AGNC Investment's Q3 2016 Income Statement And EPS Projection - Part 2

| About: AGNC Investment (AGNC)

Summary

I am projecting AGNC will report a modest net gain on derivative instruments and other securities for the third quarter of 2016.

Interest rate payer swaps/U.S. Treasury securities with shorter tenors/maturities experienced a modest - material increase in rates/yields which positively impacted valuations.

I am projecting AGNC will report a minor decrease in company expenses when compared to the prior quarter as the company internalized its manager, ACMM on 7/1/2016.

My projections for AGNC’s net income and earnings per share for the third quarter of 2016 are stated in the “Conclusions Drawn” section of the article.

AGNC’s notable increase to net income when compared to the prior quarter is mainly a result of the positive relationship between fixed-rate agency MBS prices and most derivative instrument valuations.

Author's Note: PART 2 of this article is a continuation from PART 1 which was discussed in a previous publication. Please see PART 1 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; and 3) gain (loss) on sale of agency securities, net. PART 1 helps lead to a better understanding of the topics and analysis that will be discussed in PART 2. The link to PART 1's analysis is provided below:

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

Focus of Article:

The focus of PART 2 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: 4) "gain (loss) on derivative instruments and other securities, net" (including four "sub-accounts"); and 5) "compensation expense." PART 2 will also discuss AGNC's projected net income (loss) and earnings per share ("EPS") 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).

4) Gain (Loss) on Derivative Instruments and Other Securities, Net:

- Estimate of $229 Million; Range ($21)-$479 Million

- Confidence Within Range = Moderate to High

- See Boxed Blue Reference "4" in Tables 4 and 6 Below Next to the September 30, 2016 Column

Projecting AGNC's gain (loss) on derivative instruments and other securities, net account is an analysis that involves several sub-accounts. This includes making assumptions within these derivative sub-accounts during the current quarter. One will never "fully" know management's derivatives activities for any given quarter until results are provided to the public via the company's quarterly SEC submissions. However, one can understand management's overall derivative strategy and make a projection on these derivative sub-accounts using the balances that were represented at the end of the previous quarter. Such a detailed analysis is wise to perform due to the typical events that unfold in regards to MBS prices, the fixed pay rate on newly created interest rate swaps, and U.S. Treasury yields. I believe this detailed analysis is critical once again in regards to the third quarter of 2016. When using this methodology, along with deciding specific quarterly assumptions, I have typically provided highly accurate projections within this account over the past several years.

Now let us take a look at AGNC's gain (loss) on derivative instruments and other securities, net account. I show my projection for this figure in Table 4 below. All past (ACTUAL) sub-account figures within Table 4 are derived from AGNC's quarterly SEC submissions via the company's 10-Q or 10-K where applicable. All projected (ESTIMATE) sub-account figures within Table 4 below are calculated and derived from multiple tables/charts that will not be shown within this particular article.

Table 4 - AGNC Quarterly Gain (Loss) on Derivative Instruments and Other Securities, Net Projection (All Sub-Accounts)

Click to enlarge

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

Within AGNC's gain (loss) on derivative instruments and other securities, net account is the following four material sub-accounts that will be discussed below:

a) TBA MBS

b) Interest Rate Swaps

c) Interest Rate Swaptions

d) U.S. Treasury Securities

Each of the four material derivative sub-accounts listed above will be separately analyzed and discussed in corresponding order of the blue references under the "Ref." column in Table 4 above.

a) TBA MBS (Net Long Position as of 6/30/2016):

- Estimate of $55 Million; Range ($70)-$180 Million

- Confidence Within Range = Moderate to High

- See Black Highlighted, Blue Referenced Sub-Account "a)" in Table 4 Above Next to the September 30, 2016 Column

Let us first briefly get accustomed with this type of derivative instrument. Typically, AGNC uses a combination of both long and (short) TBA MBS contracts during any given quarter. AGNC enters into TBA contracts with a long position where it agrees to buy, for future delivery, MBS with certain predetermined prices, face amounts, issuers, coupons, and stated maturities. AGNC enters into TBA contracts with a long position as an off-balance sheet means of investing in and financing MBS. Since TBA contracts with a long position are ultimately an extension of the balance sheet, this increases AGNC's "at risk" leverage. AGNC enters into TBA contracts with a (short) position where it agrees to sell, for future delivery, MBS with certain predetermined prices, face amounts, issuers, coupons, and stated maturities. Since TBA contracts with a (short) position are ultimately a reduction of the balance sheet, this decreases AGNC's at risk leverage.

