American Capital Agency Corp.'s Dividend Sustainability Analysis (Post Q2 2014 Earnings) - Part 2

Aug.25.14 | About: American Capital (AGNC)

Summary

AGNC’s on-balance sheet leverage ratio (excluding TBA positions) went from 5.9x as of 3/31/2014 to 5.0x as of 6/30/2014 which led to the decrease in quarterly ERTI.

AGNC had a cumulative capital loss carryforward of ($1.37) billion as of 6/30/2014 which will have both positive and negative impacts on the company’s future dividend sustainability.

AGNC increased the company’s net long TBA MBS position during the second quarter of 2014 which led to the increase in net dollar roll income.

AGNC has implied all net dollar roll income generated from the company’s net long TBA MBS position would be an added component to dividend distributions when favorable business conditions persist.

My final dividend sustainability conclusion along with my buy, sell, or hold recommendation for AGNC, MTGE, and NLY is stated in the “Conclusions Drawn” section of the article.

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 an initial discussion on the dividend sustainability of American Capital Agency Corp. (NASDAQ:AGNC). 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:

American Capital Agency's Dividend Sustainability Analysis (Post Q2 2014 Earnings) - Part 1

This two-part article is a very detailed look at AGNC's dividend sustainability. I perform this detailed analysis for readers who anticipate/want such an analysis performed each quarter. For readers who just want the summarized conclusions/results, I would suggest to scroll down to the "Conclusions Drawn" section at the bottom of the each part of the article.

Focus of PART 2 of Article:

PART 1 of this article mainly analyzed AGNC's past and current performance regarding the company's quarterly ERTI, quarterly ERTI and net dollar roll, cumulative UTI, and cumulative UTI and net dollar roll figures (including two tests being performed). PART 2 discusses some additional topics/trends to consider in a general net rising (and falling) interest rate environment and the impacts these scenarios would have on the future dividend sustainability of AGNC. PART 2 also includes a brief discussion of AGNC's sister company American Capital Mortgage Corp. (NASDAQ:MTGE) and the company's closest sector peer Annaly Capital Management, Inc. (NYSE:NLY) regarding the same topics/trends.

Additional Topics/Trends to Consider in a General Net Rising (and Falling) Interest Rate Environment Regarding the Future Dividend Sustainability of AGNC:

In addition to TEST 1 and TEST 2 performed in PART 1 of this article, some recent topics/trends that will affect the future dividend sustainability of AGNC should also be addressed. These topics/trends will consider a general net rise in mortgage interest rates/U.S. Treasury yields over the next several quarters. In certain instances, I may also talk about the effects of a net decline in mortgage interest rates/U.S. Treasury yields and the effects that scenario would cause to the specific topic/trend being discussed. I believe the following four topics/trends should be addressed regarding a general net rise (and decline) in mortgage interest rates/U.S. Treasury yields and the impact on the future dividend sustainability of AGNC: 1) continued realignment of the company's MBS portfolio; 2) hedging costs (in particular periodic interest costs on interest rate swaps); 3) taxation impact of a net realized gain (loss) on MBS sales; and 4) taxation impact of a net long TBA MBS position.

1) Continued Realignment of AGNC's MBS Portfolio:

The first topic/trend to discuss is the continued realignment of AGNC's MBS portfolio regarding the company's 15 and 30-year fixed-rate agency holdings and the coupons held within each maturity. During the second and third quarters of 2013, AGNC made the conscience decision of lowering the company's exposure to 30-year fixed-rate agency MBS holdings which are more price sensitive to interest rate movements when compared to the company's 15-year fixed-rate agency MBS holdings (when comparing similar coupon rates). However, as AGNC felt mortgage interest rates/U.S. Treasury yields were "overextended" during the fourth quarter of 2013, the company "eased-off" on the continued proportional shift from 30-year fixed-rate agency MBS to 15-year fixed-rate agency MBS.

As mortgage interest rates/U.S. Treasury yields began to reverse course and net decrease in the first quarter of 2014, AGNC became slightly more aggressive regarding the company's proportion of 15 and 30-year fixed-rate agency MBS holdings (when including off-balance sheet TBA MBS positions). As of 6/30/2014, AGNC's management team had continued to take a more "aggressive posture" regarding the company's proportional share of 15 and 30-year fixed-rate agency MBS holdings (when compared to the end of the prior quarter). This was a direct result of the modest net decline of mortgage interest rates/U.S. Treasury yields during the first and second quarters of 2014.

