American Capital Agency Could Cut Its Dividend

| About: AGNC Investment (AGNC)


AGNC pays a 12.2% annual dividend.

AGNC had a book value of $22.09 per common share at Q1E 2016. The estimate from May 31, 2016 was $22.36 per common share.

The drop in interest rates and other issues discussed herein seem to dictate a significant dividend cut announcement at the Q2 2016 earnings report July 27, 2016.

I hope this article will help readers grow and/or maintain the monies in their portfolio(s).

American Capital Agency Corp. (NASDAQ:AGNC) is a blue-chip mortgage REIT. On May 17, 2016, AGNC appointed Gary Kain CEO. This is an additional title to his CIO and president designations. He is also now on the board. But he is not Chairman. Gary Kain is considered one of the gurus of the MBS world, especially the Agency RMBS world. Since Gary Kain has been effectively running the company for some time, this should not be a big change. It should only allow a proven manager to have more control. That is probably good. However, the recent slide in the yield on the 10 year US Treasury Note may make Gary Kain's job a lot tougher. AGNC currently pays a 12.2% annual dividend.

Usually when the yield on the 10 year US Treasury Note goes down, it tends to be good for AGNC. It usually means that the value of AGNC's Agency RMBS rise along with the fall in rates. The TBA positions (Dollar Roll Income generators) normally do better too. AGNC is able to pocket the dividend payment (from the TBA MBS). Plus AGNC gets to effectively buy back the securities at a price lower than their price at the beginning of the Dollar Roll deal, since they likely gained value as the yield on the 10 year US Treasury Notes decreased. By looking at the following charts, readers can see that the above has generally been the case in Q2 2016.

The one year chart of the 10 year US Treasury Note yield is below.

Click to enlarge

The chart shows the move from 2.27% on December 31, 2015 to 1.77% (-50 bps) on March 31, 2016 to 1.47% on June 30, 2016 (-30 bps). Since then, the yield has moved back up to 1.57% on July 22, 2016 (+10 bps).

I will use the 30 year FNMA 3.5% MBS as an approximation of AGNC's portfolio behavior (see chart below).

This MBS closed at 103.16 on December 31, 2015. It closed at 104.91 (+1.75) on March 31, 2016. It closed at 105.47 on June 30, 2016 (+ another0.56 in Q2). It has since come back a little to close at 105.22 on July 22, 2016. These moves roughly parallel the moves of the 10 year US Treasury Note yield.

How is this bad then? The first reason it is bad is that the annualized net interest rate spread virtually has to go down as the yield on the 10 year US Treasury Note falls, especially if the financing costs are staying the same or growing. For instance, the net interest rate spread (with Dollar Roll Income) fell from 1.38% for Q4 2015 to 1.31% for Q1 2016. One would expect it to be still lower in Q2 2016. One might guesstimate 1.25%; but it could well be much lower. Remember the Dollar Roll Income is likely to be less, as the value of the TBA securities will have risen less. Plus the net interest rate spread without Dollar Roll Income should be less too due to lower interest rates.

On top of that the prepayments generally go up in the spring and summer. Therefore the prepayment costs should be expected to be higher for Q2 2016 than in Q1 2016. They should also be higher year over year, as interest rates were higher in Q2 2015. All of this paints an ugly picture for monies available to pay dividends for Q2 2016 and beyond. The above would also tend to make the swaps move more than the MBS values. That would mean the swaps would be that much more costly hedges for Q2 2016. Remember AGNC is hedging against a rise in interest rates, so the hedges lost both time and principal value in Q2 2016. All told I could see the net interest rate spread with Dollar Roll Income move down as far as 1.10% (possibly further). We will have to wait to see, since I do not have the data on the Dollar Roll trades in Q2 2016. Still "let the buyer beware" ("caveat emptor" -- a Latin saying that goes back at least to Roman times).

Book value has moved approximately the way one would think it would. The book value was $22.09 per common share at the end of Q1 2016. By April 30, 2016 it was up to $22.44 per common share. By May 31, 2016 it was down a bit to $22.36 per common share. Using the FNMA 3.5% MBS as an approximate guide, the value of that has gone from 104.62 on May 31, 2016 to 105.47 on June 30, 2016 (+0.85). This seems likely to lead to a positive gain in book value in the month of June 2016 for Q2. That should make investors happy.

