Despite being an investor in Annaly for years, I have accepted the summary presentation of what affects an mREIT as a given. I have to admit that it was not until I wrote my earlier article that I did spend the time on trying to work out the details myself. Mind you, I did do my due diligence as far as history, disclosures, financials, management, etc., prior to my first buy. Yet, I accepted the common wisdom blindly when it boiled down to what affects a REIT's profitability.
In my prior article, I discussed the not-so-surprising observation that the well-being of the Real Estate market can have some negative correlation with the fortunes of Annaly. I also discussed the (somewhat unexpected) fact that the 30-year treasury seem to have correlation with Annaly's fortunes, as shown in this chart.
You see, what I was taught in the ancient past was that 10-year instruments are used to hedge the 30-year MBSs -- called expected life, hence shielding the mREIT from the 30-year exposure. To confirm, the company's lastest disclosures indicate a mainly repurchase agreement (REPOs) debt liability with an average debt maturity of around 200 days. Interest rate on very short maturities did not really change much since the Fed action began in the wake of the Financial Crisis. The disclosures indicate that the bulk of the assets (MBSs) are fixed rate very long term (20+) maturities.
What I failed to see in the past -- and this article is an attempt to address that failure -- is that the long term-short term interest rate spreads are not the most important factor. Actually the levels of the current market interest rates of each maturity term, and in particular the longer-term market rates, are far more important.
If you are as surprised as I was, let me walk you through the logic. For those who are more analytically inclined, I have included a section at the end of this article that goes over am Excel model, albeit simplistic, of a simple term-swap; that is when shorter maturity debt is exchanged with a longer term one.
As a note, even though my focus is on Annaly (NYSE:NLY), this discussion addresses how to analyze any mREIT, be it Annaly or American Capital Agency (NASDAQ:AGNC). Personally, I now believe that by accepting that spreads are the focus, I failed to protect my investment.
The logic goes as follows. An mREIT works, normally, by using short term debt to buy longer term, higher paying mortgage backed securities ((MBSs)). Depending on the sophistication of the organization, they can engage in different forms of derivative arrangements to either enhance or insure the basic portfolio. The basic portfolio can be simply viewed as being "long" longer-term bonds (MBSs), and "short" short-term debt securities. In Annaly's case, the disclosures indicate that the bulk of their MBS portfolio is fixed rate with 20+ years to maturity and the bulk of their debt is REPOs. Since no financier would pay more for insurance than their main income, we can concentrate on the main portfolio, and, unless it is a gambling organization, we can safely assume for the purposes of this discussion that the impact of the insurance should be a few percentage points of the cashflow.
As bonds, including MBSs, have a certain coupon associated with them, the income the company gets is based on that coupon payment. The spread in interest rate between coupons received for long term and those paid for short term debt - minus hedge or insurance -- gives you the net, or effective coupon payment you receive. Clearly, if the spread between the debt securities narrows, then the income that the mREIT receives will suffer. Conversely, if the spreads widen, the company makes more. This is what we are taught, and it makes perfect sense.
The above is also how I fell into the trap! You see, I forgot to include GAAP and FASB. Accounting standards mandate that losses/gains on equity be included in the total P&L. Not to fault the company, in the above mentioned disclosures they do list all such details. It is that I was fixated on the spread and not actual Fair Market Value of assets and liabilities.
Ignoring the hedge and insurance, proper accounting principles state the equity of the company is not only based on the "real equity" or what we, the investors, paid the company for IPO and secondary offerings plus whatever reserves the company accumulated over the years. Equity also must include obligations - that would be the short term debt and interest rate swaps. Also, equity must include assets - that would be the long term debt purchased.
If interest rates on the short term rise, then your obligations drop in value! After all, your "existing" obligations can be treated as fixed-rate bonds -- read it cashflows, and these drop in value if equivalent market rates rise. Yes, you pay more for the coupons of new issues and debt, but you have just made money on the fair market valuation of existing obligations.
What will happen if long-term interest rates stayed the same and short term rates went down? That will increase the spread to your advantage. Yet, the rise in the fair market price of the short-term debt -- your obligations, will more than offset that. Hence, in the net, you will lose money.