As such, AGNC's weighted average coupon ('WAC') increased (decreased) by 3 basis points ('bps') during the second quarter of 2014 (which was a positive sign; 3.63% versus 3.60%). Including TBA MBS positions, AGNC's proportion of 15-year fixed-rate agency MBS holdings decreased from 50% to 39% during the second quarter of 2014 (based on par value). In comparison, AGNC's proportion of 30-year fixed-rate agency MBS holdings increased from 48% to 59%. Due to the attractive off-balance sheet dollar roll implied financing rates (versus the on-balance sheet repo financing rates), AGNC's net interest rate spread increased (decreased) by 41 bps during the second quarter of 2014 (which was another positive sign; 1.84% versus 1.43%). To show the compositional changes to AGNC's MBS portfolio during the second quarter of 2014, Table 5 is presented below.

Table 5 - AGNC MBS Portfolio Quarterly Compositional Changes (6/30/2014 versus 3/31/2014)

Click to enlarge

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

Side Note: Table 5 above includes the quarterly changes in AGNC's net long (short) TBA MBS position but omits all coupons that had immaterial valuation changes. As such, the coupon balances shown above will not directly equal the "subtotal" balances because I had to hide some coupons so both the 15 and 30-year fixed-rate agency MBS holdings would fit within one table.

Using Table 5 a reference, when comparing AGNC's MBS portfolio as of 6/30/2014 versus 3/31/2014, the company had a net par value increase (decrease) in its 15-year fixed-rate agency holdings with a 2.5%, 3.0%, 3.5%, and 4.0% coupon of ($0.5), ($4.9), ($1.2), and ($0.3) billion, respectively. When all 15-year fixed-rate agency MBS holdings were combined, this was a quarterly net par value increase (decrease) of ($6.9) billion. In comparison, AGNC had a net par value increase (decrease) in the company's 30-year fixed-rate agency MBS holdings with a 3.0%, 3.5%, 4.0%, and 4.5% coupon of $2.6, $1.5, $3.0, and $0.5 billion, respectively. I believe Table 5 above shows AGNC aggressively decreased the company's 15-year fixed-rate agency MBS holdings while increasing its 30-year fixed-rate agency MBS holdings during the second quarter of 2014.

However, since AGNC's on-balance sheet leverage ratio (excluding TBA positions) went from 5.9x as of 3/31/2014 to 5.0x as of 6/30/2014, the company increased (decreased) its on-balance sheet fixed-rate agency MBS portfolio by ($3.2) billion during the second quarter of 2014. This reduction explains the slight decrease in the company's net interest income during the second quarter of 2014. Specifically, AGNC's "cash/coupon interest income" was $541 million during the first quarter of 2014. However, due to the average reduced on-balance sheet MBS portfolio during the second quarter of 2014, AGNC's cash/coupon interest income decreased to only $494 million. When calculated, this was a quarterly increase (decrease) of ($47) million. Simply put, the decrease in cash/coupon interest income negatively affected AGNC's quarterly ERTI during the second quarter of 2014.

However, AGNC reported an increased "cash/coupon interest income and net dollar roll" during the second quarter of 2014. AGNC's cash/coupon interest income and net dollar roll was $589 million during the first quarter of 2014. Due to AGNC's increased net long TBA MBS position during the second quarter of 2014, the company's cash/coupon interest income and net dollar roll increased to $632 million. When calculated, this was a quarterly increase (decrease) of $43 million. Simply put, the increase in cash/coupon interest income and net dollar roll positively affected AGNC's quarterly ERTI and net dollar roll during the second quarter of 2014.

From a future dividend sustainability perspective, the fixed-rate agency MBS portfolio realignment into a greater portion of 30-year fixed-rate agency holdings should be seen as a positive sign as AGNC's WAC, yield, and net interest rate spread increased. AGNC's TBA MBS portfolio will be further discussed, on a "standalone" basis, later in the article.

2) Hedging Costs (Periodic Interest Costs on Interest Rate Swaps):

The second topic/trend to discuss is AGNC's hedging costs associated with the company's derivative portfolio in a general net rising (and falling) interest rate environment. I believe there is one specific hedging cost that should be addressed which has a direct impact on AGNC's quarterly ERTI/ERTI and net dollar roll. This hedging cost is AGNC's "periodic interest costs on interest rate swaps" expense.

