Mortgage Mayhem

Plus: Survival Bias, Softbank, Diem Done

Welcome to another issue of Net Interest, where I distil 25+ years experience as an investor and analyst in financial services into a weekly email. It’s coming up to our one year anniversary and I’m grateful to all subscribers for joining me on the journey – there are now 17,500 of us here. Please do get in touch via Twitter or by hitting Reply if you’re reading this on email. And don’t forget to invite your friends, family and coworkers to sign up too!

Mortgage Mayhem

A few months ago, I wrote a very popular piece about the American mortgage product. Mortgages are pretty standard fare in the world of finance, but the American version is special: it grants its user a free option to refinance if they can get a cheaper rate elsewhere. Last year, interest rates fell and a record number of borrowers took up that option to remortgage at the new low rates. 

For companies that originate mortgages, it was boom time. The largest, Rocket, saw its revenues rocket (sorry) by 3x versus the year before. It took the opportunity to IPO, coming to the market in August with a $36 billion valuation. Others followed. At the beginning of the year there was just $4 billion of mortgage sector market cap available for public investors to trade in; by the end of March 2021 there was over $70 billion, with United Wholesale Mortgage, Home Point Capital and LoanDepot all joining Rocket in the public markets. 

At the time of their listings, these companies cautioned that record mortgage volumes of 2020 may not be sustainable but through market share gains they would prevail. Sadly, market share is a zero sum game. As rates tick back up again and the mortgage boom begins to peter out, a price war has erupted. It’s always been a boom-bust market, the mortgage market – for reasons we’ll discuss – but this time the downturn has come very quickly. In the past month, these four mortgage debutantes have lost a quarter of their market cap. 

On his first quarter earnings call this week, the Chairman, President & CEO of United Wholesale Mortgage, one of the largest mortgage companies in America, gave it to investors straight:

Will margins compress? Absolutely, they will compress. They usually go from all-time highs to all-time lows. That's what happens in the mortgage business… We didn’t expect them to go down to this point in this part of 2021, but we’re excited about it because we’re going to still produce a lot of income where our competitors might not. And at the same time, we’re going to take market share and show you that we are the elite mortgage company.

In other words, we’re all going down, but we’re going to crush the competition along the way. There are a lot of zero sum games at play in the finance industry; mortgage lending is a particularly brutal one. Ostensibly, this is a piece about the American mortgage lending industry, but it’s also a piece about how players operate in a hyper-cyclical market. 

Here’s how it works.

Financing the American Home

There are three reasons to take out a mortgage in America.

  • You’re buying a home. In 2020, there were around 80 million homeowners in the US; that number typically rises each year alongside household formation. Mortgage demand from home buyers moves up and down with the drift of the economy but it’s fairly predictable. Last year, home buyers collectively took out 4.9 million new mortgages, accounting for around $1.4 trillion of volume.

  • You’ve already got a mortgage but you can now get a better deal. Mortgage rates last year fell from 4.1% before the pandemic to 2.9% at the lows. This put around 20 million homeowners “in the money” – borrowers with sufficient credit score and equity in their home that they could reduce their interest rate materially by refinancing into a new mortgage. Many couldn’t be bothered (more than you’d think), but lots took the opportunity to jump on lower rates, driving a fresh wave of demand for mortgages. It’s through this mechanism that monetary policy makes its way into consumer wallets. 

  • You want to take some cash out of your home. Rising home prices boosts the value of equity in your home and a new mortgage may be a way to extract some of it. Typically, cash-out refinancings account for between 10% and 40% of mortgage volumes. Last year, American homeowners tapped their homes for equity at the highest rate in 14 years. 

The combination of these three factors came together last year to create demand for nearly 12 million new mortgage loans at a value of $3.8 trillion. The last time the mortgage market was this hot was back in 2003, but 2020 surpassed it. With a big chunk (59%) coming from refinancings, it’s a difficult trick to repeat without rates dropping further. Sure enough, the Mortgage Bankers Association projects that volumes this year will be down by 14% before declining another 30% next year. 

In the aftermath of the global financial crisis, if you wanted a mortgage you’d go to your bank. In 2010, banks controlled around 90% of the market, with Wells Fargo, Bank of America and JPMorgan between them accounting for over half the market. It hadn’t always been like that. Before the crisis, specialist mortgage lenders like Countrywide dominated but a combination of bankruptcy and takeover concentrated share in the hands of the big banks. Wells Fargo, the biggest, celebrated the trend at its 2010 investor day:

For the first time in mortgage industry history, we have the opportunity to compete on a level playing field with competitors who are publicly owned and who have largely the same regulators and who are not monolines trying to drive up their market share or their stock price by doing irrational things.

