Citigroup: Can Jane Fraser Save Citi? (NYSE:C)

Michael Corbat’s inability to correct underlying issues identified in multiple consent orders, going back years, has caused significant pain for Citigroup (NYSE:C). In 2021, Michael was replaced by Jane Fraser, and in early 2022, she presented her Investor Day plan to finally fix those problems.

Her plan includes massive spending to implement the systems and controls they need, which has caused income and ROTCE to plunge. Buybacks have been substantially put on hold due to rising capital constraints.

In this article, I’ll look at Fraser’s plan to see if she can get ROTCE to her plan goal of 11-12%, and what it means for the stock and buybacks.

The Root Cause of the Pain

In 2012, when Michael Corbat took over, he immediately put his plan in place to fix the company coming out of the GFC of 2008. It consisted of dumping all the losing assets into an entity called Citi Holdings which would be reported separately and slowly whittled down over the years, focusing on utilizing huge deferred tax assets (DTA’s) stemming from the losses incurred from the GFC, reducing their expenses (i.e. headcount) and their efficiency ratio, and returning capital to shareholders via buybacks.

Shortly before Corbat came on board to fix things, Citi was hit with a consent order from the Fed (see here) indicating that it lacked effective systems of governance and internal controls. Then, in 2014, Citi failed its CCAR test for the second time in three years (see here), and the Banamex fraud came to light (see here). Next, in 2015, Citi paid $1.3B in fines in conjunction with the Fed’s cease-and-desist order relating to its foreign currency trading violations (see here). In 2017, after a change in tax rates, Citi wrote off $23B in DTA’s that it for years said had value. In the 2017 Annual Report, Corbat effectively declared victory and implied his restructuring was complete.

From 2015 to 2019 Corbat bought back 1.3B shares totaling $77B and at an average of $70 per share. Most of the buying was done from 2017-2019, after his victory lap, when the stock traded between $60 and $80.

Problem solved? Unfortunately, no.

In August 2020 Citi accidentally sent Revlon $900M it shouldn’t have. After that, in October 2020, the Office of the Comptroller of the Currency stepped in and penalized Citi $400M for not fixing its long-standing management deficiencies (see here). The OCC indicated the deficiencies included risk management and internal controls, among other items, and that if Citi didn’t fix the problems in a timely manner the OCC could use their right to make changes to senior management and the board.

And just like that … poof!! The board of directors show Michael Corbat the door and promote Jane Fraser to be the new Ms. Fixit. Investors learned the painful truth that Corbat’s five-year plan to restructure and fix C after the 2008 financial crisis, in which C’s stock plunged from $400 to $8, was all hocus pocus smoke and mirrors based almost entirely on stock buybacks and headcount reduction.

Fraser, who took over in March 2021 when the stock was around $68, took the helm with the task of saving the ship, and started to develop another plan. The stock bounced around a bit while she mulled things over and then in March 2022, when the stock is around $62, she proudly presented her Investor Day restructuring plan. The reception was tepid at best.

What did investors learn from the consent order and the Investor Day plan? We learned that Citi’s risk and controls, and reporting systems were, apparently, being done using green ledger paper and pencils, along with the help of a finely tuned abacus. And that Michael Corbat seemingly did nothing substantive to fix it.

The stock, reflecting the high uncertainty associated with the plan, including substantial execution risk, proceeded to tank further throughout 2022 with some help from the Russia-Ukraine war and inflation battle (higher rates), and again in 2023 when the tremors of a banking crisis emerged. The stock recently fell to $38 and change, where it traded back in August 2009, and down 44% since Fraser took over. Basically, we’re back to square one after the GFC, where everything still needs to be fixed. Nothing has apparently been accomplished in the last 14 years.

So there’s the pain investors are feeling. It’s entirely understandable. The only question now is whether or not Fraser can fix the underlying problems and get things turned around.