Typically, in a general net rising interest rate environment, AGNC will have an elevated "hedging coverage ratio". From a book value ('BV') perspective, an elevated hedging coverage ratio would combat the erosion in MBS valuations if mortgage interest rates/U.S. Treasury yields sharply increase over a relatively short period of time. However, from a future dividend sustainability perspective, an elevated hedging coverage ratio (in this example referring to the portion attributable to interest rate swaps) would be seen as a negative sign. Since an elevated interest rate swaps net (short) position means higher hedging costs under the current environment of extremely low LIBOR, quarterly ERTI/ERTI and net dollar roll and cumulative UTI/cumulative UTI and net dollar roll would be negatively impacted (generally speaking).

Typically, in a general net falling interest rate environment, AGNC will have a reduced hedging coverage ratio if the fear of an increase in mortgage interest rates/U.S. Treasury yields is relatively low. However, it currently appears most of the mREIT sector (including AGNC) has maintained a relatively high hedging coverage ratio for the potential/eventual increase in mortgage interest rates/U.S. Treasury yields. From a BV perspective, the elevated hedging coverage ratio partially mitigates the increase in MBS valuations during a general net falling interest rate environment (which is what occurred during the first and second quarters of 2014). Again, from a future dividend sustainability perspective, the continued elevated net (short) interest rate swaps position (hence a higher periodic costs on interest rate swaps expense) would be seen as a negative sign.

AGNC had a net long (short) interest rate swaps position of ($47.9) billion as of 6/30/2014. The quarterly net notional balance change for this specific derivative sub-account was a net (short) increase of ($1.5) billion or 3% of AGNC's interest rate swaps net long (short) position of ($46.4) billion as of 3/31/2014. AGNC slightly increased the company's overall interest rate swaps net (short) position during the second quarter of 2014 for two main reasons.

First, even though AGNC increased (decreased) the company's on-balance sheet fixed-rate agency MBS portfolio by ($3.2) billion during the second quarter of 2014, management also increased (decreased) the company's TBA MBS portfolio by $3.9 billion. When calculated, AGNC's total fixed-rate agency MBS portfolio (both on and off-balance sheet) increased (decreased) by $0.7 billion during the second quarter of 2014 (discussed in the first topic/trend above). As such, a slightly higher net (short) interest rate swaps position was warranted if AGNC wanted to keep the same general hedging coverage ratio. AGNC slightly decreased the company's net duration gap from 1.2 years as of 3/31/2014 to 1 year as of 6/30/2014. Since AGNC slightly decreased the company's net duration gap, generally a higher total net (short) hedging position is warranted.

Second, AGNC continued to reduce the company's net (short) swaption position during the second quarter of 2014. AGNC had a net long (short) interest rate swaptions position of ($7.0) billion as of 3/31/2014 (based on the notional balance of the underlying interest rate swaps). AGNC had ($1.3) billion of interest rate payer swaption additions and $1.6 billion of interest rate payer swaption exercises, expirations, or terminations during the quarter. AGNC also increased (decreased) the company's interest rate receiver swaptions by $0.8 billion during the quarter. When calculated, AGNC had a net long (short) interest rate swaptions position of only ($5.9) billion as of 6/30/2014. Therefore, AGNC slightly increased the company's net (short) interest rate swaps position in part due to the continued reduction of its net (short) interest rate swaptions position during the second quarter of 2014.

In close correlation to the slight increase in AGNC's net (short) interest rate swaps position, the company's net periodic interest costs of interest rate swaps expense also slightly increased when compared to the prior quarter. AGNC had a net periodic interest costs of interest rate swaps expense of $83 million during the first quarter of 2014. This expense slightly increased to $87 million during the second quarter of 2014. The slight increase in this expense should be seen as a small negative sign regarding AGNC's future dividend sustainability due to the minor decrease in the company's quarterly ERTI/ERTI and net dollar roll.

3) Taxation Impact of a Net Realized Gain (Loss) on MBS Sales:

The third topic/trend to discuss is indirectly related to the first topic/trend discussed above (continued realignment of AGNC's MBS portfolio). AGNC aggressively sold a material proportion of the company's 15 and 30-year fixed-rate agency MBS holdings during the latter-half of 2013. Due to material MBS price decreases across most 15 and 30-year fixed-rate agency holdings throughout the latter-half of 2013, AGNC recorded a net realized gain (loss) of ($1.40) billion during the third and fourth quarters of 2013 (based on GAAP methodologies). After a reclassification of the company's timing differences between the cash and accrual methods of accounting, AGNC recorded a net capital gain (loss) of ($1.79) billion during the third and fourth quarters of 2013 (based on IRC methodologies).