There are two main factors that impact this derivative sub-account's valuation in a given quarter. The first factor is the dollar roll income (expense) generated on the net long (short) TBA MBS position. The second factor is the realized valuation gain (loss) upon the "settlement" of all TBA MBS contracts and the unrealized valuation gain (loss) on all contracts that have yet to be settled during the quarter (one example is a "re-rolled" TBA MBS position).

For the second quarter of 2016, AGNC reported a TBA MBS total net valuation gain of $108 million. When broken out, AGNC reported "net dollar roll" ("NDR") income of $44 million and a TBA MBS net valuation gain of $64 million. As of 3/31/2016, AGNC had a net long TBA MBS position of $5.8 billion (based on notional amount). AGNC had a net long TBA MBS position of $6.8 billion as of 6/30/2016. When calculated, AGNC increased the company's net long TBA MBS position by $1.0 billion during the second quarter of 2016.

As will be discussed further in PART 3 of this article, MBS prices across most coupons had minor fluctuations during the third quarter of 2016. As such, AGNC's TBA MBS portfolio likely also only experienced minor valuation fluctuations during the third quarter of 2016.

Through interpreting management's comments via several prior investor presentations and earnings conference calls, I have made the assumption AGNC only slightly altered the company's net long TBA MBS position as of 6/30/2016 during most of the third quarter of 2016. Through a detailed analysis that will be omitted from this particular article, when combining the company's projected quarterly NDR income of $45 million and a quarterly net valuation gain of $10 million, I am projecting AGNC's TBA MBS had a total net valuation gain of $55 million for the third quarter of 2016.

b) Interest Rate Swaps (Net (Short) Position as of 6/30/2016):

- Estimate of $145 Million; Range ($105)-$395 Million

- Confidence Within Range = Moderate to High

- See Purple Highlighted, Blue Referenced Secondary Sub-Accounts "b)" in Table 4 Above Next to the September 30, 2016 Column

Let us first discuss the recent history of this derivative sub-account which will lead to a better understanding of my projected total net valuation gain for the third quarter of 2016. AGNC had a net (short) interest rate swaps position of ($35.1) billion as of 6/30/2016 (based on notional amount). AGNC had interest rate payer swap additions of ($2.6) billion and interest rate payer swap expirations or terminations of $5.6 billion during the second quarter of 2016.

There was only a minor amount of activity within AGNC's interest rate swaps during the second quarter of 2016 for two main reasons. First, as was discussed in PART 1 of this article, AGNC only slightly altered the company's on-balance sheet MBS portfolio during the second quarter of 2016 (based on par value). Second, due to the Federal Open Market Committee's ("FOMC") continued rhetoric regarding a delay to a subsequent increase to the Federal ("Fed") Funds Rate, management believed the risks associated with the fixed-rate agency MBS market were relatively unchanged going into the third quarter of 2016. Due to AGNC's continued "cautious" risk management strategy, the company's hedging coverage ratio only decreased from 83% to 79% during the second quarter of 2016.

Using Table 4 above as a reference, there are two secondary sub-accounts to discuss when projecting a total net valuation gain (loss) for AGNC's interest rate swaps. The first secondary sub-account is AGNC's "net periodic interest costs of interest rate swaps" expense. If one recalls, this figure was first discussed in AGNC's interest expense account during PART 1 of this article. In regards to AGNC's interest rate swaps net (short) position as of 6/30/2016, the company had a weighted average fixed pay rate of 1.64% and a weighted average floating receive rate of 0.64%. When excluding forward-starting interest rate swaps, this weighted average fixed pay rate was 1.51%. When calculated, AGNC's weighted average fixed pay rate (when including forward-starting interest rate swaps) decreased (19) basis points ("bps"). In addition, AGNC's weighted average floating receive rate increased 2 bps.

