Guess what year the following words were published: "Today, banks are under siege. … A new technology for lending … threatens the industry's very franchise."

If you guessed 2015, that's certainly reasonable, but also off by a millennial lifetime. Those words come from a 1988 book called "Breaking Up the Bank," in which Lowell L. Bryan, then a McKinsey consultant, argued that the industry should "unbundle, or break apart, the bank into its component functions," with each line of business legally separated from and unsubsidized by the others. (The technology he deemed threatening was securitization, then in its infancy.)

Today, banks are again said to be under siege and threatened by new lending technology (this time from online platforms). Many bankers are worried they'll be "unbundled" in a different way than Bryan envisioned — by startups picking off the most profitable, customer-facing parts of their business.

But history suggests that unbundling is a periodic, temporary phenomenon and that the disadvantages of a monoline business ultimately will lead to rebundling, whether it's banks acquiring nonbanks or nonbanks converting to banks. Just as biologists believe that diverse groups of crustaceans independently evolved into crablike form (a process known as carcinisation), it may be that nature wants to evolve a bank.

There are at least two arguments for rebundling as an eternal recurrence. The first is that during rough times, it's difficult for a lending business to survive without the stable funding sources available to banks, such as retail deposits guaranteed by the Federal Deposit Insurance Corp. and backstops like the Federal Reserve's discount window. The second is that surviving lenders have their own incentives to diversify and become more banklike.

"At the beginning of any credit or economic cycle, funding is available, losses are low, and you see the formation of new companies designed to hive off some part of banks' lending business," says Todd Baker, the managing principal at Broadmoor Consulting and a former banking executive and lawyer.

Since 1981, he's watched monoline credit card issuers, subprime auto lenders, mortgage bankers and storefront installment lenders come and go.

The nonbank model worked fine until "credit caught up with you significantly," Baker says. Then, "all of a sudden you had large losses," and liquidity would dry up. But, "ideally you sold out well before that."

A rare exception to the typical scenario is Capital One, a former monoline card lender spun off from Signet Bank 20 years ago. After the 1997 Asian and 1998 Russian debt crises, Capital One realized funding could be a major weakness and remade itself into a retail bank. The transformative 2005 purchase of Hibernia angered Capital One shareholders at the time, since nonbank lending during good times is a higher-return business, says Baker. But the move would prove prescient several years later.

"Not one of the major finance companies made it through the 2007 and 2008 crisis," Baker says. MBNA and Household were sold to banks beforehand. Countrywide had to be rescued by Bank of America during the downturn. CIT Group, American Express, GE Capital and Ally Financial (which is the former GMAC) submitted to bank holding company regulation to gain access to bank liquidity. Countless mortgage shops simply failed.

Baker expects a similar fate for marketplace lenders and other online platforms that have garnered so much buzz and investment capital in the last two years. The institutional sources of money they rely on to finance their loans will flee when loan performance worsens or regulators crack down, he predicts. Even those companies that don't hold loans on the balance sheet will be unable to originate and sell loans at gains fat enough to cover their fixed costs, he contends. The lucky ones will be acquired by banks, which will benefit from their innovative, millennial-standard technology.

One thing might be different this time: The more unusual a marketplace lender's niche, the harder it will be to find a bank buyer and to get risk-averse regulators to approve a deal. "Finance companies were bought by banks because regulators were willing to let them do that," Baker says. "They were willing to let HSBC own Household. Would they do that today? My concern is this will not be an easy task."

Marketplace lenders beg to differ on their long-term prospects, of course. Mike Cagney, chief executive of Social Finance, argues that its platforms provide real-time feedback on loan buyer demand, allowing them to adjust rates rather than get caught holding underwater loans like the finance companies of the past. Yet even Cagney foresees companies like his becoming more banklike, albeit under happier circumstances.

SoFi, which offers student, mortgage and personal loans for affluent millennials, is looking at adding a deposit product, which could be FDIC-insured or may have private insurance. The goal is not to provide a source of funding for SoFi's loans, which would continue to be financed on its market platform, but rather to meet its customers' every financial need.

"I do think you're going to see folks like us and others go after that side of the balance sheet, some for funding, but others to offer a comprehensive solution set so you never have to think about going to Wells Fargo," Cagney says.

A startup can offer the same products and services as a bank, but is not hamstrung by disparate legacy systems that don't talk to each other, leading to absurd situations like depositors having to reenter all their information when they apply for a mortgage from the same bank, he says.

Similarly, Jared Hecht, CEO of the online small-business loan platform Fundera, who gave a talk on unbundling last year at (where else?) the South by Southwest technology conference, nevertheless sees a business imperative for rebundling.

Lending products are "inherently not sticky," Hecht says. "It's difficult to instill a sense of loyalty in the customer." Hence startups are "beginning to think more holistically" about how to serve customers. "You don't just want to be a loan, a blip of a transaction over the course of someone's life."

David Brear, who holds the enviable title of chief thinker at Think Different Group, a consulting firm in London, wonders if marketplace lenders, known for their fast turnaround, can maintain the same quality of service as they branch out into other areas.

"Banks didn't start in the way they are today, incredibly complex with 500 different products," Brear says. "They started by doing something well, then moved on to the next thing. I like to joke that the grass was well cut, but then when they were looking at other things, it got overgrown."

In 10 to 15 years, hot U.K. startups like the short-term lender Wonga, the online investment management service Nutmeg and the mobile-first Starling Bank might face the same challenges banks do today, Brear says. "Would I start a bank from scratch like HSBC today? It's unlikely," he says. "But I can see why they evolved the way they have."

So can Bryan, who retired from McKinsey in 2012. He says the cycle of unbundling and rebundling is driven by regulation at least as much as by market forces. Nonbanks prosper in areas where banks cannot compete, often because regulations prevent the banks from doing so.

"Regulation is one of those facts you have to deal with, just like in nature you have to deal with the fact that there's a mountain range there or a desert," says Bryan, who now runs his own advisory firm and serves on corporate boards. And when regulations change, businesses must adapt to the new environment to survive.

Given the "overregulated" state of banking today, "I'd expect we're going to see an awful lot of innovation and new business models created outside the big banks, until they get big enough to create problems and regulation changes," Bryan says. "Then they'll be acquired by the big banks."

So a word of caution to would-be disruptors: Banks as we know them today may not be pretty, but then neither are crabs.

Subscribe Now

Access to authoritative analysis and perspective and our data-driven report series.

14-Day Free Trial

No credit card required. Complete access to articles, breaking news and industry data.