So much for the idea that small is beautiful. If you thought the advent of the Internet, the spread of cheap and efficient information technology, and the growing fragmentation of the consumer market were all going to help smaller companies thrive at the expense of the slow-moving giants of the Fortune 500, apparently you were wrong. In the wake of the three giant mergers--Citicorp-Travelers, NationsBank-BankAmerica, and Banc One-First Chicago--in the banking industry over the last two weeks, the conventional wisdom has quickly adjusted. Bigger, it seems, is now definitely better.

The banking mergers follow a year that saw $1 trillion in merger and acquisition activity in the United States alone. It's easy to overestimate the significance of that number, which is inflated by the high price of most U.S. stocks. In fact, over the last two decades, M&A activity has stayed relatively constant as a percentage of the value of the stock market as a whole. What's really important is not the size of the M&A boom but the fact that merge-bent companies are forging ahead undaunted despite all the evidence that the vast majority of acquisitions do not add value to the economy as a whole. Don't bother them with the facts. They're ready to buy.

What we're witnessing is a wonderful exercise in two-track thinking by the people running U.S. corporations and by the shareholders investing in them. On the one hand, there's a general rejection of the old Fordist, vertically integrated corporation. From Sara Lee divesting itself of its manufacturing operations to the Big Three auto companies outsourcing their parts production to Marriott running hotels that other people actually own, the idea of doing as little real work as you have to is much applauded by Wall Street today. On the other hand, though, there's also a general acceptance of the corporation that promises to do everything, such as Time Warner and Viacom in media or the new Citigroup, with its financial supermarket, in banking. For a short time--I think it was when IBM started laying off tens of thousands of workers--it seemed that the ideas of economies of scale and scope had been permanently discredited. Apparently not.

In and of itself, this is no bad thing. Economies of scale--which mean that a product becomes cheaper to produce the more of it you make--do exist in industries such as automobile and semiconductor production. And economies of scope--which mean that it's cheaper to produce or distribute many different products rather than just one--obviously exist in supermarkets and Wal-Marts, as well as arguably in a company such as GE. If a supermarket just sells milk, it will probably be less successful than if it sells milk, eggs, cereal, bread, and so on.

Saying that size sometimes matters, though, is not the same as saying it always matters. And even if size does have its benefits, a merger may not be the best or lowest-cost way to reap them. Look at the arguments in favor of the Citigroup merger. In the first place, Citigroup will now be the "largest financial institution in the world," which will make it able to acquire other companies and make expensive investments. But Citicorp and Travelers were already huge. Perhaps Citicorp couldn't afford that slamming new computer system before it was part of Citigroup, and now it'll be able to shop with impunity, but somehow I doubt it.

Citigroup will also be more "diversified," which means that its earnings every year won't depend as much on volatile sectors, as Travelers' did on the trading of Salomon Bros. and Citicorp's did on income from Asian markets. But if shareholders in Citicorp wanted to diversify, all they had to do was buy shares in Travelers, and vice versa. They're not any more protected now than they would have been then.

In essence, after all, what the Citigroup merger involves is Travelers shareholders trading half their shares--actually it's a little more than half, since Travelers officially acquired Citicorp--in Travelers for shares in Citicorp. The idea that that swap created value in itself is nonsensical. The deal is justified only if the merged entity will grow faster, and be more profitable, than the two separate ones would have independently. And so we come to synergy.

In the Citigroup case, the synergy is supposed to come from the two companies selling each other's products. Now you'll be able to get property insurance from Travelers when you take out a mortgage from Citibank, or your Travelers agent will be able to badger you into getting a Citibank credit card after he's done badgering you about life insurance. Citigroup will be your one-stop-shopping place, your financial supermarket.

As nearly everyone has already pointed out, the record of financial supermarkets is dismal. Sears failed with Dean Witter, Citicorp failed in the late 1980s when it tried to expand operations, and Sandy Weill--head of Travelers--failed miserably when Shearson joined with American Express. And the concept of one-stop shopping does seem oddly ill-suited to financial services. It's not just that, at least in the United States, there's no evidence that people prefer national banks to local ones (if anything, there's a long history of distrust of large banks). It's also that people take their financial decisions more seriously than they do their choice of toilet paper, and that they are unlikely to pick up a home mortgage just because their insurance agent is hawking it. At a time when people comparison-shop for almost everything, Citigroup's privileging of convenience seems oddly outdated.

Even if cross-selling makes sense, though, it's not clear why Citicorp and Travelers had to merge. Citicorp could have agreed to market Travelers insurance or Salomon brokerage services, and vice versa, without actually becoming one company. For the deal to make sense, the costs of working together in a contractual relationship would have to be greater than the costs of merging. But the record in this regard is not comforting. In fact, if there's one thing that is unequivocally true about M&A activity, it's that companies dramatically underestimate how much it will cost and how long it will take to make two companies--with their attendant managerial hierarchies, corporate cultures, and computer systems--into one.

That's hardly surprising when one considers how little time companies take before agreeing to merge. The Citigroup deal, from beginning to end, took less than five weeks. The $60 billion NationsBank-BankAmerica deal took three weeks, and the companies did "due diligence" in three days. How due, exactly, could that diligence have been? It's true that if a company lets a potential merger candidate look too closely, and the deal then falls through, it may have given away its secrets for nothing. But when you realize that people take more time to decide what car they're going to buy than Weill and Citicorp's John Reed took to decide on a $70 billion deal, a little uneasiness might be in order.

The truth is that everything--from the way investment banks are compensated to the rubber-stamp role of boards of directors to the dominance of publicity-hungry CEOs--is set up to make mergers attractive. And it's hard work to determine the success of mergers, because you have to compare the performance of the new company against the projected performance of the previously independent companies. That makes it easier just to assume success in the absence of complete disaster. But the chairman of a large Midwestern bank put it best when he recently said, "You get big because you're better. You don't get better because you're big."