Nobody is Coming to Save Us

In early 2024, I stood on stage at HousingWire and committed to a number in front of a large audience.

$10,500.

Specifically, to get Princeton’s cost to originate a loan down to $10,500.

I explained how the math behind it was simple.

The quarter before, the average lender lost $2,109 on every loan it closed, roughly $10,500 of revenue against $12,500 of cost. I wanted our cost under our revenue. So I put a number on it. And I said it in front of a room of people because I am the kind of guy who, once I say a thing out loud in front of people, gets it done no matter what it costs me.

And the reason I was so eager to push myself to do this was that I had spent 2022 and 2023 waiting for rates to come down, volume to come back, and revenue per loan to climb up and bail me out.

When none of it came. I realized that running the company as if it might is how I lose it.

But$10,500 was never really about cutting costs.

What I wanted was a cost structure that gave me room most lenders do not have: pay originators about 50 basis points more than average on market pricing, or pay them average comp and give their borrowers about 50 basis points better pricing. Either one, I figured, would attract and retain the kind of productive originators we wanted to build around.

That was the dream, at least.

The question now is, did it work

This week proved that it did.

UWM announced a net loss of $451.9 million, suspended its dividend, and raised $2.05 billion of additional capital.

Its stock fell sharply.

Better Mortgage lost roughly $100 million in the first half of 2026 while generating roughly $100 million in revenue.

Put simply, its expenses were about twice its revenue.

Turns out Better has been losing money every year since 2020.

Naturally, its founder and CEO was subsequently transitioned out of the role.

loanDepot has now lost money for 18 consecutive quarters. That is four and a half years. Its stock also trades more than 90% below where it went public in 2021.

PennyMac has managed to stay profitable, but its stock is still down sharply.

This is scary stuff, and it is not good for any of us.

And these are only the data points we have from the handful of public mortgage companies. What we don’t know is what is happening behind the scenes at all the private ones.

I listened to an earnings call this week and scribbled down two things that were said:

“The mortgage market has been tough the last five years.”

“We’ve been talking about rates dropping for a while, and it hasn’t happened.”

Here’s why I wrote those lines down:

Because back in late 2023, I decided to build the business for a high-rate, low-origination market.

The logic was simple. If rates stayed high and I was not prepared, I could go out of business. If I prepared for the worst and rates came down, I would win even more.

Thank goodness I planned for the worst instead of hoping for the best.

But did it work?

Ignorance is bliss. But the team that sees reality the best wins.

So the way I hold myself accountable externally is through Richey May benchmarking. Around 100 mortgage companies submit their actual financials, and the report puts Princeton’s revenue and expenses next to everyone else’s.

This week our Q2 benchmark came back.

Let’s start with revenue:

Princeton’s total production revenue was $11,172 per loan. The benchmark average was $12,215. We took in $1,043 less per loan than the average company.

In basis points, that was 294 for us compared with a benchmark average of 340. On average, our pricing was 46 basis points better.

If you are reading that the way I am, you are probably already asking the obvious question:

“So are you just losing money?”

No.

As a matter of fact, we are making money.

In the same quarter, our total cost to originate a loan was $10,550, or 278 basis points. The benchmark average was $11,311, or 314 basis points.

A conventional way to read this would be to see revenue 46 basis points below the benchmark as a problem, and to start asking how to capture more from each loan. But I see it the other way. The way I see it, lower revenue per loan is only a problem when your cost structure needs the revenue. Ours does not, so taking less means more of the economics reach the borrower through better pricing.

There is an old piece of mortgage sales advice that applies here:

“Top producers do not sell price; they sell outcomes.”

There is truth in that line.

Borrowers need expertise, communication, certainty, and a lender who can actually get the loan closed. So the cheapest loan is not always the best loan.

But that advice can also hide a false choice. Think about it:

Why should the borrower have to choose between a competitive rate and great service?

And why should a top originator have to explain away worse pricing to prove their value?

Great originators sell outcomes. They should not have to ask the borrower to overpay for them.

I believe you can have great rates and great service. After all, selling great rates doesn’t mean you are a bad salesperson. It means you are bringing more value to your borrower on the thing that matters most to them.

Our goal at Prineton was to eliminate that tradeoff. Give the borrower better pricing, give the originator strong compensation, and give both of them great fulfillment, all while still running a profitable company.

That is something only a lower cost structure makes possible.

Better pricing helps our originators win more loans.

