I think Discover is trading at an attractive absolute and relative valuation. DFS generates a ROE above 20% but is trading for less than 10x earnings, with all of those earnings being returned to shareholders. Discover is a high quality business that deserves to trade a multiple closer to the general market than to banks that often struggle to obtain sufficient profitability (as measured by sub 10% ROE). Between modest loan volume growth, dividends, share repurchases, and a potential multiple re-rating, DFS shareholders are likely to be rewarded with an attractive IRR over a long holding period.
Discover Financial originated at Sears in 1985 as part of Dean Witter. Dean Witter merged with Morgan Stanley in 1997, with Discover spun off as an independently traded company in 2007. DFS is a bank that holds the loans of Discover-branded credit cards, and a smaller balance of personal loans (Direct Banking segment). DFS also owns the Discover Network that processes transactions for Discover-branded cards, the PULSE debit card network, and the Diners Club international network (Payment Services segment).
Discover makes almost all of its money from the domestic credit card portion of its Direct Banking segment, with payment services generating just $0.1B of DFS’s $3.6B (3%) in 2015 EBT. While the payment networks potentially have some strategic/option value (especially if volumes can be increased), I do not include any additional value other than the small current income stream when thinking about DFS valuation.
Discover is the smallest of the 4 major U.S. credit cards/networks, with the lowest amount of cards in circulation (9%) and spending volume, leading to a 4% of market share by purchase volume. Visa and MasterCard have 82% of market share by cards in circulation and 70% by purchase volume. American express has 25% of market share by volume with only 9.1% of cards in circulation due to much higher spending customers
Visa and MasterCard don’t own the loans generated by their cards, but just collect the high margin fees from the use of their networks. These companies are rightly recognized as great businesses, with outstanding 15%+ ROAs, growth profiles, and minimal cyclicality. They deserve to trade at the mid to high 20x earnings multiples they are valued at.
While American Express and Discover are not truly outstanding businesses like Visa and MasterCard, they are still high quality businesses that in my opinion deserve to be valued at higher multiples than what they currently trade at. AXP used to consistently trade at 15x+ earnings until it recently fell out of favor given its issues of losing the Costco account. DFS on the other hand, has consistently traded at an 8-12x P/E multiple. This is probably in part due to the disadvantages DFS has compared to its competitors: V/MA has better network scale, AXP has a higher quality customer base, and banks have lower funding costs. Despite these disadvantages, DFS has compounded earnings and EPS at 19% and 20% for the last 8 years. This is due to the attractive economics of credit card lending and excellent management/capital allocation at DFS.
The high interest rates charged for credit cards create much higher than average net interest margins. Even after accounting for the higher loan losses and lower leverage, credit card lending is one of the highest ROA activities in banking. American banks typically earn a 0.9-1.0 ROA generating a 9-10% ROE. DFS typically earns a 2.5% ROA and 20% ROE, 2.5x and 2.0x higher, respectively.
But despite superior profitability and return of cash to shareholders, DFS is valued at or below normal bank multiples of 10-12x P/E. This is not a balance sheet issue as DFS (like most banks nowadays) has a solid financial position and performed well on the recent FED stress tests. DFS also has higher growth opportunities than normal banks who struggle to put their deposits to work.
Earnings have been down the last couple of years due to increased benefits and operating costs in a competitive environment. That headwind should be easing going forward, with Discover improving to number 1 in customer satisfaction surveys ahead of AXP. U.S. merchant acceptance has also increased a lot of the last several years, with Discover now reaching 9.3m merchants, almost on par with V/MA with 9.5m merchants. DFS’s acceptance is well above AXP’s 6.9m merchants, which is lower due to AXP’s higher merchant fees. With acceptance almost on par with V/MA, market share growth could increase if more customers switched cards for Discover’s excellent cardholder benefits (although I am not counting on it). I think DFS should be able to increase loans ~3-4% annually helped by the tailwind of increased credit market share at the expense of cash.
DFS is trading at a low multiple in part because it is seen as “over-earning” given the currently low levels of net charge offs. NCOs are correlated with unemployment levels which will go up in a recession. A lot of banks are similarly trading cheaply due to worries about a recession in the near future. While NCOs are currently low at 2.2%, compared to a normalized level of ~4%, DFS should still be able to compound value from here given a long enough holding period. Partially offsetting reduced income from increased NCOs will be increased “other income” from fees such as late fees as well as increased lending rates on remaining balances, as well as reductions in operating costs. Also offsetting increased allowance for loan losses will be continued growth in loan balances and the reduced share count from repurchases.
A positive attribute of DFS’s high ROA is that it has strong earnings power that can be used to increase capital levels if needed. Although DFS’s stock price would certainly decline in an economic downturn, with its strong earnings power, funding position, and balance sheet, DFS’s liquidity/solvency should not be at risk. Even in a severe recession with 9% charge-offs, DFS should remain profitable (although at a much reduced level, like the 2008-2009 period).
DFS’s low stock price combined with its propensity to buy back stock allows for double digit EPS growth even with modest loan growth. For example, in a scenario of unchanging NCO levels and interest rates, 4% annual loan growth could translate into 5% earnings growth given leverage to fixed operating costs. After funding the dividend and equity for loan growth, DFS would be able to purchase 6.6% of outstanding shares at $55/sh, which would translate to 12.5% in EPS growth (plus a 2% dividend). At today’s low prices and high level of buybacks, 5-10% EPS growth can continue even with meaningful increases in NCOs to more normalized levels over a couple years.
Even later in the cycle, a 12x P/E on 2017 earnings is not unreasonable for a 20%+ ROE stock. This would result in a $73 stock price, or a 40% increase in a year. In a downside scenario, DFS could trade down to 1.4x TBV or $35/share for 33% downside, with the price recovering as earnings recover in subsequent years. In the previous recession, DFS traded well below book value, but I would not expect that to happen again, as investors now have a track record of DFS’s profitability through the cycle. In a future recession I would expect DFS to trade more in-line with AXP’s multiple.
I do not hold a position with the issuer such as employment, directorship, or consultancy. I and/or others I advise hold a material investment in the issuer's securities.
Rising rates will help net interest margins
DFS has never gotten credit for its high ROE potential and I don't know a catalyst that will change that. DFS could still handily beat the market for many years without a re-rating Ipif like tobacco stocks, DFS continues to buyback it's stock at low prices while modestly growing earnings.