Armour Residential (ARR) released its monthly portfolio update this week. Clearly, the month of June was a challenging one for the entire mREIT sector. Mortgage-backed securities prices fell about 7% from the beginning of May to the end of June and have recovered about 25% of that drop. (see chart)
The basic mREIT business model dictates using funds borrowed in the short-term repurchase market as the fuel to make purchases of long-term MBS. The willingness of repo counterparties to lend is a key factor, and that willingness fell during the mini-blowup in the bond market in late June/early July.
Armour's leverage ratio fell from 7.7x in May to 6.1x in June, well below the company's target of 8x to 9x. ARR's total portfolio value went from $23.7 billion to $18.2 billion, and I estimate that $5 billion of the drop was due to divestitures, with about $500 million due to lower values on owned securities.
Some of the liquidations were forced by their repo counterparties. Looking at ARR's list of counterparties, the company's exposure to Bank of America (BAC) dropped by $750 million in one month. Balances with other repo partners dropped much less and the average "haircut" (collateral) that ARR is required to maintain increased only slightly in June to 4.99%. The small amount of the increase indicates that ARR's repo partners were reducing the size of their repo books, not looking for larger guarantees on securities that are guaranteed by Fannie/Freddie/Ginnie to be "money good."
Looking ahead, the company's leverage ratio will be a key driver of earnings and the dividend. Maintaining the monthly dividend of $0.07 will be more difficult from the lower asset base, and we would expect that ARR will steadily "re-lever "as we move throughout the summer and the bond market regains its sanity.
The other key driver of earnings is prepayment speed, and refinancing activity continues to plummet across the U.S. mortgage industry, a strong positive for ARR and the mREITs. The refi index as reported by Freddie Mac fell again this week and now is revisiting levels not seen since early 2011. ARR's portfolio's prepayment profile fell slightly in June to 11%, and I expect that number to continue to decline in the coming months.
Since ARR pays its dividends prospectively, the $0.07 rate will hold through September. We'll find out in early August (when ARR's second quarter earnings are released) how well they predicted the second quarter earnings. They estimated in a press release last month that the company would have "in excess of $15 million in undistributed taxable REIT income "at mid-year. That acts as a cushion for second half dividends, and the slower prepayment speeds will help offset the big decline in portfolio value. The fourth quarter dividend rate will be announced in mid-September -- it's too early to call it now -- but August's portfolio update will be a must-read.
We are seeing how ARR and its mREIT competitors navigate through choppy waters. ARR's performance in June was less than stellar, and they need to convince their repo partners that all is well. ARR bought back about four million shares of stock in May, but haven't bought any since.
As noted in previous columns their management is not in favor of large buybacks, because they necessarily increase leverage. Well, ARR is underleveraged, so that rationale is less valid. At a recent conference Co-CEOs Scott Ulm and Jeffery Zimmer told me to "watch what insiders do" as a sign of confidence as opposed to corporate buyback levels.
They haven't bought any in the open market (there were insider acquisitions in early July related to the vesting of stock grants) since May 29, when the stock price was much higher than its current level. They really need to step up their game and show shareholders that they are committed to ARR.