FNMA: developing a recapitalization framework

My original FNMA post was focused on understanding the GSE model, why it failed, what happened, and where we go from here. It tried to explain how Administration incentives and legal constraints could shape the sequence of events over the next 15 months. I suggested that the junior prefs offer a superior risk-reward vs. the common shares in most recapitalization scenarios, but I didn’t fully walk through how that math could change based on different assumptions. The goal of this post is to lay out the analysis in a bit more detail and better understand the range of potential outcome for the investable securities. I’ve also refined certain inputs, so the math will be slightly different.

First, let’s talk about capital – this is an area where my original post was probably too conservative. As a reminder, I envisioned Fannie targeting core capital at 3.50% of total assets. This target was informed by the FHFA’s proposed capital framework in 2018. This framework is effectively a modified version of Basel III, where Fannie would be subject to both (i) a leverage ratio and (ii) a risk-based capital ratio. Fannie would be required to meet the most stringent of these two ratios at any given time. The FHFA’s report suggested that the constraining metric would be the risk-based capital requirement. The FHFA lays out its calculation for Q3 2017 on page 72:

This analysis suggests that Fannie’s core capital requirement would be $115bn at Q3 2017, or 3.43% of total assets (leading to my original 3.50% core capital assumption). The nuance is that Fannie’s risk-based capital requirement has likely declined as a percentage of total assets since Q3 2017. The drivers include: (1) increased utilization of credit risk transfers, (2) lower risk surcharges due to the continued declines in Alt-A and other higher risk exposures, and (3) a decline in the deferred tax asset disallowance. Similar to Basel III, the FHFA’s framework limits the amount of core capital credit that an institution receives for deferred tax assets. Deferred tax assets are only valuable if an institution can generate sufficient pre-tax income to utilize tax loss carry-forwards before expiration, and this prospect can deteriorate at exactly the time an institution needs capital the most – a recession. As a result, some limitation of a DTA’s contribution towards core capital is logical. Fannie’s DTA balance is much lower today vs. Q3 2017 though ($12.5bn vs. $30.5bn). As a result, the DTA’s capital disallowance is likely <$5bn today, much lower than Q3 2017’s $19.9bn.

When factoring in additional benefits from credit risk transfers and lower risk surcharges, I would expect that Fannie’s current risk-based capital requirement is 2.50-3.00% of total assets. The risk-based capital ratio should remain the constraining capital ratio under the FHFA’s proposal. To be conservative, I’ll assume that Fannie targets the upper end of my newly estimated range and raises enough capital to reach 3.00% of total assets. This is 50bps lower than my prior assumption, reducing my anticipated primary issuance needs and dilution to common holders. I’ll continue to assume that a recapitalization will feature multiple primary issuances over a 3 year time-frame.

As a reminder, this illustrative recapitalization analysis is purely hypothetical. A recapitalization may occur under a very different structure or not at all. The significant amount of uncertainties makes this exercise highly imprecise. Despite being closer than ever to recapitalizing and crystallizing value, this remains a highly probabilistic exercise. It’s exactly this reason why the situation remains compelling though. Many investors are uncomfortable underwriting scenarios with incomplete information, leading to excess uncertainty discounts. If investors are willing to invest with incomplete information – when properly compensated – there often exists a recurring edge.

My updated base case recapitalization math is below:

The above analysis effectively says: if Fannie deems the senior preferred repaid, achieves its 3.00% target capital ratio through a $58.6bn multi-year primary issuance at an average price of $8.00/share, converts the junior prefs into common shares at a 10% discount, and trades at 1.50x BV when properly capitalized, then: (i) Treasury will realize $106.6bn net profits on its Fannie investment, (ii) current common holders will realize a 35% IRR from current prices, and (iii) the converting junior prefs will realize a 37% IRR from current prices. Participants in the primary raise should also be incentivized to participate by achieving a 16% IRR to my estimated run-rate trading value.

They key variables are the average issuance price (and related dilution) and FNMA’s run-rate P/BV multiple. I’ll lay out some P&L sensitivities based on those two metrics below:

My original post argued that a recapitalization would be aided by a willingness of the junior prefs to convert into common. This would benefit public optics and messaging to new holders in the capital raise. For this to occur, however, I would expect junior pref holders to capture additional economics by demanding a conversion price at a modest discount to the initial capital raise price – I’ve modeled a 10% conversion discount. There are important implications from this transaction structure.

As we know, the current common holders are bearing unknowable dilution risk. To the extent the average issuance price comes out lower, common holders will incur more dilution and generate lower returns. The common will make money in most realistic recapitalization scenarios, however common holders can’t avoid the fact that their key value driver (dilution) is currently unknowable and outside of their control.

On the other hand, the juniors prefs are in a much better position regarding dilution. If FNMA’s stock price remains low ahead of a capital raise, then the junior prefs will nonetheless be approaching negotiations from an advantaged bargaining position and try to negotiate a favorable conversion ratio based on prevailing prices. As a result, the junior prefs would actually benefit from a lower prevailing stock price (all things being equal), because pref holders would capture a higher ownership percentage in the pro forma company. This distinction is key and it creates an extremely different risk profile vs. the common.

To summarize the junior pref value proposition vs. the common:

  • Base case recapitalization scenario: junior prefs do well, comparable to common
  • Recapitalization with weak share price: junior prefs do well, materially better than common
  • Recapitalization with strong share price: junior prefs do well, modestly worse than common
  • Senior preferred claim is preserved at SCOTUS: junior preferred retains optionality/path for par, common is significantly impaired

In summary, although there remains significant uncertainty going forward, I continue to believe that the investment proposition of the junior prefs remains (i) absolutely attractive and (ii) superior to the common under most realistic scenarios.

Under what set of assumptions would an investor prefer the common at current prices? One scenario is that the junior prefs remain outstanding as opposed to converting to common. This would likely “cap” the junior pref upside near par and allow current common holders to retain a larger ownership stake in the pro forma company. I continue to believe that a junior pref conversion into common is more likely, but this is certainly another possible scenario.

Hopefully this post better frames the value proposition of the various investable securities.