Understanding the Fannie Mae and Freddie Mac G-Fee Report

November can be quite the busy time for housing finance in our nation's capital. Hot on the heels of the U.S. Department of Housing and Urban Development issuing their report on the Federal Housing Administration and HUD Office of Housing 's MMI Fund to the US Congress , the Federal Housing Finance Agency has issued their annual G-Fee report.

Unlike the FHA review, which is forward looking, the FHFA's report is retrospective in nature, and details 2021. Although 2021 seems like ancient history, is there value in understanding what Fannie Mae and Freddie Mac were charging back then? Absolutely.

If you've clicked on this, I'll assume you are familiar with the roles that these two GSEs play in the mortgage market, what G-Fees are, etc. If this isn't the case, please reach out - I'm more than happy to spend a few minutes sharing knowledge with others.

Two other housekeeping items: 1) As I referenced in my writeup on the FHA's MMI Fund report, I'm going to present something that is slightly abridged. Consider this a sample, and please contact me if you'd like to really go in depth in terms of understanding the report and the strategic implications for your business; and 2) All charts are coming directly from the report. You should also grab your own copy, so that you can decide whether you agree with me or not: https://www.fhfa.gov/AboutUs/Reports/ReportDocuments/GFee-Report-2021.pdf

Let's begin.

Executive Summary

For those who just want to have some numbers that they can use in conversation or as an input, here they are:

That's basically it. The upfront portion charged gets converted to an ongoing via division by a loan's present value multiplier - think of it like you would buying up/down a mortgage rate. At the end of the day, 0.56% of the average conventional mortgage's interest payment is going to pay for Fannie or Freddie guaranteeing payments to mortgage backed security (MBS) investors.

The rest of the summary starts to do some breakdowns by different attributes, but we'll come back to that shortly using the charts that are available later in the report.

The Cost of Providing the Guarantee

So contrary to appearances prior to the Great Recession, the GSEs do not have a license to print money. There are costs involved that need to be considered:

1. Expected losses. Fairly self explanatory, as not everyone is able to pay their mortgage, and defaults do happen.
2. Capital. Whether you're paying for expected losses, or unexpected losses in a stress scenario, that's going to require capital. Capital doesn't come for free - there's a cost to it, although cost of capital becomes a bit of an abstract concept given the GSE's current situation with regards to capital and being in conservatorship.
3. General and administrative expenses. Also known as the cost of doing business.
4. 10 bps to the U.S. Department of the Treasury . Remember 2011? Adele's "Rolling in the Deep" was at the top of the Billboard charts. We also got a payroll tax cut extension into 2012. In order to pay for a two-month extension of this tax cut, every single mortgage acquired by the GSEs since April 1, 2012 has had a 10 bps annual fee to cover the cost of that payroll tax. As a result of the Infrastructure and and Jobs Act of 2021, that 10 bps lives on until 2032. Some of that may indirectly somehow impact the world of housing finance, but I digress.

The report notes that capital is indeed the greatest cost out of these, but as I mentioned, cost of capital is a bit of an abstract concept. Unless you're an artist, abstract concepts do not lend themselves to running a business. Thankfully, there is the Conservatorship Capital Framework (CCF), which was used to determine capital requirements for each loan, with the FHFA then setting a targeted rate of return on that capital.

Now, the CCF isn't a "rule", but it basically provided a capital regime by which you could actually run the GSEs. Subsequently, the FHFA enacted the Enterprise Regulatory Capital Framework (ERCF), and as of 2022, that is being used as the GSEs run their business. It's important to call this out, because it means that what goes on in 2022 and later is going to reflect a different capital regime. How different? That's complicated, and the CCF has not been disclosed in detail, so your guess is as good as mine.

Why does all this matter? Because using the GSE loss estimates and the CCF, you arrive at the "gap". The gap is an estimate of profitability compared to the target rate of return. Confusingly, a positive or zero gap means that the loan meets or exceeds profitability targets, while a negative gap means that the targeted rate of return has not been achieved - even if the loan is still profitable. So as to avoid confusion later: a positive gap means exceeding the target rate of return, while a negative gap means it doesn't hit the targeted rate of return.

Does it make sense to have loans with a negative gap? Sure! It's not a situation where, "they can't all be winners" but rather certain loans are more mission oriented, so so a negative gap is okay, because that'll be offset by a positive gap elsewhere. On their books of business as a whole, the gap needs to be positive.

