Competing against banks
While we are working on the next fat thematic pitch, we wanted to look at the various industries that compete with the traditional banking sector.
Three of the current KEDM themes directly compete with banks: payments, futures commission merchants (long vol), and mortgage servicing / origination (existing home sales).
Ever since the GFC, taking market share from banks has been as easy as Fat Tony stealing lunch money from the kid who couldn’t stand up for himself.
It wasn’t just that banks were held back by their own corporate inefficiencies. After the GFC, banks got squeezed by a wave of regulatory changes. Basel III and stress tests forced higher capital and liquidity buffers. At the same time, the Dodd-Frank Act added stricter leverage rules, tougher mortgage and consumer protections, and empowered the CFPB to enforce compliance aggressively. Balance sheet-heavy, lower-return businesses became harder to justify, risk appetites narrowed, and growth took a backseat to stability.
Let’s start with payments. Every time you are charged a $30 international wire fee, both on the sender and receiver side, with an extra 50 bps for FX on top, you are reminded why bank payments are asking to be disrupted. This is a classic case of Joseph Schumpeter’s creative destruction. Disruption typically comes from new entrants with little to lose rather than from established firms dismantling their own cash cows. Everyone who runs a Wise account alongside a Chase bank account needs little further explanation of who comes out on top.
Moving on to Futures Commission Merchants, we have often shown the graph of the number of registered FCMs in the US, which has been cut in half over the last 2 decades. This graph only shows half the story. If you need to trade futures and you are too small to deal with JPMorgan or Goldman Sachs, your only real options are MRX, SNEX, and ADM. Banks are tied to Basel IV capital requirements, while the FCA and CFTC regulate MRX and have less stringent capital requirements. The non-bank FCMs have been clear winners.
Source: FIA.org
And thirdly, in the world of mortgage servicing and mortgage origination, we are seeing a similar story. The banks’ share of both mortgage servicing and origination has fallen from north of 90% in 2009 to roughly 45% today.
Source: Pennymac Investor Presentation
Post-crisis rules turned mortgage banking into a capital hog for depositories. Basel III effectively penalizes mortgage servicing rights, while repurchase risk from Fannie Mae and Freddie Mac, and a decade of CFPB scrutiny, have made the business a low-return, high-friction exercise. For large banks such as JPMorgan Chase and Wells Fargo, conforming mortgages gradually shifted from core activity to balance sheet filler.
Non-banks filled the gap. Players like Rocket Mortgage and Mr. Cooper operate without bank capital constraints and run the model as a scale-manufacturing, hedged servicing business.
Figure: Mr. Cooper eating the banks’ lunch. Share price performance until the acquisition by RKT. Source: Bloomberg
After Mr. Cooper was acquired by Rocket (RKT), PennyMac (PFSI) became our favorite play in the sector. Admittedly, Rocket provides a lot more torque to a recovering origination market, especially in refinance. However, RKT’s CEO is intent on telling everybody that he’s from Silicon Valley and that he took the job at a mortgage to disrupt the mortgage industry. Meanwhile, the most Silicon Valley-like thing he brought to the company is excessive Stock-Based Compensation (we wrote about this in ‘A bear market in dumb stuff’). He is also juggling two transformative acquisitions. We like RKT as a thematic trade, but we struggle to underwrite this as a long-term investment.
Meanwhile, PFSI has quietly become the lowest cost servicer in the industry. Running a servicing business well is akin to running a call center cost-efficiently. Rule #1 is: you make sure the homeowner has no reason ever to call you. Rule #2: If they do call you, you handle the call as quickly as possible. 9 minutes on the phone instead of 10 minutes saves you millions. In a world of AI, you can achieve significant efficiencies, which require scale. That’s why a couple of non-bank, scaled mortgage servicers came out as winners.
But PSFI has one additional advantage. The entire industry uses a software owned by ICE (the exchange, not the masked bounty hunters) called Black Knight and shares some of the economics with them. PFSI has created its own internal software. Even though ICE alleged that PFSI effectively copied Black Knight’s system, for which ICE was awarded $155m in damages, PFSI was still allowed to roll out the only competing software to what had once been a monopoly. PFSI doesn’t need to share economics with anyone, which makes its future cost advantage even larger than Mr. Cooper’s.
For those who want the tl;dr: This cost advantage has gotten so large that it makes sense for banks and mortgage REITs to pay PFSI as a subservicer of loans, which would open up a capital-light growth opportunity. To kickstart this business, PFSI has just announced the acquisition of Cenlar.
For those who want the math: PFSI’s OPEX is roughly 0.047% of a mortgage’s Unpaid Principal Balance (UPB). COOP was closer to 0.07%. Traditional banks are well north of 0.10%. In a world where Mortgage Servicing Rights pay ~0.13% (0.20% minus 0.07% amortization), the cost difference is huge.
Mortgage origination volumes are set to recover in 2026, and we would love to own a mortgage originator to play this theme. While mortgage servicing tends to be hurt when rates fall due to higher prepayments, Mortgage origination and other activities more than make up for this. In other words, it is okay if the mortgages you service are prepaid, as long as you are the one selling the mortgage.
Source: Pennymac Investor Presentation
In our prior housing write-up, we noted that PFSI (then trading at 1.8x P/B) was too expensive. It has since been rerated to 1x on the back of a weak Q4 report, in which they noted that lower interest rates did accelerate mortgage originations. Still, a more competitive origination market resulted in lower margins. The rumor is that UWMC’s CEO (another strong personality in the industry) would rather destroy his own margins than see his number one market position be relinquished to RKT.
This eliminates some of the upside, but not all of it. PFSI is still a financial that targets a 15% ROE in 2026, trading at 1x BV, and has upside beyond that if either their subservicing takes off or we do get a strong mortgage origination recovery. You don’t currently pay for that optionality.
But circling back to the “stealing the banks’ lunch” narrative, this too might come to an end. In the end, everything goes in cycles.
In a recent speech, Federal Reserve Vice Chair Bowman argued for lower capital requirements for banks involved in the mortgage market to encourage greater bank participation and lower interest rates for consumers.
We noted in our write-up on existing home sales that reviving the US housing market is a key policy goal for the Trump administration. Increased bank participation might be one of the tools in their toolbox.
This was followed by a Federal Reserve memo last week proposing a 4.8% reduction in the CET1 capital required of larger banks.
It remains to be seen which rule changes will be implemented in the end, but it seems clear in which direction the current economic powers are thinking. Maybe taking their lunch money won’t be as easy in the future.