This brings up an important topic which I believe some readers/contributors are having a hard time understanding when it comes to an agency mREIT's business model. There have been a couple articles over the past several months in regards to the ramifications of a "flattening" yield curve regarding a company's "net spread" income. In a nutshell, these types of comments/articles basically concluded since short-term interest rates are rising while long-term interest rates are decreasing, mREIT companies will automatically suffer due to shrinking spreads between interest income from MBS and interest expense mainly associated with repurchase agreements (borrowing costs). However, when it comes to the mREIT sector (especially agency mREIT companies), I believe this generalized conclusion is overly simplified and doesn't consider some important factual data/other variables at play.

If one recalls, during PART 1 of this projection analysis (see link above), I stated AGNC's quarterly interest expense (when it comes to the company's outstanding borrowings) has increased by $2 million-$13 million over the past three quarters. Simply put, short-term interest rates have continued to gradually increase during this timeframe which has negatively impacted repurchase agreement interest rates. While this is certainly a negative factor to highlight, I believe "the other side of the equation" continues to not be discussed and/or understood.

I believe a good analogy for this discussion is a peanut butter and jelly sandwich. To make this sandwich, one needs BOTH the peanut butter and the jelly (fairly logical). In regards to agency mREIT companies (and other sector classifications as well), during most interest rate cycles (including the one the U.S. market is currently in), management teams have hedges in place to mitigate interest rate risk. Due to the typically higher durations of agency MBS (versus non-agency MBS, CMBS, etc…), management of this type of mREIT classification typically prefers to have a higher hedging coverage ratio in place (say versus hybrid or multipurpose mREIT peers) since these securities have enhanced sensitivity to movements in interest rates.

So, getting back to my analogy, repurchase agreements are the peanut butter and interest rate payer swaps (or similar types of hedges) are the jelly. Simply put, when discussing/analyzing the mREIT sector (especially when it comes to agency mREIT companies), one really cannot have one side of the equation (peanut butter; repurchase agreements) without the other (jelly; interest rate payer swaps or similar type of derivative instrument).

A few readers/articles I referenced earlier were basically stating mREIT companies are going to be materially negatively impacted from the rise in short-term interest rates while long-term interest/mortgage rates continue to decrease or hold relatively steady (a flattening of the yield curve). Again, while I agree there is a possibility in SOME instances some minor negative outcomes could occur under this scenario, I would also highlight when short-term interest rates rise (or remain relatively unchanged) while long-term interest/mortgage rates move lower (or remain relatively unchanged), a company's hedging costs will typically move lower. There are two main reasons why this occurs.

First, the floating receive rate of a company's interest rate payer swaps has a 100% correlation to LIBOR (an "exact pairing"). As stated in PART 1, LIBOR has a very high correlation to the Fed Funds Rate/repurchase agreement interest rates under most scenarios. However, with that being said, there was a positive de-coupling between the two rates during the third quarter of 2016. In a nutshell, the increase in LIBOR outpaced the increase to repurchase agreement interest rates (when comparing similar tenors/maturities). While this is likely a short-term event, this positive de-coupling is FACTUAL in nature. Setting aside this recent positive de-coupling, as short-term interest rates rise, so does the floating receive rate of an interest rate payer swap. Simply put, this mitigates a decrease in net spread income (even when it is not precisely shown within a GAAP income statement).

Second, all newly created interest rate payer swaps would have a fixed-pay rate that is notably lower when compared to existing payer swaps. As such, a decrease in the intermediate and/or long-end of the yield curve (which is what has recently occurred; an agency mREIT's hedges usually match to this part of the yield curve) would be mitigated by a reduction in a company's "net pay rate." Simply put, this also mitigates a decrease in net spread income.

