Viewpoint: Internet-Only Banking Is Down But Shouldn't Be Counted Out

A few short years ago, banking futurists were predicting that Internet banking would completely displace brick-and-mortar banking.

But the number of branches in the United States has not decreased, and the typical Internet-only operation is struggling for profitability. Most large banks are integrating their Internet-only subsidiaries back into the traditional bank, and many pure-play Internet banks are establishing physical branches.

I disagree with the current conventional wisdom that Internet-only banking is not a viable strategy. Though most banks will eventually offer a hybrid, the Internet-only model could still prove viable.

I do not make this statement lightly, because it stands somewhat at odds with the results of recent research in which I compared the financial performance of six pure-play Internet banks and thrifts chartered between 1997 and 2000 against "traditional" banks and thrifts chartered at the same time. (Preliminary tests on a larger set of pure-play banks and thrifts have generated similar results.)

Pure-play Internet banks had lower-than-average physical overhead, and their assets grew faster than average. But on the downside, profits were significantly lower than at the traditional banks.

A number of cost and revenue factors contributed to these lagging profits. High salaries at the pure-play banks more than offset the savings on physical overhead; at the risk of oversimplifying, computer programmers earn more than bank tellers. In addition, both noninterest income and deposit funding were relatively low at the pure-play banks - evidence that the Internet channel is a poor vehicle for developing core customer relationships and cross-selling financial services.

But past returns do not necessarily reflect future performance. The future for pure-play Internet banking could be rosier than the past.

• First, demographics are on the side of Internet banking. We all have (usually younger) friends or co-workers who swear by it.
• Second, click-and-mortar banks share some of the Internet-only banks' weaknesses. Richard Sullivan of the Kansas City Federal Reserve Bank recently analyzed banks that operate both branches and transactional Web sites. These banks incurred significantly higher noninterest expenditures, and relied significantly more on purchased funds, than did non-Internet banks. This implies that online delivery has certain performance problems, even at banks that complement Internet delivery with branches. As banks conduct a greater percentage of their transactions online, the playing field between pure-play banks and click-and-mortar banks may level.
• Third, online banking is relatively new. Not only must banks master a constantly changing menu of new technologies, systems, and platforms, but they must learn by experience how these innovations affect customer demand, product quality, and interbank rivalry.

Customers are also learning how these new delivery channels work and what combination of branch transactions, kiosk transactions, and home computer transactions suits them best.As banks and their customers progress, online banking will gradually become more efficient and convenient. Given that pure-play customers are the most intensive users of the online banking channel, Internet-only banks are likely to benefit disproportionately from these gains.
In general, economists are wary of making predictions. The one I make here is clearly out of step with the current industry wisdom. But several years ago, when industry experts were writing off brick-and-mortar banking for dead, a contrarian prediction would have proved quite prescient.

Mr. DeYoung is a senior economist at the Federal Reserve bank of Chicago.