Arlington Asset Investment: A Constructive View On Preferreds

Summary

• The mortgage REIT preferreds space went through a roller coaster ride in 2020 which has opened up attractive opportunities in the sector.
• Arlington Asset Investment preferreds, particularly the 8.25% Series C - AAIC.PC - continue to trade very cheap relative to its credit metrics.
• Very likely this divergence is caused by investors worried about the common dividend suspension and a relatively large book value drop over Q1 last year.
• One thing to watch out for is whether the economic leverage starts creeping up back to double-digit levels.
• Looking for a portfolio of ideas like this one? Members of Systematic Income get exclusive access to our model portfolio. Get started today »

This article was originally published on 17-Feb.

The mortgage REIT preferreds sector went through quite a round trip in the past year but managed to finish in the green, in aggregate. In this article we take a look at the Arlington Asset Investment (NYSE:AAIC) preferreds. Our main takeaway is that the preferreds continue to trade at attractive levels relative to the underlying portfolio, particularly, the 8.25% Fix/Float Series C (AAIC.PC), trading at a 9.3% stripped yield with a 2024 call date.

In our view, the suspension of the common dividend and the larger than average drop in book value over the first quarter relative to other agency-focused mREITs (the portfolio was 98% in agencies at the end of 2019 and is less so now) have caused many investors to avoid the company's securities. This appears to have created an opportunity as the preferreds yields are very attractive relative to the company's portfolio based on economic leverage, the allocation to agencies and equity / preferreds coverage. One thing to watch out for is whether the portfolio leverage starts creeping up to double-digit levels as this caused much of the book value drop in the first place.

A Look At AAIC Preferreds

Arlington Asset Investment has not had the smoothest 2020. Unusually, given its agency-focused portfolio at the end of 2019, the company suspended its common dividend in March and said it would evaluate its preferreds dividends (though this latter comment came to nothing as preferreds dividends continued to be paid). This was despite the company's book value change in the first quarter being not a million miles away from other agency-focused mREITs at -33% vs. -26%. Other mREITs that suspended their dividends had book value changes over the quarter of over 50% on average.

While worrying on the face of it, this dynamic is actually attractive for AAIC preferreds investors on two fronts. First, the fact that the common dividend remains suspended is a red flag for some income investors which allows the preferreds to trade fairly cheap. And secondly, the company continues to retain cash which supports the equity coverage of the preferreds.

This is probably why AAIC preferreds are trading at elevated yields in the sector. The chart below takes the highest-yielding stock of each issuer. This is not quite apples-to-apples as some are fixed and others are fixed-to-floating (which would cause their current stripped yields to trade at elevated levels in anticipation of a significant step-down in their reset yields). That said, it's still pretty useful and shows that AAIC preferreds aren't particularly loved.

Source: Systematic Income

Of course, the elevated yields of the preferreds could be there for a reason. To understand if this is the case we need to look at the company's portfolio.

When looking at mREIT preferreds, commentators can focus on the wrong features of the financial make-up of the company, treating these entities, which are essentially portfolios of financial assets as widget manufacturers. For example, focusing on the company's income and whether it's significantly above the preferreds dividends gets a lot of focus. In reality, this metric is not particularly useful for a couple of reasons.

First mREITs have very volatile earnings compared to traditional companies so this metric is going to swing around quite a bit, producing more noise than signal.

Secondly, if an mREIT had to source cash to finance its preferreds distribution it would simply sell a tiny portion of its portfolio (the bulk of which tends to be very liquid) or borrow more cash via repo. A "regular" company would not find this as easy since the "stuff" that it owns, e.g. equipment, plants etc., are highly illiquid and such companies don't operate in financial markets on a daily basis, making it more cumbersome to borrow cash from its lenders.

