Death of a Business Model: NetBank

Includes: C, FBC, ING
by: Christopher Whalen

And they give you cash, which is just as good as money. (Yogi Berra, "Berra at the Barber")

Updated -- News that Citigroup (NYSE:C) will see earnings tank in Q3 may not be welcome news to the Buy Side, but please don't be surprised. Look for similar good news from other Sell Side dealer banks with exposure to derivatives and asset conduits. As the effects of the collapse of the secondary market in structured assets rolls through the system, many organizations which have grown dependent on non-interest earnings will need to rebalance their skewed business models.

While the Buy Side adjusts what we see as still toppy earnings expectations for financials this year and next, this week we ponder yet another "surprise," namely the demise of the online banking business model. On Friday, the Office of Thrift Supervision closed NetBank, Inc. (Nasdaq: NTBK) and appointed the FDIC as receiver, which immediately announced a purchase and assumption of some of the thrift's $2.5 billion in total assets and $2.3 billion in total deposits by ING Direct, the wholly owned subsidiary of giant insurer ING Groep NV (NYSE:ING).

After the FDIC was appointed receiver, the publicly-listed parent of the federal savings bank immediately filed bankruptcy. Of note, the bank's depositors above the insured limit are being haircut at 50% pending the disposition of $1.1 billion in assets retained by the FDIC. The loss to the bank insurance fund is so far estimated at $110 million. In announcing the transactions, the FDIC noted that "the failed bank was an Internet bank and did not have any physical branches." This fact, however, only partly explains the failure of NetBank and, with it, the online banking model as an independent business.

Back in 1989, NetBank was known as the Allatoona Federal Savings Bank, a $15 million asset savings bank. By the end of 1997, what was by then known as the Atlanta Internet Bank had grown to $83 million in assets, but was not consistently profitable. At the end 1998 NetBank was born and had reached almost $400 million in assets, the majority of which were deployed in loans, but loans with a gross lending return of just 479bps, three standard deviations below peer.

Within just a year, assets had tripled to $1.2 billion, but the bank still was just barely profitable and was spending 86 cents to generate a new dollar of revenue. Some 63 percent of assets were in loans, but the gross yield at just 495bps was still well-below peer, a red flag in any event but an especially bad metric for a bank with such a high efficiency ratio. By year-end 2001, NetBank had reached $2.8 billion in assets, supported by over $1 billion in Federal Home Loan Bank advances.

NetBank's assets grew over the next five years to a peak just over $5 billion in mid-2005, when the bank's ROA was 0.07% and ROE 0.91%, actually one of the bank's better annual performances. By then the bank's efficiency ratio was 148%, meaning the bank was spending $1.48 for every $1 in new revenue. This bank was dying, not from bad loans but from an irrational growth strategy, supported by $1.2 billion in FHLB advances at the end of 2005.

By the end of last year, NetBank's efficiency ratio was 227%, but assets had dropped to $3.6 billion. The bank reported losses equal to 4% of assets and 49% of equity. The losses continued to mount into 2007, so that the bank had just 2% equity to assets at the end of June 2007.

To us, the remarkable thing about NetBank is that investors and bank managers continue to believe, even today, that the Internet banking model is viable as a stand alone business. Looking through the peers of NetBank in the Internet category, it is difficult to find an institution that is even close to asset peer levels in terms of asset and equity returns. Brings to mind the comment by author Martin Mayer that "there are no economies of scale in banking."

Branches and people remain an important part of the business model for American bankers, for all institutions large and small. Remember that scale does not seem to be an issue with Internet banks. In fact, the largest player in the Internet bank category, ING Direct, which assumed the insured deposits of NetBank, continues to turn in truly awful results. With an ROA of 0.33% and an ROE of 4.13%, ING Bank FSB has been consistently below peer for the past five years.

The gross loan yield of 447bps as of June 2007 illustrates that ING Bank, at 1.5 SDs below peer, is not able to demand the kind of loan pricing that makes the high expenses of the Internet model truly worth while. And ING's thrift unit has an excellent efficiency ratio of just 55%, but even those frugal, kind-hearted Dutch, who eschew credit cards and predatory loans, cannot make a bank profitable with most of its assets tied up in MBS and mortgage loans.

Last year, when we suggested to our colleagues at the American Banker that ING Bank was one of the worst performing thrifts in the US and that the bank could be under supervisory guidance were it not for the large, well-capitalized parent, a representative for ING replied that the bank's equity returns were not so bad because only half of its capital was funded with equity at the parent level, the rest apparently with debt.

That comment still reminds us of the classic AFLAC television ad where Yogi Berra expounds on the difference between cash and money. Or to paraphrase the white duck: "Ahhh!"

One of the better performers in the Online banking segment, EverBank in Jacksonville, FL, managed to report ROA of 0.93% and ROE of 13% as of June 2007, albeit with an efficiency ratio of 86%. Some 75% of the $4.1 billion asset bank is deployed in real estate loans that boast a gross yield of 563bps. The bank's loan default experience is thankfully low, but profitability seems to hinge on the bank's non-interest income, which has consistently exceeded net interest income. Of interest, EverBank will purchase about $700 million of mortgages held by NetBank as part of the FDIC's resolution.

Another example is Flagstar Bank (NASDAQ:FBC), a $16 billion asset online banking operation based in Troy, MI. The bank's ROA peaked at over 3% in the middle of 2003, but has since fallen steadily to its current rank of nearly a full SD below peer in terms of asset returns.

At June 2007, Flagstar reported ROA of 0.34% and ROE of 5.45%. Run rate defaults at the end of Q2 were 19bp on the loan portfolio comprised largely of real estate loans and which boasted a gross yield of just 597bp. The bank's 95% efficiency ratio and heavy dependence on net-interest revenue for earnings is not reason for great comfort and joy.

The moral of the story is that no matter how fast a bank grows assets, if the business model is not sufficiently broad and consistent, then stable profits will be elusive. This rule applies to all banks, large and small. The fact that many online banks seem to have a very hard time achieving profitability suggests to us, at least, that this business model may not be viable -- except as a loss-leader within a much larger organization. Look at the proliferation of online banking arms of the major banks and it seems reasonable to ask how a free-standing Internet bank will survive in the long-run.

For investors, the example of the closure of NetBank illustrates a more profound lesson, namely that when the capital of a federally insured bank falls below a certain level, the primary regulator can appoint the FDIC as receiver, leaving the publicly listed parent holding company no alternative to bankruptcy.

While the larger banks in the US like C still may be "too big to fail," banks the size of NetBank apparently are not -- as that unfortunate institution's shareholders and uninsured depositors discovered this past Friday. But it is interesting to note that while the bank's jumbo retail depositors are getting haircuts, the FDIC apparently repaid all of the bank's brokered deposits in full, regardless of size.

(Author's note: Our colleagues at the FDIC tell The IRA that none of the brokered deposits repaid in the NetBank resolution were above the insured limit!)