Mortgage REITs And Interest Rate Risk II: Strategy Shifts By 2 REITs

by: James Shell

I wrote this article a while back giving some analysis of the "quantitative risk" associated with some of the mortgage REITs that are a topic of great interest at the moment.

This industry has blossomed since the 2008 real estate meltdown mainly because of the historically unprecedented situation of large amounts of money coming into the marketplace at essentially zero interest.

Since these companies are borrowing money on the short term and loaning it out over the long term, they need to use various hedging techniques to reduce risk or potentially benefit from an interest rate change. These companies disclose their strategy in their quarterly reports, under the heading "Management Discussion of Qualitative and Quantitative Risk."

No one ever talks about this, but it is extremely important because these funds are basically trading in a commodity, money. The difference in performance from one to another is based to a substantial extent on the management's ability to use leverage to lower borrowing costs, and hedging to protect against risk.

This risk reduction strategy is management's theory about the future, and if we investors know what that is, we can adjust our own investing strategy more knowledgeably.

Company Name Symbol Current Price Div Yield
American Capital Agency Corp AGNC 30.28 16.51
Armour Residential REIT ARR 7.13 18.51
Chimera Investments CIM 2.99 14.71
CYS Investments CYS 13.61 14.69
Dynex Capital DX 9.45 11.85
Invesco Mortgage Capital IVR 17.83 14.58
American Capital Mortgage MTGE 22.04 16.33
Annaly Capital Management NLY 16.6 13.73
Two Harbors TWO 10.46 15.29

We've been working with a little portfolio of the highest-yielding mortgage REITs, which are paying high dividends at the moment and are tempting to even conservative investors because rates of return on other financial instruments such as CDs are unusually low.

In the previous article, we went through the quarterly reports of this group, and developed a risk sensitivity curve for each. Here is how it works: In the "Discussion of Qualitative and Quantitative Risk" the management states the calculated effect on NOI that will occur with various changes in interest rates. These can be plotted so that we can see how the management has set up the company's risk sensitivity.

We can use this information to pick up any ideas about what this group of bright, highly paid money managers is thinking about the future direction of interest rates.

Here is the summary table based on the setup as of 12-31-2011: note that Chimera and Armour have still not issued their fourth-quarter and annual reports.

Percent Change in Operating Income with Changes in Interest Rates
Interest Rate Change, Basis Points
Company +100 75 50 25 -25 -50 -75 -100
AGNC -3.1 -0.4 -2 -13.2
CYS -17 -11.17 -5.61 2.2 4.41 5.52
DX -3.7 -1.6 1.5 2.9
IVR 8.75 15.06 -21 -44.57
MTGE -10.9 -4 -1.9 -11.5
NLY -8.5 -4.89 -1.87 3.6 7.88 11.9
TWO 1.3 0.7 1.6 2.4

For the most part, the interest sensitivity curves for these companies are roughly the same as they were last quarter with two exceptions.

(Click charts to expand)

The first one is AGNC:

In the third quarter (blue line) AGNC had its hedging structure set up in such a way so that there would be a decrease in net income if interest rates went either up or down. In its case, the effect would have been pretty substantial, nearly 10% of NOI would have been lost if rates had changed in either direction.

You can see that at some point in the fall, the management shifted its strategy (red line). It still has the basic philosophy that interest rates will be stable. It has also reduced the overall riskiness of the portfolio, and also, it's now almost completely hedged against a 50 bp rate increase.

The second one, probably even more interesting, is Two Harbors:

In the third quarter, the company's portfolio was set to benefit if interest rates went down. A decrease in interest rates (which in fact actually happened) would result in an increase in NOI. Now, the strategy has changed. The management is telling us that it is expecting some change, but it could be in either direction. The TWO hedging structure (red line) is now set up so that NOI will increase with either a decrease or increase in interest rates.

Why doesn't everybody do this? The biggest reason is that there is some expense involved: These financial instruments, like all forms of insurance, come with some cost. The management has to optimize the cost based on the probability of an event, whether it be an increase, decrease, or even remaining the same. So, the TWO management is tolerating a little higher hedging cost to protect against and/or benefit a little from a change in either direction.

Note: These graphs are not on the same scale. TWO is overall much less sensitive to risk changes than AGNC, a 50 basis point decrease would get TWO an additional 1.5% increase in income but it would cost AGNC 2.5%.

The remainder of the companies did not change very much between the third and fourth quarter:

CYS would potentially benefit from a possible interest rate decrease:

Dynex: Similar to CYS, would benefit from a decline in interest rates.

Invesco: The management has adopted a risker strategy. It has set its hedging strategy to benefit greatly from an increase in interest rates, and it will be costly if it goes in the other direction.

MTGE: Interestingly, MGTE is run by basically the same management group as AGNC, and the shape of the curve was similar for both in the third quarter, but they did not change MTGE's curve. The basic assumption here is that interest rates will stay the same, the effect will be a little less if they go down rather than up, and they are tolerating more risk in this fund, which can invest in commercial and non-agency backed mortgages.

MTGE has a portfolio of about $1.7B compared to AGNC's $54B. You can envision this as a strategy to offer more consumer choice, a more conservative and bigger AGNC plus a riskier, smaller, but possibly more interesting MTGE with basically the same interest rate assumptions.

Finally, Annaly: The management kept the strategy the same, and it will do well if interest rates continue to go down.

One last graph: This was the interest risk distribution graph, which gives a little better idea of the relative riskiness of these companies:

This is a plot of the relative effects of a 50 basis point increase or decrease on each of these companies. I've plotted the Q3 data (blue) and the Q4 data (red) to give an idea of the relative change. The farther away from the center the higher risk and/or possible reward. You can see Invesco at the far lower right, will make a lot of money if interest rates go up, or lose a lot if they go down, and they did not shift too much from quarter to quarter. I've also plotted the major shifts in strategy, the red arrows.

The closer to the center of the graph the less risky, and you can see that AGNC shifted much closer to the middle in the fourth quarter. TWO is now the lone company that occupies the upper right quadrant: the company now anticipates some change in either direction, and they still are not very far from the risk-free center. The only other major shift would be CYS, which actually added a little risk to the hedging structure. The company will get better performance if interest rates go down.

So with the exception of CIM and ARR, we now have a complete picture. A couple of interesting strategy shifts took place in the last quarter, and since all investing is a theory about the future, we can now have a much better idea of where we as investors want to be on the risk/reward graph.

The world is chaotic, and there are no guarantees on anything.

Do with this information what you will.

Disclosure: I am long AGNC, CIM, CYS, TWO, IVR. I still have an open order to go long on ARR on a pullback and also, am working on an exit strategy for CIM.