Behind the juggernauts of American Express (NYSE:AXP), Visa (NYSE:V) and MasterCard (NYSE:MA), but well above the big U.S. banks, lies a Chicago-based company that has achieved some remarkable progress since its IPO in 2007.
That company is Discover Financial Services (NYSE:DFS), which has no shortage of bullish SA contributors supporting it. Born within Sears (NASDAQ:SHLD) in 1985 (yeeks!), adopted by the Dean Witter family, and kicked out into the streets by Morgan Stanley amid the horrors of the financial crisis, DFS's corporate history reads like a hard-knock childhood.
This article is for long-view investors, and will mostly serve to support DFS's current valuation, but should leave a few bullish nuggets. Skip over the credit-focused, stress-test gibber-jabber mid-way down if it makes you queasy.
First, let's look at DFS's ROE performance relative to AXP, Capital One (NYSE:COF), and V and MA:
Source: Company disclosures
Spending most of my time on core banks, I sort of got watery eyed when I saw that DISCOVER had strung together FIVE years of 20% and higher ROEs. God Bless America - DFS is an achiever!
DFS, AXP and COF are fundamentally different companies than V and MA, because V and MA don't take credit risk, but are rather interchange fee toll collectors. V and MA rely on balance sheet strength far less because of these points (ROE is not even a fair comparison with MA in the picture - P/E would probably be a better indicator). The main point is that it's remarkable that both DFS and AXP (both banks) have cleared the 20% ROE hurdle for (at least) five consecutive years. It's well understood that few, if any, traditional banks can come close to that feat today.
If we believe that ROEs are important to driving Price/Book multiples (and we do), DFS is enjoying significantly better valuations relative to traditional large banks because of its high ROEs. The graph below shows DFS below the line (i.e. has a Price/Book discount), but not by much. AXP, DFS and COF are green dots. Mega banks are the red dots, and everyone else is blue. Every one of the banks below is part of the Fed's stress test program. In summary, the graph might point to some room for DFS's valuation to expand in terms of P/Book, but don't get too excited.
Data source: YCharts
Second, let's look at loan growth since the crisis. It would seem the market is discounting the company's growth, instead focusing on AXP's very high portfolio credit quality relative to DFS. DFS's book is only modestly behind AXP in measures such as charge-off rates and non-performers.
DFS has grown loans well excess of AXP, and at about the same rate of COF. There will be hard-nosed credit folks who look at rapid growth in credit as a problem, but in this case, my takeaway is that DFS is doing solid execution, solid branding and the company's emphasis on customer service is working. If the growth rates shown in the graph below were in commercial real estate, I'd be worried, but the behaviors of card borrowers tend to be more predictable.
Third, let's consider the performance of DFS in the Dodd-Frank Act Stress Tests (DFAST). The chart below shows Common Equity Tier 1 (CET1) before and after runs of the DFAST in each of the past three tests. The graph avoids explaining what is happening using the CET1 ratio, and instead deals in hard dollars ($ billions), which aims to provide a more intuitive feel for money available for return to shareholders.
The graph needs some explaining - so bear with!
The total size of the bars represents CET1 (in $) at the beginning of the DFAST cycle. For DFS at the beginning of the 2016 test, that number was $9.9 billion. Under a nine-quarter DFAST, some of that capital would typically be expected to dissolve in credit write-offs and other market stresses. And there will be some capital remaining, which has to be sufficient to meet the CCAR minimum thresholds. The amount that is left above the minimum is technically available for distribution over the quarters within the nine-quarter test period.
In the graph, "Minimum" (in blue) is the minimum capital needed to be in place at the end of the DFAST stress cycle. It is simply calculated as the end of DFAST cycle risk-weighted assets x 4.5%.
"Depletion" (in red) for most banks would be the CET1 capital in place at the beginning of the cycle that burned up over the nine-quarter test. For DFS, no CET1 burns up, but rather a small amount, $300 million, is actually created (Annoyingly, because of Excel graphs are what they are, that simply shows up below the zero line representing the x-axis).
"Cushion/distributable" (in green) is the CET1 capital in place at the beginning of the cycle, and was not depleted over the test. It provides management with essentially the total capacity of what could be distributed through dividends and buybacks.
The bold percentages are the amount of "depletion" divided by the total beginning of test period CET1.
Focus eyes below on the tiny "-$0.3 billion" in "depletion" of CET1. Repeating, the negative value of "depletion" means that the company actually came out of the 2016 DFAST with $300 million MORE in CET1 than when it began the test. Only a handful of 33 banks tested finish the test with more CET1 than at the beginning of the nine-quarter test cycle.
Source: Fed DFAST reports
The punch line is that DFAST not only doesn't dent any of DFS's equity, but it also leaves the company with $6.5 billion of distributable capital, plus $300 million extra, while still meeting the $3.7 billion it would need to hold a CET1 ratio of 4.5% - the minimum required to pass CCAR.