However, as MBS prices began to reverse course in January and February 2014 (which caused past cumulative unrealized losses to decrease), AGNC recorded a net realized gain (loss) of just ($19) million during the first quarter of 2014. After a reclassification of the company's timing differences between the cash and accrual methods of accounting, AGNC recorded a net capital gain (loss) of $102 million during the first quarter of 2014. This same general trend continued during the second quarter of 2014 when AGNC recorded a net realized gain (loss) of $22 million and a net capital gain (loss) of $310 million. As of 12/31/2013, AGNC's cumulative net unrealized gain (loss) on MBS sales balance was ($1.09) billion. As of 6/30/2014, this balance materially reversed course as was now $225 million. For BV purposes, this was a material positive sign.

Now let us move on to the taxation characteristics of AGNC's net capital loss for 2013 (termed "cumulative capital loss carryforward") and show how this balance will have direct impacts on the company's future dividend sustainability. To highlight these impacts, Table 6 is presented below.

Table 6 - AGNC Cumulative Capital Loss Carryforward Analysis

Click to enlarge

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

Using Table 6 above as a reference, due to the fact AGNC did not report a material net capital gain for the first and second quarters of 2013, when management reported the company's net capital gain (loss) of ($1.79) billion during the third and fourth quarters of 2013, a new scenario regarding quarterly ERTI/annual REIT taxable income ('AREITTI') occurred. For the 2013 calendar tax year, AGNC had capital losses (GAAP's equivalent to realized losses on MBS sales) in excess of the company's capital gains (GAAP's equivalent to realized gains on MBS sales). In other words, AGNC recognized a material net capital loss for the 2013 calendar tax year. This 2013 net capital loss is "carried-forward" for up to five calendar tax years. As such, AGNC's cumulative capital loss carryforward from 2013 has both positive and negative implications regarding the future dividend sustainability of the company. The future movements of mortgage interest rates/U.S. Treasury yields will inherently dictate whether positive or negative implications come to fruition regarding the future dividend sustainability of AGNC.

First, let us assume mortgage interest rates/U.S. Treasury yields net increase during a calendar tax year. In this example, all future net capital losses will continue to be "reversed-out" of quarterly ERTI/AREITTI and AGNC's cumulative capital loss carryforward would continue to increase. Still using Table 6 as a reference to show an example, this occurred within AGNC's fourth quarter of 2013. AGNC's cumulative capital loss carryforward was $849 million as of 9/30/2013. However, since AGNC recorded an additional net capital ('gain') loss of $936 million in the fourth quarter of 2013, the company's cumulative capital loss carryforward increased to $1.79 billion as of 12/31/2013. As such, AGNC's ERTI/AREITTI will continue to be positively affected in this scenario because all net capital losses, within a calendar tax year, will continue to be reversed out. All net capital losses would be added to the cumulative capital loss carryforward balance first established in the third quarter of 2013.

This first example would actually help AGNC's future dividend sustainability because all future net capital losses will NOT decrease AGNC's quarterly ERTI/AREITTI for the next five calendar tax years. Do not get me wrong, having net capital losses is never a great scenario for any company. However, purely from a future dividend sustainability perspective, this scenario is positive in nature because future net capital losses will not decrease quarterly ERTI/AREITTI (unlike if a cumulative capital loss carryforward did not exist). Just know a net capital loss is negative for BV purposes.

Side Note: It should be noted each annual capital loss carryforward balance will keep its monetary amounts for capital loss carryforward timetables and expirations. This is an important concept to understand regarding AGNC's future quarterly ERTI/AREITTI and cumulative UTI implications.

Second, let us now assume mortgage interest rates/U.S. Treasury yields reverse course and net decrease during a calendar tax year. If this specific scenario occurs, MBS prices will typically increase. As such, the probability of a net capital gain on AGNC's MBS portfolio inherently increases. In this example, all future net capital gains will have to first be "offset" against AGNC's cumulative capital loss carryforward. Therefore, all net capital gains, up to the cumulative capital loss carryforward balance, would also have to be reversed-out of quarterly ERTI/AREITTI. Once again using Table 6 above as a reference to show an example, this occurred within AGNC's second quarter of 2014. AGNC's cumulative capital loss carryforward was $1.68 billion as of 3/31/2014. Since AGNC recorded a net capital ('gain') loss of ($310) million in the second quarter of 2014, the company's cumulative capital loss carryforward decreased to $1.37 billion as of 6/30/2014.

