Assessing AGNC Investment's Results For Q2 2021 (Includes Dividend Projection Through October 2021)

Summary

• On 7/26/2021, AGNC reported results for the second quarter of 2021. AGNC reported a comprehensive loss of ($488) million and a non-tangible BV as of 6/30/2021 of $17.39 per common.
• In this article, I will discuss my previous account projections versus actual results. I will also discuss AGNC’s dividend sustainability through October 2021.
• I will also provide my thoughts about AGNC’s MBS and derivatives portfolios as of 6/30/2021 and discuss trends that have occurred during July 2021 impacting the sector.
• AGNC’s BV performance slightly underperformed my expectations while the company’s net spread + dollar roll income modestly exceeded my expectations.
• My current price target and recommendation are stated in the “Conclusions Drawn” section.
• I do much more than just articles at The REIT Forum: Members get access to model portfolios, regular updates, a chat room, and more. Learn More »

deepblue4you/E+ via Getty Images

Introduction/Recap:

On 7/26/2021, AGNC Investment Corp. (NASDAQ:AGNC) reported results for the second quarter of 2021. AGNC reported a net loss of ($411) million, an other comprehensive loss (“OCL”) of ($77) million, a comprehensive (total) loss of ($488) million, a non-tangible book value (“BV”) as of 6/30/2021 of $17.39 per common share, and a tangible BV as of 6/30/2021 of $16.39 per common share. As of 3/31/2021, AGNC had a non-tangible and tangible BV of $18.72 and $17.72 per common share, respectively.

In PART 1 of my prior AGNC I/S article and PART 2 of my prior AGNC I/S article, I projected the company would report the following amounts in relation to the second quarter of 2021: 1) a net loss of ($325) million; 2) an other comprehensive loss (“OCL”) of ($10) million; and 3) a comprehensive loss of ($335) million. I also projected the company would report a non-tangible and tangible BV as of 6/30/2021 of $17.65 and $16.70 per common share, respectively.

As such, as a whole, AGNC’s reported non-tangible and tangible BV was a minor (greater than 1.0% but less than 2.5%) underperformance when compared to my expectations. When looking at all of AGNC’s income statement accounts, this minor underperformance was “scattered” within several accounts. There were no notable surprises within each individual income statement account.

This article will take a look at each of AGNC’s income statement accounts and compare them to my previous projections. Regarding my personal performance, I project/provide all my quarter-end BV fluctuations for the entire mortgage real estate investment trust (mREIT) sector prior to any company reporting quarterly results. Also, once my quarter-end BV projections are “established”, I do not subsequently go back to revise estimates after sub-sector peers start to report (nor should any analyst to remain creditable; some analysts do this which is unfortunate). Simply put, it is not easy to provide this type of accuracy which is why myself, along with Colorado and his team making up the REIT Forum, are the analysts on Seeking Alpha (or any other web-based platform) who provide this type of continuous, labor-intensive service for readers/subscribers on Mortgage REITs.

Within the next section of this article, I will summarize my prior articles’ account projections and compare each amount to AGNC’s actual results. Where applicable, I will also provide an explanation on the variance. I will list AGNC’s accounts in the same order as projected within my income statement and EPS projection article (see link provided above).

AGNC Actual Versus Projected Results:

To highlight my projected account figures versus AGNC’s actual reported results for the second quarter of 2021, Table 1 is provided below. Table 1 shows AGNC’s consolidated statement of comprehensive income (loss) from a three-months ended timeframe. I provide AGNC’s last five quarters so readers can better compare and contrast each quarter’s results.

Table 1 – AGNC Three-Months Ended Consolidated Statement of Comprehensive Income (Loss) (Q2 2021 Actual Versus Projected)

(Source: Table created by me, partially using data obtained from AGNC's quarterly investor presentation slides)

First, let us compare AGNC’s interest income account. My projection for this account was $290 million. In comparison, AGNC reported interest income of $249 million. As such, AGNC’s interest income was a minor underperformance when compared to my expectations. As discussed within PART 1 of my AGNC I/S projection article (link provided above), there are two secondary sub-accounts that makeup this particular account.

The first component of AGNC’s interest income is the company’s cash interest income sub-account. Mainly due to AGNC’s strategy of basically “rolling off” the company’s on-balance sheet agency mortgage-backed securities (“MBS”) portfolio while increasing its off-balance sheet net long “to-be-announced” (“TBA”) MBS position during the quarter, its cash interest income very slightly underperformed my expectations. I projected AGNC would generate cash interest income of $465 million. In comparison, AGNC generated cash interest income of $453 million. When calculated, this was a variance of only ($12) million.

The second component of AGNC’s interest income is the company’s premium amortization, net sub-account. There are two sets of percentages that directly impact the amount of expense for this sub-account. During the second quarter of 2021, AGNC experienced a weighted average conditional/constant prepayment rate (“CPR”) of 25.7% versus 24.6% during the first quarter of 2021. This was fairly close to my projection of 23.5% during the second quarter of 2021.

In addition, as of 3/31/2021, AGNC had a lifetime CPR of 11.3%. With the approximate 20 basis points (“bps”) decrease in rates in 30-year fixed-rate mortgages during the second quarter of 2021, most would agree an adjustment higher was likely (especially since the lifetime adjustment was nearly (15%) below the actual weighted average percentages as of 3/31/2021). Keeping this in mind, I projected AGNC would estimate a lifetime CPR of 12.5% as of 6/30/2021. A key assumption here is that I believe mortgage rates/long-term U.S. Treasury yields will gradually move higher over time. As such, this will “tamp down”, to some extent, lifetime CPR forecasts. In comparison, AGNC increased their lifetime CPR to 11.6% as of 6/30/2021.

