Banks 1Q/25
+ Citi / Ally 1Q update
~15 min read. Couple things first:
This email might be truncated. Click the “Banks 1Q/25” title to open on the Substack site. (I also sometimes edit things; the site always shows the current version.)
For new readers who subscribed for stock ideas: there’s a bunch on Ally and Citi below. Both look attractive today and might warrant your research. (Citi is not really a “beginner” bank though, so if this is your first bank pitch…). Also, do your own work, I am not your fiduciary. My original Citi report is here. Ally is here.
Content herein: I was going to write about the all the money center banks and the banking system, but the article got unruly. There’ll be a little on banks overall, then an update on our Citi & Ally positions only. I still follow prior holdings like WFC, BAC, & JPM… but we’ve limited time. Plus, not much changes every 3 months.
Bank Earnings — Broad Conclusions
Major banks and regional/online banks like Ally are well capitalized, highly liquid, and generally have high-quality loan portfolios; it is a low-risk time to be a bank shareholder. High loan quality is due to lending discipline from the financial crisis, and strong underlying client financials; consumers and businesses both have stronger than usual balance sheets today.
The only issue I see brewing is that private equity transaction multiples have been rising for 10+ years. Loans to private equity (& private credit) are the single largest book on many banks’ balance sheets. You can’t pay higher and higher multiples for businesses forever. There could be a reckoning, though it’s my understanding most of the bank term loans are moderate risk, and many of the other loans are things like capital call loans, which are quite low risk but tough to explain in to you in one line. In all, banks are likely to lose less than average in the next recession. If you go back through history, there usually must fist be a yearslong accumulation of stupid behavior, then, some event sets the kindling alight. There’s not much dry wood lying around waiting to be struck by lightning today.
Banks’ profitability is generally strong, with most good banks earning teens returns on capital today. They also generally have ~20%+ more capital than regulatory requirements (and I think the requirement itself is higher than needed). They’re sitting on mountains of liquidity, too. E.g., Bank of America has $2 trillion in deposits and only $1 trillion in loans. Much of the differences is invested in Treasurys, ultra-high-grade bonds, and agency-insured mortgage securities. Executives tell us that client balance sheets are also generally strong, as we said, so the banks are in a good place.
Other than this, for two years we have been waiting for lending to pick up after clients absorb higher interest costs in this new interest rate regime and for this industry to return to its 5% through-cycle growth. We are still waiting.
I’m concerned impacts other than credit losses, though. E.g., tax changes in the current or next US administration. Raising the US corporate rate somewhat would still keep the country competitive globally, but would help the government reduce its budget deficit. That’d impact corporate profits.
The regulatory winds have shifted, and we haven’t talked regulation for a while, so let’s cover that today.
Regulatory backdrop:
Capital/liquidity requirements: As a reminder, we’re still waiting on clarity around the AFS opt-out for smaller banks. Ally falls under this. Management’s running the business as if the opt-out will be taken away, so it’s building capital to fill the hole left by unrealized losses in the AFS bond portfolio, rather than buy back stock. The bonds are mainly “risk-free”, but are fixed rate, so they declined in value when the Fed raised interest rates a couple years ago, hence the unrealized loss. If the opt-out sticks, Ally will buy back more stock, and the price is currently well below what we believe is intrinsic value. At Citi, we hope self-help fixes — plus regulatory change you’ll see below — will lead regulators to reduce its SCB (stressed capital buffer), which would lower its capital requirements and allow it to return more capital to shareholders, while permanently increasing its returns on the remaining capital. These changes look more probable today than in 2024. Why?
Changes: in March, Trump replaced Fed Vice Chair for Banking Supervision Michael Barr with Michelle Bowman. Bowman’s the polar opposite of Barr. The new bank watchdog is more light-handed. Banks are overcapitalized (in my opinion, based on scenario modeling) and have tons of liquidity (in my opinion), and they’re lobbying for less draconian rules. This administration is more business-friendly and more likely to listen. It would be nice to see a sensible regime shift. So far we don’t know what’ll happen, and change may take years. However, whatever happens should be a lot better than what Barr wanted. Barr is a lawyer, not a banker, and his idea was for already-overcapitalized (my opinion) banks to hold even more capital; he reasoned in that link above, that this blunt instrument is a panacea. It isn’t. Ample capital didn’t help Silicon Valley Bank. What regulation needs to do is (1) prevent banks from doing large volumes of stupid things and (2) ensure they’re sufficiently liquid, well-capitalized, and profitable to weather a bad, unpredictable storm. Then the system will still be able to readily finance economic growth, without the downside of periodically blowing up.