This second example would actually hurt AGNC's future dividend sustainability because all future net capital gains (up to $1.79 billion over the next five calendar tax years) will NOT increase AGNC's quarterly ERTI/AREITTI. Purely from a future dividend sustainability perspective, this scenario is negative in nature because future net capital gains will not increase quarterly ERTI/AREITTI (unlike if a cumulative capital loss carryforward did not exist). Just know a net capital gain is positive for BV purposes though.

Therefore, I believe this topic/trend has both positive and negative impacts regarding the future dividend sustainability of AGNC and will cause a "buffer" per se. I believe additional MATERIAL increases or decreases to AGNC's dividend should not occur over the foreseeable future as a result of this specific topic/trend because BOTH net capital gains (losses) will be reversed out of quarterly ERTI/AREITTI.

4) Taxation Impact of a Net Long TBA MBS Position:

The final topic/trend to discuss is indirectly related to the first and third topics/trends discussed above (continued realignment of AGNC's MBS portfolio and taxation impact of a net realized gain (loss) on MBS sales). As was discussed earlier, AGNC increased (decreased) the company's on-balance sheet fixed-rate agency MBS portfolio by ($3.2) billion during the second quarter of 2014. This directly caused a decrease in AGNC's cash/coupon interest income and ultimately quarterly ERTI. However, an offset to AGNC's reduced on-balance sheet MBS portfolio was the company's continued increase in its off-balance sheet TBA MBS portfolio during the second quarter of 2014. As of 3/31/2014, AGNC had a net long (short) TBA MBS position of $13.9 billion. AGNC had a net long (short) TBA MBS position of $17.8 billion as of 6/30/2014. When calculated, AGNC increased (decreased) the company's net long TBA MBS position by $3.9 billion during the second quarter of 2014.

For readers who are unfamiliar with this type of derivative instrument, I will briefly describe what a TBA MBS contract is. 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 an extension of the balance sheet, this increases AGNC's "at risk" (both on and off-balance sheet MBS) 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. Regarding all TBA contracts, there are three possible actions to take as the settlement date approaches. To focus on the specific impacts of AGNC's current TBA MBS portfolio (regarding the company's future dividend sustainability), I will only be discussing one of the three possible actions below.

On or within a rather close timeframe to the settlement date, AGNC may decide to "re-roll" certain TBA contracts with a long position by moving the settlement date out to a later timeframe by entering into off-setting (short) positions. This is known as a "pair-off". AGNC net settles the paired-off existing long and newly created (short) TBA contracts in cash and simultaneously purchases similar yet new TBA contracts with a long position which have settlement dates further out into the future. This specific transaction is known as the "dollar roll". The TBA contracts with a long position that are purchased with future settlement dates are usually priced at a discount to regular MBS in the current month (dependent on market conditions). This discount is referred to as the "price-drop". The price drop is the economic equivalent of "net interest carry income" on the underlying MBS over the roll period. Net interest carry income is just a fancier term meaning interest income less implied financing costs (funding the underlying MBS through repurchase ('repo') loans). This is also referred to as AGNC's "net dollar roll income (expense)" generated when a TBA contract with a long (short) position exists.

As initially discussed in PART 1 of this article, due to the specific nature of TBA contracts (type of derivative instrument), the IRC treats all dollar roll income (expense) as a reduction (addition) of one's cost basis versus quarterly ERTI/AREITTI. As such, all related dollar roll income (expense) is classified as an additional capital gain (loss) on the TBA contract. Since AGNC had a cumulative capital loss carryforward of $1.68 billion as of 3/31/2014 (as stated in the third topic/trend above), the company's net dollar roll income (expense) of $138 million during the second quarter of 2014 was offset against this figure. This would generally be negative regarding AGNC's future dividend sustainability because all future generated net dollar roll income would have to be first offset against the company's material cumulative capital loss carryforward and not directly benefit quarterly ERTI/AREITTI.

However, per my interpretation from a quoted statement by AGNC's CIO Gary Kain in PART 1 of this article, the company will be treating net dollar roll income differently than the "technical" tax treatment per the IRC. Based on Mr. Kain's comments during the past two quarterly earnings calls, I believe as long as favorable business conditions persist, AGNC will base the company's dividend payout level on the earnings of the entire MBS portfolio (both on-balance sheet and off-balance sheet portfolios) for the foreseeable future. This means both AGNC's quarterly ERTI and net dollar roll income.

