100% of our revenue from this article in the first month will be donated to a charity. In February, we will be supporting 2 Blondes All Breed Rescue.
We've had 2 other mortgage REIT articles come out lately and we wanted to create one more. First, so that we could cover several more stocks. Second, so that we could help investors understand the importance of leverage.
We're going to cover some mortgage REITs in this article. We provided a more extensive view of the sector and our top picks in a recent RapidFire update on mortgage REITs for subscribers.
Here are the 5 companies we will be covering in this article:
Source: This screenshot comes from our "Common Share Spreadsheet" tool which is available to all subscribers
For readers who missed our last 2 articles, we covered these 5 mortgage REITs in the article from 2/20/2019:
AGNC Investment Corp. (AGNC)
MFA Financial (MFA)
New York Mortgage Trust (NYMT)
Granite Point Mortgage Trust (GPMT)
Arlington Asset Investment Corporation (AI)
We also covered these 5 mortgage REITs in the article from 2/26/2019:
ARMOUR Residential REIT (ARR)
Cherry Hill Mortgage Investment Corp (CHMI)
Chimera Investment Corporation (CIM)
Capstead Mortgage Corporation (CMO)
Ellington Residential Mortgage REIT (EARN)
How to analyze mortgage REITs
I want to talk about the real fundamentals for a moment. Let's step away from some of the accounting games and simplify the process.
Remember that mortgage REITs are kind of like ETFs. If we remove preferred equity from the equation (functionally, it is another form of debt), we can look at mREITs simply. They are like a highly leveraged bond fund.
To be clear, that is precisely what they are.
The portfolio produces gross interest income and the financing creates an interest expense. If hedging is done through LIBOR swaps, it also creates an interest expense. If it is not done through LIBOR swaps, it creates a "realized gain or loss on derivative" that is not included in any "interest expense" category.
So, how much can the mREIT afford in dividends? Well, if we assume all hedging costs are passed through net interest expense (not even close to realistic), then the answer is the net interest income minus the operating expenses. The portfolio creates net interest income and that income has two places to go: Owners or Workers.
Much like any other business, the cost of workers (be it internal or through an external management agreement) is a direct reduction in profits. In general, internal management tends to be cheaper and have better alignment of interests with shareholders.
Consequently, when I'm evaluating the long-term profitability of a mortgage REIT, I want to look at the operating expenses as one important factor. Generally, that creates a benefit for larger mortgage REITs, but there are other factors that can favor smaller mortgage REITs. For instance, the smaller mortgage REITs can grow book value more rapidly through issuance if they trade above book value. The smaller mortgage REITs can also position their portfolio more intricately, so the benefit of great management should be easier to spot.
How to Think About It
Higher expenses are generally worse. In some cases, a smaller mortgage REIT can get around that with great management. In other cases, they can't. Generally speaking, external management makes it more difficult to get around this because even if the REIT moves over book value, issuing new equity doesn't provide scale on their operating expenses as quickly.
How leverage can change
When we talk about a mortgage REIT issuing new equity, we often assume that it will be invested in the same manner as the prior equity. However, that isn't always the case. In some cases, a mortgage REIT may be investing more heavily in different assets.
We're going to use Dynex Capital (DX) as an example. They recently raised equity which is likely to be invested in agency RMBS and agency CMBS. This capital will be leveraged at a higher level than the existing capital.
Why would they change the leverage? Part of their portfolio is invested in derivatives such as "IO Strips", which stands for "Interest Only". These assets are generally carried at much lower leverage levels than would be used for agency RMBS or agency CMBS.
This can be confusing to some analysts who attempt to simply model the same leverage, the same yield on assets, the same cost of funds, and the same operating expenses onto any new equity.
A real-world example
These ideas of changing "Spreads" may seem theoretical or complex, but they aren't that difficult once you get used to them. Unfortunately, there is very little introductory material on the topic. We have a real-world example from the Dynex Capital earnings call.
The last analyst on the call (comes on about 23:00 minutes) suggests that if spreads come in (tighten), then DX wouldn't be covering the dividend on the incremental shares. This question was so bizarre, management had to ask for clarification on his question.