However, with some changes in overall MBS/investment portfolio composition, this ultimately led AGNC to report a combined premium amortization expense of ($204) million for the second quarter of 2021 (including a ($70) million “true up” expense adjustment on the lifetime forecasted CPR). In comparison, I projected AGNC would report a combined premium amortization expense of ($175) million for the second quarter of 2021. When calculated, this was a minor variance of ($29) million. When both sub-account variances are combined, this calculates to the ($41) million underperformance within AGNC’s interest income account.

Second, my projection for AGNC’s interest expense account was $20 million. In comparison, AGNC reported interest expense of $17 million. I consider this basically an exact match. When calculated, this was a decrease of ($12) million when compared to the first quarter of 2021. Over the trailing twelve-months, AGNC’s interest expense account has decreased from $134 million to just $17 million with a notably less decrease in portfolio size. When calculated, this was an (87%) reduction in interest expense over the course of only one year (which was previously correctly projected). AGNC continues to benefit from the continued decrease in repurchase agreement rates. The relationship between repurchase loan rates and the London Interbank Offered Rate (LIBOR) was originally discussed within my income statement projection article (see links provided above).

For readers mainly focused on net spread metrics, AGNC reported net spread + net dollar roll (“NDR”) income (when excluding any “catch up” premium amortization) of $399 million or $0.758 per common share for the second quarter of 2021. When compared to the prior quarter, this was a decrease of ($0.006) per common share. In comparison, I projected AGNC would report net spread + NDR income of $360 million or $0.685 per common share. As such, I would classify this as a modest outperformance. I would also point out the institutional analysts’ consensus average for this specific metric was only $0.655 per common share for the second quarter of 2021. Simply put, a notable outperformance when compared to the consensus average. For 9 consecutive quarters, my/our projection has been more accurate versus the institutional analysts’ consensus average. A detailed “breakout” of this calculation/metric was provided within PART 1 of my I/S projection article (link provided above).

This mainly gets back to AGNC heavily utilizing the company’s off-balance sheet net long TBA MBS position during the quarter (a bit more than I anticipated). As a direct result of this strategy, AGNC had another strong quarter of NDR income. Last quarter, AGNC reported elevated NDR income of $154 million. Price drops (forward discounts) on TBA MBS remained attractive across most lower coupons during the quarter. That said, valuations on generic TBA MBS notably decreased within these coupons during the first quarter of 2021 (as previously shown in past AGNC income statement projection articles). As such, I assumed AGNC wanted to remain a bit more cautious with the company’s net long TBA MBS position in case interest rates continued to net increase during the second quarter of 2021. Lowering this position, in a rising interest rate environment, would ultimately lower the severity of losses. So, I projected a minor net decrease regarding AGNC’s net long TBA MBS position from $24.8 billion as of 3/31/2021 to $23.5 billion as of 6/30/2021. However, AGNC reported a larger net long position of $27.6 billion as of 6/30/2021. I projected NDR income of $135 million during the second quarter of 2021 while AGNC was able to generate NDR income of $162 million (from the larger net long TBA MBS position; including some compositional changes). The remainder of the net spread and dollar roll income variance was due to a combination of very slightly higher weighted average coupons (due to on-balance sheet asset composition) and very slightly lower current period hedging costs (by only $6 million).

Third, my projection for AGNC’s sales on investment securities account was a net realized gain of $50 million. In comparison, AGNC reported a net realized gain of $25 million. Considering AGNC’s agency MBS/investment portfolio was valued at $65.5 billion as of 3/31/2021,a ($25) million variance within this account should be considered a very minor underperformance (nearly an exact match).

Fourth, my projection for AGNC’s derivative instruments and other securities account was a net loss of ($670) million. In comparison, AGNC reported a net loss of ($618) million. With that said, AGNC reported a “miscellaneous” derivative net valuation gain of $62 million which was directly tied to a portion of the company’s non-agency MBS sub-portfolio which were considered off-balance sheet. This is the first time since I have covered AGNC the company had “forward-settling” non-agency investments at quarter-end. This caused a notable decrease (proportionally speaking) in the fair market value (“FMV”) of this specific sub-portfolio as of 6/30/2021 when compared to 3/31/2021.

However, that decrease is a bit deceiving because these investments will soon be on the balance sheet. I actually projected/assumed this portion of AGNC’s MBS/investment portfolio would be valued within the company’s on-balance sheet MBS/investment portfolio (again, that account is discussed later in the article). As such, I believe, for comparative purpose, this non-agency MBS valuation gain of $62 million should be reversed from AGNC’s derivative instruments and other securities account. Again, this is strictly for comparative purposes to match to my projected figure. When AGNC’s non-agency MBS off-balance sheet valuation gain of ($62) million is reversed, the company would have reported a derivative instruments and other securities net loss of ($680) million. This adjusted variance is only a ($10) million underperformance when compared to my expectations. Due to the complexities involved within this particular account when it comes to valuation fluctuations, there is typically a larger variance when discussing AGNC’s derivatives portfolio versus most other accounts. As such, I would consider this basically an exact match.

