Nonbank Update: PennyMac Financial Services

New York | In this latest edition of the Premium Service of The Institutional Risk Analyst, we take a look at the Q3 2020 results for PennyMac Financial Services (NYSE:PFSI), which is the manager of the REIT, PennyMac Mortgage Investment Trust (NYSE:PMT). PFSI is one of the best performers in the residential mortgage industry. The firm's stock routinely traded at or above book value even before the current low interest rate environment.

Mortgage Group: ACGL, AGNC, AI, BKI, BXMT, CIM, CLGX, COOP, ESNT, FAF, FBC, FMCC, FNF, FNMA, IMH, LADR, MFA, GHLD, NLY, NRZ, NYMT, OCN, PFSI, PMT, RKT, RWT, STWD, TWO

At the end of last week, PFSI was trading at or about 1.3x book in the New York equity markets, more or less the same multiple as that enjoyed by money center JPMorgan (NYSE:JPM). The stock trades on a beta slightly above “1” and its implied credit default swap (CDS) spread was ~ 140bp at the end of last week vs 41bp for JPM. It is important to maintain perspective, especially on credit.

Has this apparent equivalence in equity market valuation multiples between JPM and PFSI always existed? No, indeed it has not. Even with PFSI now showing a total portfolio of $400 billion in unpaid principal balance (UPB) of loans, it remains a tiny finance company compared with a money center bank. Thus the triple-digit CDS spread and "BB" equivalent rating from Moody's.

The current high tide in the market valuations of PFSI and other nonbank mortgage lenders and servicers is caused by the actions of the Federal Open Market Committee, but the peak in new 1-4 family lending margins was probably back in Q2 2020. While we have seen three quarters in a row of nearly $1 trillion in total mortgage backed securities (MBS) issuance in the US this year, spreads are falling rapidly. The disclosure from PFSI’s Q3 2020 earnings deck is shown below.

The obvious takeaway from the chart is that business volumes have exploded 5x since the start of 2020, mostly on the back of the open market purchases of Treasury paper and agency and government MBS by the Federal Reserve. Points:

The profitability of PFSI and other mortgage lenders is, of course, wonderful to see after years of adverse conditions for nonbank lenders. Beneath the surface, however, there are some large and important flows of liquidity that are neither well-understood nor adequately disclosed in filings with the SEC.

Going back to this summer, we wrote about the need for large government lenders like PFSI to finance the advance of principal and interest to bond holders when borrowers elect to request forbearance under the Cares Act. Nobody in Congress, it seems, thought to ask how the US would avoid a debt default if private servicers could not make the contractually required payments on “AAA” rated Ginnie Mae securities. See our comment in National Mortgage News: ("Election won’t impact mortgage industry like CARES Act problems will").

Looking at PFSI, the remarkable thing is that advances on delinquent loans had basically not gone up at all through June, this despite the fact that the level of delinquency on its government portfolio has risen and Cares Act forbearance has also clearly increased by billions of dollars. Advances started to rise in the third quarter, but still not nearly enough to explain how PFSI is funding advances on its FHA loan and Ginnie Mae MBS exposures. Says PFSI:

“Servicing advances outstanding were approximately $346 million at September 30, 2020 versus $237 million at June 30 – Advances are expected to continue increasing over the next 6 to 12 months – No P&I advances have been made in 2020, as prepayment activity continues to sufficiently cover Ginnie Mae’s requirement.”

That last sentence from the PFSI earnings materials is very significant, thus our added underline. Translated into plain terms, it says that the escrow float from loan prepayments on the $215 billion UPB PFSI GNMA book are being used to finance the missed loan payments due to Cares Act forbearance and other delinquency.

Like the rest of the industry, the FHA delinquency numbers for PFSI (15%) are large. The delinquency data suggests mid-triple digit millions of dollars is being financed by PFSI off the books using escrow funds that ultimately belong to bond holders and Ginnie Mae. And there is no actual disclosure from PSFI as to exactly how much of the firm’s default advances are being financed with bondholder escrow funds.

When we compare the treatment of the escrow issue in the IPO filings of AmeriHome and Caliber Home, for example, the difference is striking. While none of these issuers has provided quantitative data regarding the use of escrows to fund default advances, there is at least a frank statement of the fact and cautionary language regarding future liquidity risks as and when these escrow balances are no longer available.

If you look at the 100% plus expansion of bank advance financing for PFSI needed to accommodate new lending volumes over the past 9 months, would warehouse lender banks be willing to also accommodate a large shift in default advances w/o cutting back on production volumes?

