Mortgage REITs And Interest Rate Risk

 |  Includes: AGNC, ARR, CIM, CYS, DX, IVR, MTGE, NLY, TWO
by: James Shell

When we did the article last week on Invesco (NYSE:IVR), we pointed out that part if not all of the terrible price performance of this fund was predictable, based on the "Quantitative and Qualitative Risk" discussion in its last quarterly report, to the extent that anyone could forecast what was going to happen to interest rates last fall.

Since this information is available for all of the mortgage REITs, today's project is to take a look at more of them, see what can be learned about the sensitivity of these funds to changes in interest rates, and allow a little bit of extra consideration before a prospective investor gets into an investment in this industry.

To update, we have been working with a little portfolio of the highest yielding Mortgage REITs as follows:

Company Symbol Price Yield
DYNEX CAPITAL INC. DX $9.10 11.89%
Click to enlarge

This industry is in focus at the moment because of its high dividend yield. These names are in the business of lending out money for mortgage loans, some residential and some commercial, and because they are classified as REITs they need to distribute at least 90% of their income to shareholders. This is a highly profitable business right now, because borrowing costs are at multi-decade lows. These companies further enhance their profits by leverage, and also by hedging their borrowed funds by participation in derivative markets.

We were picking on Invesco because last fall, when all of the euro chaos happened, long-term interest rates and mortgage rates fell, and the income of this fund was negatively impacted, as was the stock price. Exactly the same thing happened to Chimera.

(Click charts to expand)

Click to enlarge

The sensitivity of each of these companies to changes in interest rates is published in the Qualitative and Quantitative Risk section of each company's 10Q. Here is the one for the 9-30-2011 10Q for Invesco:

Click to enlarge

What it basically says is "if long-term interest rates go down, we're going to end up having much lower earnings." As we pointed out, this is exactly what happened.

This information is available for all of these companies, and I have gone through and collected this data as follows:

Percent Change in Operating Income with Changes in Interest Rates
Interest Rate Change, Basis Points
Symbol +100 75 50 25 -25 -50 -75 -100
AGNC -9.8 -7.7 -9.7 -17.5
ARR 2.57 4.64 -7.16 -20.43
CIM 1.65 1.33 1.02 -5.99 -8.36 -10.8
CYS -15 -9.68 -4.84 0.97 1.94 2.9
DX -4.9 -2.1 2.1 3.5
IVR 9.7 15.79 -20.09 -38.69
MTGE -10.1 -4 -1.6 -7.6
NLY -7.8 -4.68 -1.95 3.13 6.79 10.17
TWO -1.9 -1 2.2 2.4
Click to enlarge

You can see where this would have been good information to have before buying Invesco. The management predicted that if interest rates went down 50 basis points, which they did, Invesco would have a 20% decrease in income, and since income is tied directly to the stock price, the stock price went down that much per quarter for the second half of last year.

The companies also provide the interest rate effects on net asset value:

Percent Change in Asset Value with Changes in Interest Rates
Interest Rate Change, Basis Points
Symbol +100 75 50 25 -25 -50 -75 -100
AGNC -0.8 -0.2 -0.2 -0.8
ARR -0.15 0.15 -0.67 -1.92
CIM -4.7 -3.14 -1.58 1.6 3.23 4.88
CYS -1.8 -1.08 -0.48 0.51 0.87 1.05
DX -1.2 -0.5 0.5 1.1
IVR -0.73 -0.23 -0.17 -0.66
MTGE -1.3 -0.5 0.1 -0.4
NLY -0.2 -0.07 0 -0.02 -0.08 -0.14
TWO -2.3 -1 0.1 0.2
Click to enlarge

This one is less interesting because the changes are really not all that great, and also, the net asset value is not what's driving the stock price, with the possible exception of

It's a little easier to see what is going to happen with a graph. Here's the NOI graph for Invesco:

Click to enlarge

You can see that if interest rates had gone up slightly IVR would have been just fine. The company had its structure of hedging set up to benefit from a slight interest rate increase.

