The capital plans of Citigroup, HSBC and several other large, foreign bank holding companies were rejected by the Federal Reserve yesterday. Most of the media reports about this significant rebuttal for Michael Corbat, Citi’s chief executive, focus on operational issues, but a more obvious explanation is that the Citi business model is more risky. Indeed, you cannot really compare Citi to either Wells Fargo, Bank of America or JPMorgan. Allow me to convey a polite “Duh” to the media and analyst community.
Looking at the most recent Fed Y-9 performance report for Citi, the first thing that should strike you is that Citi’s loss rate is 4x that of the large bank peer group. Citi’s business model is more focused on subprime than the average for the peer group. Net losses on average loans and leases were almost 2% for Citi at the end of 2013 vs just 45bp for the large bank peers. So given this fact, you can understand why the Fed is more critical of Citi’s loss rates and capital plan than with other banks.
Another clue for the media is the composition of Citi’s assets. Over 20% of Citi’s loan book are in credit cards vs just 2% on average for the large bank peers. Indeed, the only good large asset peer for Citi is HSBC, which also has a higher loss rate and more of a subprime, consumer focus than do the other 90 banks in the large bank peer group.
Larry Lindsey noted on CNBC this morning that HSBC never took a dime of bailout money from UK authorities, but the fact remains that the bank’s loss profile is far more aggressive than that of most large banks. In the extreme stress test scenario posed by the Fed, both Citi and HSBC would likely see the highest loss rates of any large US banking operation.
Another metric to give you a sense of the relative riskiness of the Citi business vs. other banks is loan commitments, which are 2x the average for the large bank peer group. Another way to think of this metric is “loss given default,” because an obligor could draw on the commitment and then file bankruptcy. Citi has historically had a much higher loss given default than other large banks, in part because of the large credit card book. There are actually four time series for loan commitments reported by banks on the form Y-9 and Citi is higher than other large banks in every category.
So when you read about Citi’s rejection by the Fed in terms of capital planning, remember that the bank’s business model is significantly different than that of BAC, JPM and especially WFC. No credible analyst would even compare Citi with these large asset peers. Indeed, going back to the Fed capital stress tests of several years ago, the best domestic peer for Citi in the large bank peer group is Capital One Financial, a business which is mostly subprime credit cards.