The other remarkable aspect of the Q3 2020 earnings from PFSI was the discussion of the risk and return of the firm’s Ginnie Mae MSR, as shown in the chart below:

At present, the strong performance of the lending side of the house has more than offset the losses in UPB due to mortgage prepayments. PFSI reported that “in 3Q20, MSR fair value decreased modestly.” That said, as with Rocket Companies (NYSE:RKT), we urge investors and risk managers to cut the “fair value” PFSI MSR in half to truly reflect the torrential prepayments visible in the markets today.

The US mortgage market is the only bond market where securities may be issued above par and with the embedded call option already in the money. This is the grim reality that faces PFSI and other lenders, and also owners of assets such as MBS and MSR.

The borrower getting a 30-year mortgage at 2.97% today may be writing a 2.25% or 2% loan in a year’s time thanks to the "go big" strategy of the FOMC. For owners of MBS, the prospect for 2021 is a continuation of negative returns. Watch the spreads between the on-the-run 2s and 2.5% UMB coupons and the growing market in 1.5% coupon MBS in the GSE market and eventually in Ginnie Mae as well.

At the same time as spreads are narrowing, the rising delinquency rates visible in government loans are likely to continue rising across the board into conventional loans and other real estate asset classes. COVID and the related economic dislocation does not merely impact consumers, but also carry a very real cost in terms of operating expenses and financing for lenders and investors. So far, like the rest of the industry, PFSI has managed to keep its loss mitigation expenses under control, even reduced servicing costs YOY, and has seen a decline in forbearance under the Cares Act.

“Of the 16% reduction in forbearance related to reperformance: – 9% were or became current – 7% were FHA Partial Claims or completed modifications,” PFSI reports. PFSI has also jumped into Ginnie Mae early buyouts or “EBOs” in industry parlance, an activity usually prohibited for nonbanks due to funding costs.

We understand that PFSI finances their EBOs via a bank partner. The purchase of EBOs by PFSI is up 10x YOY, from $293 million in Q3 2019 to $2.7 billion in Q3 2020. Indeed, PFSI is the largest buyer of EBOs after Wells Fargo (NYSE:WFC). Banks tend to be the largest buyers of EBOs due to their funding advantage and the fact that reperforming loans need not be resold.

Source: FDIC

“EBO loan-related revenue increased significantly to $170.2 million as a result of loss mitigation activity on loans emerging from forbearance while related expenses were modest as most of the loans bought out returned to performing status immediately,” PFSI reports, a truly remarkable result. It is fair to say that a good portion of those delinquent loans modified today and sold back into a Ginnie Mae MBS at 105 or 106 by PFSI will be falling back off the default waterfall next year.

PFSI reported in Q3 2020 earnings: “$2.7 billion in UPB of loans were bought out of Ginnie Mae securities in conjunction with loss mitigation activities – 79% related to FHA Partial Claims, which must remain current for a minimum of six months to be eligible for resecuritization – 21% modifications, which may be resecuritized immediately.”

Even as PFSI saw a $90 million drop in MSR cash flows, it reported a remarkable $170 million profit on EBOs. The term “cherry picking” comes to mind with the EBO performance of PFSI and other large GNMA issuers. Note too that third quarter delinquency numbers for FHA loans were a disaster, quoting one industry observer, with delinquency averaging 15.80% nationally, Texas at 18.29%, Maryland at 18.56%, and Louisiana at 19.56%.

“At September 30, 2020, 9.0% of the loans in PFSI’s predominantly government loan portfolio were delinquent and in forbearance; elevated levels of reperformance and Beginning Period Forbearance Ending Period Forbearance resecuritization are expected to continue into 2021,” PFSI reports to investors. As we noted in our earlier comment (“The Bear Case for Mortgage Lenders”), the large delinquency visible in FHA loans could eventually become a significant source of risk for PFSI and other nonbank seller/servicers operating in the GNMA market.

We believe that investors and risk managers need to carefully monitor PFSI’s funding and liquidity, on the one hand, and the performance of its EBO and FHA loss mitigation activities, on the other, for signs of increased stress. At the end of Q3 2020, PFSI reported record results. A year from now, depending upon whether Congress does the right thing for the industry with respect to the Cares Act, PFSI and other large Ginnie Mae issuers could be facing a very different and very difficult operating environment, and this even with record low mortgage interest rates.