So, if you have a theory about the direction of long-term interest rates, you can pick the fund that is set up to take advantage of it, and if you are correct (big IF) you can optimize the stock price while still taking advantage of the nice dividend stream.

Here's the graph for AGNC:

Click to enlarge

As we were saying the other day, the risk-reduction strategy in this case favors the interest rates remaining the same. If rates had gone either up or down, there would have been a pretty substantial effect on AGNC's net income.

Let's take a look at the others:

Click to enlarge

ARR is set up like Invesco, and will benefit from a slight interest rate increase, but will not suffer quite so much in the event of a decrease.

Click to enlarge

DX is just the reverse. It will benefit if interest rates go down, and suffer if they go up.

Click to enlarge

MTGE is set up similarly to AGNC, which is not surprising since it is operated by the same management group.

Click to enlarge

Two Harbors is set up more like DX but notice that the sensitivity is a lot less: A 1% increase in interest rates would only result in a 2% decline in income, where in the earlier case, DX would go down by 5%. This is important information as well, it tells us that some of these funds are inherently riskier than others.

Click to enlarge

Chimera is more like Invesco: Set up in such a way that a decrease in interest rates would cause a reduction in income and once again, that is exactly what happened.

Click to enlarge

You can see the potential booby trap with CYS too. A 100 basis point increase in interest rates will cause a pretty major 15% decline in income. The management is obviously expecting stable to lower interest rates going forward.

Click to enlarge

Finally NLY: you can see why everyone loves this fund, and why it did well last year. The management correctly forecasted the direction of long-term interest rates, when the rates went down, it was successful.

One more graph: I've plotted the sensitivity to positive and negative interest rate changes on an XY graph. Note that it is not mutually exclusive that a fund that a fund do well or poorly when interest rates go in either direction.

Click to enlarge

Here's how it works: If you want to select a fund that won't change very much no matter what direction rates go in, you want to stick to the two funds closest to the middle, TWO and DX.

You can see that if you think interest rates will go up, you want to select either IVR, CIM or ARR since they're all going to benefit with increases in interest rates, but note that six out of the nine funds will do a little better if interest rates go down, and if nothing else, this is a measurement of the expectations of some highly paid money managers who are in the business of forecasting this. The farther to the upper left the better you will do if interest rates go down, so it is back to NLY and CYS. We remember that AGNC and also MTGE are set up to do worse if interest rates go either up OR down, hence that is why they are in the lower left hand side of the graph, AGNC being the riskier of the two because it is farther from the center.

A couple of other points

Diversification is good. If you own IVR, like I do, you are also delighted that you own CYS, which I also do, to balance out the risk.

Also, these risk sensitivities change over time. The management has some highly trained financial engineers who are always adjusting the portfolio of derivatives to accommodate the next theory about what will happen to long-term interest rates. So, when the next round of 10Q announcements come out in the next couple of weeks, we'll have to update the graphs and see what adjustments were made. Perhaps with rates at historical lows, some of these companies have made some adjustments to the upside.

Also, as they point out, the risk curves are based on the companies' asset structure at the time of the report, and if they add assets, which is what most of them are doing as fast as they can, the curves can and will be different.

So, what to make of all of this

A lot of highly trained people are structuring the system of derivatives of each of these companies to try to optimize the portfolio based on some expectation of what long-term interest rates are going to do. Sometimes this is successful, as has been the case with NLY and a few of the others, and sometimes not so much.

But, we know where we can find this information, and curious investors like us, with a similar theory about the future can make a lot better informed selections if we use tools like this to understand the future effects of unforeseen events.

As we are so fond of saying, the world is chaotic, and there are no guarantees on anything. Also, all investment is a theory about the future. In this case, it is really interesting to know what the professionals are thinking, and we can adjust our strategy on that basis.

Disclosure: I am long IVR, CYS, CIM, AGNC, TWO. I also have an open limit order to go long on ARR on a pullback