Recent articles on Seeking Alpha pertaining to mortgage real estate investment trusts [mREITs] focus on their loss in book value [BV] in the 2nd quarter of 2013. These articles ignore or downplay the likely increase in mREITs' dividends resulting from the new interest rate environment. In this article, I analyze the historical dividends of a specific mREIT - Annaly Capital Management (NYSE:NLY) - to quantify how Treasury interest rates affected them over the course of 15 years, and estimate what can be expected in the near future.
The recent loss in the BV of most mREITs made some authors question the viability of their business model, and a few even suggested impending bankruptcies. There is no doubt that a steep rise in long-term interest rates damaged BV and translated into double-digit price drops for many mREITs' shares. However, a bit of historical perspective will illustrate how this rise in long-term interest rates may quickly benefit mREITs by improving their interest income. A rosier picture requires, however, a slower pace in the interest rate increase.
Recent changes in the Treasury yield curve
Since the beginning of May, following the release of the minutes of the April 30th-May 1 FOMC meeting, and especially after the speech by Federal Reserve Chairman Ben Bernanke on May 22, the yield spread has increased significantly from about 2.38% on May 1 to 3.18% on June 28. That's an almost 34% increase in barely 2 months.
Yield spreads and the mREITs' dividends
The business model of mREITs (borrowing money at short-term rates to purchase mortgage-backed securities that earn interests at long-term rates) is expected to make their dividends rise along with the long-term Treasury interest rates if the short-term ones stay put: dividends should increase when the yield spread increases. The questions are: is that generally true, and how quickly and by how much does the dividend adjust after rates change?
The Fed is not expected to increase short-term rates until 2015, and long-term rates are up sharply since the mention in Bernanke's speech of a possible quantitative easing slowdown later this year. On the short term, the BV of most mREITs did take a serious hit (especially agency-only mREITs). However, historically this environment of rising yield spread has indeed been good for mREITs' dividends. To illustrate this, I will look at the dividend history of the biggest mREIT there is, Annaly Capital Management and its $12b market capitalization. Until its recent acquisition of Crexus, Annaly was a pure agency mREIT. It was launched in 1997, and therefore provides a long-term perspective. I will compare NLY dividends with the spread between the yield of the 20-year Treasury bonds and the 3-month bills.
The following figure shows the historical dividend of NLY (normalized and offset to fit on the same plot as the yield spread) and the yield spread as a function of time. I went back to October 1997, and performed a quadratic interpolation between each quarter (NLY distributes its dividend on a quarterly basis). I used the 20-year and 3-month rates because the Treasury Department website provides them for the entire time interval considered, unlike other yields (e.g. the Treasury suspended issuing 30-year bonds for 4 years in 2002). The yield spread curve was smoothed (10-day moving average) to make the plot more readable.Click to enlarge
The relationship between yield spread and quarterly dividend is obvious on this figure. If history is to repeat itself, it follows that the current rate environment should translate into higher dividends. It might take a few quarters, but on the figure the change in dividend following a change in spread does appear to be fairly quick. On average, the quarterly dividend rose by about 4.6 cents in the quarter following a rise by more than 0.75% of the yield spread. The 3rd quarter of 2003 is an anomaly: the dividend dropped from $0.60 (2nd quarter) to $0.28, despite a small rise in yield spread. The likely culprit is a spike in constant prepayment rates [CPR]: from 2nd quarter 2002 to 1st quarter 2004, the quarterly CPRs read as 25%, 34%, 33%, 41%, 44%, 48%, 37%, and 31%. The drop in dividend coincides with the 48% spike. In general, a high CPR decreases the yield of bonds that Annaly has purchased at a premium. Such high CPR rates are unlikely to resurface any time soon.
Will the current interest rate environment last? Bernanke's declarations imply that short-term rates should not change until the end of next year at the earliest. Long-term rates should either increase at a slower rate than these past 2 months or stabilize. Jeffrey Gundlach, head of DoubleLine Capital LP, implied just this when he recently claimed that "The momentum of higher interest rates is slowing."
