This article was sent to subscribers early on Thursday 03/02/2016, when shares of Western Asset Mortgage Capital opened at about $10.31. As of 03/06/2017, shares are trading at $10.03. The analysis remains accurate.
Western Asset Mortgage Capital (WMC) is a logical short going into earnings. Their report is due today, 3/6/2016, after close. The tiny mortgage REIT has some significant flaws and the fierce beating given to Orchid Island Capital (ORC) a couple weeks ago shows that the lofty market values are still able to contract quickly on a terrible earnings release. With the relatively small size of WMC, it gets very little coverage from analysts. Due to its exceptionally high operating expenses relative to equity for an mREIT primarily invested in RMBS, there is little chance of it delivering strong returns to shareholders. Combine that with shares trading slightly above my estimate of Q4 book value and we have the recipe for a movement lower.
What You Need to Know
WMC has a couple unique factors. Among the most prevalent are their high operating expenses (around 4% to 5% of total equity per year) and their portfolio structure. WMC runs both pay-fixed and receive-fixed swaps, also known as payer and receiver swaps. That in itself can throw off some analysts, but netting the positions is actually quite easy.
However, they also run a significant portion of their portfolio through the TBA (to be announced) markets. Despite their emphasis on clear reporting (on one of their first slides in the presentation), I didn’t see the actual rates on the TBA contracts. Granted, time is limited and their 10-Q is over 100 pages. For reference, on average it takes me 30 to 120 seconds to find this data on most mREITs.
WMC runs a significant amount of positive duration in their portfolio in an attempt to boost net interest income. That is necessary to have any chance of sustaining the dividend because operating expenses + dividends comes out to nearly 17% of the book value I’m projecting for Q4.
For anyone wondering, no, 17% is not a reasonable return on book value (before operating expenses) for mREITs in this environment. Not even remotely reasonable.
For extremely rough ballpark values, we can use management’s indicated exposure to interest rates:
This data was not provided in the investor presentation (it should be), and was relegated to near the very end of the 10-Q. The percentage change in portfolio value is not the same as the change in net asset value. The difference is leverage.
The losses they were exposed on rising rates are surprisingly static. Normally a mortgage REIT’s losses on an interest rate movement would accelerate as the rate movement became larger. For instance, +100 basis points usually results in a loss that is materially more than twice the loss on +50 basis points. That is a function of “negative convexity” you will hear about occasionally. WMC’s calculations don’t show that, instead, we appear to have a roughly linear loss.
During Q4 2016 there was a bit of a twist in the curve so the increases at different parts of the yield curve were not equal. However, we can roughly ballpark the situation as being about a 75 basis point increase. This is a ROUGH ballpark. Based on management’s stated exposure to rate changes of +50 and +100, we can approximate an impact at 1.5% of portfolio value. Let’s simply adjust that for the leverage.
Leverage is reported at 5.2x on balance sheet and 7.1x including TBA. The way they are calculating leverage appears to be:
Consequently, if we want a rough ball park we would need to multiply 1.5% by the 5.24, but we would also need to include another 1 for the value of equity. That brings us to the following ballpark estimates:
Once we adjust for TBA’s being a necessary inclusion and start from 7.1, we add 1 more for equity and it puts us at an estimated 12.15% of BV lost. That sounds brutal.
Starting BV for Rough Calculations
BV at the start of Q4 2016 was $11.48. Multiply by .88 (for a .12 loss) and you would get $10.10. Recent share price is $10.35. This is enough to say WMC would make no sense whatsoever as a long play. If I were only looking for long plays, I would just stop the research here. However, some of my subscribers love a great short play. I’ll finish the analysis for them.
Near the start of the article I mentioned that the contract rates on the TBA packages did not appear to be included in the financial statements. I might have missed them, but I believe it is very rare for me to miss fundamentals on a mortgage REIT. This is my wheelhouse.
Consequently, I estimated the TBA exposure by using the notional value of outstanding contracts and the average price impact across the 30-year fixed-rate FNMAs with coupon rates of 3.0, 3.5, and 4.0. You’ll see these values worked in as I come out to the expected BV per share.
I put together a quicker model for this in excel rather than the using the more complex models I’ve designed. I wanted to use this simpler model because many subscribers have expressed interest in understanding more of the “behind-the-scenes” on how mREITs work. My normal model for valuations runs around 500 rows to track valuation over time. It works pretty well for estimates, but it isn’t great for demonstrations.
Note: I’ll be placing red boxes around the critical values for the calculations.
I normally revalue the entire portfolio with a huge model, but I am running some rougher numbers on WMC. Consequently, I used Treasury rate movements rather than revaluing every LIBOR swap:
That gives us a simple rough estimate for movements. We have a similar chart for the most significant exposures on agency RMBS. This one already incorporates the par value of securities at given rates and maturities.
Since 20-year fixed-rate securities don’t offer the same kind of easy liquidity for measurements, I’m simply modeling the price movement on a 20-year as being 2/3rds of the movement on a 30-year. It is important to point out that perfect precision on mREIT forecasts is not possible. We are simply trying to be in the right general area on an mREIT that gets virtually no coverage from other capable analysts. Note that I didn’t say “no coverage”, just that there is a lack of coverage from analysts that know what they are doing.
