Seeking Alpha

Residential Mortgage Default Outlook - A June Update

Includes: ESNT, MTG, NMIH, RDN
by: Western Edge
Western Edge
Long/short equity, special situations, debt, macro

Update to my outlook for default rates based on the latest data.

Assessing a conservative downside case.

Despite it being a volatile week for mortgage insurers, they are still in a very healthy position.

This article is meant to capture my macro outlook for mortgage default rates and the consequential investment thesis for my ongoing coverage of the private mortgage insurers: MGIC Investment Corporation (MTG), Radian Group (RDN), Essent Group (ESNT), and NMI Holdings (NMIH).

New Updates on Default Trends

There has been a slew of new developments since my last article on this topic. I'll begin by eating a bit of crow in being overly optimistic in how low default rates will remain through May and June for the PMIs. I was thrown off by NMIH's reported continuation of default rates through May, which has turned out to be an anomaly that has since corrected itself. Since then, MTG, RDN, and NMIH have filed SEC documents that served as notice to their reinsurance ILNs that default rates had reached a trigger point that would halt the amortization of their respective ILNs.

Based on the data that has now been filed, we can lay out estimates for the PMIs as follows (I am using a rough estimate for NMIH based on its ILN trigger notices):

(Source: Compiled by author based on public documents)

Another new development was Black Knight's announcing that forbearance rates had slightly ticked up for the week ending June 20th from 8.7% to 8.8% after seeing forbearance rates level off and even decline over the last few weeks. It's possible that we've reached a relatively stable rate of forbearances and would need to see additional improvements in unemployment before seeing any additional improvement.

The anecdotal evidence side is a bit murkier. I've seen several surveys which range from "all clear" signals to "we're doomed," and I suppose these variations have a lot to do with what kind of ax the publisher has to grind. Here are two great examples:

  1. "Americans Are Less Anxious About Paying Their Bills"
  2. "Half of U.S. homeowners struggle with mortgage due to COVID-19, consider selling home"

So, according to the first source (the Federal Reserve):

About 12.6% of Americans say they were worried about being able to make a minimum debt payment in May, down from a seven-year high in April (16.2%), according to a survey of consumer expectations released by the Federal Reserve. The 12.6% figure is in line with percentages from December, prior to the pandemic that sent a shock wave through the economy.

But according to the second source (NAR and OnePoll):

A survey of 2,000 American homeowners found that 52 percent are constantly concerned about making their mortgage payment on time. Forty-seven percent of the poll say they’re considering selling their home because they can’t afford their mortgage anymore.

These results clearly cannot both be correct. I'll admit I'm biased, but I would put more faith in the Fed's process, and I think an 8.8% national forbearance rate lends further weight to their side.

Updated Default Projection

With all of this in mind, I've updated my forecast for where I think reported default rates could go in June for the PMI industry:

(Source: Compiled by author based on Black Knight data and FNMA 10-Q)

Without a further deterioration in macro conditions, I would expect default rates to level out near 7% in the near term. But investors are naturally more worried about the uncertainty of what happens going forward, i.e., what is the worst-case possibility?

In my mind, there are three key risks going forward that could lead to a further increase in default rates:

  1. Expiration of supplemental unemployment benefits due to occur at the end of July.
  2. A further increase in the unemployment rates and continuing jobless claims after recently leveling out.
  3. Decline in housing prices.

When we lay out the weekly count of mortgages in forbearance against weekly continuing unemployment claims, we can see they are highly correlated (as would be expected).

Left axis: Continuing jobless claims

(Source: Federal Reserve Economic Data and Mortgage Bankers Association)

I think it is reasonable to assume that forbearance requests and defaults should not see a dramatic increase without a causal underlying increase in unemployment. The bullish bias in me even suspects it's possible that forbearances have been "sticky" even as continuing claims come down, i.e., there is not a strong incentive to voluntarily remove forbearance after regaining employment. You can just let it run its course.

Downside Scenarios

In my attempt to stress-test default rates, I am going to assume that everyone receiving continuing unemployment claims (all 19.5 million, per the latest data) will not be able to pay for housing once the supplemental benefits expire. I'm also going to assume that 60% of that population are renters (a conservative estimate, to my mind, given a greater share of service sector/lower wage jobs have been lost during this recession). This would result in 4.2 million loans in forbearance jumping to 7.8 million. I'll then assume a 20% jump in unemployment claims due to additional potential job losses and apply the same logic on 60% being renters. The table below summarizes my stress test outcomes:

(Source: Compiled by author)

This exercise suggests to me that a worst-case scenario could push default rates to 12-15%. For reference, MTG's default rates peaked at 18.4% during the GFC housing crisis, with prime defaults at 13.3%. My personal opinion, however, is that we will receive additional fiscal stimulus that gives additional help to the jobless before the supplemental benefits expire, and that we've seen peak unemployment and that it will slowly improve over time (a U-shaped recovery).

Updated Price Targets

For the benefit of everyone following the mortgage insurers, I've updated my price target model below. To address the risk of a housing price decline, I'm making a very simple assumption that the national median housing price declines by 5% as a result of all of the forced sales (median prices were $231,000 in 2019, according to Experian) and adding the resulting $11,550 loss to the firms' respective average payout per claim in the bear scenario. FWIW, the MBA currently projects housing prices to increase by 4% in 2020 and 2.9% in 2021.

In an effort to be more realistic, I am now flexing the P/T BV ratios under the 3 scenarios, but I am confident that once we obtain additional clarity over time that these ratios will go back to their pre-COVID-19 averages.

(Source: Compiled by author)

I continue to expect +75% upside for the PMIs from current levels over the next 12 months.

Disclosure: I am/we are long MTG, RDN, ESNT, NMIH. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.