Capital One not the ONE (Fool On the Hill) September 2, 1999

An Investment Opinion

Capital One not the ONE

By Dale Wettlaufer (TMF Ralegh)
September 2, 1999

It's tough to beat the combination of really good companies with really volatile stock prices. One that has fit the bill over the last couple years is Capital One Financial (NYSE: COF). Having crashed in 1997 after the company reported that customers were taking down their revolving debt and generating fewer highly profitable over-limit and late fees, the former Signet Bank unit went on a rampage and shot up just about 500% in the next two years.

When it did bottom out in 1997, it was cheap and widely misunderstood. Few on the sell-side had anything more enthusiastic than an outperform on it, many analysts had holds on it, and its mass customization, information-based strategy wasn't as widely appreciated as it is today. Further, mailbox-spamming competitors such as Advanta, which has since sold its credit card business, had blown up within the preceding year and bank-takeover enthusiasm had died out by the summer of 1997. At its nadir, Capital One was trading in the $10 range (split-adjusted), or almost one-quarter of where it's trading today at $37 1/8.

So here we are two years later. Consumers have been reliquified by lower mortgage rates and delinquencies, defaults, and bankruptcies have all been on the wane. Being a nation with a marginal propensity to spend most of its extra dollar of income, Americans also went out and charged up more stuff on their cards. Industry net charge-offs (gross charge-offs less recoveries) topped more than 24 months ago and installment debt as a percentage of disposable personal income has also risen only slightly since 1996 (according to Deutsche Bank Alex. Brown, Credit Card Quarterly, July 1999). Due to increase in lending to sub-prime credits, I would argue that this datapoint is actually skewed and that revolving credit to personal income for the standard customer has probably decreased somewhat.

Capital One, meanwhile, has worked itself into the number seven position in card lending and continues to show growth and profitability. Its super-efficient data management strategy has allowed the company to customize its products on the fly, which the credit card industry business model allows you to do, since the product can differ for each customer that comes along and since the product can be re-priced theoretically on a daily basis.

Capital One's data mining capabilities allow it to detect erratic consumer behavior and get out ahead of those customers that look like they will overextend themselves. Data mining also allows a credit card company to avoid yield-hunting credit card borrowers, who take advantage of low teaser-rates and then move on. If someone has 10 credit card accounts and carries a big balance on one of them, the company will notice that and probably let you roll over a balance onto one of its cards at a low rate. It also allows a company to predict which customers are more likely to respond to offers that will generate higher amounts of fee income from the use of their card.

All of this combined makes Capital One the GEICO or Dell of its industry -- it's able to construct a business model that takes advantage of all the inherent attractive qualities of the industry while other companies stumble on the inherent disadvantages of the industry. In other words, Capital One has a formula to take market share in established markets, avoid market share in established markets when that's not a value-creating proposition, and exploit new markets such as sub-prime revolving credit.

Recently, however, the entire credit card sector has come down, partly as a result of interest rate fears, which can impact revenues, credit quality, asset growth, and thus net income and net income growth. Of course, increased interest rates also can effect market-wide equity valuations. More importantly of late, however, Bank One (NYSE: ONE) pre-announced that it has totally blown it in its credit card business and that its profits for that unit will fall $480 million to $530 million short of expectations for the second half of 1999.

The reasons given for this include management's opinion that competition in credit cards has heated up. Most analysts were skeptical of this, however, and many chalked it up to lack of execution on Bank One's part. Higher cost of capital from securitization (and probably higher prepayment rates effecting securitization assumptions and thus gains on sale and excess income from trusts), deteriorating fundamentals in the prime credit card market, and lack of highly differentiated marketing (including product pricing and composition) all play a part in these company-specific factors that knocked down Bank One about 23% on August 25.

