This isn’t going to be a buy rating.
We’re going to cover Dynex Capital (DX) in this article. We provided a more extensive view of the sector and our top picks in a recent update on mortgage REITs for subscribers. DX still trades at a slight discount to book value, but that puts them within our neutral range. Shares significantly outperformed peers over most of the recent measurement periods. Q3 2019 results were positive. Scott Kennedy covered DX’s Q3 2019 earnings in a recent article to subscribers. Here’s a small section of Scott’s update:
On 10/31/2019, Dynex Capital Inc. reported the company’s earnings results for the third quarter of 2019. The following was DX’s actual versus my previously projected BV as of 9/30/2019:
DX’s Actual BV as of 9/30/2019: $18.07 per common share
My Projected DX BV as of 9/30/2019: $17.50 per common share (range $17.10 – $17.90 per share)
When calculated, DX reported a BV increase of 2.2% while I projected the company would report a decrease of (1.0%). As such, I classify this is a modest (at or greater than 2.5% but less than 4.0% variance) outperformance which was slightly outside my $17.10 – $17.90 per share range. There were two main factors for this outperformance which both stemmed from active portfolio management.
First, as DX was trading at a notable (at or greater than 10%) discount to current BV during the quarter, management was extremely aggressive repurchasing outstanding shares of common stock. During the third quarter of 2019, management repurchased 1.7 million shares of common stock at a weighted average price of $14.65 per share. To put things in better perspective, this was 7% of DX’s outstanding shares of common stock as of 6/30/2019. Simply put, this was a notable amount of share repurchases within just one quarter. This directly equated to $0.24 of BV accretion during the quarter. In comparison, I projected DX would report BV accretion through share repurchases of only $0.05 which still would have been more than what I anticipated from most sector peers.
Second, DX actively managed the company’s risk management strategy during the quarter. DX terminated/cancelled 57% of the company’s existing interest rate payer swaps during the quarter (based on notional value as of 6/30/2019) and entered into new swap positions toward the shorter-end of the yield curve. This served two purposes. First, since interest rate payer swaps with a shorter tenor/maturity have shorter projected cash flows over the life of the contract, this notably benefited the company during the quarter when it comes to reporting less severe valuation losses. Second, DX was able to lower the company’s weighted average fixed pay rate on the company’s interest rate payer swaps. As of 6/30/2019, DX’s interest rate payer swaps had a weighted average fixed pay rate of 2.04%. As of 9/30/2019, DX’s weighted average fixed pay rate declined to just 1.64% Simply put, this was a notable decline for just one quarter’s worth of activity. In comparison, I did not anticipate DX “rotating” over half of the company’s interest rate payer swaps portfolio into more favorable contracts during the quarter. In comparison, I anticipated a rotation of 15%-20%.
Outside those two notable factors, DX’s agency residential mortgage-backed securities (“RMBS”) and commercial mortgage-backed securities (“CMBS”) portfolio performed as expected. DX’s CMBS portfolio has certain prepayment features which are very favorable in the current interest rate environment (purchase price premium protection). As a whole, DX’s agency and non-agency interest only (“IO”) also basically performed as expected – with the preference, once again, currently being in commercial securities.
Next, the following was DX’s actual core earnings vs. my previous projection for the third quarter of 2019:
DX’s Actual Q3 2019 Core Earnings: $0.48 per common share ($0.43 per share Q2 2019 Actual)
My Projected DX Q3 2019 Core Earnings: $0.42 per common share
When calculated, DX reported a quarterly core earnings increase of $0.05 per common share when compared to the second quarter of 2019. This was a very nice “bounce back” when compared to a decrease in core earnings of ($0.10) per common share during the second quarter of 2019 when compared to the first quarter of 2019. In comparison, I projected a quarterly core earnings decrease of ($0.01) per common share. As such, DX’s core earnings also were a positive surprise/outperformance in my opinion. One main factor regarding this outperformance was, as detailed above, DX’s rotation out of over half of the company’s existing less attractive interest rate payer swaps. Other, more minor factors included a slightly larger net dollar roll (“NDR”) income on DX’s to-be-announced (“TBA”) MBS position during the quarter, a proportionately modest decrease in the company’s general and administrative expenses, and a slightly more favorable Eurodollars future position (again active portfolio management) when compared to my projections.