Let us now "test" this line of reasoning when it comes to AGNC's first and second quarters of 2016 (where short-term interest rates net increased while long-term interest/mortgage rates net decreased). In PART 1, I provided a discussion where it was determined AGNC's repurchase agreements had an interest rate increase of 15 and 2 bps during the first and second quarters of 2016, respectively. However, I would also refer readers to the calculation that was performed earlier in this specific derivative sub-account's projection when it comes to AGNC's interest rate swaps. During AGNC's first and second quarters of 2016, the company's weighted average fixed pay rate (when including forward-starting interest rate swaps) decreased (6) and (19) bps while its weighted average floating receive rate increased 22 and 2 bps, respectively. When combined, this was a decrease of (28) and (21) bps regarding AGNC's then current period hedging costs which trumped the 15 and 2 bps increase in the company's weighted average repurchase agreements interest rate during the first and second quarter of 2016, respectively.

So, as mentioned in PART 1 and confirmed here through factual evidence as support, AGNC's "cost of funds" (combination of borrowing and hedging costs) actually net decreased during the first and second quarters of 2016, even with a notable flattening of the yield curve. This is an extremely important point to mention/highlight which some readers/contributors continue to not fully understand. Simply put, the recent rise to AGNC's (or any agency mREIT's) interest expense has been mitigated (in some cases trumped) by the decreased current period hedging costs of a company's derivative instruments. I bring up this topic solely because I want to make sure readers/other contributors understand the FULL relationship within this specific sector.

Now let us briefly discuss the impact of lower long-term interest/mortgage rates when it comes to an agency mREIT's interest income (in some cases a hybrid or multipurpose mREIT as well). Again, I believe a few key points should be addressed. Simply put, I do not believe a flattening yield curve is a "doomsday scenario" when it comes to mREIT companies for several reasons.

First, the secondary agency MBS market is one of the largest investment spaces in the world (also one of the most liquid; typically second to U.S. Treasuries). There are various choices companies can decide upon, when it comes to coupons/products, when investing in the secondary agency MBS market. For instance, if management is worried about decreasing long-term/mortgage interest rates, management could purchase "seasoned/vintage," higher coupon MBS versus newer, less seasoned MBS which would typically have lower coupons. While this would likely increase a company's cost basis (higher premiums associated with a higher stated coupon under most scenarios; exceptions apply) and also increase prepayment risk (specified pools can mitigate this risk), this strategy would offset decreasing interest income. In other words, companies could sell some lower-coupon MBS (which would have appreciated in value if long-term interest/mortgage rates net decreased) and invest in some higher-coupon MBS. This would also mitigate losses associated with lower coupon MBS in a rising interest rate environment in the future.

Second, let us hypothetically say management actually goes ahead and redeploys capital into lower coupon MBS. While this would likely negatively impact interest income to an extent, in a net decreasing interest/mortgage rate environment, this would also directly lead to enhanced valuation gains on the lower coupon MBS (when compared to higher coupon MBS). As such, a company would record enhanced realized/capital gains which would likely trump the reduction in interest income (which also has a more favorable taxation treatment for shareholders).

Third, which I alluded to earlier, if there is a net decrease in long-term interest/mortgage rates, an agency mREIT's cost of funds rate will also have a high probability of decreasing. The AGNC example I provided above factually proves this notion. Dependent upon a company's weighted average maturity on its derivatives portfolio (including several other factors), a net decrease in long-term interest/mortgage rates would typically have an offsetting net decrease in hedging costs.

So, in the end, I know some people try to "simplify" the mREIT model. However, with the discussion above as support, I believe the simpler one tries to portray this sector, the higher the probability of an omission of various key factors/variables. This would generally lead to a higher probability of inaccurate assertions/analysis. I hope this discussion helps readers/other contributors have a better understanding of the mREIT business model. In the end, that is one of the main reasons why I write articles here at Seeking Alpha. I want readers to get a better understanding of the companies/sectors they invest in so they can make more informed decisions. Now let us get back to a projection for this derivative sub-account.