Capital One, meanwhile, appears to be exactly on track. Its risk-adjusted revenues as a percentage of managed receivables have gone from middle-of-the-pack performance to class-of-industry in only a few years. Risk-adjusted revenues = total revenues less managed net charge-offs:

               1996     1997      1998    Q1 1999

MBNA:          7.57%    7.48%     6.96%     7.11%
Metris:        20.4     14.3      11.2      14.1
Providian:     8.65     9.78      13.2      17.1
Discover:      7.64     8.71      9.31      8.55
Capital One:   8.85     9.37      12.9      16.1

Average:       10.6%    9.93%     10.7%     12.6%

Raw data according to Deutsche Bank Alex. Brown
Return on equity (ROE) has been maintained in the mid-20% range with a much better return on assets than most competitors and less leverage than competitors use. Growth has eclipsed the competition, as well. From fiscal 1996 to 1998, revenues grew at compounded yearly rate of 39%, net income compounded at a 33% rate, and EPS compounded at a 31% rate. For the first half of 1999, revenues grew 61% and EPS grew 25%.

There's part of the hang-up right now for investors in Capital One. The company thinks of itself as an information-based marketer and not solely a credit card company. As such, it has been reinvesting excess profits in its America One program, under which it resells to retail customers wireless airtime that it buys wholesale. Due to increased competition in wireless telecom, the company's customer acquisition costs and lifetime value of customer are not where they should be. While Capital One will continue to invest in programs under which it can market products leveraging its mass customization and service platforms, it has committed to pulling back from the commodity wireless segment and will go more for niche segments in this market.

It still projects it will lose another $57 million in the second half on America One, however. Using CIBC World Markets' Steve Eisman's (who is The Man in this sector, in my opinion) analysis of the situation, Capital One will earn $1.70 in 1999 with these losses factored in. Compared to its more innovative peers, the company is priced at a discount:
                        Mean EPS estimates
                   1999...P/E         2000...P/E
Capital One       $1.72...21.8       $2.15...17.4
Providian          3.57...22.8        4.87...16.7
Metris             2.01...13.5        2.51...10.8
This gives little value to the option to shut down the program, which Capital One could exercise at any time, at a very small strike per Eisman's analysis that the company would not have to take a write-off since it's expensing everything up front on the America One program. Without America One in earnings, Eisman estimates 1999 net income at $2.24, putting Capital One at 16.7 times that estimate and further reducing the forward multiple into the sub-15 times earnings range.

At a current earnings yield of 5.7% on the 2000 estimate, a potential earnings yield approaching 7%, a potential ROE of above 30%, and a potential return on assets (ROA) above 5%, the company would be undervalued relative to almost any larger financial services company and certainly any financial services company executing on its level. Put it this way, this is more like a Fifth/Third Bancorp (Nasdaq: FITB) type of performer selling at a mega-regional bank rollup type of valuation.

The catalyst to make this stock move higher in the short term is further guidance from the company that it would further scale back its America One initiative, come up with a new market to penetrate, or give further detail to its plans to attack subsegments of the wireless market. While this company has always played it pretty close to the vest on the specifics of its strategies, it knows how to communicate pretty clearly with investors on the broad outlines of its intentions.

The sooner it reaches break-even or profitability in America One or the sooner it exercises its option to shut off that business, the sooner the stock will move up. A 6% earnings yield on a fiscal 2000 estimate of 30% EPS growth off Eisman's core $2.24 1999 EPS estimate would yield a target of 16.7 times year 2000 EPS of $2.91, or a $48.50. As Bank One has put market psychology clearly on the defensive, due to worries that it will come into the subprime market and create competitive problems, one could wait for a more advantageous buypoint in the $32 range where 50% upside to intrinsic value is implied by the above. 30% upside to implied intrinsic value from here is available, however. Finally, off the intrinsic value, the company can compound shareholder value at 20% per year for five years.

Key points:

1. Capital One is the GEICO or Dell of its industry. It can continue to outperform flailing competitors.
2. Capital One is not just a credit card company. Using its extensive databases and proprietary methods of dissecting consumer behavior and mass customizing products, it can innovate in established markets and excel in new ones.
3. Management is not irrational. If America One is not a value-creation opportunity, it will deploy resources elsewhere.
4. Its core credit card business is the class of the industry. It's more efficient than competitors and shows superior risk-adjusted returns on shareholder capital. Its competitive advantages allow it to take market share.
5. The company is undervalued and will be reappraised higher by the market when management updates investors on its America One plans. The very real option to shut that down is currently being valued at zero, which is a mispricing by the market.