Discover Financial Services (DFSWI), the credit-card spinout from Morgan Stanley (MS) set to be completed at the end of the month, began trading on a when-issued basis last week. Here’s my early, back-of-the-envelope take:
After I read the company’s initial registration filing when it came out in March. I came up with an earnings estimate for 2007 of between $1.70 and $1.75 per share. Assuming a 14 multiple, that implies a stock value of around $25, compared to the $31 it’s trading at now, when-issued. But Morgan Stanley/Discover put out an 8K Thursday that provided more details and history on Discover’s managed-basis earnings.
Even if more optimistic numbers are included (that is, credit costs stay lower than I’ve assumed, share count is lower than it initially seemed, and management is aggressive in using excess capital to buy back stock), the current share price is still about 14 or 15 times out-year numbers, hardly a bargain. But this is not the easiest company to value: there aren’t many similar companies to compare it too, it’s in the midst of lowering the credit risk profile of its loan portfolio, and it has an oddly configured business model.
So while Discover’s current $31 when-issued price might seem high, it may not be unreasonable. Take a look at some other metrics, and you’ll see my point:
On a sum-of-parts basis, a 10-to-12 P/E multiple on the card business, and well over 20 times ''normalized'' third-party network earnings doesn’t seem out of bounds, and gets you to $30. But remember that the card business includes a good portion of all the company’s merchant acquiring and network fees. On a comparable basis, the fairest comparison is to the only other closed-loop issuer, American Express, which trades at just over 16 times 2008 earnings—but with a much superior franchise. On a portfolio basis, the stock is trading at a 20% premium to receivables, which not out of line for prime card portfolios. Discover tends to have stickier accounts than its competitors, but also tepid new account growth. Also, this implies little value to the company’s network business.
While I don't think the when-issued price is unreasonable, none of the above makes the stock particularly attractive here. Among ''other'' conclusions, this would seem to bode well for the Visa IPO, due in late 2007 or early 2008.
There are a few sources of upside to my prior estimate of $1.70 to $1.75 for 2007, which could get the numbers closer to $2.00 per share for 2008. Despite lackluster receiveables growth and ongoing credit “normalization.” How?
First, Discover's initial filing used the Morgan Stanley share count (adjusted for a 2-for-1 split) for pro forma estimates. Morgan subsequently disclosed that about 8% of those shares are management-owned restricted stock units that will not be counted at an independent Discover. That math is easy! This adds 8% to my estimates. Second, Discover will be overcapitalized when the spin occurs. Tangible equity to managed assets will be around 9%, so the company could buy back 10% to 15% of its stock over the next 18 months. (For perspective, Capital One has long said it’s comfortable at a TETMA ratio of 7%, and that was before it did all its bank deals). A 15% share buyback would add 5% to 8% to EPS, given the lower average share count over that period. Third, the credit environment remains relatively benign. I'm not incorporating this (and neither is Discover’s guidance), but it’s no big secret that Discover has decided to trade growth for better credit quality over the past few years following their disastrous subprime foray. Net chargeoffs are now running at 4.0% in the U.S., not the lowest in the industry, but certainly at the lower end. Until recently, Discover’s NCOs typically ran at the high end of non-subprime issuers’. My estimates had assumed a 50-basis-point rise by the end of the year, and another 50 to 100bp in 2008. (As a point of reference, contractual charge-offs remain low.) In the most recent quarter, NCOs rose by 19bp sequentially, to 4.0% from 3.8%, but delinquencies fell by 34bp, to 2.97% from 3.31%. This is likely a mix of seasonality and shift to bankruptcy-related losses, as bankruptcy filings continue to trend up.
One other point: Discover is actually run relatively efficiently, both from an operating and marketing standpoint. They previously disclosed that they actually expect net savings as a public company, due to over-allocation of corporate overhead for Morgan Stanley. I would not anticipate heavy back-end savings there in the near-term.
Still An Attractive Takeout Target
However, the real source of earnings upside and value will be to a potential acquirer that wants a closed-loop system. While Discover is still the ''Avis'' network, it is still a rarity, and would give one Charlotte-based bank leverage against the other associations, and another Charlotte-based bank some assets and operating/marketing scale. I rate Bank of America (BAC) the most likely acquirer, given that it has recently indicated it has an explicit interest in starting or acquiring its own network(s) for debit and/or credit. Discover has both. Because of this, I would be surprised if Discover is still independent in two years.
Despite the non-compelling valuation at present, this is still worth watching. Looking back to the past few spin-offs in financial services, they have typically traded down at least 15% post-spin, as legacy holders re-assess their “new” position.