AGNC Investment's Q3 2016 Income Statement And EPS Projection - Part 3

| About: AGNC Investment (AGNC)

Summary

I am projecting AGNC Investment Corp. will report a minor net unrealized loss on available-for-sale securities for the third quarter of 2016 (minor fluctuations for MTGE Investment Corp. and Annaly Capital Management as well).

This is due to the fact most 15- and 30-year fixed-rate agency MBS coupons had minor-to-modest price fluctuations during the third quarter of 2016.

However, this three-part article also highlights to readers the heightened importance of understanding the relationship between AGNC's derivatives portfolio and the company's MBS portfolio regarding changes in quarterly valuations.

My projection for AGNC's comprehensive income for the third quarter of 2016 is stated in the "Conclusions Drawn" section. I also include my buy, sell, or hold recommendation.

AGNC's encouraging quarterly results are mainly a result of the positive relationship between fixed-rate agency MBS prices and most derivative instrument valuations during the third quarter of 2016.

Author's Note: PART 3 of this article is a continuation from PART 1 and PART 2 which were discussed in previous publications. Please see PART 1 and PART 2 of this article for a detailed projection of AGNC Investment Corp.'s (NASDAQ:AGNC) income statement (technically speaking, the company's "consolidated statement of comprehensive income") for the third quarter of 2016 regarding the following accounts: 1) interest income; 2) interest expense; 3) gain (loss) on sale of agency securities, net; 4) gain (loss) on derivative instruments and other securities, net (including four "sub-accounts"); and 5) compensation expense. PART 2 also discussed AGNC's projected net income and earnings per share ("EPS") amounts. PART 1 and PART 2 help lead to a better understanding of the topics and analysis that will be discussed in PART 3. The links to PART 1 and PART 2 are provided below:

AGNC Investment Corp.'s Q3 2016 Income Statement Projection - Part 1 (Including My Buy, Sell, or Hold Recommendation)

AGNC Investment Corp.'s Q3 2016 Income Statement and EPS Projection - Part 2

Focus of Article

The focus of PART 3 of this article is to provide a detailed projection of AGNC's consolidated statement of comprehensive income for the third quarter of 2016 regarding the following accounts: 6) "unrealized gain (loss) on available-for-sale ("AFS") securities, net"; and 7) "unrealized gain (loss) on interest rate swaps, net (upon reclassification)". PART 3 will also discuss AGNC's projected other comprehensive income (loss) (OCI/(OCL)) and comprehensive income (loss) amounts. For readers who just want the summarized account projections, I would suggest to scroll down to the "Conclusions Drawn" section at the bottom of the article.

By understanding the trends that occurred within AGNC's operations during the third quarter of 2016, one can apply this information to sector peers as well. As such, the discussion/analysis below is not solely applicable to AGNC, but to the fixed-rate agency mortgage real estate investment trust (mREIT) sector as a whole. This includes, but is not limited to, the following fixed-rate agency mREIT peers: 1) ARMOUR Residential REIT Inc. (NYSE:ARR); 2) CYS Investments Inc. (NYSE:CYS); 3) Annaly Capital Management Inc. (NYSE:NLY); and 4) Orchid Island Capital Inc. (NYSE:ORC). In particular, PART 3 provides a discussion of fixed-rate agency MBS price movements which all of the sector peers listed above are currently heavily invested in.

6) Unrealized Gain (Loss) on AFS Securities, Net

- Estimate of ($45) Million; Range ($295)-$205 Million

- Confidence Within Range = Moderate to High

- See Boxed Blue Reference "6" in Tables 10 and 11 Below, Next to the September 30, 2016 Column

Projecting the company's unrealized gain (loss) on AFS securities, net account is an analysis that includes several assumptions and variables that need to be taken into consideration. Since this account is the summation of the quarterly unrealized valuation changes within AGNC's MBS portfolio (by far the largest asset class on the company's balance sheet), a wider projection range should be accompanied with this specific account. The same assumptions used within AGNC's gain (loss) on sale of agency securities, net account (see PART 1 of article), and gain (loss) on derivative instruments and other securities, net account (regarding the company's TBA MBS position; see PART 2 of article), will be applied when analyzing this account.

