What changed in this filing

/stocks — the workspaceLOADING

Item 2. Management’s Discussion and Analysis

2020-11-05 compared with 2020-08-07 · 146 added, 148 removed, 560 unchanged (34% of the section changed)

Management's Discussion and Analysis of Financial Condition and Results of Operations

Management's Discussion and Analysis of Financial Condition and Results of Operations ("MD&A") is designed to provide a reader of AGNC Investment Corp.'s consolidated financial statements with a narrative from the perspective of management and should be read in conjunction with the consolidated financial statements and accompanying notes included in this Quarterly Report on Form 10-Q for quarterly period ended June 30, 2020.

Our MD&A is presented in six sections:

24 unchanged sentences

Trends and Recent Market Impacts

On March 11, 2020, the World Health Organization declared COVID-19 a global pandemic (the "Pandemic"), with President Trump declaring the health crisis a national emergency shortly thereafter on March 13, 2020.

Stay-at-home orders, school closures and widespread business shutdowns adversely affected worldwide economies and sparked concerns that containment measures could lead to severe unemployment and a contraction of U.S.

gross domestic product output at rates not experienced since the Great Depression.

The ensuing financial market turmoil had a material negative impact on our business and financial results for the first quarter, resulting in a first quarter economic loss on our tangible common equity of -20.2%.

Market conditions improved considerably during the second quarter, as unprecedented monetary and fiscal stimulus drove a broad rebound in most asset categories.

Market liquidity improved substantially after declining to unprecedented levels in March.

Equity markets retraced much of their March decline, and credit spreads for most fixed income products recovered a substantial portion of their first quarter widening.

Since the peak of the financial crisis in March, the Federal Reserve’s ongoing purchases of Agency RMBS have been supportive of the Agency mortgage market.

The performance of Agency RMBS, however, was somewhat mixed during the second quarter.

Lower coupon generic Agency RMBS outperformed higher coupon generic securities, with the latter being negatively impacted by elevated prepayment concerns.

Meanwhile, specified Agency RMBS significantly outperformed generic securities as specified pool "pay-up" values recovered a substantial portion of their March decline, which was the primary driver of the increase in our tangible net book value for the second quarter.

For the second quarter, our tangible net book value increased 9.5% to $14.92 per common share as of June 30, 2020, from $13.62 per common share as of March 31, 2020.

Including dividends per common share, our economic return on tangible common equity was 12.2% for the second quarter.

After a steep drop in rates during the first quarter, interest rate volatility was muted during the second quarter.

The 10-year U.S.

Treasury rate ended the second quarter nearly unchanged from March 31, 2020, while the short end of the yield curve shifted lower as the U.S.

Federal Reserve communicated its intention to hold the federal funds rate near zero for several years.

Contrary to earlier expectations and despite widespread Pandemic related lockdowns and social distancing measures, refinancing activity and housing turnover were robust during the second quarter and led to a significant increase in prepayment rates for the quarter.

Our weighted average portfolio CPR increased to 19.9% for the second quarter, compared to 12.2% for the first quarter, while our weighted average projected CPR for the remaining life of our securities increased to 16.6% as of June 30, 2020, from 14.5% as of March 31, 2020 and 10.8% as of December 31, 2019, as mortgage rates declined to historically low levels during the quarter.

During the second quarter, we reduced our holdings of higher quality specified pools after their relative outperformance earlier in the quarter and we reduced our holdings of higher coupon, generic RMBS in favor of lower coupon, new production Agency RMBS.

We also shifted a larger portion of our holdings from 30-year to 15-year fixed rate Agency RMBS.

As of June 30, 2020, the weighted average coupon on our fixed rate securities, inclusive of TBA securities, was 3.40%, compared to 3.62% as of March 31, 2020 and 3.60% as of December 31, 2019.

Additionally, the combination of heavy origination volume and large-scale Fed purchases during the second quarter favored the TBA dollar roll market.

With implied financing rates in the dollar roll market well below comparable repo levels for most coupons, we increased our average TBA position to $15.7 billion for the second quarter, from $7.5 billion for the first quarter.

As of June 30, 2020, our net TBA position totaled $20.5 billion.

The repo markets continued to perform very well during the second quarter.

The Fed’s open market repo operations, strong demand for Agency RMBS collateral and the Fed’s apparent willingness to keep the fed funds rate near the zero bound range, have led to a meaningful repricing of repo across all tenors.

