I've covered Armour Residential (ARR) and other mortgage real estate investment trusts before, and this week I had a private conversation with Armour's chief financial officer, James Mountain.
Armour has been selling mortgage-backed-security holdings -- something that naturally increases its hedge position. The firm is clearly in the position of "battening down the hatches" with its focused on liquidity. That is clearly pacifying its repurchase-agreement counterparties, because the amount of collateral Armour is required to hold -- the haircut -- has not increased materially, while its net hedge position has.
As far as the stock price is concerned, management has told us to watch what they do. Well, I inferred from my conversation with Mr. Mountain that the company is not buying MBS at today's prices, and it is not repurchasing shares. Cash is king there, and while that may help Armour to navigate treacherous waters, co-CEOs Scott Ulm and Jeff Zimmer have not been making open-market purchases of Armour shares after they'd done so during the spring at much higher prices.
So, the following question applies here: If management doesn't think the stock is cheap enough to buy, why should we? We at Portfolio Guru own Armour shares for the dividend, and the one positive factor here is that the firm's prepayment expenses have dropped dramatically. In fact, they've fallen to a near-cycle low of 9% in July, and that trend should continue: This morning's figures from the Mortgage Bankers Association showed another huge decline in refinancing activity across the U.S.
Other things being equal, slower prepays support Armour's earnings and, by implication, its dividend. It also increases the cash flow of the firm and makes the preferred shares more creditworthy. (We own the preferred shares, as well (ARR.A), which today yield 9.6% -- a 14% discount to face value.)
As far as the common stock is concerned, Mr. Mountain did not offer any guidance on the fourth-quarter dividend (Armour pays its dividends prospectively on a monthly basis). The lower asset base will likely offset the impact of lower prepayment speeds.
So let's consider all of this together: the uncertainty of that payment stream; the market's hatred for the mREIT sector; and our own desire to keep receiving Armour dividends in order to fund purchase of other securities, mainly in the energy sector. In view of all of this, I am using an options-writing strategy that I call "harvesting" against the Armour positions of my firm's clients. The strategy is too involved to describe, given the allotted space here -- but, if you would like further details, please contact me here.
This aside, one cannot engage in a reasonable discussion about the mREIT sector without mentioning the biggest mREIT of them all -- the New York Federal Reserve.
The third round of quantitative easing, announced to universal fanfare on Sept. 12, 2012, mandated that the Federal Reserve would buy $40 billion of agency MBS per month. But did you know that, on that date, the New York Fed's system open market account (SOMA) was already holding $843 billion worth of agency MBS?
Yes, the New York Fed has been buying these securities since January 2009, and it has been reinvesting the proceeds from monthly MBS payments into more MBS. In essence, therefore, QE3 really amounted to an acceleration of the program, not a new endeavor. So it builds and builds -- and, as of Aug. 14, the New York Fed's SOMA held $1,299,830,964,000 worth of Agency MBS.
I picked one Fannie Mae 30-year 3.5% CUSIP at random -- 31417CLZ608 -- and, since March 6, the position has declined 7.1%. That's an $11 million drop on what had originally been a $155 million position. The New York Fed currently has more than 31,500 different CUSIPs in its portfolio (I actually looked this up). One could extrapolate, therefore, that it is sitting on an unrealized capital loss on its agency MBS holdings in the neighborhood of $90 billion. That's staggering.
The new catchphrase, then, should be: "Too big to unwind." There is just no way the New York Fed could start selling these MBS into a market that is already wobbling. Some have already moved two steps ahead and looked past the future Fed stimulus tapering to start worrying about the unwinding of the New York Fed's $1.87 trillion in holdings of U.S Treasuries. However, that analysis cannot logically be applied to its MBS holdings.
That, ultimately, is what will support the mREITs. It's also why I see value in their fixed-income securities. It's not about wonderful asset-management skills or an undying commitment to shareholder value. Rather, it's about the boys on Maiden Lane and the feckless, reckless, never-before-tried tactics of quantitative easing.