Now great fulfillment helps them protect their referral relationships and handle more volume while that increased production spreads our fixed expenses across more loans, which also means our cost per funded loan falls again, which creates even more room for pricing, compensation, fulfillment, and continued investment.

That is the flywheel:

Lower costs.
Better pricing.
More production.
Lower costs again.

The cost advantage I am talking about here is not fluff. Our operations cost was $1,016 per loan against a benchmark average of $1,436. In Q3, I think we’ll be around $750 per loan.

And before you ask, no, our lower operating cost did not come from starving fulfillment. During the same period, 66% of our loans were clear to close at least four days before closing. And on the sales side, our total sales expense, which includes originator compensation, was $7,373 per loan, and our non-sales expense was $3,177.

Together, that is the $10,550 it costs us to originate a loan.

And that was the dream when I got on stage at HousingWire. And now the independent benchmark proves we are doing it.

Our internal Q3-to-date cost per loan is already below Q2. And while I know the quarter is not done yet, that early result confirms that the Q2 benchmark was not a one-time achievement.

So the benchmark proves the economics worked. But there is another question worth asking. Did the originators we built it for actually care?

At Princeton, we made a deliberate choice about who we wanted to build for: originators producing $20 million to well over $100 million a year.

So we asked what mattered most to them. The answers were not complicated. They were the same answer all originators give to that question:

Better pricing.
Better compensation.
Better fulfillment.
Better transparency.

So of course, every mortgage company promises some version of those. The difference is that what we built here at Princeton is not just a promise. We built the economics to deliver it, and, as is evident above, we are willing to show the numbers.

And to prove it, I started talking openly about it through this newsletter.

So far so good.

Actually not just good.

Great.

Our funded loan volume year to date is up 77.9% compared with the same period last year, and the growth is making the model stronger, not more expensive.

An originator or branch closing 20 loans per month generates approximately $80,000 per month of revenue to corporate after sales expenses.

Under our current staffing model, supporting that production adds approximately $30,000 of monthly expense, primarily in additional operations capacity.

That creates approximately $50,000 of additional monthly operating contribution.

More importantly, the additional production spreads our fixed expenses across more loans and lowers our cost per funded loan.

Which means that a productive originator’s production does not disappear into an expanding corporate bureaucracy.

But the strongest proof is what happened with two groups that represent about $250 million in annual production.

I tried to recruit both of them last year.

Both of them chose other companies to originate for.

I will admit, the other companies had better pitches. But a recruiting pitch can promise almost anything. It’s the economics underneath the pitch that decide what a company can actually deliver, and for how long.

So, as I guessed would happen, what was promised did not hold up.

Both groups left and joined Princeton this year.

I kind of felt like the backup date they wish they had chosen first. And that is okay because they made the same decision originators are asked to make every day. And it’s a tough one, especially when you are choosing between competing promises without being shown the economics underneath them.

At the end of the day, the benchmark proves the cost advantage exists. Our growth proves originators value what it delivers. And the boomerang proves that a better story is not always a better platform.

But I am not writing any of this to take a victory lap.

I am writing it because most originators are asked to make one of the biggest decisions of their working life (where to work) with almost none of this information in front of them.

And I think it’s important for every originator to understand that you do not earn more just because a recruiter quotes you a bigger comp number. You earn more when you can win more loans, close them reliably, keep your referral partners, and keep more of the economics you create. I also think it’s important for every originator to understand that every one of those gets easier on a platform that actually costs less to run.

So here are three questions you should ask yourself, your current company, or anyone recruiting you. Even me.

1. What is your total revenue per loan?

2. What is your total sales expense per loan?

3. What is your total non-sales expense per loan?

Then ask to see the benchmark data because those are the only numbers that tell you whether the compensation you are being offered is actually sustainable, as well as whether the pricing is real or if the company is carrying more corporate overhead than its production can support.

Rate, comp, service, and the company’s profitability are not four separate subjects.

They all come out of the same place: the cost structure. And a company with a bad one can only fake a piece of it at a time, and then, only for a while. But a company with a good one does not have to fake any of it.

The market never rescued us like it was supposed to.

Rates did not fall enough to fix the business.

Revenue per loan did not suddenly expand.

The only thing that worked for us was building a model for the market that actually existed instead of the one we kept hoping would arrive.

Hopefully now you can see why the $10,500 goal was never really about cutting costs, but about creating enough room to deliver more to the borrower and the originator while still running a profitable company.

Rich Weidel
CEO, Princeton Mortgage