We recently put our money where our mouth is buying some mREITs. Nibbling might be the better term since we are still a bit queasy about leverage interest rate risk but valuations and current yield (and volatility) levels became too enticing. In particular, we purchased Annaly (NLY) at around $11 average based on our previous analysis (here and here; ok, blatant page-pump). We also purchased a few other names in the space along with some non-mREITs that seemed to get dragged down with the whole sector sell-off. In particular, we took a stab at owning another mREIT, American Capital Agency Corp. (AGNC), after we determined the equity was trading at a decent discount (> 90%) to where we saw book value even if mortgage rates were to increase another 50 bp to 5%. While, in that scenario, book value would decline more than the increase in interest income, we feel that a 75c/qtr dividend is sustainable going forward. That would be a pretty big decline from the current dividend of $1.05 and represents what we refer to as a steady-state dividend - one that can be held with modest 25 bp /quarter increases in interest rates. The soon to be announced Q3 dividend may be higher but even with $3/annum, and our suggested 12% yield, a $25 price target looks doable.
Shown above are selected interest rate data from the Federal Reserve's H-15 report along with volatility data from the CBOE Interest Rate Swap Volatility Index. One can see the dramatic increase in interest rates that occurred in Q2 when (conventional) mortgage rates increased 90 bps (.9%) but, importantly, 10-year swaps and treasuries moved by less (75 and 65 bps, respectively) and volatility increased. All this led to not only interest rate damage in the mREIT sector but also spread damage. Even those players, like AGNC, that supposedly have a decent hedging program in place, were exposed to the carnage. Interest rate volatility, which is quoted in bp/annum and based on 1yx10y swaptions, increased also, further hurting the MBS basis as many mREIT portfolios lean short interest rate volatility to gain yield. All told, the moves led to some ugly performance. Recently, however, things are a bit different. Rates have gone up, yes. But notice that mortgage rates have not gone up as much as either swaps or Treasuries leading to basis compression (good for mREITs w/hedges). Volatility remains elevated but all indications would imply that mREITs should have more stable hedge performance this quarter. (We know, it ain't over yet!).
According to AGNC's recent 10-Q, at the end of Q2, their NAV would contract 2.8% if rates rose another 50 bp and 5.9% if rates rose 100 bp. Note that is pretty linear indicating decent convexity hedging in place. As we mentioned above, there is reason to believe that hedges were effective this quarter so those numbers are believable albeit curve moves (steepening) will have an effect. End Q2 book value, netting out preferred and fully diluted, was in the ballpark of $25/share. Even assuming a 3% decline in NAV, associated with a 5% mortgage rate, would put book at $24.25 so we see a conservative bottom of $24 per share book value even with a modest 50bp rate increase. Of course rates have only edged up slightly to date and hedges look to be effective so we could see more promising numbers at the end of this month (Q3) if rates hold. Today's 10-year note treasury auction stopping 2.25 bp through the market on top of the $50 bn Verizon bond deal being oversubscribed is indicative, in our opinion, of a top being formed in yields. We might bump up to 3 or 3.25% in 10-year Treasuries in the near-term but we don't foresee a large rate sell-off repeating itself before the end of the FED's ZIRP sometime in late '14 or 2015.
Looking at AGNC's income statement, they earned about $545 mn in interest last quarter with a hefty $5bn principal paydowns (scheduled and prepay). Assuming reinvestment at yields 100 bp higher, they should earn closer to $565mn this quarter. Note this increase in interest income is dwarfed by the Q2 equity decline of $600mn illustrating that portfolio rollover is helpful but limited. We should see about $375mn in income after expenses, tax, and preferred share payments, leading to about 80c per share for the quarter at a 90% payout. This is conservative and what we would term steady-state since it does not include any cap gains/losses. A more modest payout of 75c/quarter would be sustainable even if rates were to increase at a 25 bp /quarter clip - our expectation. Our previous analysis, looking at duration and convexity risk and how its priced by the market, implied a 12% yield should be demanded of such a risky investment, which would imply a $25 pps near book. At its current price of $23, we are trading at a discount to where we see fair value and book, even with conservative assumptions.
AGNC has considerable cash to withstand a further modest book erosion without having to "lock-in" losses by selling investments to cover repo margin calls. For investors willing to accept the 13% yield, with our assumed $3/annum distribution, this is your main risk, notwithstanding mark-to-market risk. Of course, if one believes that interest rates are going to have another upheaval then one should not buy any mREITs or, for that matter, any long-duration cash flows (i.e., most stocks). We are bit more sanguine and we expect that interest rates will rise gradually from here and have more of an impact on the so-called shorter end of the curve in the 2-5 year sector implying curve flattening and a lesser impact on MBS. We also have faith in AGNC's management in navigating these tricky times. It seems that DoubleLine Capital's Jeff Gundlach has warmed up to the sector also and we share his tepid optimism.