I would also point out valuing AGNC’s hedging portfolio involves projecting four material derivative sub-accounts (currently TBA MBS, interest rate swaps, interest rate swaptions, and U.S. Treasury securities) and several other minor derivative sub-accounts. In addition, AGNC’s derivatives portfolio had a combined net notional balance of ($73.7) billion as of 6/30/2021 (excluding TBA MBS). While no one has a “crystal ball” per se regarding future events, being able to project all these derivative sub-accounts, before any sector peer provides quarterly results, takes a great deal of expertise in my opinion. This includes fully understanding how to value all these derivative instruments and correctly deciding on specific assumptions that one believes coincided with management’s overall risk management strategy during any particular quarter.

I projected AGNC would report a quarterly net valuation gain (loss) of $335, ($395), ($300), and ($310) million regarding the company’s TBA MBS, interest rate payer swaps, interest rate payer swaptions, and U.S. Treasury securities, respectively. These projections were provided in the linked article above. In comparison, AGNC reported a net valuation gain (loss) of $396, ($400), ($313), and ($363) million regarding the company’s TBA MBS, interest rate payer swaps, interest rate payer swaptions, and U.S. Treasury securities, respectively. As such, AGNC’s interest rate swaps and swaptions were basically an exact match, the company’s TBA MBS were a minor outperformance (from a larger net long position; discussed earlier), and its U.S. Treasury securities were a minor underperformance (from a slightly larger net (short) position during the quarter). Additional discussion regarding AGNC’s derivatives portfolio will be provided later in the article.

Fifth, let us discuss AGNC’s on-balance sheet MBS/investment portfolio. Continuing to use Table 1 as a reference, regarding AGNC’s “unrealized gain (loss) on investment securities measured at fair market value (“FMV”) through net income, net” (see boxed blue reference “5a”) and “unrealized gain (loss) on available-for-sale (“AFS”) securities, net” (see boxed blue reference “5b”) accounts, I projected the company would report a combined net realized/unrealized gain of $40 million. In comparison, AGNC reported a combined net realized/unrealized loss of ($105) million. Due to the sheer size of AGNC’s on-balance sheet MBS/investment portfolio as of 3/31/2021 and 6/30/2021 of $65.5 and $59.8 billion, respectively, I believe a combined ($145) million variance is a minor underperformance (considering the “rotation” of assets through specific specified pools and coupons).

In addition, remember the $62 million off-balance sheet non-agency MBS miscellaneous valuation gain AGNC recorded within the company’s derivatives portfolio (discussed earlier). When that specific gain is reclassified to this account (which is where I projected this allocated gain would occur), this net variance is reduced to ($83) million. Furthermore, when considering the $61 million off-balance sheet TBA MBS outperformance in AGNC’s derivatives portfolio (discussed earlier), in actuality AGNC’s total/combined MBS portfolio (both on- and off-balance sheet positions) underperformed my expectations by only ($22) million. This is basically an exact match. A detailed analysis regarding AGNC’s fixed-rate agency MBS portfolio is provided in the next section of the article.

Finally, my projection for AGNC’s compensation expense (formerly management fees) and operating expense accounts was $16 and $9 million, respectively. In comparison, AGNC reported compensation expense and operating expenses of $12 and $10 million, respectively. As such, when combined, nearly an exact match (not too concerned of the “placement” of expenses between the two accounts). Now, let me briefly highlight some quarterly compositional changes that occurred within AGNC’s MBS and derivatives portfolios.

MBS Portfolio Considerations:

AGNC slightly-modestly decreased the company’s on-balance sheet MBS portfolio while slightly increasing (proportionately speaking) its off-balance sheet net long TBA MBS position during the second quarter of 2021. AGNC’s non-tangible “at-risk” (total) leverage increased from 7.3x as of 3/31/2021 to 7.5x as of 6/30/2021 (both figures slightly higher regarding tangible leverage ratios). This fractionally higher leverage was mainly due to the decrease in the value of AGNC’s combined investment and derivatives portfolio (lower equity valuation directly equates to higher leverage). To show the underlying compositional changes to AGNC’s combined on- and off-balance sheet fixed-rate agency MBS portfolio during the second quarter of 2021, Table 2 is provided below.

Table 2 – AGNC Fixed-Rate Agency MBS Portfolio Quarterly Compositional Changes (6/30/2021 Versus 3/31/2021)

(Source: Table created by me, including all calculated figures and percentages)

Using Table 2 above as a reference, when comparing AGNC’s portfolio as of 6/30/2021 versus 3/31/2021, the company had a net par value increase (decrease) in its 15-year fixed-rate agency MBS holdings with a 1.5%, 2.0%, 2.5%, 3.0%, 3.5%, and 4.0% coupon of ($2.1), ($4.2), ($0.1), ($0.1), ($0.1), and ($0.1) billion, respectively. When all 15-year fixed-rate agency MBS holdings are combined, this was a quarterly net par value decrease of ($6.7) billion. AGNC had a combined net par value decrease in the company’s 20-year fixed-rate agency MBS holdings of ($0.1) billion. AGNC had a net par value increase (decrease) in the company’s 30-year fixed-rate agency MBS holdings with a 2.0%, 2.5%, 3.0%, 3.5%, 4.0%, and 4.5% coupon of $1.0, $7.0, ($0.2), ($1.3), ($1.5), and ($0.9) billion, respectively. When all 30-year fixed-rate agency MBS holdings are combined, this was a quarterly net par value increase of $4.0 billion (rounded).

