This article was commissioned by Byrne Hobart a few weeks ago for his excellent site, The Diff. Byrne has kindly provided permission for me to now share it with my audience.

Economic conditions in the United States have rarely been so good: unemployment is at record lows, wages are rising and asset prices are buoyant. By some measures, the US is growing at the fastest pace in 20 years.

But real estate is a mixed bag: construction and sales of new homes have soared, even as sales of existing homes have entered a deep freeze.

Between mid-2022 and earlier this year, existing home sales fell from an annualized rate of almost 7 million sales per month to just 4 million, a record pace of contraction and an overall drop in magnitude only slightly shy of the 2008 financial crisis. There are currently only about 600,000 homes on the market, compared to 1.5 million before the pandemic. Prices remain close to record highs, but given such thin liquidity, are virtually meaningless.

Golden handcuffs

The freeze in transactions is a function of interest rates.

Homeowners borrowed and repriced about $3trn worth of mortgage debt (half of the entire outstanding amount) in 2002-22—either through purchases or refinancings—at emergency-level low interest rates. Today, one third of mortgage debt carries an interest rate below 3%.

Contrast that with the going rate on a 30-year fixed rate mortgage: at the time of writing, above 7%. Since existing homeowners can’t ‘port’ their mortgage to a new home, or sell it to the would-be buyers of their home (except in rare circumstances), moving houses entails losing a cheap mortgage and resetting at a much higher rate. Borrowers are “locked-in” by the golden handcuffs of their cheap mortgages. 

This dynamic has always existed in the US housing market, but—given the swing from rock bottom rates to the highest borrowing costs in a generation, and in such a short amount of time—this mortgage lock-in effect has never been so strong.

A mortgage originated at a lower rate than prevailing rates is worth less than par—this is an unrealized loss to the lender and an unrealized gain to the borrower. Unsurprisingly, pandemic-era borrowers are unwilling to lose their collective $700 billion worth of these gains. 

It’s tempting to dismiss the mortgage lock-in effect as an unfortunate side effect of the otherwise borrower-friendly way homes are financed in the United States. In most countries, mortgages are funded by banks—and specifically, bank deposits. Because banks pay floating rates on deposits, they offer mortgages with floating rates (usually after a short period of fixed payments). 

By contrast, American mortgages are stamped with government guarantees, bundled together, and sold as mortgage-backed securities to pension funds, insurance companies, and other investors who have an appetite for long-term assets that match their long-term liabilities. 80% of US mortgages are fixed for 30 years.

In one line of thinking, mortgage lock-in amidst rising interest rates is a small price to pay for the luxury of long-term borrowing.

The cost of mortgage lock-in

In fact, the cost of mortgage lock-in is anything but small.

Mortgage lock-in prevents home prices from adjusting to the shock of higher financing costs. This is, in particular, a burden for the ~2 million Americans who are first time home buyers every year. Millennials are the biggest cohort of buyers these days, and in 2022, 70% of them were first time buyers.

The sprawling edifice of American intervention in the mortgage market—from the Federal Housing Administration (FHA), to the government-sponsored enterprises (GSEs) like Fannie Mae, Freddie Mac, and Ginnie Mae, to Federal Home Loan Banks (FHLBs)—are premised on the goal of making home ownership more widely available. But amidst rising rates, mortgage lock-in is an impediment to achieving that goal.

If the stickiness of the asset is a problem for prospective first time buyers, the inflexibility of the liability is a problem for existing owners. Aside from locking owners into their current homes, the rigidity of mortgages in the face of rising interest rates means that an increasing number of borrowers are ‘underwater’ or very close to it—they hold a mortgage that is close to, or exceeds, the value of their equity in their home.

Mortgage lock-in has prevented home prices falling as much as they should have, but even the recent modest 11% peak-to-trough fall—combined with mortgages that remained fixed in value—had by early this year resulted in 40% of homebuyers who bought in the summer of 2022 holding less than 10% equity in their homes (these were, of course, an unfortunate cohort of who bought at the peak—for all homeowners, the most recent data shows that less than 1% are underwater).

Mortgage lock-in also has less obvious but more harmful, widespread, and long-lived effects. Not least: it exacerbates declining levels of geographic mobility. Americans are staying in their homes longer. They are evincing a declining willingness to move neighborhoods, cities, or states in order to find work that better matches their skills. By giving Americans a powerful incentive not to move, mortgage lock-in contributes to a handful of modern-day macroeconomic problems, including anemic productivity growth and neo-feudal levels of income inequality.

But this needn’t be an intractable problem. The only other country in the world that finances mortgages with bonds—Denmark—is a model for how policy changes could alleviate mortgage lock-in and other problems with the American housing market.

The Danish model

Mortgage lock-in doesn’t exist in Denmark because borrowers can buy back their loans at market prices in the secondary market.

This strategy is compelling when rates have risen and mortgages trade below par, and contrasts with the US, where borrowers can get out of their mortgages at par but no lower.

In the US, banks originate mortgages but then sell them onwards to GSEs for bundling into mortgage backed securities. But Danish mortgage finance operates on the ‘balance principle’: bank lending is funded by the issuance of bonds which precisely match the cash flows of the underlying mortgages. Danish banks retain ownership of mortgages, including their credit risk. These remain on their balance sheets within ring-fenced ‘cover pools’. 

Securitization of these assets is not via mortgage backed securities but instead via ‘covered bonds’. Cash flows pass directly from borrowers to covered bond investors, with the mortgage bank acting as servicer.