However, the net interest rate spread with Dollar Roll Income almost certainly went down (see reasons above); and it may mean that the $0.20 per common share per month dividend may get cut for Q3 2016 and beyond. Readers also need to take into account that the net long TBA position was only $6.0B on March 31, 2016. This was substantially lower than the average of $8.1B for Q1 2016. If AGNC maintained an average closer to $6.0B in Q2 2016, then Dollar Roll income should be down roughly 25% on that basis alone. Since the 10 year US Treasury Note yield fell a lot more in Q1 (-50 bps) compared to Q2 (-30 bps), readers should also expect less income due to appreciating TBA Agency MBS. The TBA income amounted to $0.15 per common share in Q1 2016. A nearly 30%-50% cut to that income would be $0.05 to $0.08 less per share contributed to the monies available to pay the dividend (the net interest rate spread and Dollar Roll Income).

Using the FNMA 3.5% MBS as a comparison basis, the gain in book value of that MBS itself from Q1E to Q2E was only +0.56 compared to the +1.75 gain seen from Q4E 2015 to Q1E 2016. Hence the Dollar Roll Impact on the net interest rate spread and Dollar Roll Income could be a lot more than I have suggested above. 0.56 is only 32% of 1.75. Looking at the smaller Dollar Roll gains on that basis alone, the loss to Dollar Roll Income might be as high as -$0.10 per common share. Plus don't forget the lower amount of TBA positions that AGNC was long at Q1E 2016.

In addition the net interest rate spread without the Dollar Roll Income was $0.36 per share in Q1 2016. This by itself should go down as the interest rates have decreased. Such losses of interest income should subtract from the Q2 2016 total of monies available for dividends due to Q2 2016. This too argues for a dividend cut for Q3 2016 and beyond. I am not going to try to calculate this effect here; but it could be several cents subtraction from the approximately $0.36 per common share of Q1 2016.

The one thing arguing against the above scenario is that the basis spread contracted in Q2 2016. This would tend to reverse some of the above negative effects. However, I am not going to try to calculate that in this article. I will wait for the company result.

Now remember that the monies available for a dividend were $0.52 per common share in Q1 2016. The dividends declared were $0.60 per common share per quarter (or $0.20 per common share per month). In other words AGNC was running a deficit in Q1 2016 in order to pay its dividend. If you subtract a further roughly -$0.10 per common share (or even more) from the monies available to pay the dividend (the net interest rate spread and Dollar Roll Income), then the resulting $0.42 per common share (or perhaps less) is far below the level necessary to pay the dividend declared in Q2 2016 of $0.60 per common share. That seems almost sure to mean that the dividend will be cut significantly for Q3 2016. As a ballpark figure, one might think it would be cut to $0.45 per common share per quarter (or $0.15 per common share per month). We will have to wait on AGNC's earnings report on July 27, 2016 after the market closes. Still investors may want to have some clue about what may happen before then. That is one reason for this article. I hope the article helps.

The two year chart of AGNC provides some technical direction for a trade/investment.

Click to enlarge

The chart above shows that AGNC bottomed near the beginning of 2016. It has risen steadily since then. This means that the stock price of $19.64 per share as of the close on July 22, 2016 is much closer to the $22.36 book value estimated by AGNC as of May 31, 2016 (about a 12.2% discount to book value). This is still a lot. However, interest rates have gotten so low that there will be even more book value losses when interest rates do eventually rise; and some think a Fed Funds rate hike is back on the table for September 2016. Regardless of the stock price discount to book value, the dividend seemingly has to be cut significantly if management is being responsible. If they are not being responsible, that brings even worse worries into play.

As I have proposed above, the dividend may see a cut from $0.60 per common share per quarter to $0.45 per common share per quarter. That would translate into about a 9.2% annual dividend. That would still be good; but it would seem almost sure to cause some investor selling. For these reasons I rate AGNC a HOLD at this time. I do not know what aggressive traders will do. However, if you want to be safe in the near term you could sell AGNC.

NOTE: Some of the fundamental fiscal data above is from Yahoo Finance.

Good Luck Trading/Investing.

Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.

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.