To show the correlation between AGNC's net long (short) TBA MBS position and level of net dollar roll income (expense), Table 7 is presented below.

Table 7 - AGNC Net Long (Short) TBA MBS Position and Dollar Roll Income (Expense) Analysis

Click to enlarge

(Source: Table created entirely by myself, partially using AGNC data obtained from the SEC's EDGAR Database [link provided below Table 6])

Since net dollar roll income is a reduction of AGNC's cost basis on TBA MBS per the IRC, a portion of AGNC's dividend will be classified as a "return of capital" ('ROC') when the company's cumulative UTI is exhausted. For shareholders, the portion of AGNC's dividend attributed to ROC will be nontaxable in the year of distribution and is used to decrease a shareholder's cost basis in the stock (resulting in a higher potential long-term capital gain/less of a long-term capital (loss) when eventually sold). Readers should understand this notion because I believe this adds an additional "layer" per se to the company's dividend sustainability analysis.

As such, this specific topic/trend should be seen as a positive for AGNC's future dividend sustainability as long as favorable business conditions persist. Also, it should be noted that if management decides to "settle" a portion of the company's net long TBA MBS position, these off-balance sheet derivative instruments would become regular on-balance sheet MBS which generate interest income which is considered quarterly ERTI per the IRC.

Brief Discussion of Additional Topics/Trends to Consider in a General Net Rising (and Falling) Interest Rate Environment Regarding the Future Dividend Sustainability of MTGE and NLY:

Regarding the first two additional topics/trends discussed above, MTGE and AGNC had pretty similar MBS and derivative portfolios as of 6/30/2014 (proportionally speaking). One minor exception between the two companies was that MTGE had approximately 16% of the company's MBS portfolio in non-agency holdings (when including the net long TBA MBS position) whereas AGNC is basically a "pure" agency mREIT.

Similar to AGNC, MTGE also shifted the company's fixed-rate agency MBS portfolio back into a higher portion of 30-year fixed-rate agency holdings when compared to the prior quarter. Including TBA MBS positions, MTGE's proportion of 15-year fixed-rate agency MBS holdings decreased from 41% to 35% during the second quarter of 2014 (based on par value). In comparison, MTGE's proportion of 30-year fixed-rate agency MBS holdings increased from 56% to 62%.

Also similar to AGNC, MTGE increased the company's net (short) interest rate swaps notional balance during the second quarter of 2014. MTGE had a net long (short) interest rate swaps position of ($4.1) billion as of 6/30/2014. This was a net (short) increase of ($0.1) billion or 1% of MTGE's interest rate swaps net long (short) position of ($4.0) billion as of 3/31/2014. As such, MTGE's periodic interest costs on interest rate swaps expense could negatively impact the company's quarterly ERTI/ERTI and net dollar roll in future quarters if mortgage interest rates/U.S. Treasury yields were to rise once again and MTGE adds to the company's net (short) notional balance to mitigate BV erosion. However, MTGE may keep the company's hedging coverage ratio at a slightly lower level when compared to AGNC now that the acquisition of RCS, a fully-licensed mortgage servicer (also known as a "mortgage servicing rights" ('MSR') company), acts as a indirect hedge per say. As of 6/30/2014, AGNC and MTGE had a hedging coverage ratio of 88% and 86%, respectively.

Regarding the third topic/trend, when compared to AGNC, MTGE also had a very similar taxation impact from the company's net capital loss in 2013. As was the case with AGNC in 2013, MTGE had capital losses in excess of capital gains. MTGE recorded a net capital loss of ($195.2) million for the 2013 calendar tax year. Similar to the taxation methodology for AGNC, capital losses in excess of capital gains are NOT deductible from MTGE's ERTI/AREITTI and are allowed to be carried forward for up to five calendar tax years. As was the case with AGNC, MTGE's current situation has both positive and negative implications regarding the future dividend sustainability of the company. The future movements of mortgage interest rates/U.S. Treasury yields will inherently dictate whether positive or negative implications come to fruition regarding the future dividend sustainability of MTGE. This could actually help MTGE's future dividend sustainability because all future net capital losses will NOT decrease MTGE's quarterly ERTI/AREITTI for the next five calendar tax years. However, this could also actually hurt MTGE's future dividend sustainability because all future net capital gains (up to $195.1 million over the next five calendar tax years) will NOT increase MTGE's quarterly ERTI/AREITTI.

