When you invest in an actively managed bond portfolio, you are investing in the manager. That goes double when they use leverage and hedging.
Corollary, and a question:
Understanding the basic portfolio strategy is important, but it’s unreasonable to expect a specialist to explain what they spent a career learning in five minutes or less.
So why do investors believe they can or should understand all the details of financial companies before investing, while not holding themselves to the same standard for manufacturing, software, retail, or any other investment?
Now that we have that out of the way, we can begin to understand these “simple” financial companies.
As I pointed out a couple of weeks ago, the opportunity may be coming to buy these portfolios of government-backstopped mortgage-backed securities (MBS) at a discount. In a couple of cases, it’s already true.
Still, there are several risks that hang over these companies.
The business itself almost disappeared a year ago when the short-term repo financing market dried up. Only extraordinary action by the Fed allowed holders of MBS to get their positions financed, and even then, the Fed danced right up to edge of their legal authority by providing repo financing to Wall Street dealers.
As secondary borrowers, the amREITs were at the mercy of the banks and dealers to get their positions financed. The ultimate threat to existence was averted, but every amREIT responded by dropping their leverage ratios substantially. Today they range from 5 times leverage to around 8x.
On the positive side of the coin, over the past year, short rates went to zero, and repo rates followed them down. Today, the quoted financing rate for an MBS holder agreeing to buy the security back a month from now (a repo, or repurchase) can finance their position for just 1.5 basis points (0.18% annual rate). That’s about five basis points less than 1-month LIBOR, which is hovering just under 0.25% per annum.
If you didn’t have to worry about details like rollover risk, price risk, or principal paydown risk, owning MBS that yield 4% and financing them for less than a quarter of a percent seems like a pretty easy way to make money. It would be a license to print money if you could operate at 8x to 12x leverage (like a bank) and fund these things at ridiculously low rates.
In fact, it is so attractive that most banks don’t feel particularly driven to make business loans. Why take the risk? Sure, you could charge 8%, or maybe even more, but why not take the sure thing at 4% and lever it up 10 times?
Several reasons pop into my head for not doing this all we can, although I have to say that the boys at AGNC are looking at the situation and making hay while the sun shines, which is also why they are paying a 22% dividend and trading at a hefty premium to liquidation value. In other words, today, it works for them.
Others have been through a few credit cycles and have reacted at the other end of the spectrum, holding lower-yielding MBS that don’t have nearly as much price risk if the mortgage rate goes up, and also only levering themselves 5x, so the hit to capital from a decline in value isn’t nearly so drastic.
The truly conservative amREITs are even taking out insurance against the mortgage repo rate shooting back up in a Fed tightening cycle. And it’s not a all-or-none thing. You can choose how much of the refinancing risk to hedge, how much price risk to hedge, and if you really want to spend some serious money, how much spread risk to hedge.
That last one, the risk the yield spreads widen apart from other rate moves, is the hardest and most expensive.
Truly being fully hedged against all three of these risks would cost so much for the insurance that you would more than wipe out the positive earnings from holding the MBS.
It stands to reason when you think about it — MBS are a huge efficient market, even today. (There are still more than $5 trillion of them, after all.) The hedge markets, especially the private swap and option contracts that lock in spread, are far less liquid, and each deal is customized between two institutions. Naturally the bid/offer spread is wider, as is the cost of entering the trade or unwinding it.
Among the half dozen amREITs now in the market, the one I wrote about last week (Anworth) is living the truism that “no good deed goes unpunished.” They hedge out virtually all of their funding risk. That means that in 2004 and 2005, when they bought brand-new GNMA 5/1 ARMs, they were going to receive a fixed rate for five years, so they swapped fixed-for-floating, agreeing to pay out a fixed rate to receive LIBOR back each month over those five years. Today they still have some of those swaps, and they’re paying out 3.5% or 4% to receive a piddling 0.24% back. Ouch!