A bit on credit and spending, then we’ll move on to Citi:
Credit remains strong. Lending discipline continues to be good and there isn’t much funny business going on (in my opinion). Loan loss reserves (provisions) are already adequate as well. Given the banks’+clients’ strong financial positions, I expect many banks may break even or turn a profit during the next recession.
Delinquencies (DQs) lead losses, since loans become delinquent when borrowers first fail to make a payment. DQs stopped rising after the 2023-2024 post-COVID normalization period. For example, credit card delinquencies are down at Wells, and remain low. See pg 10 of Wells Fargo’s earnings supplement, if you like detail. The story’s similar elsewhere. Commercial office, the hardest hit area on bank balance sheets today, is also seeing declines in the total loss provision and provision rates.
Spending remains strong: credit card purchase volume was +6% at Wells, +7% at JPMorgan. The consumer’s still spending.
Mortgage lending is picking up with new loan originations +26% at Wells despite the business being downsized recently.
Capital markets activity is still good, with for example investment banking +~10% at Wells, and higher Markets trading activity for clients. Trading was driven mostly by derivatives hedging activity with clients, many of whom are trying to protect capital given higher uncertainty in the economic outlook. Banks’ market-making businesses are counter-cyclical.
Commercial/corporate lending growth is still very slow given the recent interest rate shock. See FRED data. The vast majority of commercial loans are floating-rate, so mid-sized companies are paying a lot more interest these days and are still absorbing this hit to profits. (Big companies, by contrast, have access to bond markets, and mainly issue long-term fixed-rate bonds, so they take less of a hit.)
Dry powder: Many banks, like JPMorgan, are sitting on tons of “dry powder.” That’s excess capital and liquidity they want lend out, but don’t have enough demand to do so. JPMorgan’s CET1 capital ratio is >15%, several % higher than required. The same is true of its liquidity levels. Rather than repurchase shares, the bank’s holding excess cash and waiting for loan growth. Many banks have ample capacity to safely lend more, but can’t right now, and are waiting.
On to our positions.
Citi
Our deep value investment in Citi’s turnaround looks like it’s working; the stock still doesn’t reflect this.
Reiterating our thesis: Citi is a collection of average-to-excellent banking lines, which are under-earning their potential. The new CEO, Jane Fraser (and team), are unlocking this and are about halfway through restructuring the company. The bank’s earning ~10% on equity while we think it’s capable of ~14%. In the low $60s, the stock trades at ~half of what we think is fair value ($120+), and 2/3rds of its $90+/share in tangible book value.
It’s not the best business in the world, but its quality is under-appreciated. For example, Citi runs the world’s #1 international corporate treasury bank, a business which benefits from scale advantages, earns 25% returns on capital, took decades to build, and will not be replicated any time soon. It runs the international treasury, financing, and payments operations for nearly all the Fortune 500 (incl. sophisticated clients like Alphabet; yes: Google’s cross-border stuff runs on Citi), plus their thousands of global suppliers and buyers. Citi’s businesses include one of the world’s largest wealth managers, serving ultra-high-net-worth families, and a leading global investment bank. Many of these, like wealth management and investment banking, are “good neighborhoods,” meaning every decent competitor earns good returns on capital because the industry’s competitive structure is inherently attractive. Our report goes over why all this is the case (spoiler: it’s because the customer relationships are sticky).