As I have highlighted in various prior mREIT articles, typically fixed-rate agency MBS portfolios with higher coupons “mitigate” the severity of valuation losses in a rising interest rate environment. This strategy also partially offsets any notable rise in borrowing costs. This trend occurred throughout 2018 and most of 2019. However, typically fixed-rate agency MBS portfolios with lower coupons “enhance” the amount of valuation gains in a decreasing interest rate environment (generally less prepayment risk when all other characteristics are held constant). As shown in Table 2, AGNC decreased the company’s 15-year fixed-rate agency holdings with a 1.5% and 2.0% coupon while increasing its exposure to 30-year fixed-rate agency holdings with a 2.0% and 2.5% coupon. In a rising interest rate environment, this strategy would generally lead to less severe MBS valuation losses. In a falling interest rate environment, this strategy would generally lead to less enhanced MBS valuation gains. Again, each quarter is “unique” in its own way so these are just general tendencies (more for educational purposes). Now let us analyze AGNC’s derivatives portfolio as of 6/30/2021.

Derivatives Portfolio Considerations:

During the second quarter of 2021, AGNC maintained the company’s very high hedging coverage ratio. Still, to show the compositional changes to AGNC’s derivatives portfolio during the second quarter of 2021, Table 3 is provided below.

Table 3 – AGNC Derivatives Portfolio Quarterly Compositional Changes (6/30/2021 Versus 3/31/2021)

(Source: Table created by me, partially using data obtained from AGNC's quarterly investor presentation slides [link provided below Table 1])

Using Table 3 above as a reference, AGNC had a hedging coverage ratio of 98% as of 3/31/2021. AGNC’s hedging coverage ratio “trickled” down to 97% as of 6/30/2021. When I projected AGNC’s quarterly valuation fluctuations within the company’s derivatives portfolio, I assumed management would slightly decrease its hedging coverage ratio to 90% as of 6/30/2021. This mainly consisted of a lower net (short) U.S. Treasury securities position. This directly led to some additional net valuation losses within this derivative sub-account when compared to my expectations (as discussed earlier).

As I have highlighted in various prior mREIT articles, typically a derivatives portfolio with a higher hedging coverage ratio mitigates the severity of BV losses (or enhances BV gains) in a rising interest rate environment. However, this particular strategy is detrimental to BV in a declining interest rate environment. The second quarter of 2021 was a great example of this statement.

This past quarter, clearly AGNC’s very high hedging coverage ratio led to some severe BV losses. That said, I anticipate AGNC will continue to maintain a more “elevated” hedging coverage ratio over the foreseeable future (75%-100%) as the risk of rising mortgage interest rates/U.S. Treasury yields over time remains at least modest. In July 2021, this strategy has not paid off.

Conclusions Drawn:

Readers have continued to request that I provide these types of “update/follow-up” articles showing how my previously disclosed quarterly projections “stacked-up” to AGNC’s actual results (continue to be the only contributor/team to provide this type of projection analysis/insight via either subscriber-based or “free to the public” articles). I believe the analysis above accomplishes this request.

When all applicable accounts are combined, AGNC reported the following investment/derivative instrument net valuation fluctuation, net funding expense, net loss, and comprehensive loss figures versus my projections:

My Projected Quarterly Net Valuation Fluctuation within AGNC’s Combined Investment and Derivative Instrument Portfolios: ($580) Million (Shown in Bottom Part of Table 1)

AGNC’s Actual Quarterly Net Valuation Fluctuation within the Company’s Combined Investment and Derivative Instrument Portfolios: ($698) Million (Shown in Bottom Part of Table 1)

My Projected Net Funding Expense (Repurchase Agreement and Other Debt + Net Pay Rate on Interest Rate Swaps): $45 Million (Shown in Bottom Part of Table 1)

AGNC’s Actual Net Funding Expense (Repurchase Agreement and Other Debt + Net Pay Rate on Interest Rate Swaps): $36 Million(Shown in Bottom Part of Table 1)

My Projected Quarterly Net Loss: ($325) Million (Earnings Available to Common Shareholders of ($0.67) Per Share)

AGNC’s Actual Quarterly Net Loss: ($411) Million (Earnings Available to Common Shareholders of ($0.83) Per Share)

My Projected Comprehensive Loss: ($335) Million (Comprehensive Loss Available to Common Shareholders of ($0.69) Per Share)

AGNC’s Actual Comprehensive Loss: ($488) Million (Comprehensive Loss Available to Common Shareholders of ($0.97) Per Share)

When including projections within AGNC’s equity section of the balance sheet, this ultimately led to the company reporting a non-tangible and tangible BV of $17.39 and $16.39 per common share versus my projection of $17.65 and $16.70 per common share, respectively. As such, AGNC’s quarterly BV decrease was a minor underperformance when compared to my expectations and was well within my projected skewed range (skewed to the downside).

Moving to the third quarter of 2021 (through 7/27/2021), I am projecting AGNC’s BV has fluctuated (3.0%) – (1.0%) as a continued negative relationship between MBS pricing and derivative instrument valuations has occurred in July 2021. This excludes the upcoming accrual of AGNC’s July 2021 dividend of $0.12 per common share.