When all factors and assumptions are taken into consideration, I project AGNC will record a decrease in the company's net periodic interest costs of interest rate swaps expense for the third quarter of 2016. This is due to the following three assumed interest rate swap factors during the quarter: 1) relatively unchanged average notional balance as AGNC remained fairly cautious regarding its risk management strategy; 2) minor - modest decrease in the weighted average fixed pay rate on all newly created contracts when compared to terminated/settled contracts; and 3) modest increase in the average floating receive rate on all contracts. Through a detailed analysis that will be omitted from this particular article, I am projecting AGNC had a net periodic interest costs of interest rate swaps expense of $60 million for the third quarter of 2016. This calculates to a decreased expense of ($9) million when compared to the prior quarter. This should be seen as a cautious projection (possibility of an even higher reduction within this secondary sub-account).

The second secondary sub-account to discuss relates to the net valuation gain (loss) on AGNC's interest rate swaps. With the exception of very long tenors/maturities, there was a modest - material increase in the fixed pay rate of all interest rate payer swap contracts during the third quarter of 2016.

The fixed pay rate on interest rate swap contracts with a tenor/maturity towards the shorter-end of the yield curve had a modest - material net increase in valuations during the third quarter of 2016. The fixed pay rate on interest rate swap contracts with a tenor/maturity towards the longer-end of the yield curve only had a minor net increase in valuations during the third quarter of 2016.

Through a detailed analysis that will be omitted from this particular article, I am projecting AGNC's second secondary sub-account had a net valuation gain of $205 million for the third quarter of 2016. When both secondary sub-accounts are combined, I am projecting AGNC's interest rate swaps had a total net valuation gain of $145 million for the third quarter of 2016. This should be seen as a cautious projection (possibility of an even higher valuation gain if certain strategies were implemented during the quarter).

c) Interest Rate Swaptions (Net (Short) Position as of 6/30/2016):

- Estimate of ($7) Million; Range ($32)-$18 Million

- Confidence Within Range = High

- See Pink Highlighted, Blue Referenced Sub-Account "c)" in Table 4 Above Next to the September 30, 2016 Column

Let us first briefly get accustomed with this type of derivative instrument. Interest rate swaptions are options to enter into underlying interest rate swap contracts. Whereas interest rate swap contracts have no initial "up-front" costs (gains and losses are incurred as interest rates fluctuate over the life of the swaps), interest rate swaptions have implicit up-front costs (similar to an option contract; generally speaking).

Let us discuss the recent history of this derivative sub-account which will lead to a better understanding of my projected total net valuation loss for the third quarter of 2016. AGNC had a net (short) interest rate swaptions position of ($1.1) billion as of 6/30/2016 (based on the notional balance of the underlying interest rate swaps). AGNC had no interest rate payer swaption additions and $0.7 billion of interest rate payer swaption exercises, expirations, or terminations during the second quarter of 2016. As of 6/30/2016, AGNC's interest rate payer swaptions had a weighted average of 2 months until expiration with an underlying interest rate swaps weighted average tenor/maturity of 6.7 years.

I am projecting AGNC exercised, had expired, or terminated interest rate payer swaptions for a total net valuation loss of ($7) million for the third quarter of 2016 (expired worthless). The probability of AGNC's interest rate swaptions being worthless continues to be very high. This is due to the fact AGNC's interest rate swaptions had an underlying weighted average fixed pay rate of 3.38% as of 6/30/2016. This continued to be notably higher when compared to current interest rate payer swaps of a similar tenor/maturity.

d) U.S. Treasury Securities (Net (Short) Position as of 6/30/2016):

- Estimate of $35 Million; Range ($115)-$185 Million

- Confidence Within Range = Moderate

- See Dark Blue, Brown, and Teal Highlighted, Blue Referenced Secondary Sub-Accounts "d)" in Table 4 Above Next to the September 30, 2016 Column

Let us first briefly get accustomed with this type of derivative instrument. AGNC purchases (or sells short) U.S. Treasury securities and U.S. Treasury security futures to help mitigate the potential impact of changes in MBS prices (hence the valuation of a majority of the company's investment portfolio). AGNC borrows securities to cover U.S. Treasury (short sales) under reverse repurchase agreements. AGNC accounts for these derivative instruments as "security borrowing transactions" and recognizes an obligation to return the borrowed securities at fair market value ("FMV") based on the current value of the underlying borrowed securities.