The reverse happens on your existing assets, which are the long term MBSs you bought with your equity and debt. That is, if long-term interest rates rise, even though your receipts on new debt rise, you will end up losing more on the valuation of the assets you already have on your books. Conversely, if long-term rates drop, then you will end up more than making for the loss of new receipts by the gain on the existing assets.
An interesting scenario arises, though, when the short term financing expires, while the longer term one is still on the books. That means, you have to finance new debt at the then current rates. In this situation, the new debt will be "fairly priced" at purchase. Hence, the effect on profitability is due to cashflow (spreads). As we have been discussing how the cashflow due to coupons is dwarfed by the interest rate movements and such effect on fair market value, you can see that cashflow impact is much more manageable than the impact of interest rates on valuation of existing assets and liabilities.
In short, with minor regard to the spread - a very strong statement indeed - your mREIT will mainly make money on dropping long-term interest rates. Yes, widening spreads help, but only if the change of the value of existing - on the books -- assets and obligations do not offset that.
How can a REIT stay in business if they do not renew their balance sheet? If rising new short-term debt is coinciding with a dropping long-term, that will squeeze the new swap margins. In the meantime, their overall "profit," which includes cashflows as well as equity loss and gain, should be blossoming enough to over compensate for such margin drop. Yet, you would expect that they will do less swapping if the incremental margin does not cover cost. That is healthy deleveraging.
To cap the qualitative part of the article, the typical question would be: how would I trade this?
I have gone on record stating that I believe that long term interest rates should be rising to around 2-3% above inflation. This is what the Fed is working hard to achieve, and I believe that in the next couple of years, they should be able to. Unfortunately for the Fed, most of human history is deflationary. Further, the Japanese experiment does not reflect well on quantitative easing. That is, for the longer term, interest rates are going to go back down, possibly close to zero.
As far as my investment in Annaly is concerned, I believe it is going to be a hard environment for the next couple of years. Yet, afterwards, I believe that they will be rewarded well if they manage to build a marginally profitable high-interest rate portfolio. Yes, they will need to manage their redemptions well during that deflationary period that may follow, but dropping interest rates, after the Fed shock is absorbed, should reflect well on the book-value of their existing investments.
Hence, I am not looking to add to my holdings in the near future, despite the highly depressed price. My trigger to add would be a stabilizing interest rate environment, preferably with dropping long-term rates.
To carry this discussion further for those who are analytically inclined, let me walk you through a somewhat simplified model that I created. I assumed that I have a $100 equity on 1/3/2011, the first trading day of that year. At that time 5-year yield (FVX) was 2.02%, while the 30-year (TNX) was 4.4%. I (hypothetically) bought a (fresh) 30-year bond for $100 at that yield, and simultaneously sold $100 of a (fresh) 5-year bond, with my equity as collateral. Effectively, this is how an mREIT works: you finance the purchase of long term, higher yielding debt, with short-term, lower interest rate borrowing.
This image summarizes this simple sheet.
The sheet itself can be found as a link on this page on my personal blog site.
To get to a 5-fold leverage limit, I went ahead and (hypothetically) entered into similar arrangements, at the then current fair yield. This was repeated on the first day of each following half-year. On 1/2/2013, I (hypothetically) purchased the last (fresh) 30-year bond at 3.05% yield, and sold a (fresh) 5-year bond at 0.76% annual yield. This is again how you would expect a REIT to work: as debt matures, they will need to either buy more assets or finance more debt. At this point I would have a net of $5.69/period guaranteed semi-annual coupon payments, and my on-the-books equity would have been $135.14. Yet, that is a net loss from the previous period, mainly due to rising interest rates. Also, note that you would have lost money on 7/1/2013, again due to rising interest rates.
Seven different scenarios for 12/31/2013 are then proposed. I have used the first scenario, which assumes interest rates on that date to be similar to 11/26/2013, as a reference. The last row in the sheet indicates how the other six scenarios would have fared compared to this reference scenario. You will see that the only alternative scenarios that will make you money involve dropping long-term rates or higher short term, even with shrinking margins!
The most important simplification here is that I used only "then Net Present Value" of remaining cashflows in my valuation. Further, I assumed that all bonds are at 100 at the time of purchase.
In effect, as simple as this model is, it strengthens my view that traders of mREITs should make long-term interest rates a primary focus.
Disclosure: I am long NLY. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.