Moving to dividend metrics, I currently believe AGNC will declare the following monthly dividends through October 2021:

Dividend for August 2021 – October 2021 (Paid in Each Subsequent Month): $0.12-$0.13 Per Common Share (90% Probability)

When compared to last quarter’s AGNC assessment article, this range tightened by $0.01 per common share and the probability remains unchanged. Worst-case, AGNC would declare a stable dividend per share rate of $0.12 per common share during 2021. When it comes to the mREIT sector, when the worst-case scenario is merely an unchanged dividend per share rate, investors should interpret that as a pretty good/bullish situation.

My BUY, SELL, or HOLD Recommendation:

From the analysis provided above, including additional catalysts/factors not discussed within this particular article, I currently rate AGNC as a SELL when I believe the company’s stock price is trading at or greater than my projected non-tangible CURRENT BV (BV as of 7/30/2021), a HOLD when trading at less than my projected non-tangible CURRENT BV through less than a (10%) discount to my projected non-tangible CURRENT BV, and a BUY when trading at or greater than a (10%) discount to my projected non-tangible CURRENT BV. These ranges are unchanged when compared to my last AGNC article (PART 1 of this analysis).

Therefore, I currently rate AGNC as a HOLD.

As such, I currently believe AGNC is appropriately valued from a stock price perspective. My current price target for AGNC is approximately $17.20 per share. This is currently the price where my recommendation would change to a SELL. The current price where my recommendation would change to a BUY is approximately $15.50 per share. Put another way, the following are my CURRENT BUY, SELL, or HOLD per share recommendation ranges (the REIT Forum subscribers get this type of data on all 20 mREIT stocks I currently cover on a weekly basis):

$17.20 per share or above = SELL

$15.51 - $17.19 per share = HOLD

$13.81 - $15.50 per share = BUY

$13.80 per share or below = STRONG BUY

Along with the data presented within this article, this recommendation considers the following mREIT catalysts/factors: 1) projected future MBS/investment price movements; 2) projected future derivative valuations;and 3) projected near-term dividend per share rates. These recommendations also consider the eight Federal (“Fed”) Funds Rate increases by the Federal Open Market Committee (“FOMC”) during December 2016-2018 (a more hawkish tone/rhetoric when compared to 2014-2016), the three Fed Funds Rate decreases during 2019 due to the more dovish tone/rhetoric regarding overall monetary policy as a result of recent macroeconomic trends/events, and the very quick “plunge” in the Fed Funds Rate to near 0% in March 2020. This also considers the previous wind-down/decrease of the Fed Reserve’s balance sheet through gradual runoff/partial non-reinvestment (which began in October 2017 which increased spread/basis risk) and the prior “easing” of this wind-down that started in May 2019 regarding U.S. Treasuries and August 2019 regarding agency MBS (which partially reduced spread/basis risk when volatility remained subdued). This also considers the early Spring 2020 announcement of the start of another round of “quantitative easing” (“QE”) that includes the Fed specifically purchasing agency MBS (and “rolling over” all principal and interest payments into new agency MBS) which should bolster prices while keeping long-term/mortgage interest rates near historical lows (which lowered spread/basis risk for quite some time when volatility remained subdued). This also includes the eventual “taper” of the Fed’s most recent QE program regarding its monthly purchases of $80 billion of U.S. Treasury securities and $40 billion of agency MBS. This taper will likely begin to occur in late 2021-early 2022 and market speculation around this future event has already caused a recent rise in spread/basis risk.

Each investor's BUY, SELL, or HOLD decision is based on one's risk tolerance, time horizon, and dividend income goals. My personal recommendation will not fit each reader’s current investing strategy. The factual information provided within this article is intended to help assist readers when it comes to investing strategies/decisions.

Current/Recent mREIT Sector Stock Disclosures:

On 3/18/2020, I once again initiated a position in AGNC at a weighted average purchase price of $7.115 per share (large purchase). This weighted average per share price excludes all dividends received/reinvested. On 6/2/2021, I sold my entire AGNC position at a weighted average sales price of $18.692 per share as my price target, at the time, of $18.65 per share was surpassed. This calculates to a weighted average realized gain and total return of 162.7% and 188.6%, respectively. I held this position for approximately 14.5 months.

On 1/31/2017, I initiated a position in New Residential Investment Corp. (NRZ) at a weighted average purchase price of $15.10 per share. On 6/29/2017, 7/7/2017, and 12/21/2018, I increased my position in NRZ at a weighted average purchase price of $15.775, $15.18, and $14.475 per share, respectively. When combined, my NRZ position had a weighted average purchase price of $14.912 per share. This weighted average per share price excluded all dividends received/reinvested. On 2/6/2020, I sold my entire NRZ position at a weighted average sales price of $17.555 per share as my price target, at the time, of $17.50 per share was surpassed. This calculates to a weighted average realized gain and total return of 17.7% and 41.2%, respectively. I held this position, on a weighted average basis, for approximately 20 months.

On 9/22/2020, I once again initiated a position in NRZ at a weighted average purchase price of $7.645 per share. On 1/28/2021, I increased my position in NRZ at a weighted average purchase price of $9.415 per share. When combined, my NRZ position has a weighted average purchase price of $8.53 per share This weighted average per share price excludes all dividends received/reinvested.