Asking whether an offering is accretive to earnings is a great step in the analytical process, but this is a textbook example of how it shouldn't be done.
The analyst stated he was using a 224 basis spread on investments and assuming 6.5x leverage. Remember that "224 basis points" is equivalent to 2.24%.
Let's do that math quickly: 2.24%*6.5=14.56%
That would be exceptional, but the math has several flaws:
The math provided offered nothing for the assets financed through equity.
The only spread that sounds like "two hundred twenty-four" would be the metric DX reported as "Adjusted Net Interest Spread" which was 124 basis points. Oops! Note: 1.24%*6.5=8.06%. Add 3.42% (yield on assets) to get 11.48%. Compared to the most recent BV of $6.2, 11.48% would be $.71.
This math assumes precisely no change in net spread or in leverage.
Instead, there is a simpler way to do the math.
Management previously stated (in the same call) that the expected return on new equity invested in agency CMBS was in the range of 12% to 15%, and as of "today" (the earnings call date), it was towards the lower end of this range.
Let's go with 13% since 13% is towards the lower end of that range.
To determine if the dividends on additional shares would be covered, we can simply divide the dividend rate of $.72 by 13%:
The minimum proceeds needed for DX to have enough new equity coming in to cover the dividend on the additional shares would've been $5.54.
There you have it. In one math question which can be done with a pencil and paper, we can easily determine the answer to the question. It is also worth noting that spreads were slightly wider prior to the call (when DX actually issued the new equity), so at that point, it may have been more appropriate to use 13.5% rather than 13%.
It is important to point out the benefit of internal management. Because of internal management, raising equity doesn't raise operating expenses. If operating expenses were tied to equity, then we would need to reduce the expected return by the increase in operating expenses.
A couple questions about mortgage REITs
One of our subscribers recently asked us 2 great questions we would like to cover.
1) Why do you add the yield on assets to the yield obtained after leveraging and hedging? If the REIT used the capital to get leverage, buy those assets and hedge them, where does this additional equity/assets (and the yield on them) come from?
Imagine a REIT with 9x leverage - that means $9 in debt per $1 in equity.
This REIT might have $100 in assets and $90 in debt (using small numbers for example).
If the REIT is earning $3.00 in gross interest income, that would be 3%, calculated as $3.00/$100.00 = 3% Yield on Assets.
If they spent $1.80 on interest expense that would be calculated as $1.80/$90.00 = 2% Cost of Funds.
The spread between the 2 is simply: 3% - 2% = 1%
The leveraged spread would be 1% * 9 = 9%.
However, when we look at the net interest income divided by the equity, we would start with $3.00 in gross interest and subtract $1.80 in interest expense:
$3.00-$1.80 = $1.20
Since we established that equity was $10, this would be $1.20/$10.00 = 12%
This can be referred to as net interest income on book value, or more frequently as "Gross ROE". This comes before any overhead expenses.
Since we know the "Gross ROE" is 12% in this example and the leveraged spread is 9%, we can tell that we still need to add 3% for the difference.
So, why does it work that way? Because the yield on assets is calculated using the average outstanding level of interest-earning assets, in this case, $100.
The cost of funds was calculated using the average level of interest-bearing debt, which was $90. Consequently, when we simply subtract the two to reach the net interest spread, the value we are really finding is: "How much additional net interest income can be generated by each turn of leverage?"
Note: A "Turn of Leverage" refers to taking the debt/equity ratio higher by one. So, for instance, the difference between 8x and 9x is "One turn" or "One turn of leverage"
2) The Book Value includes the value of all assets - both the MBS and the hedges (swaps, futures) on them - right?
Generally correct. There have been some extremely rare cases where book value doesn't correctly capture things, but for all practical purposes, the answer is yes. If something isn't captured effectively in book value, we try to avoid covering that REIT. Note that the hedges might be assets or might be liabilities, depending on the value assigned to them. For instance, if you enter a swap at 2% and rates decline to 1%, your swap would be in an unrealized loss. That would show up as a liability. If rates rose to 3% the next year, it would turn into an asset. Regardless, in both years, it should be reflected in the book value.