We’ll talk about the quarter overall and in terms of Citi’s 5 major business lines:
Overall revenue, loans, & deposits: revenue was +3% y/y. Deposits were -2%, lagging peers (~+2%). Loans were +4%, above peers given Citi’s Markets and credit card exposure. Citi’s deposits are more price-sensitive (“high beta”) vs. other banks, because they’re mostly held by corporate customers’ treasuries. So, Citi quickly repriced deposits when the Fed cut rates, which preserves its its NIM and revenue more than other banks since most commercial bank loans are floating-rate.
Cost: -6% y/y, mostly due to the absence of restructuring & severance, with some productivity saves. Citi invested saves into technology as it’s automating ancient processes such as manually-generated reports to regulators (these can be thousands of pages of data & analysis).
Citi expects costs to fall slightly the next couple years. Costs would likely rise less than revenue in a mid-single-digit loan & deposit growth environment. The combination of these two should get us to ~mid-teens returns on capital. Clearly, if the industry doesn’t return to mid-single-digit growth in the next few years, it will be harder for Citi to meet our (and its) targets, since it’ll be fighting cost headwinds (mainly wage inflation, since wages are a bank’s top noninterest expense).
Capital & liquidity: similar to previous quarters i.e. very strong. CET1 capital is 13.5%, well above regulatory requirements. Other metrics similar. Citi’s pulled capital out of its worst-performing businesses and optimized others. It’s still exiting or winding down low-returning, international consumer banks, and released capital from Markets (see below). The bank can then return this capital to us via buybacks, or re-allocate it to growing, high-returning businesses like TTS (see below).
Returns: the bank earned 9.1% return on tangible capital, well below peers and Citi’s potential, but there’s loads of evidence for improvement below.
Next, performance by segment:
Services segment: Citi’s diamond, Treasury and Trade Solutions (TTS) — the corporate treasury and international finance business — gained 65bps market share y/y, and has been gaining share the last several years. Securities services (which holds things like stocks and bonds in custody and does the trade settlement processing) gained 100bps market share. Cross border transaction values were +5%, and USD clearing volume was +8% — evidently global commerce continues in spite of trade war fears. We might see a lull if China-US trade falls, since Trump is walking back many tariffs except those with China.
Polishing the diamond: Services is improving the client value proposition & moat. More countries were added to its same-day/instant payments settlements, which the vast majority of banks do not or cannot offer.
(To reiterate from our report: Services earns >$6.5 billion annually, grows 5%+ through the economic cycle, does 25% returns on capital, and makes only very-high-credit-quality loans such as inventory loans, accounts receivable factoring, and equipment loans, all secured against assets. Services alone is worth ~$120 billion, Citi’s entire market value. It is by far the world’s #1 corporate treasury and international finance bank, is heavily entrenched with customers and will not be unseated by competitors, and is a low-credit-risk form of banking. You could think of Services as protecting our investment’s downside risk.)
Markets (securities trading) was up 12% on higher client trading and prime brokerage (e.g., hedge fund) client balances. Tangible equity declined to $50 billion from $54 billion. Management had previously said they’re looking to reduce the capital required to run Markets, so this is evidence of execution. (This said, the allocated equity shifted to the Corporate segment as there’s yet no change in Citi’s overall regulatory capital requirements. Our hope is the regulatory “Stressed Capital Buffer” or SCB is reduced in time. The stress tests are each June.)
Investment banking continues to be strong, slightly gaining market share, and still ranked #5 globally. Banking is benefitting from the same tailwinds at other banks, where flat/falling interest rates have re-invigorated some dealmaking. A bad trade war may put a pause on this, mind you. We can’t predict.
Wealth is on fire. Revenues are 16% y/y, of which 11% is attributable to net client inflows (or organic growth from new clients and existing clients adding more; not from market appreciation in client’s existing balances). The segment’s efficiency ratio (1 minus the profit margin) is decreasing, to 78% today. We expect this to fall further, potentially to ~70%. Andy Sieg is doing what he promised when Jane Fraser hired him: accelerate organic growth, get operating leverage on that growth, and deliver margins in-line with the industry’s ~20-30%. Keep it up, Mr. Sieg!