Fraser’s Investor Day Plan

Fraser’s plan consists of three core components (summarized from her slide presentation here):

1. Building Stronger Synergies within its 5 key business segments
2. Building a Modern, Efficient and Simpler Bank
3. Creating a Winning Team and Culture that focuses on Excellence

Her first component focuses on getting rid of underperforming international consumer segments, improving business mix by focusing more on low capital / high ROTCE segments (mostly Services and Wealth), and generally growing revenue across all segments at a 4-5% CAGR, either through scaling up, or taking market share.

She’s made substantial progress on the divestiture piece and will likely make more progress soon, but has only made moderate progress on the business mix and growth objectives.

Her second component focuses on what I think is the most important thing: fixing the controls and risk management systems and investing in automation and scale. The heart of this component is getting rid of the consent orders and making sure they don’t happen again. If they don’t get this right, nothing else matters.

There is no real way to know, for sure, whether they are getting this fixed or not. They’ve spent an extra cumulative $21B in operating expenses the last three years over what their annual run rate used to be. Their current annual run rate is burning roughly $10B more a year than what they spent in 2020 and prior years. That’s a lot of cash. So based on what they’ve spent so far, and continue to spend, one would hope problems are being fixed in a definitive and decisive way. It is certainly a good sign that the trading consent order from 2012 was recently removed (although it took 11 years).

A key part of this component is that if, and when, they get the risk and controls and automation finished, they can then begin to use those improvements to scale back on systems and people and become a leaner firm. They indicated that they’d have more details in the Q4 2023 release on how they plan to do that starting in 2024. They have indicated already that they have eliminated some layers of management already.

Her third component is focused on making sure they have a high caliber, diverse and accountable team. I don’t know why they didn’t in prior years, but if they don’t have that now, then yes, they should make sure they do.

In summary, I think Fraser’s plan looks solid, is rather ambitious, and checks all the boxes… at least on PAPER. They have mid-term (3-5 years) and long-term goals (5 plus years) associated with the plan. One mid-term goal is ROTCE of 11-12%, which is a MUST if they want to get a PE multiple over 8x EPS anytime soon.

My concerns about Fraser’s plan are three-fold: 1) executing it successfully in a timely manner is very hard to do and fraught with risk, 2) reducing expenses, after they’ve been set in place, is VERY hard to do as well, and 3) reaching their 11-12% ROTCE mid-term goal will require significant buybacks that will be hard to do if their capital continues to be constrained as it is now.

In my opinion, the key to hitting their financial goals and raising ROTCE back to where it was is going to be getting both their headcount and their efficiency ratio back down. Here is where their headcount and efficiency ratios have been recently:

Headcount (left axis) and Efficiency (right axis)

As can be seen above, when Corbat pushed headcount down, the efficiency went down as well, and when Fraser pushed headcount up, efficiency went up. The impact of headcount on operating expenses is similar, as would be expected:

Headcount (left axis) and Operating Expense (right axis, in millions)

The upward curve of the expenses above is the problem that they need to tackle when you hear them talking about “bending the curve” on their conference calls. By bend, they mean get it to come back down. As I indicated earlier, they plan to release further details in that regard in their Q4 2023 earnings release. The impact of these expense management decisions on income and ROTCE likewise show their importance:

Income (left axis, in millions) and ROTCE (right axis)

As can be seen above, when expenses go down, income goes up and so does ROTCE. The relationship between ROTCE and the stock price is also clear:

Stock Price (left axis) and ROTCE (right axis)

None of the above is rocket science. Of course if you push down expenses you will get higher income. Of course if income goes up you get higher ROTCE. Of course if ROTCE goes up you get a higher stock multiple and stock price. These are all obvious to just about everyone.

But the real point here that I’m trying to make is that hitting Fraser’s ROTCE mid-term target of 11-12% will likely require that she does in fact successfully reach her risk, controls, automation and scale goals, AND that they produce the efficiencies and scale that she promises all this spending (i.e. “investing”) will obtain AND that having obtained it, she will be ABLE to in fact reduce headcount and redundancies. I mean I’m estimating that by the end of this year she will have increased headcount by 60,000 people from where it was at the end of 2019. That’s a lot of people to suddenly think about eliminating or cutting back significantly.

It’s a very tall order.