We hope these questions helped you better understand how mortgage REITs work. Let's jump into some analysis.
Dynex Capital is one of the more attractive options. Shares are trading around book value again, but DX is one of the few mortgage REITs running agency CMBS. The "C" stands for commercial. These securities can be hedged more effectively against interest rate risk, though credit spread widening can still impact book values. We are very moderately bullish on DX based on the fact that they should outperform most of their peers.
DX announced a share offering late January. The offering should be accretive to earnings, but dilutive to book value. Click HERE for the prospectus. The prospectus included a very thorough preview of Q4 earnings.
DX - Q4 book value
Our models called for a significant decrease in book value per share for almost every residential mortgage REIT. Book value was down about 2% more than projections, which is close enough given widening of spreads in the last quarter. The prospectus stated:
Book value per common share of approximately $6.02 at December 31, 2018 versus $6.75 per common share at September 30, 2018 and book value per common share of an estimated $6.18 at January 25, 2019.
By issuing shares around $5.90, the impact on book value will be negative. However, it won't be significantly negative. We don't start encouraging REITs to be aggressive with buybacks until they see discounts of at least 10% and especially around 15% or greater. When the share price is slightly below book value, the impact of issuing or repurchasing shares is quite small.
DX - earnings vs. book value
In covering mortgage REITs, we often indicate that management should attempt to maximize TER (Total Economic Return). The TER is the combination of the change in book value per share and dividends per share. Many investors confuse it with TSR (Total Shareholder Return) which is the change in the share price plus the dividends per share.
For most mortgage REITs, maximizing book value and maximizing future earnings will always point in the same direction. The reason book value and earnings are tied so closely for most mortgage REITs are the external management agreements. When a REIT is externally managed, raising equity (issuing new shares) leads to higher management fees.
DX - internal management
Dynex Capital has internal management. When they increase total equity, there is no material impact on operating expenses.
When equity increases but operating expenses are roughly flat, the REIT achieves greater scale on their operating expenses. It becomes more efficient. A more efficient REIT warrants a higher price-to-book ratio, so the long-term impact on share price is positive.
DX - managing a portfolio
Ideally, when spreads are widening, management of a mortgage REIT will be purchasing more assets and hedging them. The idea is to simply lock in a larger portion of the investing opportunity when spreads are more favorable for investing (wider). If the spreads tighten again, the mortgage REITs book value climbs because of the wise investing decisions that were made. The following chart from the Dynex Capital presentation demonstrates the idea in practice:
Notice how at the end of Q4 there was a very material increase in the size of their portfolio? That is what we are looking for mortgage REIT managers to do. They sit on a portfolio that is relatively flat in size for a few quarters, then they raise leverage when spreads increase.
The ideal time to be buying new assets was right around the end of the year, as demonstrated in the charts below:
When the lines on those charts move higher, it represents the optimal periods for investing. The idea here is that the mortgage REIT intends to buy the asset and to hedge the majority of their interest rate risk. Consequently, they are simply earning the spread between the two rates. While the ideal time for buying assets was right around the end of the year, DX didn't offer shares until a few weeks after that.
Why would they wait? They wanted to get a better price on the shares they were selling. This is a common practice for most REITs as they want to compare the price of the stock with the yield they can achieve by investing the capital.
Finally, we want to look at how DX shifted their total exposure to interest rates over the quarter:
When a mortgage REIT hedges out the duration exposure, they can run higher leverage and focus on the spread between assets and liabilities. If rates stay wide, net interest income looks great. If spreads tighten, book value rises.
AG Mortgage Investment Trust
AG Mortgage Investment Trust (MITT) delivered some disappointing 4th quarter results (down around 10%, bouncing back about 2% since). They followed the earnings release with a stock offering below the implied recent book value. If they were internally managed, that could be quite positive for operating expenses, but they are externally managed.
We view MITT as currently being overpriced.
