Capital One is now a fairly large bank that primarily focuses on Credit Cards, Consumer Banking, and some Commercial. They are not a globally systematically important bank that offers full spectrum services (such as investment banking, wealth management, etc) so this is not a JPMorgan or Bank of America type of business model.
In 2024, Capital One announced the acquisition of Discover Financial for $35.3B in an all stock deal. In the present, this acquisition is being integrated nicely and could be the next unlock for the company.
Capital One is expected to deliver about $20 in earnings per share this year, projected to grow at least mid single digits for a few years and the stock price is about $200 a share. The bank is also well capitalized with a CET1 ratio at 13.7% (above regulatory requirements) and mid double digits return on tangible equity.
At about 10x earnings, Capital One has roughly a 10% earnings yield. The shareholder return depends on how much of those earnings must be retained to fund growth versus how much can be returned through dividends and buybacks. If COF can grow EPS at mid single digits while returning a portion of earnings and excess capital to shareholders, returns can reach the mid teens even without multiple expansion.
The market values Capital One as a credit card lender exposed to credit cycles. The Discover acquisition introduces a new earnings stream: payment network economics. If investors begin valuing Capital One less like a traditional subprime credit card issuer and more like a hybrid of sorts with the payment network, there is even more potential upside from multiple expansion.
**COF’s 3 Levers to Drive Returns**
**1. Discover Synergies**
Discover isn’t exactly a credit card that turns heads when you put it down on a table. It doesn’t have any aura, I mean nobody at the table is like “damn, you have the Discover it Cash Back Credit Card?”.
The real value here is that Discover actually owns their underlying payment card network.
Assume you buy an item for $100 from Apple using a Capital One card. Typically, \~$2.50 will go to the various players in the system. A simplified version is as follows:
|**Player**|**Example Company**|**Cut**|**Role**|
|:-|:-|:-|:-|
|Merchant|Apple|$97.50|Receives the sale proceeds after fees|
|Issuing Bank|Capital One|$1.80|Provides the credit card and assumes risk|
|Payment Network|Visa|$0.20|Routes and processes the transaction|
|Merchant Acquiring Bank|Chase|$0.50|Connects the merchant to the payment network|
In the new Discover world, Capital One lets say saves the $0.20 on the payment network fee. Apply that to the $600B of purchase volume on Capital One credit cards. Every 10 bps of fees Capital One can save is worth $600M annually. In the example above at 20 bps that Visa was taking, that is worth about $1.2B in cost savings. This example is over simplified, but Capital One itself originally modeled $1.2B of 2027 network synergies.
What Capital One is aiming for with this acquisition is owning the toll road.
1. Migrate a lot of Visa and Mastercard network cards to Discover
2. Acquire new merchants to reach parity with Amex, Visa, and Mastercard
3. Save per swipe compared to using another network
**2. Capital Buffer Optionality**
This next one took me a while to wrap my head around and I will explain it as best as I can.
So first let me explain a couple important concepts:
**Common Equity Tier 1 (CET1)**
CET1 is the highest quality capital a bank holds to absorb losses and the CET1 ratio compares that against the bank’s risk weighted assets:
CET1 Ratio = CET1 Capital / Risk Weighted Assets
A bank with a 13.7% CET1 ratio is holding $13.70 of core equity capital for every $100 of risk weighted assets.
Regulators impose minimum capital requirements, and management generally holds an additional buffer above those requirements to protect against recessions and credit losses.
**Return on Tangible Common Equity (ROTCE)**
ROTCE measures how much profit a bank generates relative to the tangible common equity shareholders have invested in the business.
ROTCE = Net Income / Average Tangible Common Equity
For banks, this is an important measure of efficiency. Two banks might generate the exact same earnings, but if one requires less capital to produce those earnings, that bank produces a higher return on equity.
**The Relationship Between CET1 and ROTCE**
CET1 shows how much capital the bank is carrying, while ROTCE shows how efficiently it is generating earnings on that capital. CET1 is a core component of the overall tangible common equity used in the ROTCE calculation. Therefore, it has somewhat of an inverse relationship. More required capital, higher ROTCE denominator, lower ROTCE.
For example, if a bank earns $20 on $100 of tangible common equity:
$20 / $100 = 20% ROTCE
If the bank determines it only needs $80 of equity to safely support the same business and returns the excess $20 to shareholders:
$20 / $80 = 25% ROTCE
This becomes even more interesting if the excess capital is returned through buybacks. A buyback reduces tangible common equity while simultaneously reducing the number of shares outstanding. If executed at an attractive price, it increases EPS and ROTCE.
Bank stocks are all about levers on the balance sheet. Earnings matter, but so does the common equity the bank needs to hold against those earnings. If Capital One can safely operate at a lower CET1 ratio, capital that was previously trapped as a regulatory buffer becomes available to return to shareholders. Returning excess capital reduces tangible common equity, which increases ROTCE assuming earnings remain constant.