On 6/29/2017, I initiated a position in Cherry Hill Mortgage Investment Corp. (CHMI) at a weighted average purchase price of $18.425 per share. On 10/6/2017, 10/26/2017, 11/6/2017, 1/29/2018, 10/12/2018, 6/6/2019, 7/23/2019, 9/5/2019, 3/16/2020, and 4/6/2020 I increased my position in CHMI at a weighted average purchase price of $18.015, $18.245, $17.71, $17.145, $17.235, $16.315, $15.325, $12.435, $8.55, and $3.645 per share, respectively. When combined, my CHMI position has a weighted average purchase price of $7.735 per share (yes, my last 3 purchases were proportionately large). This weighted average per share price excludes all dividends received/reinvested. On 3/10/2021, I sold my entire CHMI position at a weighted average sales price of $10.405 per share as my price target, at the time, of $10.40 per share was surpassed. This calculates to a weighted average non-annualized realized gain and total return of 34.5% and 52.2%, respectively. I held this position, on a weighted average basis, for approximately 14 months. This calculates to a weighted average annualized total return of 44.0%.

On 8/31/2017, I initiated a position in CHMI’s Series A preferred stock, (CHMI.PA). On 9/12/2017 and 4/6/2020, I increased my position in CHMI-A at a weighted average purchase price of $25.145 and $10.945 per share, respectively. When combined, my CHMI-A position had a weighted average purchase price of $18.071 per share. This weighted average per share price excluded all dividends received/reinvested. On 6/8/2020, I sold my entire CHMI-A position at a weighted average sales price of $24.273 per share. This calculates to a weighted average realized gain and total return of 34.3% and 49.4%, respectively. I held this position, on a weighted average basis, for approximately 1.7 years. This calculates to a weighted average annualized total return of 29.7%.

On 1/29/2018, I initiated a position in Two Harbors Investment Corp. (TWO) at a weighted average purchase price of $15.155 per share. On 4/17/2019, I increased my position in TWO at a weighted average purchase price of $13.165 per share. When combined, my TWO position had a weighted average purchase price of $13.825 per share. This weighted average per share price excluded all dividends received/reinvested. On 2/3/2020, I sold my entire TWO position at a weighted average sales price of $15.355 per share as my price target, at the time, of $15.25 per share was surpassed. This calculates to a weighted average realized gain and total return of 11.0% and 25.2%, respectively. I held this position, on a weighted average basis, for approximately 13 months.

On 3/8/2018, I initiated a position in New York Mortgage Trust, Inc.'s Series D preferred stock, (NYMTN). On 4/6/2018, 4/27/2018, 10/12/2018, 12/7/2018, 12/18/2018, and 3/22/2019 I increased my position in NYMTN. When combined, my NYMTN position had a weighted average purchase price of $22.379 per share. This weighted average per share price excluded all dividends received/reinvested. On 12/8/2020 - 12/9/2020, I sold my entire NYMTN position at a weighted average sales price of $23.287 per share. This calculates to a weighted average realized gain and total return of 4.1% and 24.0%, respectively. I held this position, on a weighted average basis, for approximately 27 months.

On 10/12/2018, I initiated a position in Granite Point Mortgage Trust, Inc. (GPMT) at a weighted average purchase price of $18.155 per share. On 5/12/2020, 5/27/2020, 5/28/2020, 8/26/2020, 9/10/2020, and 9/11/2020, I increased my position in GPMT at a weighted average purchase price of $4.745, $5.144, $5.086, $6.70, $6.19, and $6.045 per share, respectively. My last two purchases made up approximately 50% of my total position (to put things in better perspective). When combined, my GPMT position had a weighted average purchase price of $6.234 per share. This weighted average per share price excluded all dividends received/reinvested. On 6/8/2021, I sold my entire GPMT position at a weighted average sales price of $15.783 per share as my price target, at the time, of $15.75 per share was surpassed. This calculates to a weighted average realized gain and total return of 153.2% and 168.7%, respectively. I held this position, on a weighted average basis, for approximately 11 months.

On 10/12/2018, I initiated a position in AG Mortgage Investment Trust Inc. (MITT) at a weighted average purchase price of $17.105 per share. On 4/17/2019 and 6/3/2019, I increased my position in MITT at a weighted average purchase price of $16.22 and $15.52 per share, respectively. When combined, my MITT position had a weighted average purchase price of $15.946 per share. This weighted average per share price excludes all dividends received/reinvested. On 5/11/2020, I sold my entire MITT position at a weighted average sales price of $2.115 per share as my price target, at the time, was surpassed (as MITT’s estimated BV as of 4/30/2020 was even lower versus my previously projected (75%) quarterly BV decrease [6/30/2020 versus 3/31/2020]). This was my first “realized total loss” within either the mREIT or business development company (“BDC”) sector since I began writing here on Seeking Alpha in 2013. With that said, my proportional allocation in MITT (versus the rest of the mREIT sector) was small.

On 6/3/2019, I initiated a position in ARMOUR Residential REIT Inc. (ARR) at a weighted average purchase price of $17.545 per share. On 9/10/2019, I increased my position in ARR at a weighted average purchase price of $16.785 per share. When combined, my ARR position had a weighted average purchase price of $16.975 per share. This weighted average per share price excluded all dividends received/reinvested. On 2/20/2020, I sold my entire ARR position at a weighted average sales price of $21.045 per share as my price target, at the time, of $20.90 per share was surpassed. This calculates to a weighted average non-annualized realized gain and total return of 24.0% and 31.0%, respectively. I held this position, on a weighted average basis, for approximately 6 months.

On 6/3/2019, I initiated a position in Invesco Mortgage Capital Inc. (IVR) at a weighted average purchase price of $15.49 per share. This weighted average per share price excludes all dividends received/reinvested. On 2/14/2020, I sold my entire IVR position at a weighted average sales price of $17.965 per share as my price target, at the time, of $17.95 per share was surpassed. This calculates to a weighted average non-annualized realized gain and total return of 16.0% and 25.0%, respectively. I held this position for approximately 8 months.

