Given the movement in mREIT share prices, and Annaly Capital Management (NLY) in particular, over the last week and particularly during Friday, many can be tempted to say "oversold" or "undervalued".
Sure, things have taken on a whiff of panic. But is the panic unwarranted? One could point to the price/book value and say "look, it's historically cheap!"
NLY Price / Book Value data by YCharts
There is, however, a problem with this. The price/book value you see above is calculated with the book value as of March 31 2013, and guess what, interest rates between then and now have risen by the largest relative amount ever (calculated as the present 10-year rate over the 10-year rate 3 months ago)!
Now, here is where I need to remind people of two things:
- One is that when interest rates rise, MBS turn into fixed income bonds of an increasingly large duration (the so-called "extension risk"). This happens because as interest rates increase, it makes less and less sense to refinance a mortgage. So prepayment rates slow down considerably. So the duration on the MBS trends towards the stated maturity which can be long as 30 years.
- The other is that when interest rates go up, the prices of fixed income long duration bonds go down.
And then there's a third thing to remind: Annaly, like many other agency mREITs, is simply a levered portfolio of MBS. Hence, at this point having shares on Annaly is the same as having a sliver of the equity of a levered portfolio consisting of long duration fixed income bonds.
In a sense, then, if one were buying a CEF (Closed End Fund) or a mutual fund doing the same as Annaly, one would probably not be willing to pay a premium for its NAV -- which is the same as the book value here -- and under the present circumstances one could probably only be willing to buy such a fund at a discount.
Now, if one were buying such a CEF/mutual fund, one would also have access to an updated NAV. It wouldn't cross anyone's mind to have to painstakingly value the fund asset per asset to know how much the NAV was per share. Yet, when buying here Annaly that's just what an informed shareholder would have to do, with the added problem that he wouldn't have access to an updated portfolio and the number of securities would be in the hundreds.
These problems aren't just theoretical. While we know the book value as of March 31, we also know that MBS have taken on a substantial beating. Take for instance a typical FNMA 30 year MBS with a 3% coupon. This is what has happened to it during the last few months (Source: Mortgage News Daily)
It has gone down in price from around 103 to around 94 over the last quarter. That's a 8.7% drop … but remember, Annaly is levered at 6.3x on its fixed income portfolio (this is just comparing the fixed income portion of its MBS portfolio to its entire book value, the total leverage is even larger). While not all of Annaly's portfolio will be on 30 year MBS with 3% coupons, if it were this alone could slash book value by 54%! Thankfully, Annaly also has other 15-year MBS with different coupons and remaining lives (but still, a 15 year FNMA MBS with 3% coupon also saw a 3.8% drop in price over the same time frame).
In short, it's not a given that NLY is cheap here. We don't know how much book value fell, and the nature of the investment is such that if rates continue higher, the book value can continue falling - indeed, it could continue falling all the way down to zero, much like any other leveraged long duration fixed income portfolio.
Now, saying this could happen is different from saying it will happen. Interest rates can also turn around due either to the pressure from Federal Reserve buying or from renewed worries about economic growth. Indeed, I'd say they're likely to turn around just like that. But that doesn't make this levered bet any less dangerous or any cheaper. NLY fell because that's what made sense - it had to fall and the panic was more than warranted.
We already saw that the last 3 months presented the market with the largest ever increase in yields (on relative terms). We can also see that NLY has mostly had performance that was a mirror of such moves. Below we have the same chart for the period where NLY has been trading (rates in blue, NLY in red). What we see in the chart are bars representing a rolling 3-month return. As we can see, they mostly move in opposite directions (there are many other factors influencing NLY, but clearly how interest rates have done is pretty powerful).
The mREIT business is brilliant
In a way, the mREIT business is brilliant. It's incredibly simple: you gather equity, turn around and lever it using short-term financing to buy long-term MBS. You earn the spread and distribute most of it in the form of dividends, while pocketing huge multi-million dollar salaries.
