On April 11, the Mortgage Bankers Association (MBA), the largest of the industry associations in housing finance, issued the results of its Annual Mortgage Bankers Performance Report. While this is a routine data release, the implications of the report are anything but.

According to the report, the “net production income” of mortgage lending firms—a key measure of the profit earned by those firms, calculated by taking all revenues related to originating a new mortgage minus the associated expenses—was a record 1.57 percent in 2020 and 0.82 percent in 2021. However, this compares to an average, since reporting began in 2008, of just 0.60 percent. In other words, mortgage lending profit margins in 2020 were an astounding 262 percent of the historic average, and still 133 percent of it in 2021. Even in a highly cyclical industry, that’s an extraordinary increase during the dislocations of the pandemic.

This pattern of profitability is evidence of an unintended flaw in the country’s system of funding new mortgages, created by how the origination process is layered between the primary market (several thousand lenders who make loans directly to homeowners) and the secondary market (which consists almost entirely of three government agencies) that provide permanent financing. These two markets are tightly connected because significantly more than half of the loans originated in the primary market are held by its lenders for only a short period, and then sold to a permanent funding source via the secondary market. Thus, primary market firms will earn as revenue the difference in the interest rate charged to borrowers versus the lower rate paid to the secondary market funder when the loans are sold. Given the mechanics of the market, this interest rate differential is reflected in primary market lenders selling loans to the secondary market for an amount larger than actually advanced to borrowers, producing what is called the “gain on sale margin.” (Unlike the MBA’s net production income measure, it does not consider expenses or other sources of revenue.)

This gain on sale margin confirms the pattern of dramatically increased profitability during the pandemic. For example, Rocket Mortgage, currently the largest originator of new mortgage loans, had a gain on sale margin of 3.19 percent before the pandemic in 2019, saw it peak 40 percent higher at 4.46 percent in 2020, and then it went back down to 3.13 percent in 2021. Loan Depot, another of the largest non-bank mortgage originators, showed a similar but more exaggerated pattern, with 2.18 percent in 2019, a spectacular increase of almost 90 percent to 4.13 percent in 2020, and back down to 2.61 percent in 2021.

The unusual primary versus secondary market structure was designed to ensure that the primary market can keep lending even in the most stressful economic or financial market periods, which it largely does. Regrettably, however, it also means that the economic benefit of federal policymaking actions cannot be assumed to be fully and promptly transmitted to the homeowners who were its intended beneficiaries; instead, some portion can be diverted to the middlemen primary market lenders, which is exactly what happened during the stress of the pandemic.

The three secondary market government agencies are (1) the government-sponsored enterprises (GSEs) of Freddie Mac and Fannie Mae, which buy the loans and then securitize them, and (2) Ginnie Mae, through which primary market lenders securitize mortgages guaranteed in turn by two other agencies (the Federal Housing Administration and the Department of Veteran Affairs). These three organizations support the issuance of what the marketplace calls agency mortgage-backed securities (MBS), which provide the permanent funding for about 70 percent of the $12 trillion of US single-family first mortgages. The investors in agency MBS consist overwhelmingly of institutional investors (domestic and foreign), banks and government-related entities, including many foreign central banks.

In 2020, when the Federal Reserve aggressively intervened to reduce interest rates in the early days of the pandemic, including on agency MBS, it meant that Freddie Mac, Fannie Mae, and Ginnie Mae were able to offer to finance mortgages at much lower interest rates. However, primary market lenders were not required to pass through those lower rates to borrowers. Instead, there is an assumption in the design of the split between the primary and secondary markets that strong competition among primary market lenders would naturally generate such a result. And this is true most of the time.

But during the pandemic, it was decidedly not true. The tsunami of refinances generated by the Federal Reserve’s aggressive cutting of interest rates overwhelmed the primary market lenders. The result was a market in which they did not feel a competitive need to fully pass through those lowered interest rates. Instead, to equilibrate supply with the massively increased demand, they raised their margins by reducing interest rates to homeowners in an amount measurably less than if the much-lowered secondary market interest rates had been fully passed through to homeowners.

The results were (1) record levels of profits in 2020 when the supply-demand imbalance was at its peak, and (2) profit levels returning to historic averages through 2021, as supply and demand came more into balance. So, it seems the impact of the reduction in interest rates was only fully passed through to homeowners a year-plus after the Federal Reserve’s policy actions. (Of note, the greatly expanded profit margins in 2020 could have opened up the industry to accusations of price gouging, as happens from time to time when there are severe and sudden imbalances between supply and demand during natural disasters. However, as the price increase was fairly well hidden—i.e. it took the form of mortgage rates going significantly down, just not as much as they would have on a straight pass-through basis—this didn’t happen.)

Interestingly, the increase in profit margins could have been even more pronounced. In October 2020, at the direction of the government, the GSEs announced that they would (with some exceptions) add a 0.50 percent up-front “adverse market fee” on their refinances. It was justified by the elevated credit losses expected to occur because of the pandemic, and took effect on December 1. Naturally, the mortgage industry complained strongly about the fee. It was eliminated on August 1, 2021 when it became apparent to the government that the anticipated losses were not being incurred. However, in retrospect, we can see that the fee kept the industry’s 2020-21 profits from increasing even more than they did (with the fee’s impact largely felt in 2021). As a result, a portion of the primary market profit margin increase benefited the GSEs instead of going exclusively to shareholders of the lenders. And as the GSEs are effectively owned by the taxpayer, this means the taxpayer captured a portion of those enhanced profits, which only seems fair since it was government policymaking that produced the conditions for the record profits in the first place.

This primary vs. secondary market friction in policymaking can also be found elsewhere, including the requirement that the GSEs meet certain affordable lending goals. This is because it is not clear how much the below-market rates offered by the GSEs on goal-qualifying mortgages are fully passed through to borrowers versus increasing the margins of the primary market middlemen.

In order to make future policymaking more effective, the ability of the primary market middlemen to enhance their margins as described above is something that those who make decisions about the design and operation of the mortgage system should think deeply about. While the industry would naturally be vociferous against any attempt to limit their margins, a program to do so—but restricted to times of great economic stress or with respect to targeted programs designed to expand homeownership—might make sense. It would be implemented through changes in the legal contracts primary market sellers enter into with the GSEs and Ginnie Mae; what is unclear is how difficult it would be to operationalize.  But it is definitely worth exploring.