USPB (Personal Banking) is doing well because it’s credit-card-centric, and card’s doing well at every bank. Balances were +8% y/y in Branded cards (where most of the revenue and profit is), 2% in private label cards. Revenues were +2%, vs. contraction at other banks who do more consumer & small business banking and are seeing NIM compression on lower interest rates. USPB lost deposit market share, with deposits off 7% y/y, but up 3% sequentially. Citi is a smaller player in consumer banking. It’s presence is mainly in large cities (New York, San Fran, etc.) to focus on credit cards and wealth management. Card charge-offs are 3.9%, a bit high for the industry at this point in the cycle.1 DQs were flat, also in line with peers.
I believe there’s more value Citi can extract from USPB, and that its ROE can rise to ~15%+, from ~9% recently. This said, USPB is Citi’s weakest segment, and if it’s not managed well, can easily bleed market share over time. Its competitive position is worse than Bank of America, say.
In a downturn, card will be Citi’s source of losses. However, the bank earns ~$15 billion pre-tax already. In a bad recession, the ~$160 billion card portfolio may lose ~$10-25 billion over ~1-2.5 years, well within that $15 billion in earning power. The bank should emerge with (nearly) all its capital, and be positioned to participate in the next economic expansion.
Shareholder returns: Citi’s likely to accelerate buybacks next ~2 years, especially after it divests its Mexican bank. This’ll be accretive us shareholders as it trades at a huge discount to tangible book value, and to our estimate of intrinsic value. A $20 billion buyback program will start soon.
Valuation: without talking through a financial model, consider Citi trades at 2/3rds tangible book value despite considerable evidence its competitive position should allow it to earn ~14%+ returns on capital, and that management is successfully executing toward this. It’s easily worth 1.5x+ tangible book value ($135+) as a 14%+ returning, mid-single-digit growth business. In a rough recession, Citi might lose ~5% of book value or even eke out a profit. This leaves us with a huge margin of safety vs. what we paid for the bank’s capital.
We continue to find Citi very attractive today, with good payoff asymmetry: it looks hard to lose money, while we can double our money if Fraser succeeds.
Why doesn’t the market agree? Many market participants look at Citi’s long-term record of operational mistakes and think it’ll walk right into a new issue. Many also believe Citi’s going to be a victim of a large, protracted trade war given it’s embedded in global trade. I don’t believe that trade war is the most likely outcome. I also believe that if it does happen, it will be reversed in time (if you look at the history of tariff-based wars, they don’t persist forever because they’re bad for all the countries involved, including the instigating country). There’s a video somewhere of Ronald Reagan apologizing for imposing tariffs and reversing his decision. Trump won’t apologize, but already we see he’s using tariffs mainly to negotiate trade deals. Finally, the US is not the only country in the world that other countries trade with. Many countries are already looking at reducing US reliance. I don’t think the market’s narrative is anywhere near the most plausible outcome, and even then, I don’t think this business is going to be permanently impaired by a trade war. Imagine we are sitting here 3 years from now. The trade war has passed, the global economy is larger, and Citi is still the top international bank. If Jane Fraser has finished cutting costs and repositioning most of Citi’s business lines, will the market still care about the trade war of 2025, or will it care about this now-more-profitable bank participating in a growing economy?
Ally
Our thesis looks to be playing out.
Management is positioning Ally’s NIMs to rise under most interest rate scenarios next ~2 years. There are several drivers. Long-time readers know the main one:
The auto loan portfolio’s yields are rising: the auto loan portfolio, Ally’s biggest book by far, yields ~9.1%. Even though interest rates have fallen, the bank’s still originating loans at 9.8% today, so the portfolio yield’s still rising. (The credit quality is also improving.) Our report goes over how Ally has the best consumer automotive lending platform in the industry, and that its wider distribution and better analytics allow it to poach the best risk-adjusted loans out of the market.