Let us discuss the recent history of this derivative sub-account which will lead to a better understanding of my projected total net valuation gain for the third quarter of 2016. AGNC had the following three derivative secondary sub-account positions as of 6/30/2016: 1) long U.S. Treasury securities of $0.1 billion; 2) (short) U.S. Treasury securities of ($2.9) billion; and 3) U.S. Treasury security futures sold (short) of ($2.0) billion. This is based on each secondary sub-account's face amount ("par"). When combining all three secondary sub-accounts together, AGNC slightly decreased the company's net (short) U.S. Treasury securities position from ($5.0) billion as of 3/31/2016 to ($4.8) billion as of 6/30/2016.

Three likely scenarios occurred within this derivative sub-account during the third quarter of 2016. If the assumption is made that AGNC switched back to a net long U.S. Treasury securities position during the first-half of the third quarter of 2016, then the company would likely have a modest total net valuation loss for this derivative sub-account (lower end of my projected range). If the assumption is made that AGNC modestly - materially increased its net (short) U.S. Treasury securities position during the first-half of the third quarter of 2016, then the company would likely have a modest total net valuation gain for this derivative sub-account (higher end of my projected range). If the assumption is made that AGNC maintained a minor - modest net (short) U.S. Treasury securities position throughout most of the quarter, then the company would likely have a minor total net valuation fluctuation for this derivative sub-account. The amount of the total net valuation change would be dependent on the timing of the net long (short) positions as yields fluctuated throughout the quarter. These three scenarios are not "every" possible scenario that could have occurred within this derivative sub-account. However, I believe these three scenarios would have caused the greatest amount of FMV fluctuations. To put things in better perspective, yields on 5-, 7-, and 10-year U.S. Treasury securities increased 13, 13, and 11 bps during the third quarter of 2016, respectively.

Since U.S. Treasury securities are one of the most liquid investments in the marketplace, AGNC trades these derivative instruments throughout the quarter. However, with that being said, I have made the assumption AGNC was not overly "bearish" on U.S. Treasury yields during the quarter. Through a detailed analysis that will be omitted from this particular article, I am projecting AGNC's U.S. Treasury securities and U.S. Treasury security futures had a total net valuation gain of $35 million for the third quarter of 2016.

As stated earlier, all remaining derivative sub-accounts within Table 4 that have not been specifically discussed above are deemed immaterial for discussion purposes. As such, these immaterial derivative sub-accounts will be omitted from any analysis even though a projected net valuation gain (loss) has been included in Table 4. This includes valuation projections on the following derivative sub-accounts: 1) interest-only ("IO") and principal-only ("PO") strips; 2) debt on consolidated variable-interest-entities ("VIE"); 3) REIT equity securities (CYS stock); and 4) put options (when applicable).

When combining all the derivative sub-accounts together (both material and immaterial), I am projecting AGNC's derivative instruments and other securities, net account had a total net valuation gain of $229 million for the third quarter of 2016.

Brief Discussion of MTGE's and NLY's Derivatives Portfolio:

I see some general similarities between AGNC and the company's affiliate MTGE Investment Corp. (NASDAQ:MTGE) regarding hedging strategies. However, there usually are a few differences between AGNC's and MTGE's derivatives portfolio as well. One minor difference was each company's TBA MBS position as of 6/30/2016 (proportionately speaking). As stated earlier, AGNC had a net long TBA MBS position of $6.8 billion as of 6/30/2016. This was equal to 13% of AGNC's on-balance sheet MBS portfolio. In contrast, MTGE had a net long TBA MBS position of $0.3 billion as of 6/30/2016. This was equal to 9% of MTGE's on-balance sheet agency MBS portfolio. Another difference to note was MTGE had a modestly lower net (short) interest rate swaps position as of 6/30/2016 when compared to AGNC (proportionately speaking). When all derivative sub-accounts are taken into consideration, AGNC and MTGE had a hedging coverage ratio as of 6/30/2016 of 79% and 70%, respectively.