As a result, our weighted average repo rate for the second quarter declined to 0.766% from 1.80% for the first quarter, and our weighted average repo rate as of June 30, 2020 declined to 0.41%, as an increasing portion of our repo have reset to lower prevailing rates.

With our near-term hedging requirements somewhat lower given the shorter duration of our assets and muted interest rate volatility, our interest rate hedge position decreased to 66% of our funding liabilities, inclusive of our net TBA position (at cost), as of June 30, 2020, compared to 70% as of March 31, 2020 and 102% as of December 31, 2019.

With swap rates declining to historically low levels, we have also continued to adjust the composition of our interest rate swap portfolio, terminating shorter-term swaps and replacing them with longer-dated swaps.

The average pay rate on our interest rate swaps decreased to 0.39% as of June 30, 2020, from 0.94% as of March 31, 2020 and 1.29% as of December 31, 2019, while the average maturity of our interest rate swap portfolio increased to 5.1 years as of June 30, 2020, from 4.5 years as of March 31, 2020 and 2.7 years as of December 31, 2019.

Our duration gap, which is a measure of the difference between the interest rate sensitivity of our assets and liabilities, inclusive of interest rate hedges, was -0.1 year as of June 30, 2020, compared to 0.0 and 0.4 years, as of March 31, 2020 and December 31, 2019, respectively.

Our aggregate cost of funds for the second quarter, inclusive of TBA implied financing rates and interest rate swaps, decreased to 0.88%, from 1.67% for the first quarter, more than offsetting lower asset yields.

Our net interest spread, inclusive of TBAs and swaps and excluding catch-up premium amortization, increased to 1.68% for the second quarter, from 1.30% for the first quarter.

The improvement in our net interest margin offset the decline in our average portfolio balance and tangible leverage for the second quarter, driving the slight increase in our net spread and dollar roll income to $0.58 per common share for the second quarter, excluding catch-up amortization, compared to $0.57 per common share for the first quarter.

Our average leverage for the second quarter declined to 8.8x tangible equity, compared to 9.9x for the first quarter.

As of June 30, 2020, our leverage was 9.2x tangible equity, compared to 9.4x as of March 31, 2020 and December 31, 2019.

Following the actions we took during the first quarter to prioritize liquidity and risk management, our liquidity position remained strong during the second quarter with cash and unencumbered Agency RMBS totaling $4.5 billion as of June 30, 2020, which excludes unencumbered CRT and non-Agency securities as well as assets held at our broker-dealer subsidiary Bethesda Securities, LLC.

As we look ahead, we believe liquidity concerns during the first quarter have given way to fundamental performance metrics.

Although prepayments are likely to somewhat negatively impact earnings, the broad availability of funding at rates near zero and muted interest rate volatility are positives for our business, and we expect the earnings environment for Agency RMBS to remain favorable for the foreseeable future despite the ongoing economic uncertainties associated with the COVID-19 pandemic.

Nonetheless, longer-term risks remain, and you should carefully review our Risk Factors and additional information regarding our interest rate and spread sensitivity in our Quantitative and Qualitative Disclosures about Market Risk located in this Form 10-Q.

Despite continued uncertainty regarding the COVID-19 virus and the broader macroeconomic outlook, financial markets posted significant gains for the second quarter in a row since the peak of the financial dislocations in March, driven in large part by the monetary support provided by global central banks.

During the third quarter, stable interest rates and incremental purchases of Agency RMBS by the Federal Reserve (the "Fed") of approximately $40 billion per month, supplementing its reinvestment of paydowns on its existing portfolio, were supportive of the Agency mortgage market.

As mortgage rates maintain historically low levels, the prepayment risk for mortgage investors remains elevated but appears to have plateaued for several cohorts.

The size of our investment portfolio, including TBA securities, for the third quarter was largely unchanged at $97.6 billion as of September 30, 2020, compared to $97.7 billion as of the end of the second quarter.

We continued to shift the composition of our portfolio to lower coupon 30- and 15-year TBA securities with our higher coupon holdings concentrated in high quality, specified Agency RMBS pools, which have slower prepayment characteristics than similar coupon generic securities.

Inclusive of TBA securities, the weighted average coupon on our fixed rate securities declined to 3.08% as of September 30, 2020 from 3.40% as of June 30, 2020.

Our investment portfolio outperformed relative to our interest rate hedges during the third quarter.

As a result, our tangible net book value per common share increased to $15.88 per share as of September 30, 2020 from $14.92 per share as of June 30, 2020, an increase of 6.4%.