A couple other drivers:
Deposit costs should fall. Ally’s deposits are “high beta”: large-ticket, online savings accounts (OSAs) that are price-sensitive, as people move this money around for the best rate. Thus, OSA rates closely track central bank rates. The Fed Funds Rate is down 100bps since last summer, while Ally’s avg deposit costs are down only 50bps. There’s room here. If you’re a customer, you know Ally’s current OSA rate is 3.6%, while its average deposit cost this quarter was 3.78%. These OSAs make up most of Ally’s deposits. So, we’ll get another ~0.2% from here. Further, there may be more from competitive action: historically, it takes time for banks like Ally to lower rates vs. what the central bank does as banks compete for depositors. Ally’s rates usually end up ~1.0% below the central bank rate. The strength of its brand has improved in time (we’ve written about this, and management has showed this directly translates to lower deposit costs), so things may be even better this time. The company’s also more focused on margins than on deposit growth today. E.g., its trying to win smaller (less price-sensitive) accounts with a larger number of customers, vs. large (more price-sensitive) accounts with fewer customers. This, combined with the auto loan portfolio, should drive a bigger wedge in its interest rate spreads, since the auto loan portfolio is by far its biggest asset, and OSA deposits are by far the biggest liability.
Securities, cash, & cash deployment should also put upward pressure on margins. The securities portfolio (US Treasury bonds and mortgage-backed securities) is yielding ~3.3%. The overnight rate is 4.25-4.5%. As Ally’s securities mature and pay back, it’s been holding this as cash, so it picks up an extra ~1% yield. Ally’s also expanding its existing commercial loan relationships into new industries to grow its commercial loan book and put that cash to use. These loans earn far better risk-adjusted returns than either cash or bonds. Thus far, the commercial and real estate loan books have shown good growth and returns, and may become its second largest book in the future. That’ll continue to help profitability. This book earns higher returns on capital than even the core automotive lending book.
Delinquencies (DQs) & losses: are a bit elevated, and would rise further in a recession. However, at Ally and across the banking industry, DQs have leveled off. Below, you can see Ally’s newer auto loan vintages are outperforming older ones. This is partly from “high-grading” (lending to better quality borrowers) the portfolio the last few years. As older loans leave its books and newer vintages season, credit performance should improve.
In all, it looks like our thesis is playing out and margins should improve. Ally should be a low-mid-teens ROE bank, up from high-single-digit returns on capital these days.
Regulatory note & unrealized losses: remember, Ally’s still holding excess capital and isn’t buying back shares. It’s not just economic uncertainty. We need that decision above, on the AFS opt-out. That’ll tell Ally how it can treat its unrealized bond losses. The amount is ~$4 billion, ~$13 per share. As we get certainty, or as these bonds mature at par and the loss unwinds and “accretes” back to us, Ally will either return that capital via buybacks (which would be accretive to us since the stock’s far below intrinsic value) or deploy the capital e.g., into commercial loans that are earning good returns for us. It’s overcapitalized if the opt-out sticks: it carries 9.5% CET1 capital, vs. management’s 9% target, and the 7% regulatory requirement. If the opt-out is removed, it’ll be a little while to rebuild capital. The removal would be phased in over many years though, so it wouldn’t matter.
Valuation: At $30, the stock’s fallen bellow its $35 in adjusted tangible book value, while it still looks like the bank’s on track to earn $4.50+ per share in a few years. It could be an interesting time for someone new to buy. I already own as much as I want.
Other notes on execution:
Deposit/pricing strategy & discipline. Deposit growth had stalled for some time, and the company talked about being more price-disciplined. On this call, Ally said it still doesn’t expect to grow the balance sheet much. The number of deposit customers is still growing well, while deposits per customer is shrinking as the company tries to capture smaller, less price-sensitive accounts. The goal is to first push margins up.
Underwriting mistakes: Ally corrected underwriting errors from last quarter, where it was being too “tight” in who it was lending to.
Commercial lending: Ally’s looking to expand its commercial lending capabilities to new industries, etc. The risk-adjusted margins and returns here are about the best the bank has — better than the auto lending business. So I look forward to Ally growing its presence in this multi-trillion dollar market. It started this business from scratch and is doing well, just like it started the online bank from scratch in 2010, only to become the largest online bank in the US.
Happy belated Easter!
— Chris
The industry’s going through a period of elevated card losses because of multiple “card vintages” “seasoning” at a similar time — but it’d be hard for me to explain this effect. Executives expect card charge-offs to decline, which kind of matches the delinquency behavior we see.