Citigroup (C)

Business Overview

Citigroup is the recovery case: a global universal bank shrinking itself on purpose while returns on tangible common equity still sit well below peers. Market capitalisation was roughly $229.8 billion as of 9 June 2026. FY2025 ROTCE was 7.7%, the lowest of the six banks covered here, against JPM at 20% and USB at 18.1%. That describes Citi as it stands, not the simplified bank the disposals are meant to leave behind.

Two things sit behind that gap. The first is that one-offs depress the 7.7%: it includes a $1.2 billion loss on the sale of Citi's Russian bank and a $726 million goodwill write-down on the Banamex disposal, and stripping both gives 8.8%. The second is the disposals themselves. Citi has sold about half of Banamex and is heading for an exit, and each simplification step removes capital and complexity from a bank whose problem has always been that it does too many things in too many places. If that ends in mid-teens ROTCE on a smaller franchise, Citi becomes a different bank to value. The P/TBV vs ROTCE guide uses a sub-8%-ROTCE restructuring bank and a 20%-ROTCE franchise as the two ends of its range.

Consolidated taxable-equivalent NIM was 2.47% for FY2025 (average interest-earning assets $2,426,751 million against $59,898 million of taxable-equivalent net interest income). That is mid-pack. Returns and credit drive the equity story here.

How Returns and Credit Diverge

The efficiency ratio was 64.7% for FY2025, total operating expenses over revenues net of interest expense. Since the ratio is cost divided by revenue, higher is worse, and 64.7% puts C between the 60% threshold and Wells Fargo's 66% at the bottom. It improved by nearly two points during the year and is heading towards a stated 60% target, but it still leaves C heavier than USB at 58.6% or JPM's 52% overhead ratio.

Credit is the outlier. Citi's net credit losses were $9,097 million on average loans of about $716 billion, so roughly 1.27% for FY2025 against 0.62% for the industry. Citi publishes that ratio only quarterly, and the quarterly reading had fallen to 1.18% by the fourth quarter. The level is a mix story more than a quality story: US branded cards and retail services carry loss rates several times those of a commercial book, and Citi has proportionally more of them. The credit-cycle guide uses C as the high end of the range.

Capital is stronger than returns suggest. CET1 on the Basel III standardised approach was 13.2% at 31 December 2025 against a total requirement of 11.6% (a 4.5% minimum, a 3.6% stress capital buffer and a 3.5% G-SIB surcharge, the extra capital the largest banks carry for being systemically important), headroom of +1.6 pp in the normal band per the CET1 guide. Citi's requirement is the steepest any US bank carries, a shade above JPM's 11.5%. A 13.2% ratio looks commanding beside PNC's 10.6% until you notice PNC only has to clear 7.0%.

Valuation Framework

At 7.7% ROTCE, or 8.8% underlying, Citi earns below the 10.5% cost of equity the model uses. The steady-state identity (P/TBV ≈ ROTCE ÷ cost of equity) describes a bank as it stands; for Citi the question is the bank the restructuring is meant to produce.

If simplification lifts ROTCE into the mid-teens on the same tangible book, returns clear the cost of equity with room to spare. If it does not, a charge-off ratio twice the industry's widens the shortfall in any downturn.

TBVPS of $97.06 at year-end 2025 is the anchor to watch through the disposals. Selling a business above the tangible book it carried adds to tangible book per share and selling below it subtracts, so the price achieved on each sale matters more than the headline that it happened. Track tangible equity and share count each quarter.

What to Watch in the Financials

ROTCE path from 7.7%. The recovery case lives here. Sub-8% returns sit below the cost of equity, so the case needs a credible route to low- or mid-teens returns. Management has said 10–11% for 2026, which is the first checkpoint.

Charge-offs around 1.27%. Direction matters more than the decimal. Drift towards 1.0% supports the recovery case; deterioration towards 1.5% undermines it, and it would show up in the provision a quarter or two before it showed up here.

Efficiency toward 64.7%. Restructuring has to show up as sustained revenue-over-expense improvement; one-off charges tucked into adjusted EPS do not count.

CET1 versus the 11.6% requirement. A 1.6 pp cushion sits in the normal band and is thinner than JPM's. Carrying the steepest requirement in the country means any future increase in the G-SIB surcharge or the stress buffer bites harder here than at WFC, which only has to clear 8.5%. For now the bar is fixed: the Fed has frozen stress capital buffers while it revises its stress models.

Key Risks

Restructuring execution. Divestitures, wind-downs, and organisational simplification can destroy value if sold below tangible book or if stranded costs linger. Quarterly ROTCE shows whether the plan is working.

Credit normalisation. At roughly twice the industry charge-off rate, Citi has less room than peers before losses eat the earnings. Further deterioration collides with a returns base already under 8%.

Complexity itself. A footprint across dozens of countries adds operational and regulatory surface area that a domestic bank like USB never has to manage, and Citi has been under regulatory orders to fix its risk and data controls since 2020. The cost of that work lands in the expense line, which is the same line the recovery case needs to fall.

Returns stuck near 8%. If ROTCE stays near 8% for several years, the recovery case fails regardless of CET1 headline strength.