The increase in book value was the primary driver of a decline in our “at risk” leverage to 8.8x from 9.2x tangible common equity as of September 30 and June 30, 2020, respectively.

Our economic return on tangible common equity was 8.8% for the quarter, which includes both the increase in our tangible net asset value and dividends declared of $0.36 per common share.

The TBA dollar roll market benefited in the third quarter from the combination of heavy origination volume and ongoing Fed purchases.

As such, implied financing rates in the dollar roll market were well below comparable repo levels for most coupons.

Given these favorable financing terms, we increased our average TBA position to $27.8 billion for the quarter from $15.7 billion for the second quarter.

As of September 30, 2020, our TBA position totaled $29.5 billion, or approximately 30% of our investment portfolio.

Our weighted average repo rate for the third quarter declined to 0.40% from 0.76% for the second quarter.

Our aggregate cost of funds, which includes the implied funding cost associated with our TBA position and interest rate swap costs, declined to 0.15% for the third quarter, compared to 0.88% for the second quarter.

The combination of strong TBA dollar roll performance and very low funding rates more than offset lower asset yields on our portfolio for the quarter and drove a significant increase in our net spread and dollar roll income, excluding "catch-up" premium amortization cost, a non-GAAP measure, to $0.81 per common share for the third quarter from $0.58 per common share for the second quarter.

(Refer to Results of Operations below for further information regarding non-GAAP measures.)

Our net interest margin, including TBA securities and interest rate swaps and excluding "catch-up" premium amortization, increased to 2.15% for the third quarter from 1.68% for the second quarter.

Looking ahead, the Fed's recently announced policy shift to average inflation targeting should cause the current low cost funding environment to persist for a prolonged period and continue to benefit our net interest margin.

Notwithstanding this, our net interest margin is likely biased marginally lower, as we expect the implied funding advantage of TBA dollar rolls to contract over time to more modest levels and our average asset yield to decline as portfolio turnover is reinvested at lower prevailing asset yields.

Our investment portfolio, excluding TBA holdings, prepaid at an annualized rate of 24.3% during the third quarter, compared to 19.9% during the second quarter.

Our weighted average projected CPR for the remaining life of our securities decreased to 15.9% as of September 30, 2020 from 16.6% as of June 30, 2020 largely due to changes in asset composition.

Our liquidity position, consisting of unencumbered Agency RMBS and cash, totaled $5.2 billion at the end of the third quarter, which excludes unencumbered CRT securities, non-Agency securities and assets held at our broker-dealer subsidiary.

Our interest rate hedge position increased to 71% of our funding liabilities, inclusive of our net TBA position (at cost), as of September 30, 2020, compared to 66% as of June 30, 2020.

Our duration gap, which is a measure of the difference between the interest rate sensitivity of our assets and liabilities, inclusive of all of our interest rate hedges, was largely unchanged during the third quarter at 0.0 years as of September 30, 2020 compared to -0.1 year as of June 30, 2020.

Our interest rate swap portfolio totaled $43.0 billion as of September 30, 2020, largely unchanged from $42.1 billion as of June 30, 2020.

We continued to alter the composition of our swap portfolio during the third quarter, shifting to SOFR-indexed swaps and extending the maturity profile to 5.3 years as of September 30, 2020 from 5.1 years as of June 30, 2020.

SOFR-indexed swaps are expected to highly correlate to changes in repo funding levels.

As of September 30, 2020, 69% of our interest rate swaps were indexed to SOFR and the remainder to OIS.

As of June 30, 2020, 16%, 79% and 5% of our interest rate swaps were indexed to SOFR, OIS and 3M LIBOR, respectively.

Our interest rate swap repositioning in recent quarters also led to a decline in our average swap cost, as the majority of our interest swaps now reflect prevailing market rates.

For the third quarter, the average pay rate on our interest rates swaps decreased to 0.15% as of September 30, 2020 from 0.39% as of June 30, 2020.

Although the ongoing economic uncertainties associated with the COVID-19 pandemic and U.S.

fiscal policy continue as risks to the financial markets and prepayments will likely remain elevated, our earnings profile should continue to benefit from the broad availability of repo funding at rates near zero, muted interest rate volatility and continued Fed acquisitions of Agency RMBS.

As such, in the current environment, we remain optimistic about our ability to generate positive risk-adjusted returns for our stockholders through a combination of dividends and book value growth over the near term.

