Second quarter earnings are out. Armour Residential REIT (NYSE:ARR) had a strong quarter and an even stronger month of July. According to yesterday's 10Q and July's monthly company update, the main story is book value increased to $26.94 from $25.45 on 05/18/16, a stellar 5.85% increase in a month and a half. ARR's strategy has been to take more risk in hopes of a higher return. This quarter the risk has paid off. In this article, I will compare ARR with American Capital Agency (NASDAQ:AGNC) to see how each of the companies are handling the low interest rate environment and where your money should be invested.
AGNC is more conservative with 100% of investments in agency mortgage-backed securities with an implicit U.S. government guarantee. The conservative investments lead to a smaller interest rate spread of 1.16% leveraged 7.7 times. AGNC states that their hedging strategy's primary objective is not to eliminate interest rate risk or lock in a particular net interest margin, but to maintain net book value within reasonable bands over a range of interest rate scenarios.
ARR, in an effort to maximize returns, has moved outside safer agency MBSs and into non-agency investments resulting in a higher spread of 1.35%, which is leveraged an eye-popping 8.95 times. The market is punishing ARR's level of aggression with a 19.5% discount to book value.
An investment with such great risk should come with a corresponding higher yield. The yield difference between ARR and AGNC is only 1.1%. With a higher spread and higher leverage, ARR should be able to support a higher yield. A back of the napkin calculation of multiplying spread by leverage and dividing by the market price to book (Spread*Leverage/(P/B)), shows that before expenses and hedges, ARR could support a 15.0% yield vs. AGNC's 10.2%. In other words, in a stable interest rate environment, ARR can more than support its dividend, while AGNC cannot.
Personally, I don't take the risk owning the common shares in either. The magnitude of leverage and minuscule spread leave little room for error. An investment with a double digit yield during a time when the ten year treasuries yield is below 2% has substantial risk in an efficient market.
Preferred without the Premium
ARR's risk is not reflected in the forward yield of the common shares, but it is in the preferred shares. Below is a table with data compiled from DividendYieldHunter.
I included Annaly Capital (NYSE:NLY) to compare the market share leader. The current yields of the preferred are similar to the difference in yields of the common shares. The key difference is the yield to worst which reflect the call risk and premiums of AGNC and NLY. AGNC-B and NLY-E are callable in the next couple of years, but each trade at a premium. ARR-A and ARR-B both trade at a discount. If ARR preferreds get called, it would have a positive effect on return. The market is pricing these securities as if they are perpetual. It remains unknown why NLY allows the A series to remain outstanding when its D series and its 7.5% coupon can trade at a premium. NLY should call the A series and issue a new series at 7.25%. It will be revealing to see if AGNC calls its 8.0% P series in 8 months.
It is unlikely that ARR calls its preferreds, but two situations could cause the shares to rally. If ARR changes course and invests more conservatively like AGNC or NLY, the existential threat for ARR would decrease causing the preferred shares to trade in line with its two competitors. For ARR-B, a share price of $26 would be an extra 11% gain on top of the fat yield.
Alternatively, the large discount to book could attract a buyer. Every mREIT is in the same situation, buying low yield assets at a premium to par. Why buy at a premium to par, when you could buy a competitors' entire book at a discount? NLY has recently purchased Hatteras and ARR bought Javelin at discounts. Both were immediately accretive to the acquirer. A buyout by NLY would lead to either redemption at par or shares trading at a premium.
I bought ARR-B with its 8.43% current yield and nearly 12.0% yield to call. I said earlier that I would not buy the common because of the high risk due to leverage. I own the preferred because of the cumulative dividend. ARR cannot pay its common shares a dividend until the preferred shares have received their cut. The worst case scenario I envision for ARR is large losses to book value and a suspended dividend. Eventually, to maintain its REIT status, dividends would be paid out to the common shares and all accumulated preferred dividends would be paid. The long term return of ARR-B should be near its current yield. My ARR-B is locked in a Roth IRA, sheltered from taxes, and set to DRIP. The Yahoo Finance graph below shows why I am uncomfortable recommending the common shares.
Interestingly enough, I am short AGNC common shares as I wrote in a recent article. Buying puts in AGNC is cheaper than shorting and paying the dividend at the current level and limits the downside of the trade. The put option helps offset the credit risk of my large positions in mREIT preferred shares. I would prefer a put option on ARR; however, the options have shorter duration and pricing reflects ARR's higher beta.
For those looking for more information regarding the credit risk of mREIT preferred shares, I point you in the direction of an article I wrote last month.
If you found the information in this article useful, I invite you to follow me.
Disclosure: I am/we are long ARR-B, AGNC-B, AND JAN 18 $20 AGNC PUTS.
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.