COOP: some quick thoughts on the stock

I don’t plan to frequently comment on short-term price movements in ideas that I’ve posted, but I wanted to provide a few quick thoughts on COOP following the 8-K this morning.

COOP’s stock has been weak since my initial post and is down -20% thus far. This would be a large move for most stocks, but it’s less material for COOP given the (1) high operating leverage in this business, (2) high financial leverage, and (3) positive skew of outcomes (in my opinion). I don’t think much has changed on my thesis, but I did want to address recent macro and covenant concerns.

Although mortgage delinquencies remain very low, the increasing trade tension and global uncertainty has pushed capital into Treasuries. This has dragged down mortgage rates, and Freddie Mac reported that the average 30 year fixed-rate mortgage dropped to 3.99% in the final week of May 2019 (the lowest level since the first week of January 2018). For perspective, mortgage rates were 4.55% at the end of Q4 2018, 4.06% at the end of Q1 2019, and temporarily peaked at 4.20% in late April.

As we know, there are two major impacts of declining rates on COOP’s business. Declining rates increase refinancing volumes, leading to both (1) increases in the CPR and run-off rate of COOP’s existing servicing UPB, and (2) a tightening in the origination industry’s supply-demand balance.

For servicing, higher CPRs reduce the expected duration of the portfolio’s FCF stream. This should result in a negative mark to the current MSR balance, higher ongoing amortization expense, and lower future servicing margins. Subservicing profitability is mostly unaffected. For lending, higher refinancing applications overwhelm existing lending capacity, allowing lenders to raise primary rates (relative to secondary rates) and increase profitability.

In short, this is exactly what we saw in Q1 2019. Mortgage rates declined ~50bps, COOP increased its future CPR assumption from 10.8% to 13.0%, and COOP took a negative $268mm fair-value mark on its MSR asset. If not for the seasonally low CPR rates in Q1, we would’ve also seen a sequential increase in the underlying amortization expense. On the flipside, COOP’s origination margins demonstrated strong sequential gains. Following Q4 2018’s incredibly weak 33bps margin, Q1 2019 margins rebounded to 77bps. The rebound in demand drove industry pricing higher, and the overall supply/demand tension was exacerbated by the fact that many lenders have recently reduced headcount and capacity (see pages 55-57 of my original presentation).

I’ve described COOP’s combination of servicing & lending as a “natural hedge,” and it appeared to work relatively well in Q1 2019. But what if rates continue to drop further (as they are trending)? Per the 10-K, internal analysis suggests that for every 25bps decline in interest rates, there would be a ~$160mm pre-tax hit on the MSR fair value (net of offsetting excess spread impacts), or ~10-11% of tangible book value. Other disclosures suggest that this move would increase servicing amortization by ~$45mm pre-tax, or ~$0.50 per share.

I think the key question is: how much would origination margins need to increase to fully offset this servicing profitability drag? Some rough math would suggest 14bps ($45mm ÷ $33bn 2019E origination volumes). Relative to Q1 2019’s margins of 77bps, we would therefore need segment margins of 91bps. As a reminder, I estimate that through-the-cycle origination pricing would produce segment margins of 87bps, so this essentially gets us back to average cycle profitability. My presentation’s 2020 earnings estimate assumes a more conservative 75bps lending margin, so we still have some slack in that estimate to the extent servicing profitability underperforms.

Is this realistic? The question essentially is: if mortgage rates drop to 3.80%, driving another mini-refi wave, could lenders push through price increases that simply get it back to average cycle profitability levels? I feel like the answer is yes, or fairly close to it.

I’ve tried to anchor my valuation approach on normalized EPS/FCF production because I think other approaches are less relevant. As a reminder, I believe that a focus on tangible book value is increasingly less relevant because (1) it excludes several sources of capital-light earnings generation that lack material attributable tangible assets (Xome and subservicing), (2) I believe COOP’s structurally advantaged competitive position should result in long-term excess returns on capital, and (3) my bottoms-up projections indeed suggest high returns on tangible equity. I see tangible book value as a financial metric, but certainly not a valuation ceiling.

With all of that said – despite arguing that COOP has a decent natural hedge on its underlying earnings power – there are capital implications to all of this. Declining rates negatively impact the MSR asset value, however there isn’t a material offsetting balance sheet impact on the lending side of things. Origination margins and earnings should indeed move higher in a declining rate environment, but we won’t see a corresponding increase in GAAP asset fair values. Put differently, declining rates should be mostly neutral to earnings and FCF production, but negative to tangible book value. If you focus on underlying FCF generation (like me), this doesn’t matter that much. If you focus on tangible book value (like many investors), this is negative.