Another metric that gets a lot of play is the amount of cash and unencumbered assets at the disposal of the company with the view that all of this can simply go to retire the preferreds in a shock event. This ignores the fact that the vast majority of mREIT borrowings are both mark-to-market and recourse to the company and in a bankruptcy these creditors will be ahead of the preferreds shareholders. Not only that but the mREIT derivative exposures are, what is known as "wrong-way," meaning they will tend to fall in value during a shock (specifically, fixed-rate payer swaps which hedge agencies will lose money as rates collapse). This means that the situation where an mREIT is most likely to go bust is when its derivative positions are in the red and when the derivative counterparties will have a large senior claim on the company, again, ahead of the preferreds shareholders.

So, if these metrics aren't very helpful let's take a look at the ones that do matter. We split the more useful metrics into macro and micro features of the companies' balance sheets and portfolios. The macro features are leverage, percentage of agency allocations and equity to preferreds liquidation preferreds coverage.

The information in the chart is, unfortunately, a mix of either Q3 or Q4 reporting as we are going through a Q4 reporting season. Economic leverage is shown on the y-axis (this does not always reflect the leverage as reported by the companies since the company definitions differ widely). The percentage of agency allocation in the portfolio is on the x-axis. The number in parenthesis next to the ticker is that equity / preferreds liquidation preference coverage - the higher the better here.

What the chart shows is that for its leverage (imagine drawing a straight horizontal line), AAIC holds a relatively high proportion of agency securities which is great to see. And for its agency allocation (imagine drawing a straight vertical line through AAIC) its portfolio has a relatively low leverage. Finally, for the combination of leverage and allocation metrics of companies close to it, AAIC has a relatively high equity / preferreds coverage ratio. This tells us that the combination of these three metrics shows the portfolio is in a pretty solid place from the perspective of preferreds shareholders.

Source: Systematic Income, SEC Filings

Micro features also are important but they have a far smaller impact on the company's book value in a crisis. They are useful, however, because they illustrate why AAIC underperformed other agency-focused mREITs in March in terms of book value.

These micro features are the composition of hedges and the composition of the agency allocation. Both of these were unfavorable for AAIC going into the market drawdown. The company does not appear to use any option-based or TBAs as hedges, preferring instruments that drive basis risk and have no positive convexity. This is not a great feature of its hedging program and, hopefully, the company takes its 2020 experience onboard.

On the agency allocation front, the company's portfolio was 100% in specified pools at the end of 2019 which was bad luck since TBAs strongly outperformed SPs in the crisis due to the Fed's direct purchases of TBAs and dollar roll financing.

To summarize, AAIC macro features are pretty attractive within the broader population of mREITs and the valuation its highest-yielding preferreds is among the highest in the sector - a nice combination to have. The micro features are not great but they are not going to be primary drivers of book value performance in the medium term.

Let's take a look at the company's two preferreds. This is how the two series look relative to each other.

Source: Systematic Income Preferreds Tool

The company also has issued two baby bonds though they look much less attractive as they are both currently callable and are trading not far from par. The benefit of being higher up in the capital structure also is less attractive for a leveraged financial entity all of whose securities could very conceivably go to zero given its lack of "hard" assets to help in the recovery process.

Investors familiar with preferreds will notice that the stripped yields of the two series are unusually different - being around 1.5% apart. Other mREIT preferreds with fixed and fixed-to-float series are much closer in their stripped yields. For example, the CHMI series are only 0.30% apart, CIM series are 0.4-0.6% apart, IVR series are trading close to each other. This could be because the floating spread of the AAIC-C is unusually high but that's not actually the case. It also could be that the AAIC-C call date is unusually long, providing investors an ability to lock in a high yield for longer, but that's also not the case as its call date in 2024 is not unusually long.

Our view here is that not that AAIC-C is priced very attractively relative to AAIC-B but that the fix-to-float series of other mREIT preferreds are priced too expensive relative to their fixed-coupon counterparts. That said, it's hard to make this evaluation on the basis of yield alone since fixed-coupon preferreds tend to be currently callable and only a handful have call dates more than four years in the future.

This could mean that investors in fixed-to-float preferreds are simply happy to settle for a similar yield to the fixed-coupon preferreds now to lock in longer call protection in the sector at the expense of a potentially large step-down in coupons when/if the preferreds float. Our view is that it's a combination of both - most retail investors are probably not aware of the large potential coupon step down. At the same time, yields in the sector are quite attractive and currently callable preferreds have been redeemed at a fast clip (four in the sector in the last couple of months by our count) so it's tempting to try to lock in decent yields for as long as possible.