It seems they spoke too soon. Two months after that investor day, the Dodd-Frank Act was signed into law. Along with other regulations, provisions in the Act placed a significant regulatory burden around banks’ mortgage business. They had to hold higher capital against mortgages they kept on their balance sheets, the regulatory treatment for mortgage servicing became more onerous and, in addition, they faced a wave of crisis-related mortgage enforcement actions and lawsuits. Non-banks were exempt from many of these constraints and as a result market share began to shift.

The biggest winner was Rocket. Founded in 1985 in Detroit, it survived the financial crisis but remained quite small – its market share in 2010 was 1.7%, most of which was in refinancings. By 2015 it had overtaken Bank of America to become the third largest mortgage originator in the US. Although it dressed itself up as a technology company during its IPO, it’s worth acknowledging that it was regulation rather than technology that gave it its leg up. A research paper published a few years ago estimates that most of the growth enjoyed by Rocket and the other non-banks in mortgage in the period up to 2015 stemmed from looser regulation, with only a third stemming from superior technology. 

Today, Rocket is the number one player in the mortgage market, with an 8% share. Banks’ share, meanwhile, has dwindled to 25%, with Bank of America not even making the top ten. (In 2008, Bank of America alone had a 25% market share following its acquisition of Countrywide.) Rocket distributes through two channels: direct – principally via its online platform – and through brokers. The competitive environment is different in each. United Wholesale Mortgage competes exclusively in the broker channel, which is also the focus of Home Point; LoanDepot’s focus, by contrast, is the direct retail channel.

Mortgage Economics

To understand the boom-bust nature of the mortgage market, we’ve got to understand a bit more about its economics. When a mortgage company makes a mortgage, it doesn’t keep it on its balance sheet like a European bank would; it sells it in the capital markets. We spoke about this in Financing the American Home. When it makes the sale, the mortgage company takes a gain – the difference between the price it originates the loan at and the price it can get for it in the secondary market. Right now the spread between these two prices is around 120 basis points, although this does vary by channel.  

Like other prices, the spread is a function of supply and demand. The problem in the mortgage industry is that it’s really difficult to strike an equilibrium between these two forces. When demand picks up as rates come down, there is usually insufficient capacity in the industry to meet it. The typical cost of originating a loan is between $7,000 and $9,000 so idle capacity can be expensive. Pricing therefore goes up, i.e. spreads widen. 

Firms respond by building up capacity but when demand tails off again, they’re stuck. Rather than quickly shed capacity, they cut prices to feed the machines they’ve built. As the CEO of United Wholesale Mortgage says, “margins… usually go from all-time highs to all-time lows.”

Right now, mortgage rates are on their way back up. They are currently above 3.0% and the Mortgage Bankers Association reckons they’ll be at 3.7% by the end of the year (but who knows, right? If mortgage companies knew the path of interest rates, they’d be able to manage their capacity a whole lot better). With higher rates, there are fewer refinance candidates – the number has already fallen by over 4 million borrowers – and the market cools off.  

Here’s what the CEO of LoanDepot said on his earnings call:

…We’re looking at $4 trillion of capacity last year. This year, we’re looking at $3 trillion. There’s a trillion dollars of excess capacity that the industry needs to shift, and everyone is always stubborn about shedding capacity, until they understand they must… Last year, we were under-capacity because originations were on fire… The entire industry hired up, and now you have excess capacity. That excess capacity has to work its way through the system.

Mortgage companies have two ways to confront the brutal dynamics inherent in their market. They can strike up a relationship with their customers, taking them beyond the transactional business of mortgage lending. This is clearly what banks tried to do. Bank of America’s strategy in buying Countrywide was waylaid by (a) a high price (b) terrible loan underwriting and (c) poor business practices, but it reflected the importance of the mortgage product as a cornerstone for customer relationships.

Some non-banks are navigating this approach, too. Rocket has launched other lending businesses and has also moved into the home buying market. “We will continue to increase the lifetime value of our clients as we expand our platform to address more and more of the important transactions in their lives.” It’s a strategy increasingly universal to banks and fintechs and it makes for a more competitive financial services landscape overall. 

Or they can go all out for scale. The mortgage lending market is currently very fragmented. The top five players have around a quarter of the market and there is a big tail behind that. Most other financial service markets are more concentrated and there is a history of greater concentration in mortgages, with both Countrywide and Wells Fargo having high shares in the past.

For investors, this can be a painful strategy to watch. The CEO of LoanDepot says now is the time to pounce:

During the times we’re at rising interest rates and decreasing volumes, this is the best opportunity for companies of scale and with differentiated assets to properly go through the pressure point and to increase market share… And the good news is the longer the price war lasts, the harder it is in competition and the better it is for our market share gain. So that's generally how it works. It's pretty simple that way, but it does create lots of earnings pressure.