Covered bond investors are highly secured: they have exclusive recourse to the segregated cover pool of assets on the issuing bank’s balance sheet, and (nonexclusive) recourse to the overall assets of the issuer.

Danish mortgage banks are specialized institutions that only issue and distribute mortgages without maturity transformation (a form of narrow banking). They are barred by law from taking deposits.

The balance principle and generally tight regulation ensure a very stable market—since its creation in 1797, the Danish covered bond market has not experienced a single default in a bond series. 

The allowance for repurchases below par is facilitated by the fact that Danish mortgage-backed bonds are pure pass-through securities: each specific mortgage can be traced directly to a bond that is traded in the secondary market. This means that when a mortgagor wants to terminate the loan, it is possible to identify the bond it was financed through and buy back an equivalent portion at the prevailing market price. 

In Denmark, mortgages are also assumable—as long as the new borrower taking on the mortgage meets the same criteria as the original borrower. In the US, with the exception of FHA and Veterans Administration (VA) mortgages, conventional U.S. mortgages have a due-on-sale clause and are not assumable except in the case of the borrower’s death. 

A mortgage buyback in practice

For example, after an interest rate increase, a Danish borrower who observes their loan trading below par in secondary markets can ask the originating bank to retire the loan from the cover pool, and repurchase the equivalent portion of the covered bond which referenced the mortgage in secondary markets.

The funds for the repurchase are provided by the bank (which is incentivized to offer the repurchase financing, since the bank holds the credit risk and the repurchase lowers the borrower’s risk profile). 

In practice, the bank extends a new loan for a smaller amount than the original loan (since it is trading below par), but at the prevailing interest rate (which is higher than the interest rate on the original loan).

In effect, the borrower ends up with a new loan with a smaller principal than the original loan, but a higher coupon. She is relieved of her original mortgage, has a new loan, and her monthly payments are unchanged.

With the buyback option available, Danes have no reason not to sell their homes amidst rising interest rates. This is clear from the supply data: the number of homes listed for sale in Denmark has declined but by less than it has in the United States. 

Other advantages of the Danish system

To be fair, mortgage lock-in is not the only explanation for the market freeze. Countries with floating mortgage rates, like Canada and the UK, have also seen steep declines in transactions. And so has Denmark, despite the flexibility in its mortgage market and its smaller decline in supply. Lock-in is clearly not the only driver here—the deterioration in affordability, sticky prices, and declining moving rates are also playing key roles. 

Still, amidst higher interest rates, the Danish model also offers other benefits, aside from the elimination of mortgage lock-in. 

When home prices fall due to higher interest rates, the value of the mortgage also falls, dampening the effect of rising loan-to-value (LTV) ratios. Fewer borrowers fall underwater and have an incentive to strategically default. Over the past year, Danes have bought back over 300 billion kroner-worth of covered bonds ($44bn)—this has allowed them to reduce their liabilities alongside the drop in home prices and ahead of a potential recession.

As researchers at the NY Fed point out, this flexibility would have helped during the 2008 crash: since interest rates can also rise from wider credit spreads, during the financial crisis, “an alternative redemption clause would have made it possible for US homeowners to redeem their loans at a discount and thereby protect the equity in their homes, and potentially avoid foreclosure.” Indeed, despite the Danish real estate market being historically more volatile than the American one, Danish foreclosures have been much lower. 

The Danish system is in a sense the best of both worlds.

Borrowers have access to long term loans financed via the bond market, with interest rate and prepayment risk residing within the portfolios of relatively sophisticated, long term investors who can stomach it.

And because banks retain “skin in the game”, underwriting standards are tighter, and banks are incentivized to act in ways that enhance the credit of the borrower—for example by suggesting they refinance opportunistically amidst falling rates, or by buying back their bonds trading below par when rates rise.

In fact, Danish banks routinely allow borrowers to refinance even if the value of their loan to the value of their home (LTV) exceeds the 80% that was required at origination. As the IMF argues, “the rationale behind this policy is that, while a mortgage bank carries the entire credit risk, but does not benefit from maintaining higher interest payments as there is full pass-through of interest payments to the investor, it can only make the bank’s position better to allow the borrower to refinance to lower mortgage payments, and hence improve the homeowner’s likelihood of meeting their payments.”

Ironically, this Scandinavian system would permit the government to have a much lighter hand in the mortgage market. This is partly because banks, rather than GSEs, would retain credit risk, and partly because default rates—and losses to the GSEs—would be lower.

Why hasn’t this happened in the US?

Even before the 2008 financial crisis, academics have been pointing out the merits of the Danish mortgage finance system and the benefits the US would reap from adopting some of its design principles. The current suboptimal American setup seems to be due to path dependency and vested interests, but also legitimate technical challenges.

Former Fed Chairman Bernanke has outlined four of these challenges.

First, since FHLBs can tap capital markets at very low rates (due to their implicit government backing) and advance these funds to originating banks, covered bonds are not economical in comparison.

Second, there has been a degree of crowding out by GSEs: “The GSEs' implicit government backing and their scale of securitization operations have made it difficult for banks to use covered bonds to finance their own prime mortgages.”

Third, American regulation is not yet sufficiently protective of covered bondholders.

And fourth, banks are disincentivized to issue covered bonds because capital requirements for these are higher compared to securitization via the GSEs.

None of these challenges are insurmountable. 7% mortgage rates present an opportunity to rethink the way we finance housing in America. A small Scandinavian country is showing us a path forward.