I pulled those values from the following slide:
Note that I haven’t boxed in anything on the several other positions. The IO and IIOs built from RMBS, whether they are listed as agency or non-agency should provide some negative duration. However non-agency RMBS and CMBS will also exhibit positive duration. In this case, for simplicity sake, I’m just focusing on the major exposures and allowing the others to net against each other. Normally for revaluing a portfolio I would build those numbers in.
How About Those Swaps
Remember that WMC has both payer and receiver swaps. I pulled the numbers on each:
Values are all in millions. So we are getting ready to find the total impact to equity.
Impact to Book Values
The next table puts together the expected gains and losses:
When we just do the agency RMBS and swaps the rough projection is for $26.14 million in losses. Divided by 41.92 million shares and we would be looking at a loss of $.62 per share. However, we still need to include the TBAs. The projected loss per share from TBAs subtracts another $.71. This may be overkill slightly since the TBAs might have been 15 year securities or they might have been rolled to higher coupon securities. On the other hand, they might also have been focused on the 30-year fixed-rate FNMA 3.0, which would have created losses greater than $.71.
After including the projected TBA loss, we come to $1.34 per share (rounding error present in the numbers). From a starting BV of $11.48, that leaves us with $10.14 if we assume that the dividend was covered. Remember that pulling anything near 17% off a portfolio before operating expenses would be extremely difficult and would require substantial positive duration. Those operating expenses get priority over dividends, so we need to be looking at both dividends and operating expenses as values to be covered with net interest income.
This projection comes an estimate of 11.64% of BV lost. How well does that match with the initial estimate? The initial estimate came in at 12.15%, so we are in the same general area. That gives us enough information to confirm that BV loss for Q4 2016 is probably somewhere in the range of 9% to 14%.
Based on the 11.64% projection, a recent price of $10.35 suggests a price to book value ratio of about 102%.
Here are the calculations for operating expenses. The ratios incorporate my estimate of Q4 2016 equity value:
Consequently, I’m calling for operating expenses in the range of 4% to 5%, which I believe is materially too high. If this were a tiny internally managed mortgage REIT they might be able to get near book value and issue a huge volume of shares which would drive up equity with little increase in costs. WMC has an external management agreement. While I believe expanding equity would reduce the ratio of costs to equity, I don’t believe they would scale nearly as fast as a solid internally managed mortgage REIT.
Rating on WMC
On WMC, I’m establishing a short rating. I believe this is one of the stronger opportunities in the sector.
Pair Trade Option
I recently called off the bullish stance on MTGE Investment Corp. (MTGE) based on the increasing interest rates, but investors that really wanted to move closer to neutral on interest rate exposure could hedge shorting WMC by buying MTGE. MTGE should be running somewhere around an 11% or so discount to book value (too small for me, but about 13% larger than I’m estimating for WMC) and operating expenses for MTGE run closer to 2% compared to around 4% or 5% for WMC.
Whenever I provide a short idea, I want to have clear catalysts for the correction.
Catalyst One - Huge BV Loss
WMC ran heavy on leverage and duration to try to pump up net interest income to cover high level of operating expenses + dividends. I don’t believe investors or analysts are recognizing the size of the upcoming loss.
Catalyst Two – Dividend Cut
WMC declares their dividends late in the quarter. The declaration for Q1 2017 should be coming up in 2 to 3 weeks. A dividend cut would not surprise me at all. It just isn’t reasonable to generate enough net interest income (including alternative factors like drop income from using TBA securities) to cover the dividend plus operating costs. Not at something around 17% of equity value per year.
Further, WMC’s total hedge portfolio had a notional balance of around $1.5 billion compared to around $2.5 billion in financing. If short-term rates move up, the net interest income is set up for a squeeze.
Remember WMC has chopped their dividend substantially a few times lately and is at risk of getting a nasty reputation with investors for consistent cuts.
I see WMC as being materially overvalued and think it is a logical target for shorting going into earnings. If the earnings release doesn’t provide the catalyst, there is a secondary catalyst opportunity within the same month. Investors wanting to take a more market neutral approach could look at using MTGE as the long portion of a pair trade. I’m comfortable with this rating going out as either a single trade or part of a pair trade. I believe WMC is now one of the weakest mortgage REITs and yet they are trading at a premium valuation. It isn’t the 125% type of premium ORC was resting on, but the high operating expenses suggest the fair discount on WMC should be significantly larger.
I believe a reasonable target for WMC would be a decline to around $9.64 with the potential to drop as low as $9.13. Based on the high operating expenses, I think those valuations would still be optimistic for shareholders. I wouldn’t want to pay over about $7.50.
Subscribers already have access to several of my best upcoming pieces. They include two pieces on ARI going into their technique for creating book value and their dividend sustainability report. A piece explaining why I’m so bearish on WMC heading into earnings (that was this one). A real-time update when the bid price on NLY-C was higher than the ask on NLY-E. It was about $.15 to $.19 per share in easy money for anyone selling C and buying E. Besides those pieces, there was also an update to my view on NLY along with my weekly series on discounts to current book value (rather than trialing, I model out all the changes in asset and hedge performance), and the best opportunities in preferred stock for mortgage REITs. If you’re interested in reading my best work before it comes to market, check out The Mortgage REIT Forum.
Disclosure: I am/we are long BMNM.
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.
Additional disclosure: Investors should do their own due diligence and consult with a professional who knows their objectives and constraints. This piece was provided to subscribers with a long enough lead time to allow them to do that.