MTGE's cumulative capital loss carryforward was $215 million as of 3/31/2014. Since MTGE recorded a net capital ('gain') loss of ($35) million in the second quarter of 2014, the company's cumulative capital loss carryforward decreased to $180 million as of 6/30/2014. When compared to AGNC, MTGE had a slightly lower cumulative capital loss carryforward per share balance as of 6/30/2014.

Regarding the fourth topic/trend, MTGE also had a similar movement within the company's TBA MBS portfolio. As of 3/31/2014, MTGE had a net long (short) TBA MBS position of $0.7 billion. MTGE had a net long (short) TBA MBS position of $1.1 billion as of 6/30/2014. When calculated, MTGE increased (decreased) the company's net long TBA MBS position by $0.4 billion during the second quarter of 2014. Similar to the taxation methodology for AGNC, all related dollar roll income (expense) is classified as an additional capital gain (loss) on the TBA contract. Since MTGE had a cumulative capital loss carryforward balance of $215 million as of 3/31/2014, the company's net dollar roll income (expense) of $8 million during the second quarter of 2014 was offset against this figure. This would generally be negative regarding MTGE's future dividend sustainability because all future net dollar roll income will have to be offset against the material cumulative capital loss carryforward balance and not directly benefit quarterly ERTI/AREITTI.

However, similar to AGNC's rhetoric, it was indicated MTGE will be treating net dollar roll income differently than the technical tax treatment per the IRC. I believe as long as favorable business conditions persist, MTGE will base the company's dividend payout level on the earnings of the entire MBS portfolio (both on-balance sheet and off-balance sheet portfolios) for the foreseeable future. This means both MTGE's quarterly ERTI and net dollar roll income. Now let us compare NLY to AGNC.

Simply put, NLY reported one of the best results in the mREIT sector for the second quarter of 2014. NLY even slightly outperformed AGNC regarding quarterly BV gains (by 0.34%). However, even though both companies reported "above average" quarterly results, AGNC and NLY had implemented differing strategies regarding each company's MBS and derivative portfolios.

First, while AGNC relied on the company's net long TBA MBS portfolio to enhance returns while keeping a high hedging coverage ratio, NLY enhanced returns by keeping the relatively same MBS portfolio but drastically decreasing the company's hedging coverage ratio. NLY had a net long (short) interest rate swaps position of ($30.8) billion as of 6/30/2014. This was a net (short) decrease of $25.9 billion or (46%) of NLY's interest rate swaps net long (short) position of ($56.7) billion as of 3/31/2014. Equally important, NLY did not increase the company's net (short) interest rate swaption balance or utilize (short) U.S. Treasury security positions to offset the material reduction of its interest rate swaps net (short) position. Granted, NLY can always materially increase this position when the risk of mortgage interest rates/U.S. Treasury yields begin to show signs of materially increasing. However, readers should understand if management does not correctly time a rise in mortgage interest rates/U.S. Treasury yields, NLY is at a much higher level of risk when compared to AGNC regarding BV erosion. It should also be noted NLY's materially lower interest rate swaps net (short) position as of 6/30/2014 will directly benefit the company's future dividend sustainability. In my opinion, this is a risky strategy. However, with that being said, higher risk can ultimately lead to higher potential rewards.

When compared to AGNC, NLY did not have the same level of net realized losses on MBS sales during 2013. However, NLY is subject to the same taxation methodologies as AGNC. Therefore, the same implications discussed in the third topic/trend above in regards to net capital losses (hence a cumulative capital loss carryforward balance) would also apply to NLY.

Regarding the fourth topic/trend above, AGNC and NLY had material differences in each company's TBA MBS portfolio. As discussed earlier, AGNC increased (decreased) the company's net long TBA MBS portfolio by $3.9 billion during the second quarter of 2014. In sharp contrast, NLY had a net long (short) TBA MBS position of ($0.7) billion as of 3/31/2014. NLY continued to have a net long (short) TBA MBS position of ($0.7) billion as of 6/30/2014. When calculated, NLY did not change the company's net (short) TBA MBS position during the second quarter of 2014. Since NLY actually had a minor net (short) TBA MBS position as of 6/30/2014, the probability of the company generating net dollar roll income during the third quarter of 2014 is relatively low.

Since NLY had core earnings of $0.30 per common share while distributing dividends of $0.30 per share during the second quarter of 2014, the company cannot offset any overpayment of core earnings by stating net dollar roll income will be a component along with quarterly ERTI (also known as ordinary taxable income).