Prior to performing an account projection analysis, let us first analyze the fixed-rate agency MBS price movements during the third quarter of 2016. Using Table 7 below as a reference, let us first analyze the 15-year fixed-rate agency MBS price movements. This will then be followed by a similar analysis (via Table 8) of the 30-year fixed-rate agency MBS price movements. Doing so will help readers understand how I come up with my projected valuations discussed later in the article.

Table 7 - 15-Year Fixed-Rate Agency MBS Price Movements (Q3 2016)

Click to enlarge

(Source: Table created entirely by myself, using MBS pricing data via private access to a professional resource [Thomson Reuters])

Table 7 above shows the 15-year fixed-rate agency MBS price movements during the third quarter of 2016. It breaks out these agency MBS holdings by "government-sponsored entity" ("GSE"). This includes both Fannie Mae (OTCQB:FNMA) and Freddie Mac (OTCQB:FMCC) MBS. As of 6/30/2016, AGNC's Ginnie Mae holdings accounted for less than 1% of the company's MBS portfolio (excluding TBA MBS positions). As such, Ginnie Mae fixed-rate agency MBS price movements are deemed immaterial for discussion purposes and are thus excluded from this table. Table 7 further breaks out the 15-year fixed-rate agency MBS price movements into the various coupons on AGNC's books ranging from 2.5% - 4.0%. The REIT holds an immaterial balance over the 4.0% coupon, and thus, these specific coupons are excluded from Table 7 above.

From the information provided in Table 7, a valuation gain (loss) can be calculated which is broken down by the various coupons. This valuation gain (loss) is shown in Table 9 later in the article. It should also be noted AGNC continually changes the company's MBS portfolio in any given quarter. As such, I must determine specific purchase and sale assumptions towards the end of my account projection analysis.

Using Table 7 above as a reference, let us look at the 15-year fixed-rate agency MBS price movements for the third quarter regarding a few of the coupon rates where AGNC held a material MBS balance as of 6/30/2016. The cumulative quarterly net MBS price movements for each coupon rate are shown within Table 7 under the "Cumulative Quarterly Change" column. For example, during the third quarter of 2016, a Fannie 15-year fixed-rate agency MBS with a 2.5%, 3.0%, 3.5%, and 4.0% coupon had a cumulative quarterly price increase (decrease) of 0.08, 0.16, (0.55), and (0.46), to settle its price at 103.55, 105.00, 105.42, and 103.06, respectively. As such, a minor price increase occurred on the 2.5% and 3.0% coupons, while a modest price decrease occurred on the 3.5% and 4.0% coupons. In addition, as of 9/30/2016, one should note the price on a Fannie 15-year fixed-rate agency MBS with a 2.5% coupon exceeded the price on a 4.0% coupon. This is mainly due to the fact long-term interest/mortgage rates have continued to "hover" near historical lows. As such, the "enticement" for homeowners to refinance has notably increased when compared to, say, the end of last year. As such, higher-coupon MBS typically have heightened prepayment risk when compared to lower-coupon MBS, due to the fact these specific securities have higher rates when compared to current rates. As such, in the current interest rate environment, the current premium market participants are willing to pay for a higher-coupon MBS typically "narrow" when compared to lower-coupon MBS (this topic was originally discussed in PART 1).

When compared to Fannie 15-year fixed-rate agency MBS, Freddie 15-year fixed-rate agency MBS had slight differences in weekly (hence cumulative quarterly) price valuations across the same coupons. As will be shown in Table 9 later in the article, the minor price fluctuations between the two types of 15-year fixed-rate agency MBS only had a minor impact on differing valuations. Now that we have an understanding of the 15-year fixed-rate agency MBS price movements during the third quarter of 2016, let us take a look at the 30-year fixed-rate agency MBS price movements.