Now here’s a question that stood out to me on the latest earnings call:
Erika Najarian
Managing Director, UBS
>Hi. Good evening. I wouldn't prolong this call if this question wasn't important. I think investors really want clarity on this. Rich and Andrew, you keep mentioning that your earnings power is expected to be the same as you anticipated when you first announced the Discover deal. I was hoping to unpack that a bit. I was looking through your disclosures. I wasn't sure what you were using for the baseline, but in 2027, consensus EPS Sorry, at the deal announcement, consensus EPS for Capital One standalone was about $21. You mentioned over 15% accretion to 2027 EPS at the announcement. That rounds up to, let's call it like $24.50 if we use 16%, 17% accretion. Last quarter, you mentioned that when you were thinking of ROTCE, you weren't thinking of CET1 all the way down to 11%. If you use 12.5% on the current share count, you can get to a mid-20s ROTCE pro forma. What is wrong with that line of logic?
Erika is essentially saying:
“The deal was said to produce >15% EPS accretion. You are telling us the earnings power you expected is still intact and you're calculating that earnings power assuming 12.5% CET1. When I do the math, I get like a mid 20s ROTCE. So what am I missing?”
Erika knows that Capital One’s earnings are going to come from the combined entity’s earnings power and how much capital buffer COF needs to hold.
This is why the difference between a 12.5% CET1 assumption and an 11% capital need matters. The same earnings produced can generate much higher returns on equity if less capital is required to support the business.
Management’s answer:
Andrew Young
CFO, Capital One
>Well, Erika, let me just unpack a couple of the assumptions that you made that don't actually tie to things that we've said. I just want to be really clear here. First of all, our assumptions that we laid out. I'd encourage you to go back when we announced the deal in February of 2024. What we said was we were taking consensus estimates for both Capital One and Discover, and we made the adjustment to Discover's loss forecast just based on the things that we had seen during diligence. With respect to ROTCE, at the time, the weighted average consensus for CET1 was 12.5%. When Rich last quarter highlighted that we're defining earnings power as ROTCE, we just wanted to remain consistent with that denominator of 12.5% for the sake of doing the math.
It is not saying that that is our target. As I answered before, in terms of what we believe our capital need is, we believe our capital need is 11%, but we're just doing the math on ROTCE at 12.5% for the sake of comparability. With respect to EPS, of course, share price assumptions have moved. There's just a number of things that have moved in terms of those assumptions, particularly as they relate to individual line items, and that is why we keep coming back to the ROTCE as our definition of earnings power.
Andrew essentially responds with:
“When we built the original Discover deal model, the capital assumption was 12.5%. So, to compare today's earnings power with what we expected in February 2024, we continue to calculate ROTCE using that same capital assumption.”
However, he also implies their actual regulatory capital need is only 11%.
Right now they are sitting on \~$516B of risk weighted assets with a 13.7% CET1 ratio. If they go from 13.7% to 11%, that’s nearly **$14B** of capital freed up. They can’t use all this at once, but it goes to show the level of optionality available for dividends and buybacks, given the entire market cap of the company is around **\~$120B** at the moment. Even say half that amount, $7B, available for buybacks would greatly benefit shareholders.
**3. Premium Cards and Moving Upmarket**
Capital One said in Q2 that it continues investing in heavy spenders at the top of the market, including rewards, lounges, and premium benefits. Fairbank described its upmarket portfolio as performing near the top of the industry's growth metrics. The Venture X card provides an attractive premium card that doesn’t function like a coupon book in the same way as the Chase Sapphire Reserve or Amex Platinum. You pay a $395 annual fee to get $300 in annual travel credits and 10,000 in points, making the breakeven much more achievable. Their lounge network also continues to grow and given that they are newer builds, they tend to be pretty modern and competitive with the other top end airport lounges.
Capital One can potentially attract premium customers, drive more spend on its own network, and achieve miniature Amex type benefits they previously did not have access to.
**Risks**
1. Inability to Acquire Merchants
1. Discover has nearly universal acceptance in the US, but internationally it does not have the same acceptance. Capital One will have to spend money to acquire new merchants and the question becomes how much do merchants really care about saving a few basis points just to support another network?
2. Credit Cycle Risks
1. Capital One is a lender and is heavily exposed to the health of the consumer. If the economy tanks, so will Capital One. There is no way around that.
2. Capital one says 73% of their credit card portfolio has a FICO above 660 and 27% is 660 or below. This is much more exposed than a company like Amex.
3. Acquisition Risks
1. It is not easy integrating a $35B acquisition and it's been 2 years already, showing how slowly this stuff moves. Consumers also will not be thrilled about having to get a new credit card and number and deal with the process of transitioning over.
**Concluding Thoughts**
Overall I think Capital One has a bright future in the long term as they continue to execute well in the core business. It is also still led by the founder, Richard Fairbank and he's done a great job of leading the company for nearly 40 years now. The Discover acquisition I believe will spur the next leg of growth and efficiency for the company. Capital One is also a neobank of sorts, they don’t rely on a ton of physical branches and are more technically capable than most banks. Fairbank was prescient to migrate the whole Capital One tech stack to AWS, which gives them some benefits from a technical standpoint that other legacy banks don’t have. If the health of the consumer deteriorates rapidly, Capital One will suffer as well, but that is simply the risk you take when buying individual stocks.