When it comes to AGNC's sector peer NLY, I see several differences that would impact the derivative sub-accounts described above. I will note a few of these differences. AGNC and NLY had a modest difference in each company's TBA MBS position as of 6/30/2016. As discussed earlier, AGNC had a net long TBA MBS position of $6.8 billion as of 6/30/2016. In contrast, NLY had a net long TBA MBS position of $12.7 billion as of 6/30/2016. Since NLY had a material net long TBA MBS position as of 6/30/2016, the probability of the company generating a modest - material amount of NDR income during the third quarter of 2016 was relatively high. Simply put, NLY continued to utilize the financing advantage of the TBA forward market more so than AGNC.

Second, AGNC and NLY had a notable difference in each company's hedging coverage ratio as of 6/30/2016. As stated above, AGNC had a hedging coverage ratio of 79%. In sharp contrast, NLY had a hedging coverage ratio of only 46%. As such, NLY was more vulnerable if mortgage interest rates/U.S. Treasury yields modestly - materially increased during the third quarter of 2016. While this specific scenario did not occur, most derivative instruments (especially across the short- to intermittent-end of the yield curve) experienced notable valuation gains during the third quarter of 2016. As such, NLY's lower hedging coverage ratio at the start of the third quarter of 2016, including an interest rate payer swaps portfolio that had a longer weighted average tenor/maturity when compared to AGNC, was a disadvantage for the company which will likely lead to a lower total net valuation gain within its derivatives portfolio (proportionately speaking). Further analysis of NLY's MBS and derivatives portfolios was discussed within the following two-part article:

Annaly Capital's BV, Valuation, And Dividend Compared To 17 mREIT Peers (Post Q2 2016 Earnings) - Part 1

Annaly Capital's BV, Valuation, And Dividend Compared To 17 mREIT Peers (Post Q2 2016 Earnings; Including Q4 2016 Dividend Projection) - Part 2

5) Compensation Expense (Formerly Management Fees):

- Estimate of $16 Million; Range $6-$26 Million

- Confidence Within Range = Moderate to High

- See Boxed Blue Reference "5" in Tables 5 and 6 Below Next to the September 30, 2016 Column

Prior to the third quarter of 2016, AGNC had a base management fee payable to American Capital Mortgage Management ("ACMM") that was paid in arrears and was equal to 1/12th of 1.25% of the company's equity balance. Equity was defined as AGNC's month-ended stockholders equity, adjusted to exclude the impact of any unrealized gains (losses) included in either retained earnings or accumulated OCI/(OCL).

However, in conjunction with the internalization of ACMM which was purchased from American Capital Ltd. (NASDAQ:ACAS) (who is in the process of being acquired by Ares Capital Corp. (NASDAQ:ARCC)), AGNC no longer pays quarterly management fees. Instead, AGNC directly pays/consolidates all compensation costs in relation to its newly acquired wholly owned subsidiary, ACMM. As such, I have changed this account's name (tentatively until confirmation) to AGNC's compensation expense.

In addition, until AGNC provides the specific line item(s) where management will account for ACMM's management fees in relation to MTGE and the company's intangible amortization expense in relation to the ACMM acquisition, I am keeping all account transactions regarding these events within this specific line item. Readers should understand these specific transactions will likely have a separate line item in the future.

I show my projection for this figure in Table 5 below. All past (ACTUAL) figures within Table 5 are derived from AGNC's quarterly SEC submissions via the company's 10-Q or 10-K where applicable. This excludes all recalculated figures and ratios.