Unfortunately, it’s not a strategy that works collectively. Rocket has a target to increase share to 25% from 8%; United Wholesale Mortgage has a target to increase share in the broker channel to 50% from 35%; HomePoint says, “we expect to continue to drive growth in our market share.” They can all afford to give it a go because they raised capital at the top of the cycle when the market was booming. 

A lot of investing is about identifying where we are in a cycle. Cycles themselves emerge because human behaviour stays the same. Sometimes it is different this time and fortunes can be made when that happens. But most of the time, the cycle keeps turning and the mortgage lending market is an apt demonstration of that.

Forwarded this? Sign up here

Share Net Interest

More Net Interest

Survival Bias

Studying successful entrepreneurs is great; I am looking forward to reading Brad Stone’s new book, Amazon Unbound: Jeff Bezos and the Invention of a Global Empire. But studying unsuccessful ones can be more enlightening. As Charlie Munger says, “All I want to know is where I’m going to die so I’ll never go there.” The trouble is that there are few books written about unsuccessful entrepreneurs. The next best thing is a parliamentary hearing. This week, Lex Greensill, founder of Greensill Capital, appeared in front of the UK House of Commons Treasury Committee to help lawmakers understand what went wrong.

The Chair of the Committee cited the piece Steve Clapham and I wrote back in July last year warning of problems at Greensill. Lex Greensill didn’t confirm if he’d read it but replied that he didn’t become concerned about the position of his business until December. His view is that the failure of his firm rests with the insurance company that denied him cover. He even used the opportunity to give the Committee a recommendation: “ of the real lessons from the failure of my firm… is that a heavy reliance on trade credit insurance is dangerous. I urge you and the Committee to consider the manner in which that is regulated, because it is fundamentally counter cyclical in its behaviour.”  

No surprise that he would deflect. The firm failed because it was riddled with conflicts of interest, carried heavy customer concentrations and grew too fast. The problem with unsuccessful entrepreneurs is that they may be less than honest.


Softbank has been subject to a lot of bashing by investors in the past (including here). But hats off to Masayoshi Son. This week the company provided an update on the investment performance of its $100 billion Vision Fund and it’s pretty good. The fund includes “many regrets” (Masa’s words) like Greensill and WeWork, but it also includes some very successful investments like Coupang, which it marks at 10.3x cost, and Doordash, which it marks at 11.9x. Overall the public portion of the fund, which now makes up 43% of the total, sits on a multiple of 3.85x cost.

Softbank had been gearing up to launch Vision Fund II when the WeWork fiasco unfolded. So it paused external fundraising and invested its own funds instead. As of earlier this month, its commitment stood at $30 billion, of which $6.7 billion had been invested. Already it has had a number of wins on the fund, in particular Beike in which it made a $1.4 billion investment that’s now worth $6 billion. It’s too early to count, but the IRR on the fund is tracking at 119%. 

This practice of principal investing to the exclusion of outside investors is commonplace in the hedge fund industry, as we discussed last week. In venture capital it’s quite rare and Softbank has indicated it will open Vision Fund II to outside capital. But as these industries converge, it’s a trend worth watching.  

Diem Done

At about the same time Elon Musk fired off the tweet that crashed Bitcoin, the Diem Association put out a press release outlining a change of strategy. We’ve discussed Diem here before. Launched as Libra in the summer of 2019, it was Facebook’s grand project to enter the crypto space. Mark Zuckerberg declared, “The idea behind Libra is that sending money should be as easy and secure as sending a message. Libra will be a global payment system.” 

The launch invited a hostile backlash from regulators around the world, who worried they might lose control of their domestic monetary systems. Libra responded by watering down some of its vision and rebranding as Diem. Our earlier piece ended with the ball in the court of FINMA, the Swiss financial regulator, from whom Diem was seeking authorisation.

It’s not clear whether that authorisation was refused (under a new chair) or whether Diem just gave up, but it announced that it’s retrenching to the US from Switzerland and will partner with Silvergate, a regulated bank, to run its currency reserve – which will be USD only rather than multiple currencies as previously envisioned.

Facebook had increasingly taken more of a backseat role – its branding doesn’t appear anywhere on the Diem website. But its fully owned subsidiary, Novi, was Diem’s consumer wallet partner: “Novi is from Facebook, so you’ll be able to access it from your favorite apps.” Through Novi, Diem would have been accessible to Facebook’s 2.7 billion users worldwide, opening up the world of crypto to a huge global market. That looks less likely now, which might be more significant for the outlook for crypto than Elon Musk’s tweet.