… maybe.” It would certainly make banks think if JP Morgan and Morgan Stanley were to merge, he speculates. The conventional wisdom has been that building investment banking business is easier than buying it. “That view has been borne out by looking at some historical examples of acquisitions. They have an uneven track record of success,” says CIBC’s Rulle. But he’s not betting on that view persisting. “These things sometimes go in cycles, and I would not take past behaviour as a necessary predication of future behaviour.” Already the Europeans have been challenging that view by buying British merchant banks. “The only reason it hasn’t happened here is Glass-Steagall.” CIBC’s position on the subject no doubt mirrors what other bankers are thinking: “We haven’t made a decision one way or the other,” says Rulle. “But we haven’t ruled anything out. |