Now, to be fair, Fraser’s plan also includes a revenue growth component that will, if successful, help drive the efficiency ratio down and ROTCE up. The plan includes 4-5% CAGR in revenue growth. But growing revenues is not something Citi has done very well for a very long time. Revenues in 2023 are simply getting back to where they were in 2011. That’s basically no revenue growth for over a decade.

I think here, and Jane Fraser and Mark Mason have hinted at this, that they are hanging their revenue growth hat, to at least some degree, on having new systems in place that allow them to offer better product offerings quicker. Is that possible? I suppose. But I suspect that Citi hasn’t been able to grow revenues for over a decade for other reasons as well. I doubt having new systems in place will be a magic bullet for revenue growth.

They can also grow ROTCE by doing more buybacks, so let’s look at their capital constraints and recent buybacks to see just how significant those could be to achieving her targets.

CET1 Capital and Buybacks

Before I dive into this section, I want to discuss a couple of things.

First, I’ve heard comments on SA frequently about how once C sells this or that international operation there will be more capital for buybacks. That’s not necessarily true and confuses Cash with Capital. Citi already has a ton of cash … that’s not the driving factor in doing buybacks. Capital is. So just because a segment is sold and generates $8B in cash, that does NOT mean it facilitates more buybacks.

If a segment with $12B in capital on the balance sheet is sold for $8B, it will reduce capital available for buybacks by $4B, even though you are getting cash of $8B. Citi has alluded to this regarding Banamex and that’s one of the reasons they haven’t been doing buybacks recently (at least not in any significant way). Here’s what they said in last year’s 10K filing:

Citi has continued to pause common share repurchases in order to absorb any temporary capital impacts related to any potential signing of a sale agreement for its Mexico Consumer and Small Business and Middle-Market Banking (Mexico Consumer/SBMM) business.

They know the hit is coming. How much is the hit? To know that we’d need to know how much they are getting for the IPO and how much capital they’ve allocated to the segment being sold. We know neither. All we know is that they’ve allocated $13B in capital to the Legacy operations and that Banamex is a good portion of that. And keep in mind that they are not selling all of Banamex, just the consumer banking piece.

Divestitures of duds are great, but the fact that they are duds in the first place makes it more likely that there will be a hit to capital than a benefit. It doesn’t mean they shouldn’t get rid of them though.

Second, we need to acknowledge that capital requirements have gone up and will continue to go up with Basel III changes expected next year. On October 1, 2022, C’s CET1 capital ratio requirement increased from 10.5% to 11.5%. On January 1, 2023, it increased again from 11.5% to 12.0%. And just recently it went up again to 12.3% after Citi performed poorly in its 2023 CCAR test. The Basel III changes expected next year are likely to further increase capital requirements. It’s been estimated that it will raise capital requirements by as much as 20% for many big banks.

Citi is holding back on buybacks to build the necessary capital to prepare for this event, which I believe is expected sometime mid next year.

So if we look at these upcoming changes, what is the potential impact on capital and buybacks? Well, if we assume the Banamex IPO results in a $7B hit to allocated capital (a little over half the $13B allocated to Legacy businesses) and a corresponding $17B reduction in RWA (estimated from their latest 10Q), and an increased capital requirement of 15% from Basel III changes, and apply them to last quarter’s capital we get the following:

This is just an estimate based on assumptions that may or may not be accurate. It’s a guess more than anything. It assumes they get rid of the management buffer portion (which makes sense once these things are known). Using these assumptions, the result is that Citi needs to either raise an additional $12B in capital or reduce RWA’s by $83B between now and sometime next year. If they are relying on raising capital, it’s basically a whole year of income they have to hold back (i.e. not use for buybacks).

If they rely on reducing RWA’s, doing so will also reduce income. A reduction of $83B, if you earn 1% or so on assets as Citi does, is an $830M reduction in income (and capital). Reducing RWA’s, while directly increasing the CET1 ratio, also directly reduces income, which reduces capital, so it’s not a one-for-one tool for raising the CET1 ratio.