Table 8 - 30-Year Fixed-Rate Agency MBS Price Movements (Q3 2016)

Click to enlarge

(Source: Table created entirely by myself, using MBS pricing data via private access to a professional resource [Thomson Reuters]; link provided below Table 7])

Table 8 above shows the 30-year fixed-rate agency MBS price movements during the third quarter of 2016. It breaks out these MBS holdings by GSE as well. As stated earlier, AGNC's Ginnie Mae fixed-rate agency MBS holdings are deemed immaterial for discussion purposes and are excluded from this table. Table 8 further breaks out the 30-year fixed-rate agency MBS price movements into the various coupons on AGNC's books ranging from 3.0% to 4.5%. The REIT holds an immaterial balance over the 4.5% coupon, and thus, these specific coupons are excluded from Table 8 above. From the information provided in Table 8, a valuation gain (loss) can be calculated which is broken down by the various coupons. This valuation gain (loss) is shown in Table 9 later in the article.

Using Table 8 above as a reference, let us look at the 30-year fixed-rate agency MBS price movements for the third quarter regarding a few of the coupon rates where AGNC held a material MBS balance as of 6/30/2016. For example, during the third quarter of 2016, a Fannie 30-year fixed-rate agency MBS with a 3.0%, 3.5%, 4.0%, and 4.5% coupon had a cumulative quarterly price increase of 0.16, 0.01, 0.17, and 0.36, to settle its price at 103.97, 105.53, 107.39, and 109.52, respectively. As such, a minor price increase occurred on the 3.0%, 3.5%, and 4.0% coupons, while a modest price increase occurred on the 4.5% coupons.

When compared to Fannie 30-year fixed-rate agency MBS, Freddie 30-year fixed-rate agency MBS had slight differences in weekly (hence cumulative quarterly) price valuations across the same coupons. Again, the minor price fluctuations between the two types of 30-year fixed-rate agency MBS only had a minor impact on differing valuations. Now that we have an understanding of the 15- and 30-year fixed-rate agency MBS price movements during the third quarter of 2016, let us take a look at how these price movements impacted AGNC's MBS portfolio regarding the company's weekly and cumulative quarterly valuations (quantified in dollar amounts).

Table 9 - AGNC Summarized Weekly and Cumulative Quarterly Fixed-Rate Agency MBS Portfolio Valuation Gain (Loss) (Q3 2016)

Click to enlarge

(Source: Table created entirely by myself, including all calculated figures and projected valuations)

Table 9 above first shows AGNC's weekly and cumulative quarterly projected valuation gain (loss) on the company's 15- and 30-year fixed-rate agency MBS holdings across all coupons (including immaterial coupons omitted from Table 7 and Table 8 above). I am projecting an "initial" net valuation gain of $11 million on AGNC's 15- and 30-year fixed-rate agency MBS holdings for the third quarter of 2016. However, as stated earlier, several other adjustments need to be performed within the REIT's unrealized gain (loss) on AFS securities, net account, before a final projection can be made. Since AGNC had a notable proportion of its fixed-rate agency MBS holdings within "specified pools" (prepayment protected MBS; mainly through the Home Affordable Refinance Program ("HARP") and low-loan balance ("LLB") securities), a quarterly valuation adjustment needs to be calculated. Through a detailed calculation that will be omitted from this particular article (beyond a "free to the public" article), I am projecting AGNC had an additional net valuation gain of $5 million for the third quarter of 2016 in regard to its specified pools. This should be seen as a cautious estimate. Due to the fact long-term interest/mortgage rates as of 9/30/2016 remained near historical lows, the premiums associated with HARP and LLB securities remained elevated across higher coupons during the third quarter of 2016. When these two figures are combined, I am projecting the company had a net valuation gain of $16 million on its 15- and 30-year fixed-rate agency MBS portfolio (see blue reference "A" in Table 9 above).

Still using Table 9 above as a reference, through a detailed calculation that will be omitted from this particular article, I am projecting AGNC had a net valuation loss of ($10) million during the third quarter of 2016 in regard to the following MBS holdings: 1) 20-year fixed-rate; 2) collateralized mortgage obligations ("CMO"); 3) adjustable-rate mortgages ("ARM"); 4) credit risk transfers ("CRT"); and 5) AAA non-agency (see blue reference "B" in Table 9 above).

