We believe the steady-state dividend for Annaly (NYSE:NLY), if interest rates stabilize at current levels, to be in the ballpark of 40c per quarter. Given the leverage, interest rate and convexity risk in the portfolio, a 12% required yield seems acceptable. Putting it together, we see fair value currently at, or slightly above, market prices. However, given the asymmetric and significant tail risk for higher interest rates, we believe investors should be patient and look for NLY to trade at a significant discount. Book value at the end of Q1 was $14.82 per diluted share, which undoubtedly declined in Q2. Our estimate is that book value declined by 75c to $1 so NLY's current price represents about a 6% discount. While that may look attractive, we recommend investors wait until the interest rate dust settles and a better picture develops on NLY's sustainable distributable cash flow. As we discuss, there is a risk that dividends will decline to 30c per quarter before rebounding to our steady state estimate. While we give kudos to management's recent moves (e.g., diversifying away from a pure mREIT strategy), we believe patience is a virtue here until at least Q2's financials are observed.
Recently, there have been a number of interesting articles regarding Annaly Capital Management and the impact that a widening net interest margin (the "spread") will have on valuation. Roughly speaking, the spread is defined as the mortgage rate minus a short-term interest rate and the argument is that NLY, and other mREITs, will take advantage of the widening spread to improve their net interest income and distributable cash flow. However, that process can take quite some time to have a measurable impact as the reinvestment capital must come from retained interest income, realized gains and/or principal repayments from the mortgage portfolio.
In one article, the correlation between the steepness of the U.S. Treasury yield curve (T10 - T2) and the price per share of NLY seemed to imply a current mispricing. In particular the argument made was that either the yield curve would need to flatten or NLY would have to richen. However, correlation does not imply causation and we believe that NLY share price is more related to the level of interest rates. In particular, the steepness of the yield curve historically has a very good correlation with the level of rates. When rates decline, the curve steepens. Hence the "usual" correlation between the yield curve and NLY - a steeper curve implies lower bond yields implies higher bond prices implies higher NLY price. However, as of late, the curve has been steepening as rates have been moving higher breaking this usual correlation. This is because of the market's view that the Fed will "taper" longer maturity bond purchases before they lift the short-term interest rate. Hence, shorter maturity bond yields (2 years and under) have not risen much at all.
This steepening of the yield curve in a rising rate environment (a so-called "bear-steepener") hurts mortgage backed securities (MBS) and hence mREITs given the interest rate risk in their portfolios. A bear-steepener is particularly nasty to the mortgage sector since forward rates, which MBSs price off of, rise even faster than the headline interest rates. Coupled with an increase in volatility, both realized and implied, it is not too surprising that the recent steepening of the yield curve has been a net-negative for NLY and mREITs in general. With the short-term rate pegged by the Federal Reserve until 2015 (their words, not mine) the usual correlation between the steepness of the yield curve and mREIT prices is likely to be the opposite of history.
Based on NLYs recent 10Q, it earned roughly $333 mn in net interest income (net of swap interest paid). It also had about $183 mn in realized gains from the MBS portfolio, which we found interesting since it also had unrealized gains on the swap portfolio (hedges) of $325mn. It is not often that you get to claim gains on both although the secret here is that, even though interest rates rose by about 25 bp in Q1, NLY had $2.9 bn in portfolio gains that were harvestable on 12/31/12, which declined to $1.9 bn as of 3/31/13. Note that unrealized gains (losses) on MBS are not in (GAAP) net income but those on swaps are. So even though interest rates rose slightly in Q1, NLY had plenty of room to harvest past gains. Net of the $183 mn in realized portfolio gains, that were used to support operations and dividends, the portfolio value declined by .8 bn in Q1 when the mortgage rate rose by ~25 bp (.25%). Given mortgage rates rose by about 40 bp since the end of the quarter, we would expect remaining harvestable gains to be minimal at the end of Q2. It is quite possible that NLY harvested further gains before the recent rate sell-off although we suspect, at most, $200 mn in gains were booked (slightly higher than Q2), leaving the MBS portfolio with about 250mn in unbooked gains currently.
Given the interest rate sensitivities shown in the 10Q, it is likely that the total portfolio, including swap hedges, lost about $700mn. We think that could be an underestimate because, while mortgage rates rose 50 bp recently, 10-year swap rates moved less. Thus, the effectiveness of hedging was diminished as a result of the bear-steepening of the yield curve coupled with an increase in implied swaption volatility, Admittedly, we could not find information on NLY's volatility hedging so we can't be sure until we see next quarter's financials. If we are right, however, a net $650mn will be owed on repurchase agreements and assuming dividends remain at $450mn (including preferreds), then a net $1.1bn will be required. Assuming, a similar $333mn in net interest income from Q1 and $200mn in harvested gains, we see a net cash need. Given NLY's substantial cash holdings of $1.8 bn, reverse repos (short-term investment) of $4.9 bn, as well as the likely principal repayments that occurred in the quarter (~18% CPR), there is little to no concern about liquidity. In fact, NLY could up the dividend by returning more equity to shareholders but we doubt it will. That said, if the dividend is maintained, a substantial portion will be a return of capital to investors as opposed to return on capital and a further dividend slice cannot be discounted.
So, the question is really - what is in store going forward? First, we think there will be little to no additional capital gains to be had on its investment portfolio unless rates decline. This means that the dividend will need to be supported by net investment income which will be on the order of 30c per share in the near term based on Q1 results although, as we stated, NLY can probably maintain a 40c dividend per quarter by harvesting gains and returning some capital. This, of course, assumes that interest rates do not go any higher. If the whole portfolio was turned over at today's higher interest rate, and MBS portfolio marked on an accrual basis, 45c per share per quarter would be the sustainable distribution. Given the payout ratio of REITs, we believe that 40c per share is a reasonable estimate for a steady-state dividend. In a recent article, we highlighted that a 12% yield is appropriate for a mREIT whose interest rate risk is amplified, via leverage, and retains negative convexity. Given that, a pps with a low 13 handle looks acceptable as a long-term buy and hold although we are cautious since interest rates are still exceptionally low at this time. We see a 1% increase in yields more likely than a decline. While NLY will probably turn over 10% of its portfolio this year at higher yields (spreads), it will take time for that to seep into dividends and we caution giving too much weight to the higher spreads argument. All in all, we believe that investors should be patient and look to buy into NLY at a deep discount, which in our view would be closer to $11.
Additional disclosure: We are short the mREIT sector via options on basket products (e.g., ETFs)