Another point of interest on the figure is that twice during the past 15 years the yield curve was inverted (short-term rates higher than long-term ones): this is most unfavorable to mREITs. Not only did Annaly not go bankrupt, but it kept paying dividends. The resilience of Annaly in the face of a changing Treasury yield environment proves that a good management team can produce value for shareholders whatever the circumstances. It should provide solace to the long-term investor depressed by the recent share price action.
Correlation between dividends and interest rates
The following figure displays the (interpolated) quarterly dividends of Annaly as a function of the yield spread (20 years minus 3 months). The linear Pearson correlation coefficient is high at 0.864 (where 0 means no linear dependence between the two quantities, and 1 means a perfect one). A higher correlation (0.877) is actually obtained for a yield spread between the 20-year bond and 1-year bill. If the anomalous 3rd quarter of 2003 is ignored, the correlation between yield spread (20 years minus 3 months) and dividend even reaches 0.9! Important too is the fact that this high correlation does not increase when we shift the interpolated dividend curve in time: the dividend adjusts, on average, very quickly (within 1 quarter) following a significant change in yield spread.Click to enlarge
If we look at the interest rates, instead of their spread, we find that the correlation between dividends and 20-year rates is -0.436, while it is -0.825 between dividends and 3-month rates. In other words, Annaly's quarterly dividend has proved much more sensitive to short-term interest rates than long-term ones. This is not too surprising considering that short-term rates have changed much more dramatically over the past 15 years than long-term ones (see the figure below).
Short-term rates deeply impact the cost of funding for mREITs. A negative correlation means that higher short-term rates hurt the dividend. On the previous figure, the impact of several rounds of quantitative easing by the Fed is clearly visible from 2009 onwards.
To summarize, even though NLY's dividend is more correlated with yield spread than with short-term rates, the high anti-correlation with the 3-month rate means that investors should pay close attention to its evolution. On the one hand, it is a good sign considering that short-term rates will likely not increase until 2015 at the earliest. On the other hand, it means that when they do increase, it may create another shock in NLY's share price (lower dividends negatively impact the share price).
Historically, two important predictors of NLY's quarterly dividend have been the Treasury yield spread (20-year minus 3-month) and the 3-month interest rate: lower 3-month rates and, especially, higher yield spread translate within a couple of quarters, at most, into a higher dividend. This dividend was increased, on average, by 4.6 cents in the 3 months following a quarterly raise in the yield spread larger than 0.75%. With the latest dividend of $0.40, that represents an 11.5% raise.
At the current share price of NLY ($12.57 on June 28, 2013), I believe that a lot of bad news is already priced in: the recent sharp drop in BV, the idea that the Fed will one day stop abusing the printing press, the recent dividend cut, the death of CEO Michael Farrell... With a high probability that the yield spread continues rising at a slower pace, or stabilizes, and that short-term interest rates stay close to zero until 2015, Annaly can be expected to benefit from this interest-rate environment and increase its dividend (maybe as early as the next quarter). For long-term investors, I believe the current price is a good entry point (especially now that Annaly is diversifying its business from a pure agency play to a hybrid one).
Moreover, it can be argued that even when the slowdown in quantitative easing starts, the Fed will try to prevent the interest rates from changing too quickly, to avoid further shock in the bond market (like the one we just experienced). A slow rise in long-term interest rates is much less detrimental to BV than a rapid one: typical interest rate swaps (one of the mREITs' hedging tools) offer some protection to the BV against slow changes in rates. Conversely, a rapid rate change may decrease the value of securities held by these mREITs to the point where an offsetting increase in the hedge instruments does not compensate the loss anymore, and margin calls suddenly appear, leaving the company with no choice but to sell some securities at a loss (thus negatively impacting its interest income). If, like me, you believe that long-term rates are likely to rise at a slower pace from now on, while short-term rates remain close to zero for the foreseeable future, then Annaly is in an ideal environment and its share price may have bottomed.
The historical Treasury yields are provided by the US Department of Treasury website, and the historical dividends of Annaly come from the company website.
Disclosure: I am long NLY. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it. I have no business relationship with any company whose stock is mentioned in this article.