On 11/22/2019, I initiated a position in Anworth Mortgage Asset Corp. (ANH) at a weighted average purchase price of $3.475 per share. This weighted average per share price excluded all dividends received/reinvested. On 12/23/2020, I sold my entire ANH position at a weighted average sales price of $2.75 per share as my price target, at the time, was surpassed. This was my second “realized total loss” within either the mREIT or BDC sector since I began writing here on Seeking Alpha in 2013. With that said, my proportional allocation in ANH (versus the rest of the mREIT sector) was very small (less than 0.5% at the time of sale).

On 11/22/2019, I initiated a position in AAIC’s Senior Notes Due 2023 (AIW) at a weighted average purchase price of $24.13 per share ($25 being par). On 3/10/2020, 3/13/2020, and 3/19/2020, I increased by position in AIW at a weighted average purchase price of $23.50, $19.75, and $9.31 per share, respectively. When combined, my AIW has a weighted average purchase price of $14.804 per share. This weighted average per share price excludes all interest received/compounded.

On 12/31/2019, I initiated a position in Arlington Asset Investment Corp.’s Senior Notes Due 2025 (AIC) at a weighted average purchase price of $24.00 per share ($25 being par). On 3/10/2020 and 3/19/2020, I increased by position in AIC at a weighted average purchase price of $23.72 and $8.71 per share, respectively. When combined, my AIC has a weighted average purchase price of $16.182 per share. This weighted average per share price excluded all interest received/compounded. On 9/2/2020-9/4/2020, I sold my entire AIC position at a weighted average sales price of $23.55 per share. This calculates to a weighted average non-annualized realized gain and total return of 45.5% and 51.1%, respectively. I held this position for approximately 6.5 months.

On 1/2/2020, I initiated a position in AAIC at a weighted average purchase price of $5.57 per share. On 1/9/2020, 3/16/2020, 9/24/2020, and 5/6/2021, I increased my position in AI at a weighted average purchase price of $5.59, $3.25, $2.53, and $3.875 per share, respectively. When combined, my AAIC position has a weighted average purchase price of $3.362 per share. This weighted average per share price excludes all dividends received/reinvested.

On 3/18/2020, I initiated a position in Annaly Capital Management Inc. (NLY) at a weighted average purchase price of $5.05 per share (large purchase). This weighted average per share price excluded all dividends received/reinvested. On 6/9/2021, I sold my entire NLY position at a weighted average sales price of $9.574 per share as my price target, at the time, of $9.55 per share was surpassed. This calculates to a weighted average realized gain and total return of 89.6% and 112.0%, respectively. I held this position for approximately 15 months.

On 4/6/2020, I initiated a position in CHMI’s Series B preferred stock, (CHMI.PB) at a weighted average purchase price of $10.65 per share. This weighted average per share price excluded all dividends received/reinvested. On 6/19/2020-6/24/2020, I sold my entire CHMI-B position at a weighted average sales price of $22.045 per share. This calculates to a weighted average realized gain and total return of 107.0%. I held this position for approximately 2.5 months.

On 10/19/2020, I initiated a position in PennyMac Mortgage Investment Trust (PMT) at a weighted average purchase price of $16.275 per share. On 10/29/2020, I increased my position in PMT at a weighted average purchase price of $14.90 per common. When combined, my PMT position has a weighted average purchase price of $15.358 per share. This weighted average per share price excludes all dividends received/reinvested.

Final Note: All trades/investments I have performed over the past several years have been disclosed to readers in “real time” (that day at the latest) via either the StockTalks feature of Seeking Alpha or, more recently, the “live chat” feature of the Marketplace Service the REIT Forum (which cannot be changed/altered). Through these resources, readers can look up all my prior disclosures (buys/sells) regarding all companies I cover here at Seeking Alpha (see my profile page for a list of all stocks covered). Through StockTalk disclosures and/or the live chat feature of the REIT Forum, at the end of July 2021 I had an unrealized/realized gain “success rate” of 96.2% and a total return (includes dividends received) success rate of 96.2% out of 52 total past and present mREIT and BDC positions (updated monthly; multiple purchases/sales in one stock count as one overall position until fully closed out). Both percentages experienced a modest increase, when compared to April-May 2020, as a direct result of the recent partial market rally to counter previous fears/panic surrounding the COVID-19 pandemic. In addition, in early April 2020, I initiated several new positions and increased several existing positions at attractive-very attractive prices.I encourage other Seeking Alpha contributors to provide real time buy and sell updates for their readers which would ultimately lead to greater transparency/credibility. Starting in January 2020, I have transitioned all my real-time purchase and sale disclosures solely to members of the REIT Forum. All applicable public articles will still have my sector purchase and sale disclosures (just not in real time).