Annaly Capital Management
Annaly Capital Management (NLY) trades above our estimate of current book value. They already raised capital in January and again in February. They remain externally managed. Since capital was raised recently, the price could still climb, but the risk/reward on common looks poor compared to their preferred shares.
We view NLY as currently being overpriced. However, we continually keep an eye on their preferred shares. NLY has some of the best preferred shares in the sector and they routinely come into our buy range.
Orchid Island Capital
Orchid Island Capital (ORC) may finally be trading around book value again. Their premium/discount range has been quite wide over the last few years. They've already provided their Q4 ending BV, which disappointed. However, our models include the impact of lower BV. We wouldn't open new positions here.
We recently had a buy alert and a trade alert on ORC:
On 12/17/2018 (5 days after our buy alert), shares of ORC plunged and we sent this to subscribers:
We bought shares of Orchid Island Capital on 12/12/2018. Then we bought shares twice more on 12/17/2018 (today) as prices plunged.
ORC's book value today should be about the same as it was on 12/12/2018. The only difference is that we have a much larger discount to book value. By our estimates, shares are priced at roughly 83% of book value.
Compared to trailing book value (ignoring the fluctuations within the quarter) the discount is larger than 20%. We normally don't highlight a discount to trailing book value, but it is convenient for this chart:
ORC has rarely exceeded this level. We see the blue line near 20% in early 2018, when ORC was facing a substantial book value loss. By our estimates, this quarter doesn't look as bad as that one, but the discount today is wider.
We saw huge discounts from late 2015 through early 2016, however that turned out to be a great buying opportunity. Their peers were also trading much lower. Remember that AGNC Investment Corp. (AGNC) is still trading ABOVE tangible book value per share. ORC is trading around .83, but AGNC is trading around 1.04 based on our most recent estimates of tangible book value per share.
ORC easily had the worst performance in the sector today, which seems to highlight that the shares were mostly held by weaker hands. In an efficient market, AGNC and ORC should trade in very close correlation.
ORC - price chart and breaking down our trades
Below is a price chart of our buy and sell dates:
Source: Seeking Alpha
Here is the final tally of returns on our positions in ORC:
Source: The REIT Forum
Shares were 11.29% of our portfolio when we snapped the image. The gains came in at (rounded) 4.9%, 7.9%, and 11%. In each case (REM) was down during the same period.
That's exceptional alpha.
Western Asset Mortgage Capital Corporation
Western Asset Mortgage Capital Corporation (WMC) has often traded at high price-to-book ratios despite external management and high operating expenses. It currently trades at one of the lower valuation ratios in the sector, but still looks too expensive. The abrupt departure of the Chief Investment Officer in December remains a question mark.
We view WMC as being overpriced.
The mortgage REIT sector has bounced back dramatically since late December when almost everything was on sale. Given the solid bounce, we've sold a few investments since then for healthy profits.
Several mortgage REITs have entered the "overpriced" category. Therefore, we recommend investors be very cautious when deploying capital. Overall, the sector appears almost fully valued.
Slightly bullish: DX
Overpriced: MITT, NLY, WMC
The importance of diversification: While most of our coverage is on REITs with far less than average risk, The REIT Forum still recommends diversifying. We invest the substantial majority of our portfolio in REITs and preferred shares. We suggest that investors choose a maximum allocation using our risk ratings combined with their risk tolerance. Each of our risk ratings connects with a suggested maximum allocation. The maximum allocations generally range from 1% for higher risk options to 6% for our lowest risk choices. By diversifying among these choices investors can build a portfolio with a less volatile value and more consistent dividend growth. We consider mortgage REITs to have a substantial amount of risk. However, several of their preferred shares have dramatically less risk on the fundamentals and lower volatility. Within the preferred shares, there are several great opportunities for buy-and-hold investors.
The REIT Forum is the #1 rated service on Seeking Alpha. We focus primarily on defensive investments with high growth potential. It is our objective to find quality investments at a discount, along with trading opportunities for the more active investors. Most of our research is on companies that are excellent investments over the long term.
Disclosure: I am/we are long AGNCB, AIC, CMO-E, DX, MFA-B, NLY-C. 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.