Transitioning to the ERCF

The FHFA is devoting a section of the report to talk about the shift to the ERCF from the CCF, and should be applauded for doing so, because this really can move the dial. If you pay attention to nothing else, please keep this in mind:

So capital requirements are going to increase. This means the following:

1. Higher average G-Fees in order to hit return targets; or
2. Lower return targets in order to avoid increasing G-Fees.

All that being said, there's a lot of tweaking that one can do to the various negative and positive gaps.

Let's look at those components they called out:

• The risk weight floor requirement essentially means that there's a limit to how little risk a mortgage can represent. At some point, increasing credit scores, decreasing LTVs, and mortgage insurance ultimately hit a wall in terms of their ability to reduce the amount of capital the GSEs have to hold on a given mortgage. Absent this, it's possible to end up in a situation where the capital necessary on a given loan could be too low for safety and soundness purposes.
• The countercyclical adjustment occurs when home price appreciation has been running significantly beyond trend. The result is that some math is applied to the LTVs of mortgages that makes them more capital intensive. This is actually going to be an important thing for the next several years, unless home prices give back most of the past few years of gains. The Urban Institute wrote a good paper about this topic in response to the FHFA's proposal to change how Credit Risk Transfer (CRT) was treated in the ERCF.
• Capital buffers. Think of these as dollars that help ensure that the GSEs can be ongoing concerns in the event that the ERCF did not otherwise require sufficient capital.

Again, it's good to know this now, because remember: next year's report is going to reflect decisions being made now.

The Gap

Now we get to the charts. Let's kick things off with the aggregate view!

As you can see, the gap had a solid amount of growth in 2021. Another way to think of this is that if they shaved 5.5 bps off of G-Fees on average, the GSEs would still be hitting the FHFA's return targets.

There are a lot of charts that follow that show the different gaps. I'm going to skip around, and focus on the ones that I find interesting or educational.

First up is loan purpose. To set the stage, let's look at the average G-Fees by purpose.

So, the big takeaway here is that cash-out refis are 10 bps more expensive than purchase loans. Rate-term refis have slightly lower G-Fees than purchases do. How does that translate to the gap?

Did you expect to see that purchase mortgages would have a negative gap? In fact, the irony is that the difference in G-Fees doesn't even explain the full gap. From a safety and soundness perspective, it probably doesn't make sense to have the GSEs load up on cash-out refis. There's also the philosophical question about whether cash-out refis are really part of the scope of what the GSEs should be doing. Those are both topics for another day, but one could make the argument that cash-out refis subsidize the purchase market.

Looking at LTVs, it becomes clear that in absolute terms, the 70.1-80% LTV mortgage borrowers are paying the most in G-Fees, while the loans with the lowest risk, namely those at 70% LTV and below are essentially paying the same as the over 80% LTV segment. Naturally, this is going to produce some interesting gap results.

As you can see, the over 80% LTV segment of the market is subsidized by the below 80% LTV segment, as remember: the gap must be positive. When we hear calls for reducing the LLPAs charged by the GSEs in the over 80% LTV segment, it sounds logical because the borrower is already paying for mortgage insurance, which is absorbing the risk above 80% LTV (and then some). Still, when you look at this chart, what it essentially is telling you is that removing the LLPAs would result in an even larger negative gap.

Stated another way, implicit in how the FHFA and GSEs look at this that mortgage insurance doesn't fully offset the extra "costs" as outlined above - even with the LLPAs that are imposed. What to make of this? Is mortgage insurance been properly considered in the calculation of the gap or is the loan loss severity benefit of mortgage insurance insufficient to offset the higher frequency of default that is seen in low downpayment mortgages? I really have no answers to this one, except to say that it will be challenging to change the current structure of LLPAs.

There are additional gap charts that look at credit scores and seller volume - I'll leave those to you to look at and interpret.

Takeaways and Conclusion

One thing that I should mention is that this report looks similar to the previous one - in fact, some of the graphs would immediately cause you to reach that conclusion. Nothing earthshaking has really changed in terms of the directionality of gaps.

The big thing to focus on here is that this is the last marker of the "old" capital regime of the CCF, and next year, we'll be seeing a report that represents the ERCF.

What I would suggest is that you file a copy of this report away and go out to the FHFA's website and pull down the comment letters regarding the ERCF - specifically the ones that talk about the impact to G-Fees. Then, when next year's report comes out, you'll be well prepared to explain what's going on in G-Fees and the gap...and if you've made it this far, you'll be familiar enough with the report to navigate it and quickly find the important insights you need.

Thanks for taking the time to read, and as always, if you're interested in something that is deeper or more tailored to your interests, please reach out.