I read one analysts’ work that said there was a $100 million negative present value on all those old hedges at ANH. The good news is that they still have the MBS, so the spread is still positive, though nothing like the 5% or higher spread they’d be collecting if they had gone “naked” back in 2005. Over time, assuming the borrowers don’t refi or sell their homes (not too likely in this lending environment), those swaps will run off and the $100 million in negative present value will come back to their shareholders’ equity.
For those who are still with me, now comes the “cheat sheet”.
Let’s start with a pair of generic amREITs, ones that don’t exactly exist, but have portfolios we can analyze easily to learn how it’s done.
Howard’s investment analysis rule number 1:
If you feed a man some fish, he might enjoy it. If you teach him to fish, he will talk about it endlessly, drink beer with his buddies, and tell the same stories over and over again, enjoying them every time.
The same is true when you do your own stock analysis, as long as you make a profit on the trade, or at least remember it that way. Just think how jealous you can make them ten years from now when you talk about your investment that paid you dividends ten times higher than CD’s. No need to say what CD’s were paying at the time.
RR – my imaginary symbol for Risky REIT.
The equity capital is $100 million. On top of that, they issued $20 million of preferred stock that pays an 8% coupon. [Investors who like to play it safe buy these, and these days don’t have to pay the full $25 face amount, so investor yield is above 8%.]
Let’s say they own Fannie Mae 5’s which trade for 103.75%. Since they finance the MBS at 8-1 leverage, they own $900 million (purchase price) of the MBS, and have a cushion if the price goes down on the portfolio.
By the way, the market only requires them to put up 10% in margin capital ($90 million), so they have a cushion. Basically, by having $120 million
The market trades hundreds of billions in MBS every week, so it’s no big deal to the RR management to buy or sell $50 or $100 million. That literally takes a couple of minutes, and won’t move the price more than a sixteenth of a percent, or $60,000 on $100 million.
RR’s management can’t go completely without hedging, but their main interest is locking in the spread they got by buying those Fannie 5’s. In today’s market, they are getting 179 basis points over swaps. So what they do is hedge half their financing position for the next three years. On that half, they agree to pay 1.76% to receive LIBOR back. Spreading that hedging cost over the entire position, they are paying 0.22% to finance half the portfolio, and 1.76% on the other half, for an average cost just under 1% per annum. Their yield is 3.60%.
Some of you may be asking: Why is the yield so low when the Fannie Mae MBS are paying 5%? Answer: that 3.75% premium in the price is assumed to be paid off over the 3.1 year average life, assuming the mortgages prepay at 13% per year today, on their way to 30% by the beginning of 2011.
Still, at 8x leverage, they are getting 3.6% + 8 * 2.6%, or 24.6% in profit. Even after paying the preferred shareholders their 8%, RR can pay its common stock shareholders 20%+.
OK, so what happens if we wake up tomorrow, and the world hates mortgages, or hated US Dollars or some other investment is pulling investors away by offering higher yields? Or maybe the Fed decided to scare the market into losing its inflation by raising rates (unlikely). No matter how it happens, let’s say that mortgage rates go up 100 basis points.
This is when RR would feel some pain. The first problem is the plain math of higher yields meaning lower prices. A simple 100 basis point move will cause the price to drop 2.80%, based on the duration of 2.8 for that MBS. But wait! There’s more!
If mortgage rates are up 100 basis points, those borrowers can’t be expected to prepay at nearly 30% per year beginning in 2012. In fact, the models for prepayment behavior suggest the terminal prepayment speed will be more like 10.5%. So now the MBS isn’t compared to the 3-year bonds any more. Now it has to compete for money with 7-year bonds. At least they have seven years to amortize the premium, but moving out to the seven year point on the yield curve costs about 165 basis points more in today’s market.
Taken together, that overnight mortgage rate shock would hit the RR portfolio for upwards of 6% of its value, lowering the price to 98.5% or so. What was common equity of $100 million takes the entire 6% hit on $900 million, so half is wiped out, overnight. Their $30 million cushion, and then some, is gone. They will be forced to take some ugly losses to meet the margin calls.
