It seems disjointed to see American Express Co. cut its contributions to its workers’ pension plans on the one hand and offer credit cards to Advanta Corp.’s rejected clients on the other. Nearly, every measure of American Express’s financial performance has worsened in recent months, yet the New York charge card company is going to take on clients from among the worst performing credit card companies in the nation? There is something wrong with this picture.
Officially, American Express OPEN “reached an agreement with Advanta to extend an invite to select Advanta customers to apply for a new small business card from American Express OPEN.” As some of you will recall, Advanta essentially went belly up, a victim of charge offs climbing more than 200% on a year-over-year basis. Advanta told its 1 million or so customers that it was firing them by cancelling their cards.
Ask American Express what is going on and you’ll get a detail-less response about a “marketing agreement” with Advanta. When I asked American Express why it is “choosing to take on these credit risks that Advanta deemed to be too great” and “how does this effort comport with American Express’s increasing credit writeoffs and state expectation that credit losses will continue to increase at AXP,” I got this response:
We are targeting select Advanta small business customers who fit the profile of our target customers. These business owners tend to be what we call “transactors.” They tend to pay off their entire balance each month, and if they revolve a balance, it tends to be a small amount.
Like everything in financial services, it boils down to risk, and indeed American Express is making a bet on the risks it will likely take on through this “marketing agreement.” When you look at American Express and Advanta, it is interesting to note that while American Express generates nearly three times the amount of revenue per account, Advanta’s provisions for credit losses per account was 46% lower than American Express’s in the first quarter of 2008. Now, this may be a reflection of conservative provisioning – or it may not. Suffice it to say, American Express will most likely make money from this marketing agreement. How much is a whole other story.
Reading between the lines of American Express’s public statements on this deal, Amex is only offering cards to ex-customers of Advanta that were current. Advanta’s charge-off rate was more than 20%, so the maximum number of accounts American Express could be pitching is around 800,000 companies. We presumed about 20% of those Advanta customers (160,000 accounts) would take Amex up on its offer. Using a revenue-per-account metric that was average of American Express’s and Advanta’s past performance, it appears as though these Advanta accounts will generate about $40 million of revenue for Amex. On that revenue, Amex will eventually have to provision around $13 million for credit losses, in the likely scenario. After compensating Advanta for the accounts (we figure Amex will pay around $20 per account, or $3.2 million), American Express should clear around $24 million in net income.
But that’s in the likely scenario. In the less likely, worst-case scenario, the Advanta accounts yield Advanta’s revenue per account and American Express’s provisions per account. In that case, American Express will lose a relatively nominal $500,000 or so. Not big numbers, for sure, but when you are cutting contributions to employee 401K accounts, it’s a number worth mentioning.
There’s a broader dynamic at play here that deserves mentioning. As Jamie Dimon, the CEO of JPMorgan Chase & Co., said last week during his company’s earnings call, all the regulatory initiatives underway “will at least reduce profitability we believe in the short run, meaning the rest of this year and mostly next year,” for credit card finance. And amid this environment, when the credit card business itself is suffering, and when American Express, in particular, is weathering subpar performance, is now the best time to be taking on Advanta’s baggage? I’ll let you decide.