They do have a lot of levers they can pull to manage their CET1 ratio but they all come down to either raising capital or reducing RWA’s. Nothing’s guaranteed, other than that higher capital requirements are coming, and that will continue to constrain buybacks. So Citi is likely going to continue to be cautious on capital levels until the dust settles and unknowns become knowns.

I do NOT think we should expect any significant buybacks until the last half of 2024 or, more likely, sometime in 2025. Through the first half of 2023 they bought back $1B ($500M per quarter average), but they also issued $720M in shares for employee compensation. In Q3 they did about the same run rate, buying back only $500M or so, and likely just a smidgen more than the amount they issued for employee compensation. So I’d expect that $500M per quarter run rate to continue, with it being roughly a bit more than what they are issuing to employees as compensation. It’s pretty close to hardly any share reduction when you factor in employee share issuance.

With buybacks essentially on hold, C will continue to trade at a substantial discount and is not likely to garner much investor interest until these things play out and they reinstate buybacks in a significant manner.

We certainly should not expect buybacks like we saw in prior years for at least another year or two:

Buyback Spend (in millions)

As you can see from the chart, Corbat really ramped up buybacks from 2016 to 2019, when the stock was trading between $60 and $80. Corbat bought back a total of 1.1B shares for $77B, or roughly $70 per share.

Fun Fact: If that capital had been saved, Citi could buy back the WHOLE company today now that its market cap has fallen to around $79B.

Ooops!

Now that we’ve looked at capital and buybacks, let’s look at whether Fraser can get to her ROTCE of 11-12%.

Can Fraser Get to ROTCE of 11-12% as Planned?

If you look at the charts above, you can see that Corbat ramped up buybacks and cut headcount from 2016 to 2019 and sure enough, ROTCE finally (finally!) broke above 10% in 2018, 2019 and 2020. What did that do for the stock? Well, in 2014 it traded at a PE of around 8 or so, and by 2017 it was back to trading at 10-15 like its peers.

Then when Fraser took over, the R in ROTCE fell precipitously from 2021 through 2023 as she increased expenses by $10B a year to fix Corbat’s risk and controls mess and deal with the underlying problems. Income and ROTCE plunged back to square one. And here we sit with C again trading at some absurd PE of 6 or something.

Can Fraser get ROTCE back to 11-12%? Mike Mayo, on the last conference call (transcript here), sure seemed skeptical:

Hi. Thanks for the follow-up. From your initial answer, Jane, I hear you with the restructuring, deconstructing Citi to global lines, delayering of management, and decluttering reporting, and when you add it all together, we’ll get some numbers in January [on expenses]. But as it relates to your return targets and efficiency targets for 2025 and 2026, consensus is about one-third below what you target.

And frankly, I have not spoken to one investor who thinks you’re going to get those targets and maybe you would want to revise those lower in some way or maybe to be determined? Or what’s your degree of conviction of getting to those targets or at least getting above your cost of capital?

Fraser, in her response to Mike, seems to think her plan will get them there, but acknowledged that macro and regulatory capital requirements have changed significantly since they set their plan and created “headwinds” to doing so:

We’re confident it [the plan] will drive the revenue growth of 4% to 5%. It’s not the primary purpose, but the org simplification is the third driver of the expense reductions that we’ve talked about. And I would also say that, when you look at revenue, expenses and the targets we’ve laid out at Investor Day, we’ve certainly had plenty of headwinds in macro regulatory geopolitics in the last couple of years.

She sounds just as confident as Michael Corbat was back in the day.

The headwinds on regulatory matters that she refers to, I believe, is the fact that when they presented their plan in early 2022, their CET1 requirement was 10.5%. In the last year and a half it has gone up to 12.3%. That’s a significant headwind change since they set their goals for sure.

And just for clarity, when she mentions the org simplification being the “third driver” of expense reduction, just note that the other two drivers, presumably, are the systems implementations that will allow them to reduce headcount, and the divestitures.