Also, RR is not 50% hedged any more. Now it’s more like 20%, and it can get pretty ugly buying protection against rates rising for seven years, even if the Fed hasn’t moved yet. If all US bond rates have moved, and that’s what caused the problem in the first place, there is a silver lining.
That silver lining is that the 3-year swap was set up when rates were lower (unless ONLY mortgages take the hit), so at least the swap can be sold off at a profit. But the profit is at most 2.8% on half the portfolio, so the hit to capital is still quite large if they take a 6% shellacking on 100% of the portfolio and only have a 2.8% profit on half of it to soothe the wounds.
Just to make life even more exciting around the Risky REIT offices, chances are that margin calls come every half hour or so in a scenario like this. Forget dodging your broker’s phone call while you figure out what to do. If you don’t pick up the phone, they’ll just sell, and send you the bill for whatever the sale doesn’t cover.
The good news is that days like that hardly ever happen. Unfortunately, if they happen infrequently enough, people behave as if they never happen.
With apologies to the classic rock lyricist team Holland/Dozier/Holland, “Mama said they’ll be days like this.”
Some of those days were the day that Lehman died (September 2008), the day that Bear Stearns Mortgage Opportunity admitted they were zombies (July 2007), the day American Home Mortgage was vivisected by margin calls (February 2007), the day LTCM imploded (September 1998), the day the Mexican Peso collapsed (July 1995), the day David Askin got liquidated (March 1994)….
So, back to my back-of-the-envelope cheat sheet.
Let’s look at Conservative REIT (imaginary symbol CR).
They have $100 million of equity, too, but only issued $10 million in preferred stock, also at 8%.
They also buy mortgages in today’s market, but they hold the lower-yielding seasoned ARMs, which they swapped out to capture a spread of only 220 basis points.
On top of that, they only use 5x leverage.
CR’s economics work out like this: Gross earnings are 5 * 2.2% (financed spread)+ 3.2% (yield on ARMs), or only 14.2%. After paying the preferred and overhead, that drops to around 13% for the common shareholders.
The first thing to notice is that ARMs just don’t move around as much in price. In fact, they seem to move about half as far to the upside and half as far to the downside as their fixed rate cousins.
The repo lenders know this, so they only require 7% margin capital if you have some nice seasoned ARMs to finance at a time when they set a 10% minimum capital requirement for those fixed rate MBS.
We call that a shorter effective duration, and it means that when the Shirelles start singing, CR are only taking about a 1% to 1.25% hit on their mortgage portfolio price. They also take the hit six times, not nine.
On their $600 million portfolio, CR takes a $6 to $9 million hit. Painful? Sure. Margin mania? Hardly. They came into the game with a portfolio that only required $600 * 0.07, or $42 million in minimum capital, with the margin clerks. They meet their calls without breaking a sweat, and without selling anything.
Now, here’s where it gets interesting:
If the event is one of those MBS-only events (there were several in that list of disasters), the lenders would be perfectly happy to let CR add new ARM mortgages at the new, wider spread. In fact, they could ramp up to 9x or 10x leverage, if they felt like this was one of those crazy stupid days that bounces back.
Imagine that! CR takes a 6% to 7% (unrealized) hit to capital, and comes out of the mess with net interest margin over 300 basis points, at 9-1 leverage, for a return on equity of — get ready — 30%+.
Their buddies down the street at RR had to sell until they raised $15 million or so in capital. Unfortunately, given the fact that they owed $103.75*0.875 = $90.78 on each MBS, and now the MBS are selling for $98, RR only raises the cash they need by selling roughly $250 million in mortgages, locking in $15 million in losses (15% of capital). The market value of their collateral has declined so far that they only have about $75 million capital to work with, so holding on to the remaining $650 million of MBS is touch and go. On top of that, the preferred shareholders have a senior claim to the tune of $20 million, or $1.6 million a year.
Overall, even though spreads for the MBS they held onto have ratcheted up from 2.6% to 3.6%, the $15 million loss and the lowered equity leaves RR with less than half the income it used to have to distribute to its shareholders.
Good things these days only come along once in a blue moon.