Also, when considering the impacts of an assumed partial conversion of AGNC's net long TBA MBS position of $6.8 billion and the company's realignment of its MBS portfolio throughout the quarter, I am projecting a net valuation gain adjustment of $10 million for the third quarter of 2016 (change in leverage; see blue reference "C" in Table 9 above).

Therefore, when all the referenced figures stated above are combined, I am projecting a total net valuation gain of $15 million (rounded) on AGNC's MBS portfolio for the third quarter of 2016. This figure is PRIOR to all sold MBS being reversed out in the current quarter. As such, the final calculation within the company's unrealized gain (loss) on AFS securities, net account, is the quarterly reversal of its MBS sales shown in Table 10 below. Remember, all MBS sales have already been accounted for within AGNC's gain (loss) on sale of agency securities, net account (discussed in PART 1 of the article).

Table 10 - AGNC Quarterly Unrealized Gain (Loss) on AFS Securities, Net and Unrealized Gain (Loss) on Interest Rate Swaps, Net (Upon Reclassification to Interest Expense) Projection

Click to enlarge

(Source: Table created entirely by myself, partially using AGNC data obtained from the SEC's EDGAR Database)

Table 10 above shows AGNC's projected total net valuation gain of $15 million on the company's MBS portfolio (see red reference "AB" in Table 10 above). This amount is highlighted in teal. The second amount shown is the REIT's projected "reversal of prior period unrealized ("gain") loss, net, (upon realization)" figure (see red reference "AC" in Table 10 above). This amount is highlighted in pink.

Therefore, after the company's projected net realized gain on the sale of agency securities of ($60) million is reversed out, its total net unrealized loss on AFS securities is projected to be ($45) million for the third quarter of 2016 (see red reference "(AB + AC) = AD" in Table 10 above). This amount is highlighted in grey.

Brief Discussion of MTGE's and NLY's Unrealized Gain (Loss) on Agency Securities, Net Account

I see general similarities between AGNC and its affiliate MTGE Investment Corp. (NASDAQ:MTGE) regarding agency MBS portfolio strategies. As such, I see somewhat a similar projection between AGNC's unrealized gain (loss) on AFS securities, net account, and MTGE's "unrealized gain (loss) on agency securities, net" account for the third quarter of 2016 (proportionally speaking). With that being said, when compared to AGNC, MTGE had a slightly higher proportional share of 30-year fixed-rate agency MBS holdings as of 6/30/2016, which needs to be considered regarding projected valuations.

In addition, MTGE also had a much larger non-agency MBS portfolio (proportionately speaking). Due to this fact, it also has an "unrealized gain (loss) on non-agency securities, net" account that needs to be incorporated into the company's financials. MTGE's non-agency MBS portfolio has different valuation methods, which are mainly based on specific indexes and simulated models which are classified as level 3 assets per Accounting Standards Codification ("ASC") 820. This includes, but is not limited to, prime, CRT, Alt-A, option ARM, and subprime holdings. The company's non-agency MBS portfolio had a conditional prepayment rate ("CPR") of 15%, a conditional default rate ("CDR") of 5%, and a voluntary prepayment rate ("VPR") of 10% as of 6/30/2016. I anticipate only minor fluctuations in all three variables during the third quarter of 2016, which would slightly impact valuations. A further discussion of MTGE's non-agency MBS portfolio is beyond the scope of this article.

When it comes to AGNC's sector peer NLY, I see one modest difference that would impact the account described above. As of 6/30/2016, only 11% of NLY's fixed-rate agency MBS portfolio consisted of 15-year maturities, whereas AGNC had 29% of its MBS portfolio in 15-year maturities (excluding TBA MBS positions). Since most 30-year fixed-rate agency MBS coupons had more favorable price movements when compared to 15-year fixed-rate agency MBS with similar coupons, NLY should have a slightly higher combined total net valuation gain on AFS securities (realized and unrealized) when compared to AGNC during the third quarter of 2016 (proportionally speaking). Slightly offsetting this factor is the fact, as of 6/30/2016, NLY had a slightly lower on-balance sheet leverage ratio when compared to AGNC. It should also be noted, management has recently broadened NLY's investment portfolio by allocating more capital into commercial debt/real estate, preferred equity, corporate debt, and, most recently, middle market ("MM") lending. In addition, NLY recently acquired a variable-rate agency mREIT, Hatteras Financial Corp. (NYSE:HTS). Generally speaking, these asset classes, when compared to fixed-rate agency MBS, performed similarly during the third quarter of 2016 (minor net valuation changes). A further discussion of NLY's MBS/investment portfolio is beyond the scope of this article.