It is a brilliant business for those running it. It has inherent risk of failure due to the leverage of a fixed income portfolio, but that risk is off-loaded to the shareholders. This kind of business will tend to live on until hit by an interest-rate black swan. The black swan naturally takes his time to arrive, so those collecting the huge paychecks will live extraordinary lives before they ever see the bird. As for those investing in this type of business - you'll have those which reap extraordinary results because they got in early, collected massive dividends and got out, and you'll have those who get hit with the black swan and see their investment be wiped out.
It is thus a brilliant business for those running the show, and mostly a crap shot for those investing in it. Sure, only snake eyes gets you and most of the time the dice will be paying decent dividends. But it's important to understand the nature of the game - although those snake eyes are rare, they do come up once in a while.
The hedging myth
Here, I have to talk about something which comes up often. That is the idea that mREITs can hedge away their risk of failure. This is a myth. The way the mREITs can avoid failure is by selling MBS down and destroying the business model. They can't hedge effectively.
Why? Because if they could, we'd be in the presence of a "free lunch". And "free lunches" are arbitraged away. Thus:
- If the cost of hedging was smaller than the spread an mREIT earns from buying MBS and financing the buys with short-term financing, many players would buy those hedges and lever up more and more since the risk inherent in that leverage would be hedged away. This would put substantial pressure upwards on the prices of the hedges until it was no longer possible to replicate the trade;
- On the other hand, if the cost of hedging was higher than the spread, other players would make up "reverse mREITs" - shorting MBS and investing the proceeds in short term instruments, while selling hedge instruments. Then they'd leverage this over and over again, putting downward price pressure on the hedges until it was no longer possible to replicate the trade.
In short, due to these mechanics, it's likely that the price to hedge the MBS exposure is close to the spread that can be earned from holding them. And the only way to effectively earn the spread as mREITs do, is not to hedge the risk entirely.
This effect is no stranger to the markets - it's how all the markets work and price assets. It's the base of the "arbitrage theory" of pricing. It can be seen in many other instances. Take for instance forex. Many could see the obviousness of borrowing in low-interest currencies like the USD or JPY and then take the proceeds and investing them in, say, BRL instruments (Brazilian Real, paying much higher interest rates). This would be similar to what mREITs do, only on a different asset.
But guess what, you couldn't do that without taking on currency risk. Because if you tried to hedge away the currency exposure, you'd wipe out the spread. Indeed, the forward currency rates you would be using to hedge away exposure are actually calculated by using the interest rates of both currencies. Again, though much more complicated, the same thing ends up happening to the mREITs.
In short, the only way for mREITs to earn the interest rate spread, is for them to take on interest rate risk. Much like the only way to earn the spread in differing currency interest rates, is to take on the currency risk.
Is it doomed, then?
As we arrive here, one would be right in asking if we're all doomed, then, if we invest in these mREITs. The answer, however, is that we aren't -- not necessarily so, at least - but we aren't doomed in the same way that we aren't doomed in throwing those dice. The dice can come up favorably. That doesn't tell us we're clever, though. It just tells us we're lucky.
Having said that, I too find that the most likely course for interest rates is, amazingly, still down. Either because the Fed will continue buying or because the economy will suck enough for those rates to head down again. That makes the mREITs more likely to recover than to go bust.
But you should understand the nature of these beasts. They're a bet, pure and simple, that the black swan isn't just around the corner this time.
Annaly fell as it did because interest rates had the largest (relative) move ever over the course of the last 3 months. This means that while price/book seems cheap, we really don't know where book value stands now but we know that it has fallen substantially over the last 3 months.
In the best of days, buying Annaly is a bet on declining long-term interest rates. A levered bet at that. So what one needs to answer is not whether one is buying cheap or Annaly is oversold. What one needs to answer is whether one really wants to make a levered bet on a long duration fixed income portfolio. Even the dividend yield is irrelevant. What is needed is the answer to that question.
And needless to say, if one doesn't know how a long duration fixed income portfolio behaves under a scenario of rising interest rates, one should have no business buying into Annaly.