Table 4 – AGNC + NLY Seeking Alpha Recommendations (1-Year Timeframe)

(Source: Table sourced from Seeking Alpha; “Bearish” indicator included by me directly from the REIT Forum’s weekly subscriber recommendation article series; week of 6/4/2021 for AGNC and week of 6/11/2021 for NLY)

Lastly, I just want to quickly highlight my/our AGNC and NLY Seeking Alpha recommendation ranges over the past year. Simply put, my/our “valuation methodology” has correctly timed when both AGNC and NLY have been undervalued (a BUY recommendation; bullish), appropriately valued (a HOLD recommendation; neutral), and overvalued (a SELL recommendation; bearish). The same holds true when going back to 2019 and 2020 (both pre-COVID-19 where I/we had a SELL recommendation on both AGNC and NLY and post the initial onset of COVID-19 where I/we had a STRONG BUY recommendation on both AGNC and NLY). A contributor’s/team’s recommendation track record should “count for something” and should always be considered when it comes to credibility/successful investing. Please disregard any minor “cosmetic” typos if/when applicable.

Analyst’s Disclosure: I/we have a beneficial long position in the shares of AAIC, AIW, NRZ, PMT either through stock ownership, options, or other derivatives. 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.

I currently have no position in AGNC, ARR, BXMT, CHMI, CIM, CMO, DX, EFC, GPMT, IVR, MFA, MITT, NLY, NYMT, ORC, RC, TWO, or WMC. CO Wealth Management is currently long AGNCO, AGNCP, ARR-C, CIM-A, CMO-E, DX, DX-C, NRZ, and NYMTM.

Comments

[dkpaintball]

Any idea on why this company has never raised its dividend? Seems odd when the income is so high

[Scott Kennedy]

@dkpaintball

Technically, AGNC probably could right now. However, estimated REIT taxable income ("ERTI") is notably lower than its core earnings for 2021. So, they are NOT "required" to pay a higher dividend.

Also, the eventual taper/rise in short-term interest rates is looming (albeit at least a year down-the-road) so the company is being cautious.

Hope that helps.

Scott

[AllStreets]

Fabulous detailed coverage as always. I've held AGNC for years and always check your articles.

To value a number of my stocks and MLPs I look at free cash flow yield as a gauge of value. I see that using Morningstar's value for AGNC's free cash flow is $1.742 billion, the number of shares outstanding according to finviz.com is 526.60 million, and the closing price on 8/13 was 16.34 yielding a free cash flow yield of 20.24% per share; that's a very healthy return. I have never seen anything in your articles about free cash flow yields, but I'm wondering whether you ascribe any significance to free cash flow yield in your analysis of mREITs?

[Scott Kennedy]

@AllStreets

Hi, Thanks for the positive feedback and sharing your thoughts.

FCF is a useful metric in most sectors but I believe it "takes a back seat" to net spread and dollar roll income for AGNC. For the other mREITs, this would be the equivalent to "core earnings" and/or estimated REIT taxable income ("ERTI"). For years, I've continued to believe that's the best metric to track when it comes to any mREIT's dividend sustainability/financial liquidity.

FCF is still a useful metric but for mREITs the importance is diminished a bit in my professional opinion. Could write a short novel on all the reasons why but not going to get into that broad topic in a comment. Possible topic for a future article but I'm still doing a lot of work on recent earnings for the 35 mREIT and BDC stocks I cover. Hope that helps.

Scott

[Illius]

I consider these the most valuable articles on all of SA. Very very authoritative and professional. Thank you, sir.

[dhturk]

Getting close to the “buy” window. However, Scott may have already updated his recommendations. Sometimes it pays to peak behind the screen.

[Scott Kennedy]

@dhturk

Hi, yes subscribers of the REIT Forum get weekly updates to my/our price targets, recommendation ranges, risk ratings, and CURRENT BV/NAV projections for all 35 mREIT/BDC stocks I/we currently fully cover.

Scott

[mmoy]

Any thoughts on Netstreit (NTST)? It's not really high-yield at 3.08% but it seems like a potential stock for The REIT Forum (maybe not your coverage though).

[Scott Kennedy]

@mmoy

Hi, if NTST isn't a mREIT or BDC, it would be outside my coverage sectors on S.A. That said, you certainly could ask CO Wealth Management Fund to take a look at it if you want. He's usually open to taking a look and providing his thoughts regarding coverage.

Scott

[Bobd524]

As usual great work Mr. Kennedy.

[Scott Kennedy]

@Bobd524

Hi Bob,

Thanks for the quick, positive feedback.

I appreciate it.

Scott

[Retire2020]

@Scott Kennedy Thanks again for the update. Always appreciated. Have a great week.

[Scott Kennedy]

@Retire2020

Hi, thanks for the continued positive feedback (remember your username). Thanks for being a loyal/long-term reader.

Scott

[dlr.rrx_00]

Any thoughts on the possibility of NRZ increasing the dividend?

[Scott Kennedy]

@dlr.rrx_00

Hi, I/we expect a dividend increase eventually. I don't believe it'll occur until the Caliber Homes acquisition closes. Hope that helps.

Subscribers of the REIT Forum get quarterly dividend projections for all 20 mREIT covered peers (and 15 BDC peers). Here's a link to the specific mREIT dividend projection article:

seekingalpha.com/...

Scott

[Income4ever aka Cyclenut]

Good read,
Long AGNC
Any thoughts on ABR, another SA Contributor specialized in Reits wrote ABR could see $13 again... also BRMK ?

[Scott Kennedy]

@Income4ever aka Cyclenut

Thanks for the positive feedback. I can't comment on ABR or STWD. Quote from my profile page: "...I cannot cover ABR or STWD in the mREIT sector due to indirect conflicts of interest...."

Regarding BRMK, I/we decided not to cover that specific mREIT. The list of mREIT and BDC stocks I cover are in my profile.

Regardless, thanks for asking.

Scott