7) Unrealized Gain (Loss) on Interest Rate Swaps, Net (Upon Reclassification)

- Estimate of $8 Million; Range $3-13 Million

- Confidence Within Range = High

- See Boxed Blue Reference "7" in Table 10 Above and Table 11 Below, Next to the September 30, 2016 Column

Now let us take a look at AGNC's unrealized gain (loss) on interest rate swaps, net account. I show my projection for this figure in Table 10 above. The company's unrealized gain (loss) on interest rate swaps, net account, consists of two "sub-accounts". The first sub-account is the REIT's "unrealized gain (loss) on interest rate swaps designated as cash flow hedges" (see red reference "AE" in Table 10 above). Since AGNC discontinued accounting for the company's interest rate swaps as cash flow hedges under Generally Accepted Accounting Principles ("GAAP") back in 2011, this balance should remain $0 for the third quarter of 2016.

The second sub-account is AGNC's "reversal of prior period unrealized ("gain") loss on interest rate swaps, net (upon reclassification to interest expense)" (see red reference "AF" in Table 10 above). Each quarter, a portion of the remaining accumulated net deferred loss prior to the change in accounting treatment is reclassified out of the OCI/(OCL) amount as it relates to AGNC's "de-designated" interest rate swaps. A reclassification of $8 million has been projected for the third quarter of 2016. It should be noted, this account will have no activity by the end of 2016.

B) Other Comprehensive Income (Loss) (OCI/(OCL))

- Estimate of ($37) Million; Range ($287)-213 Million

- Confidence Within Range = Moderate to High

- See Red Reference "B" in Table 11 Below, Next to the September 30, 2016 Column

Let us now take a look at the company's projected OCI/(OCL) and comprehensive income (loss) amounts. This information is provided in Table 11 below.

Table 11 - AGNC Quarterly OCI/(OCL) and Comprehensive Income (Loss) Projection

Click to enlarge

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

After combining the company's net unrealized loss on AFS securities of ($45) million and its net unrealized gain on interest rate swaps (upon reclassification) of $8 million, I am projecting AGNC will report an OCL of ($37) million for the third quarter of 2016 (see red reference"B" in Table 11 above).

C) Comprehensive Income (Loss):

- Estimate of $452 Million; Range $252-652 Million

- Comprehensive Income Available to Common Shareholders of $1.34 Per Share; Range $0.74-1.95 Per Share

- Confidence Within Range = Moderate to High

- See Red Reference "C" in Table 11 Above, Next to the September 30, 2016 Column

Finally, let us look at AGNC's comprehensive income for the third quarter of 2016. This is the summation of the following amounts: A) net income of $489 million (see PART 1 and PART 2); and B) OCL of ($37) million (see analysis above). Therefore, when these two amounts are combined, I am projecting the REIT will report comprehensive income of $452 million for the third quarter of 2016.

Conclusions Drawn From PART 1, PART 2, and PART 3

To sum up the analysis from all three parts of the article, I am projecting AGNC will report the following amounts for the third quarter of 2016:

A) Quarterly Net Income of $489 Million; Earnings Available to Common Shareholders of $1.46 Per Share

B) Quarterly OCL of ($37) Million

C) Quarterly Comprehensive Income (A and B Combined) of $452 Million; Comprehensive Income Available to Common Shareholders of $1.34 Per Share

AGNC's projected net income of $489 million for the third quarter of 2016 is a notable improvement when compared to a net loss of ($135) million for the second quarter of 2016. This is mainly due to the company's projected net valuation gain of $229 million pertaining to its derivatives portfolio for the third quarter of 2016. For the same account in the prior quarter, it recognized a net valuation loss of ($367) million.