Now, let’s look at the numbers and see if we can gain the same confidence she has. We’ll start with their projection for 2023. They’ve guided to $78B-$79B in revenue. I think they’ll be closer to $79B based on YTD so far. They’ve guided to $54B on expenses, which I think they’ll be close to. In addition, since they are getting rid of substantially all their remaining Legacy businesses next year, we should strip that out as well:

The Q4 projection above includes the $1.5B FDIC hit they said they’d be booking. It does not include any accruals for severance or other restructuring charges as we don’t know what those will be. Q4 EPS with the FDIC charge is going to result in an EPS figure substantially below current estimates, but I don’t know if the estimates are pre-charge or not. I presume they are. I do think Q4 has potential, with the FDIC charge and reorganization accruals, to be very ugly.

To get to the full year guidance, we back out the FDIC charge and the YTD divestiture numbers. And then from guidance we deduct the Legacy piece that is going away in the next year or so.

Now that we have a “clean” 2023, excluding Legacy and the other non recurring items, we can see whether Fraser’s plan will be achievable using her assumptions. Before we do that though, let’s look at what happens if Citi merely continues 2023 performance ($12B in annual income and small but growing buybacks) for the next couple years and see if we can get to ROTCE above 10% without lowering CET1 too much. Here’s what the “status quo” numbers look like:

The above shows that if Citi just continues to generate $12B a year in income and modestly increases its buybacks by $2B a year, it will never get ROTCE to its target if it must maintain CET1 at 13% or higher. In this status quo scenario C’s stock is probably a $40 value that can maybe get to $50 in a couple years assuming a discounted PE multiple of 7 or so.

Clearly then, something must be done to get C’s efficiency ratio down, whether that’s lower expenses, higher revenues, or a combination of the two. Let’s see if Fraser’s plan, which includes 4-5% revenue growth (I’ll use 4%), and an efficiency ratio that goes back to the mid to high 50’s, will get us to ROTCE of 11-12% without pushing CET1 below regulatory requirements. Here are the numbers:

As can be seen above, the plan does get to ROTCE of 11% in 2026 under these assumptions. But, and this is a big but, and one I think Mike Mayo likely shares based on his comments, it does so only by pushing the CET1 ratio below its current regulatory minimum of 12.3% (which is going to go up substantially after Basel III changes).

Citi is projecting in their plan that their long-term goal for their CET1 requirement is 11.5-12.0%. Is it just coincidence that the above shows that getting to an ROTCE of 11% REQUIRES a CET1 of 12% just as they have as their long term projection? I don’t know. It does suggest that they know they need that lower CET1 requirement to get to their ROTCE goal.

I’m just not sure WHY they think that 12% CET1 is an achievable goal when their current requirement is 12.3% and Basel III is expected to make it go up substantially. They can get rid of the management buffer for sure. But that alone won’t be enough. Perhaps they are thinking they can get their SCB portion, currently at 4.3% down to 3% via better CCAR performance in the next couple years. That would certainly help get them closer to 12%. But there’s a lot of uncertainty here.

All I know is that if they can’t get CET1 down to 12% they likely can’t get to 11% ROTCE either.

Overall, the goals are achievable, but they require a perfect roll of the dice. They need perfect execution on ALL the following:

• Fix the controls and risk management problems
• Remove all consent orders
• Achieve the productivity improvements that allow them to reduce their efficiency ratio
• Grow revenues at 4-5% a year when they haven’t done that in forever
• Reduce the CET1 requirement to 12% at a time when Basel III is pushing them up
• Avoid any further missteps along the way

If they can do all that, then yes, the ROTCE can hit their 11-12% target in the next 2-3 years. If they do, then I’d expect a significant increase in the multiple the stock trades at. As shown above, I think a PE multiple of 9 to 10 is possible if they make progress towards their ROTCE goal. That would mean some significant price appreciation. But there’s a lot that must happen between now and then that will take several years.

Final Thoughts

I think that if Citi does not show substantial progress on achieving its plan objectives in the next couple of years, especially in regard to reducing its efficiency ratio and ramping up buybacks, that it likely trades between $40 and $50 at best. If it can show some substantial progress in the next couple of quarters, it has the potential to trade up to the mid $50s in the next year or so. If it executes everything as planned over the next 2-3 years, it could double from here.

The downside risk is probably already baked in. Seems like there’s plenty of potential upside if you have the patience and courage to wait it out.