The graph above does not mean that DFS won't incur credit losses in its income statement. Rather, it says that the NET effect of pre-provision net revenue (PPNR) and cards' credit losses, combined, makes DFS relatively much more immune to a credit downturn than other large banks.
The same points above can be said of AXP, which also cleared the past three DFAST tests with more CET1 at the end of the test versus the beginning of the test. The same graph of AXP above (if shown) has a -3% where DFS has a -4%. In other words, AXP also survived DFAST 2016 not losing CET1, but instead, generating fresh CET1 through the test.
Now bear in mind that credit card portfolios are imputed with high losses in DFAST. The average loss rate of cards for DFS over the past three DFASTs was 14.0%. In the 2016 DFAST, DFS incurred card losses at a rate of 12.8% of the portfolio, which amounted to 82% of the overall credit losses the company would face under 2016 DFAST.
While loss rates in the portfolio held steady with the 2015 DFAST (12.7%), it was much better than the 2014 DFAST, when the loss rate was 16.4%. Directionally, this could point to either improved risk management or higher-quality card borrowers relative to 2014. Also possible, the test itself has become easier on cards since 2014.
I'm no mouthpiece for DFS, but there is something about credit cards in general that seems to make this loan type relatively impervious to DFAST, and by association, the Comprehensive Capital Assessment and Review (CCAR). Essentially, what is happening in DFAST is that credit tanks, but the PPNR hangs in and offsets the losses! Together, DFAST and CCAR amount to moving targets for banks, and hurdles that I believe are contributing meaningfully to the large bank sector's valuation discounts. More on that in a later post.
Fourth, let's consider the factor of customer loyalty. The J.D. Power survey data below is from DFS's own presentations, but we'll take the survey as a fair assessment of how well customers are liking the card services.
To be beating AXP in customer satisfaction is remarkable feat for a company that gets a fraction of the attention of its larger and more brand aware card competitors (AXP, Chase (NYSE:JPM), COF, Citi (NYSE:C)).
Finally, some commentary on credit ratings. DFS's Moody's rating is coming in at an inexplicable three notches below AXP's A3 rating (DFS's holding company rating, Ba1, is a whopping FOUR notches below AXP's rating). S&P's issuer rating on DFS comes in one notch higher than Moody's at BBB, and Fitch at BBB+.
Regarding the Moody's rating, there is little in my mind that makes DFS a below-investment-grade company.
Moody's lower rating is at least partially explained by the company's concentration in cards and the legacy of the crisis. DFS saw loss rates on its card portfolio broach the 20% mark in the crisis. Moody's disses on "monoline" business models and tends to weight diversification in revenue streams higher than the other agencies. So even though DFS's specialty in cards is tremendously profitable and superbly well diversified with loyal customers, a glowing J.D. Power survey is not something on which Moody's is going to put a lot of credence. But the proof is in the pudding - the profitability of the company is very good and the DFAST results are substantive.
Another strike against DFS in the agencies' minds is the company's modest dependence on wholesale funding - specifically structured finance, or credit card asset backed securities (ABS). DFS funds itself about 20-25% in that market. Access to the structured finance markets at the levels of DFS is more of a blessing than a curse in my view, yet the fundamental analysis side of the rating agencies won't view it that way.
How does DFS mitigate the risk that the structured finance markets shut down? By holding loads of cash. DFS carries a remarkable $33/share of cash, over $10 billion at 3/31/16 (AXP also carries a boatload of cash). Based on the company's annual funding demands, my sense is that's enough to carry it easily through the scary shutdown of the credit card ABS market.
DFS Funding Structure:
AXP Funding Structure:
Not shown in the AXP breakout is that, as of 3/31/2016, $24.4 billion of its total $55.0b in deposit funding was brokered deposits, which credit guys don't like. AXP's brokered portion of total deposits is a higher fraction (43%) than DFS. Overall, AXP's access to the unsecured term market, accounting for 33% of its funding, is largely explained by the company's higher ratings. DFS almost certainly would have better access to the unsecured market were it not for the Moody's Ba1 rating. In my estimation, DFS is well deserving of ratings higher than the Moody's ratings.
Wrapping up, the card business is highly competitive, takes a ton of branding spend and customer sweeteners spend, and can suffer from severe losses in a recession. But payment technologies are expanding the revenue pie (loan growth is proving that) and the demand for credit seems to be there even when the economy is tanking. Regulation, fines and technology risks are real issues across each of the firms mentioned. But let's face it, the Tina Fey commercials are annoying. DFS is a card company for folks who want AXP-like services, but maybe not the AXP attitude. DFS is not a card in my wallet, but for this just occasional studier of the card space, it's quite impressive what the firm has accomplished.
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.