I am also projecting AGNC will report an OCL of ($37) million for the third quarter of 2016. This projected amount is a notable decrease when compared to OCI of $382 million for the second quarter of 2016. The main factor for the projected decrease is the more subdued cumulative quarterly price increases within lower-coupon 15- and 30-year fixed-rate agency MBS holdings during the third quarter of 2016 when compared to the prior quarter.

I believe AGNC will outperform a few fixed-rate agency mREIT peers in regard to valuation fluctuations for the third quarter of 2016, while underperforming a few others. This is mainly due to its higher hedging coverage ratio and the composition of the company's MBS portfolio heading into the third quarter of 2016 when compared to its fixed-rate agency mREIT peers.

I believe four key factors to analyze within the fixed-rate agency mREIT sector this quarter are the following: 1) each company's proportion of 15-year MBS holdings versus 30-year MBS holdings; 2) each company's hedging coverage ratio; 3) each company's proportion of long-term derivative instruments versus short-term derivative instruments; and 4) each company's proportion of specified pools (for instance HARP and LLB securities). Dependent upon these factors, I believe results will vary across the fixed-rate agency mREIT sector for the third quarter of 2016.

My BUY, SELL, or HOLD Recommendation

From the analysis provided above, including additional factors not discussed within this article, I currently rate AGNC as a SELL when I believe the company's stock price is trading at less than a (7.5%) discount to my projected BV as of 9/30/2016, a HOLD when trading at or greater than a (7.5%) but less than a (15.0%) discount to my projected BV as of 9/30/2016, and a BUY when trading at or greater than a (15.0%) discount to my projected BV as of 9/30/2016. These ranges are unchanged when compared to the last time I provided a recommendation on the company (PART 1).

Therefore, I currently rate AGNC as a BUY, since the stock is trading at or greater than a (15.0%) discount to my projected BV as of 9/30/2016. My current price target for AGNC is approximately $21.25 per share. This is currently the price where my recommendation would change to a SELL. This price target is unchanged when compared to the last time I provided a recommendation on the company (PART 1). The current price where my BUY recommendation would change to a HOLD is approximately $19.55 per share. This price is also unchanged when compared to the last time I provided a recommendation on the company (PART 1).

Along with the data presented within this article, these recommendations consider the following mREIT catalysts/factors: 1) projected future MBS price movements; 2) projected future derivative valuations; and 3) projected near-term dividend per share rates. This recommendation also considers the higher probability of one Fed Funds Rate increase by the FOMC during late 2016/early 2017 (this is a more "hawkish" view when compared to early summer, when no rate hikes were projected by most market participants) due to recent macroeconomic trends/events.

Final Note: The projected amounts from this three-part article will have a direct impact on AGNC's projected book value ("BV") as of 9/30/2016. My upcoming AGNC BV projection article will be available to readers prior to the company's earnings press release for the third quarter of 2016 on 10/24/2016.

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. By reading this article, I hope readers can better understand how AGNC (or any mREIT) generates income, incurs expenses, values assets held for investment, and mitigates risk.

On 9/12/2016 and 10/7/2016, I directly increased my position in AGNC at a weighted average purchase price of $18.985 and $18.745 per share, respectively. Each purchase had the same approximate monetary value. On 11/27/2015, I initiated a position in AGNCB (preferred series of stock). On 12/7/2015, 12/9/2015, 12/14/2015, 1/14/2016, and 1/20/2016, I selectively increased my position in AGNCB. When combined, my current AGNCB position has a weighted average price of $23.215 per share. This weighted average per share price excludes all dividends received/reinvested. I currently hold 0.71% of the outstanding shares of AGNCB. Each AGNC/AGNCB trade was disclosed to readers in "real time" (that day) via the StockTalks feature of Seeking Alpha. Through this resource, readers can look up all my prior disclosures (buys/sells) regarding companies I cover here at Seeking Alpha.

Disclosure: I am/we are long AGNC, MTGE.

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: I currently have no position in ARR, CYS, FMCC, FNMA, NLY, or ORC.

Editor's Note: This article discusses one or more securities that do not trade on a major U.S. exchange. Please be aware of the risks associated with these stocks.