Are there other implications to this? Some apparently think so. I don’t have access to the report, but I believe one sellside firm has suggested that further declines in the MSR values could result in a capital impairment that is large enough to potentially trip servicing covenants. If servicer covenants are tripped, GSEs, PLS investors, and subservicing counterparties generally have the right to remove COOP as a servicer. This would be an unlikely outcome given the potential disruption to borrowers and lack of buyers for such a large portfolio, but the tail-risk is still scary.

Is this correct? Did I miss this? My presentation only briefly acknowledged this risk, and I brushed it off on page 82: “COOP’s relevant subsidiaries are well capitalized on key GSE servicer ratios.” I stand by this and think these concerns are largely incorrect. There are a variety of covenants and requirements for servicers, including minimum net worth, minimum capital ratios, and minimum liquidity balances. In Note 17 of the Q1 2019 10-Q (page 44), COOP discusses these concepts and suggests that the most restrictive covenant requires a minimum adjusted net worth of $829mm. For additional information on specific requirements, see page 86 of the 10-Q.

Critical to all of this is the fact that servicer requirements are calculated at the servicing subsidiary level (not the corporate level). COOP had $1.5bn of tangible net worth at Q1 2019, but this isn’t the capital at the servicing entity – the relevant subsidiary is “Nationstar Mortgage LLC.” Nationstar Mortgage LLC and Nationstar Capital Corporation collectively are issuers of the 2021/2022 unsecured senior notes. Nationstar Capital Corporation has no material assets or liabilities (it’s simply a co-obligor of the unsecured notes), so Nationstar Capital Corporation is effectively the issuer of the notes. All of this is readily available in the 10-Q.

With this in mind, we can look to the guarantor/non-guarantor financials to assess capital levels at Nationstar Mortgage LLC, the relevant entity for servicing covenants and requirements. And on page 50 of the 10-Q, we can see that the “Issuer” (Nationstar Mortgage LLC) actually has $2.6bn in equity at Q1 2019. This $2.6bn of equity is in significant excess of the most restrictive servicing requirement of $829mm, and provides a large cushion against any realistic asset value declines. There are some adjustments to calculate net worth under GSE ratios, however the largest adjustments are related to goodwill and intangible assets (only $225mm for COOP). The full DTA value is allowed.

Apparently this has been a controversial topic. In the 8-K filed this morning, the company would note:

In response to investor inquiries, Mr. Cooper Group Inc. (the “Company”) clarified that the financial covenants applicable to its credit facilities are defined with respect to its subsidiary, Nationstar Mortgage LLC, where equity was $2.6 billion as of March 31, 2019, as referenced in the Consolidating Balance Sheet in its first quarter report on Form 10-Q. Furthermore, the Company disclosed that so far in the second quarter, Servicing and Xome segment results are in line with expectations and Originations volumes and margins have improved over the levels reported in the first quarter and that it is financing this activity with existing facilities and cash flow from operations and remains in compliance with all financial covenants.”

8-K filed on 6/3/19

So in addition to re-affirming COOP’s subsidiary capital levels, management has also clarified that (1) servicing/Xome results are fine, and (2) origination volumes and margins have actually improved. Remember, Q1 2019 origination margins were strong at 77bps, so further sequential improvement is impressive. I do expect servicing margins to weaken sequentially in Q2, but anything near ~6bps (pre-MTM) is on-track with my estimates.

I could care less about quarter-to-quarter expectations and results, but it’s always important to use downward price movements as a reminder to refresh the thesis and re-evaluate risks. Some value investors operate in a vacuum, but I think this approach is incorrect. Markets are incredibly competitive, so it’s always important to have an explainable reason for why participants on the other side of a trade are incorrect. 5.1mm shares are currently short COOP, which admittedly is a decent number. To the extent this thesis is indeed predicated on a rate-driven servicer covenant breach, however, I’m comfortable being on the other side of that trade.

In any situation like this, you’d like to see management stepping up and buying stock. In May, Jay Bray (CEO) bought 52,910 share at $9.36 (~$500k outlay). Chris Marshall (the new CFO) also bought 110,000 shares at $9.08 ($1mm outlay). Jay currently owns ~1.8mm shares, so you can do the math on how motivated he should be to get the stock higher. KKR obviously is an aligned and deep-pocketed party here as well.

I’ve used this recent price move to add to my position.