The upshot of this is that AAIC-C is priced attractively in our view, even given the risk of a lower coupon step-down. However, investors who don't expect short-term rates to rise over this decade may be willing to go for 7% Series B (AAIC.PB) which is a fixed-coupon stock with a 2022 call date.

Takeaway

The divergence in the AAIC portfolio quality and its preferreds yields strongly suggests that investors are avoiding the company's securities given its common dividend suspension and larger than average book value drop in Q1 (for agency-focused mREITs). This makes the preferreds attractive in our view, particularly as their low dollar price leaves additional room for total return as well.

One risk to watch is what AAIC does with its leverage. Prior to entering the COVID-19 shock its leverage was very extended which is one reason its book value fell more than the average agency-focused mREIT. Another reason was that it took its leverage way down to low single digits while other mREITs kept it mostly unchanged. AAIC calculates something they call "at-risk" leverage which only looks at its repo borrowings and adds unsecured debt to equity, which understates the denominator and overstates the numerator. There's nothing nefarious here as everything is disclosed but it does make it more difficult to compare it to other mREIT leverage figures, none of whom appear to use this definition of leverage. In short, if the company's leverage starts to creep up close to double digits it may require a revaluation,

Analyst’s Disclosure: I am/we are long AAIC.PC. 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.

Comments

[Keep it Country]

Does the fiasco of the CDR preferreds make anyone gun shy buying AAIC preferreds?

[Fero.]

mReit preferreds really sucked last spring. I won't touch any of them again but good luck to those who do. I can get 9% elsewhere and not have to see the position drop 75% while suspending dividends. even if restarted later. that's more adrenaline than I like.

[CincinnatiRick]

@Fero. "mReit preferreds really sucked last spring."

Yes, that's when the real money was made. It turns out that you could have blindly grabbed most anything and emerged a hero. Let me tell you about those Hersha preferreds with their suspended dividends, But these days, $22 for AAIC prC looks to be a solid deal. I bought today.

[ADS Analytics]

Yes, the drawdowns were pretty uncomfortable (for those who were long). The other thing to consider is 1) few fund options "earn" their 8-9% yields and 2) few 9% payers are up 10% year-on-year. It's hard to escape heavy drawdowns owning yields at high single-digits. Best one can do in my view is to find securities that can remain resilient i.e. bend, not break.
ADS

[garkster]

@Fero. Yes, please do share, in retrospect, those 9% payers that did not take a major, albeit temporary, haircut to principal a year ago.

[User 29188045]

I'm long AAIC.PC and finally back in the green (way up when considering dividends over the past few years) but really the main issue here is the management team, as @Colorado Wealth Management Fund routinely points out. I'll admit they've done pretty well lately but CEO "Rock" is a bit arrogant that has made a series of dumb moves over the past few years and I'm concerned that his hubris could lead to additional risky & ill fated decisions. For now I'm holding but I'm holding my breathe as well.

[ADS Analytics]

Typically, the beta of dumb management decisions is much higher on the common than the preferred. So long as agency allocation remains high, leverage remains low-ish and common/pref coverage is high while pref yields are high it looks attractive in my view.
ADS

[Colorado Wealth Management Fund]

@ibid6677 Agree. Looking at AAIC, the coverage remains very thin. Looking at common equity to preferred equity or total core earnings projections to the amount needed to pay out preferred dividends, the coverage just isn't very strong. It certainly wouldn't be considered "high" since it is near the lowest in the sector.

If we consider the baby bonds as additional leverage that ranks senior to the preferred shares, it further weakens the coverage ratios.

[ADS Analytics]

Common / pref coverage is among the highest in the space at 6.8x - what number do you see? The baby bonds are already in the equity calculation (via liabilities) so there is nothing more to subtract. Interest coverage is a nearly useless metric for mREIT preferreds, unlike for "widget manufacturers" that hold a large amount of marketable assets that can readily be sold in the market.
ADS