DX’s conditional prepayment rate (“CPR”) on the company’s fixed-rate RMBS portfolio was basically as expected. As of 6/30/2019, the CPR within this specific portfolio 9.4%. As of (9/30/2019), DX’s CPR increased to 13.6%. In comparison, I projected DX’s fixed-rate agency RMBS CPR would be 14.0% as of 9/30/2019. As such, no surprises regarding that metric.
We’ll be talking about some charts. We’re starting with the change in tangible book value since 9/30/2018. For this purpose, we’re simply using the values reported by management from 9/30/2018 and 9/30/2019. That gives us the following chart:
Source: Author calculations
DX delivered better than average results. The book value was down 10.8%. It’s still a loss in book value, but it reflects some solid decisions from management such as using agency CMBS and buying back shares at a discount to book value.
With the first chart giving us a feel for the change in book value, we can move on to evaluating the change in share price (adjusted for dividends). If markets were efficient, we would expect a very strong connection between the two charts. If you’ve been following us for a while, you probably know that the markets will be semi-efficient. You’ll see some correlation, but it won’t be close to perfect. The $100k chart demonstrates how much an investor needed to invest on any day to have $100k today (dividends are reinvested for the model). Below we have the $100k chart for these stocks:
Source: Author calculations
We’ve added some symbols to make it easier to track the most important lines:
DX: Dark blue line with diamonds
You may notice that between 7/1/2019 and 10/1/2019 there’s a period where every line dips lower. That reflects a fall in prices throughout the sector, followed by a recovery. The REITs at the very bottom at that point: Arlington (AI), DX, and Cherry Hill (CHMI).
Consequently, we can say that those three REITs all enjoyed a much larger bounce back since late August / early September. The REITs which bounced hardest often have less upside left. This is one of the reasons we think it makes sense to post a neutral rating for DX.
DX: Dark blue line with diamonds
Throughout the chart, DX’s dark blue line has often been near the bottom. That means investors who bought DX on any of those days earned better returns than if they had bought almost any other stock in the sector. We can see that even on 10/1/2019, DX’s dark blue line was the very lowest in the sector. After being undervalued for most of the prior year, DX finally rallied to trade at a reasonable ratio. Now shares are only at a moderate discount (about 5%) to our estimate of current book value.
This is a great time for investors to be taking gains so they can reallocate into other shares. As of early December, we see Anworth Mortgage Asset Corporation (ANH) carrying a 19% discount and we see ARMOUR Residential (ARR) carrying a 14% discount.
Even though we feel DX has superior management, other mortgage REITs are trading at a materially higher discount compared to DX’s 5% discount to book value.
Another way to evaluate the change is to put the change in book value next to the change in the share price. Use percentages to make it easier for comparisons. That creates the following chart:
Source: Author calculations
We also can see that for DX, AGNC Investment Corp. (AGNC) and Annaly Capital Management (NLY) the cumulative change in book value from 9/30/2018 to 9/30/2019 is extremely similar to the change in share prices.
However, ANH and ARR still have a material disconnect. ANH and ARR are both within our buy range. That can only be rectified by those two REITs either seeing a rally in the share price or a drop in book value per share. We have Scott Kennedy modeling these book values in real time on a frequent basis. Consequently, we can detect changes in book value as they happen. That means we can see how book values should’ve changed from 9/30/2019 to 11/30/2019, even though the mortgage REITs haven’t reported those values yet.
However, if someone tells you they already can forecast the change for 11/30/2019 to 12/30/2019, that person is lying. It’s possible to forecast the values before they are reported, but not before they happen. Consequently, anyone who says the difference for ARR and ANH reflects that they will lose more book value between 11/30/2019 and some future date is out of their minds. We frequently see comments like this from investors who only follow a few mortgage REITs and who don’t carefully track the success of their predictions against the sector.
The REIT Forum is the #1 rated service on Seeking Alpha.
Bought VNQ (REIT index ETF)? Your returns track the red line. Bought PFF (preferred share index ETF)? Your returns track the yellow line. Why index? By carefully picking individual shares, we’ve been able to dramatically outperform the indexes for our sectors.
Try a Free 2-week trial today. You can use your Seeking Alpha account.
Disclosure: I am/we are long ANH, CHMI, AGNCO, AGNCN. 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.