Conclusions Drawn From PART 1 and PART 2:

In PART 1 of this analysis, I concluded the probability of AGNC being able to maintain the company's current dividend of $0.65 per share has remained relatively unchanged when compared to the prior quarter. Due to the attractive yields and net dollar roll income being generated from AGNC's net long TBA MBS position, the company has stated its dividend will continue to be based on the earnings of the entire MBS portfolio (both on-balance sheet and off-balance sheet portfolios) for the foreseeable future.

Regarding PART 2 of this analysis, AGNC's quarterly ERTI/AREITTI would be adversely affected by the following topics/trends in conjunction with a general net rising (and falling) interest rate environment: 1) an overall decrease in the MBS portfolio's WAC if the company once again increases the proportion of 15-year fixed-rate agency holdings when compared to 30-year fixed-rate agency holdings (to protect BV in a net rising interest rate environment); 2) the continued deleveraging of the on-balance sheet MBS portfolio which would directly lead to less cash/coupon interest income; 3) an elevated net (short) interest rate swaps position which leads to a higher periodic interest costs on interest rate swaps expense; 4) the reversal of all net capital gains over the next five calendar tax years in relation to the 2013 cumulative capital loss carryforward of ($1.79) billion (remaining balance of $1.37 billion as of 6/30/2014); and 5) a decrease in net dollar roll income from a reduced net long TBA MBS position.

However, AGNC's quarterly ERTI/AREITTI could be positively affected by the following trends in conjunction with a general net rising (and falling) interest rate environment: 1) an overall increase in the MBS portfolio's WAC if the company continues to increase the proportion of 30-year fixed-rate agency holdings when compared to 15-year fixed-rate agency holdings (to enhance BV in a net falling interest rate environment); 2) an increase in leverage of the on-balance sheet MBS portfolio which would directly lead to more cash/coupon interest income; 3) a reduced net (short) interest rate swaps position which leads to a lower periodic interest costs on interest rate swaps expense; 4) the reversal of all net capital losses over the next five calendar tax years in relation to the 2013 cumulative capital loss carryforward of ($1.79) billion; and 5) management's recent consideration (due to the capital loss carryforward balance for 2013) that net dollar roll income generated from a net long TBA MBS position will be treated as an additional layer of dividend support even though the IRC's interpretation states it is treated as a reduction of one's cost basis.

When combining the results from PART 1 of this article with the additional topics/trends discussed above, I continue to believe AGNC should be able to maintain the company's quarterly dividend rate of $0.65 per share throughout 2014 (even if a portion of the 2014 dividend would be treated as a ROC distribution at the shareholder level). I also currently project stable 2014 dividend per share rates for MTGE and NLY.

I continue to believe the probability of a MATERIAL dividend increase for AGNC, MTGE, or NLY remains fairly low (below 20%) going into 2015.

Currently, I believe AGNC is attractively priced at $23.42 per share as of 8/15/2014. I rate AGNC, MTGE, and NLY as a SOLID HOLD when trading at a minor (under 5%) discount to CURRENT BV. I rate AGNC, MTGE, and NLY as a BUY when I believe each company's stock price is trading a modest (at or over 5% but under 10%) discount to CURRENT BV. I rate AGNC, MTGE, and NLY as a STRONG BUY when I believe each company's stock price is trading at a material (at or over 10%) discount to CURRENT BV. As such, I currently rate AGNC, MTGE, and NLY as a BUY/STRONG BUY since each company's stock price is trading at an approximate (9% - 11%) premium (discount) to CURRENT BV (BV as of 8/15/2014).

Final Note: This article excludes a detailed analysis of AGNC's dividend range scenarios for the third quarter of 2014. The article I have written above focuses on AGNC's dividend sustainability in general, regardless of what quarter any dividend change might occur. As such, it is more of a broad discussion of AGNC's future dividend sustainability and tends to look over the next several quarters. A future article will specifically focus on AGNC's dividend range scenarios for the third quarter of 2014. This future article will be released sometime next month.

Each investor's BUY, SELL, or HOLD decision is based on one's risk tolerance, time horizon, and dividend income goals. My personal recommendation may not fit each investor's current investing strategy.

Disclosure: The author is long AGNC, MTGE.

The author wrote this article themselves, and it expresses their own opinions. The author is not receiving compensation for it (other than from Seeking Alpha). The author has no business relationship with any company whose stock is mentioned in this article.

Additional disclosure: I have no position in NLY.