Table 5 - AGNC Quarterly Compensation Expense Projection

Click to enlarge

(Source: Table created entirely by myself, partially using AGNC data obtained from the SEC's EDGAR Database [link provided below Table 4])

Using Table 5 above as a reference, I am projecting AGNC's compensation expense (formally the company's management fees expense) to decrease by ($9) million for the third quarter of 2016 when compared to the second quarter of 2016 ($16 million versus $25 million). This account has a slightly larger range this quarter due to the fact there may be certain "one-time" expenses in relation to management's internalization and the combination of several potential accounts described above.

Side Note: Two remaining accounts within AGNC's consolidated statement of comprehensive income that impact the company's net income (loss) amount are the following: 1) general/administrative expenses; and 2) income tax provision (benefit). While both accounts have been projected with Table 6 below, these two accounts are immaterial for discussion purposes and will be excluded from any analysis within this article.

A) Net Income (Loss):

- Estimate of $489 Million; Range $239-$739 Million

- Net Income Available to Common Shareholders of $1.46 Per Share (Excluding OCI/(OCL)); Range $0.70-$2.21 Per Share

- Confidence Within Range = Moderate to High See Red Reference "A" in Table 6 Below Next to the September 30, 2016 Column

Finally, let us look at my projection for AGNC's quarterly net income for the third quarter of 2016. This information is provided in Table 6.

Table 6 - AGNC Quarterly Net Income (Loss) Projection

Click to enlarge

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

Table 6 is a summation of the following consolidated statement of comprehensive income accounts for the third quarter of 2016: 1) interest income of $330 million; 2) interest expense of $107 million; 3) gain on sale of agency securities, net of $60 million; 4) gain on derivative instruments and other securities, net of $229 million; 5) compensation expense of $16 million; 6) general/administrative expenses of $7 million; and 7) excise tax of $0. Therefore, when these seven accounts are combined, I am projecting AGNC had net income of $489 million for the third quarter of 2016. After accounting for AGNC's quarterly preferred stock dividends, this would be earnings available to common shareholders of $1.46 per share.

Conclusions Drawn (PART 2):

To sum up the analysis above, I am projecting AGNC will report the following account figures for the third quarter of 2016 (refer back to Table 6 above):

4) Quarterly Net Gain on Derivative Instruments and Other Securities of $229 Million

5) Quarterly Compensation Expense of $16 Million

I am projecting AGNC will report the following net income and EPS amounts for the third quarter of 2016 (refer back to Table 6):

A) Quarterly Net Income of $489 Million and Earnings Available to Common Shareholders of $1.46 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 AGNC's projected net valuation gain of $229 million pertaining to the company's derivatives portfolio for the third quarter of 2016. For the same account in the prior quarter, AGNC recognized a net valuation loss of ($367) million.

As stated in PART 1 of this article, AGNC's OCI/(OCL) amount is part of the company's consolidated statement of comprehensive income but EXCLUDED from the company's net income (loss) and EPS amounts. As such, I suggest holding off on a "final verdict" regarding AGNC's projected results for the third quarter of 2016 until PART 3 of this article is released. In my professional opinion, I believe AGNC's "comprehensive income (loss)" amount is more important than the company's net income (loss) and EPS amounts.

My BUY, SELL, or HOLD Recommendation:

For my BUY, SELL, or HOLD recommendation regarding AGNC (including current price target), please see PART 1 of this article (link provided at the beginning of the article).

Final Note: PART 2 of this article is only a PARTIAL analysis of AGNC's consolidated statement of comprehensive income for the third quarter of 2016. As such, a "final" conclusion will not be provided yet. PART 3 of this article will just pick up where PART 2's analysis ends. PART 3 of this article will discuss AGNC's projected OCI/(OCL) and comprehensive income (loss) amounts. This will be followed by a projection of AGNC's BV as of 9/30/2016 and 10/21/2016 which will be available to readers prior to the company's earnings press release for the third quarter of 2016 in late October.

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. 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 ACAS, ARCC, ARR, CYS, NLY, or ORC.