Market Information

1 unchanged sentence

Interest Rate/Security Price 1

June 30, 2019 Sept.

30, 2019 Dec.

31, 2019 Mar.

31, 2020 June 30, 2020 June 30, 2020

31, 2020 June 30, 2020 Sept.

30, 2020 Sept.

June 30, 2020

26 unchanged sentences

$104.08 $104.28 $105.56 $105.75 $106.15 +$0.40 +$1.87

________________________________

Price information is for generic instruments only and is not reflective of our specific portfolio holdings.

4 unchanged sentences

FINANCIAL CONDITION

As of June 30, 2020 and December 31, 2019, our investment portfolio consisted of $77.1 billion and $100.4 billion of investment securities, at fair value, respectively, and $20.5 billion and $7.4 billion of TBA securities, at fair value, respectively.

The following table is a summary of our investment portfolio as of June 30, 2020 and December 31, 2019 (dollars in millions):

June 30, 2020 December 31, 2019

As of September 30, 2020 and December 31, 2019, our investment portfolio consisted of $68.0 billion and $100.4 billion of investment securities, at fair value, respectively, and $29.5 billion and $7.4 billion of TBA securities, at fair value, respectively.

The following table is a summary of our investment portfolio as of September 30, 2020 and December 31, 2019 (dollars in millions):

September 30, 2020 December 31, 2019

Investment Portfolio (Includes TBAs) Amortized Cost Fair Value Average Coupon % Amortized Cost Fair Value Average Coupon %

26 unchanged sentences

TBA securities are presented net of long and short positions.

As of June 30, 2020, 30-year TBA securities consisted of $12.9 billion long and $(26.0) million short TBA securities (at fair value) at an average coupon of 2.59% and 2.50%, respectively, and 15-year TBA securities consisted of entirely long TBA securities at an average coupon of 2.11%.

As of December 31, 2019, 30-year TBA securities consisted of $6.8 billion long and $(1.6) billion short TBA securities at an average coupon of 3.17% and 4.00%, respectively, and 15-year TBA securities consisted entirely of long TBA securities at an average coupon of 2.91%.

For further details of our TBA securities held as of each date refer to Note 6 of the accompanying consolidated financial statements.

For further details of our TBA securities refer to Note 6 of the accompanying consolidated financial statements.

TBA securities are recorded as derivative instruments in our accompanying consolidated financial statements and our TBA dollar roll transactions represent a form of off-balance sheet financing.

As of June 30, 2020 and December 31, 2019, our TBA positions had a net carrying value of $130 million and $25 million, respectively, reported in derivative assets /(liabilities) on our accompanying consolidated balance sheets.

As of September 30, 2020 and December 31, 2019, our TBA positions had a net carrying value of $76 million and $25 million, respectively, reported in derivative assets /(liabilities) on our accompanying consolidated balance sheets.

The net carrying value represents the difference between the fair value of the underlying Agency security in the TBA contract and the contract price to be paid or received for the underlying Agency security.

As of June 30, 2020 and December 31, 2019, the weighted average yield on our investment securities (excluding TBA securities) was 2.64% and 3.07%, respectively.

The following tables summarize certain characteristics of our fixed rate Agency RMBS portfolio, inclusive of TBAs, as of June 30, 2020 and December 31, 2019 (dollars in millions):

June 30, 2020

As of September 30, 2020 and December 31, 2019, the weighted average yield on our investment securities (excluding TBA securities) was 2.56% and 3.07%, respectively.

The following tables summarize certain characteristics of our fixed rate Agency RMBS portfolio, inclusive of TBAs, as of September 30, 2020 and December 31, 2019 (dollars in millions):

September 30, 2020

Includes Net TBA Position Excludes Net TBA Position

21 unchanged sentences

4.0% 17,014 17,784 18,841 93% 104.5% 4.51% 2.86% 50 17%

≥ 4.5% 9,514 10,002 10,682 98% 105.1% 5.01% 3.14% 35 19%

Total 30-year 71,222 74,179 76,943 56% 104.4% 4.23% 2.58% 42 16%

2 unchanged sentences

Specified pools include pools backed by lower balance loans with original loan balances of up to $200K, HARP pools (defined as pools that were issued between May 2009 and December 2018 and backed by 100% refinance loans with original LTVs ≥ 80%), and pools backed by loans 100% originated in New York and Puerto Rico.

As of June 30, 2020, lower balance specified pools had a weighted average original loan balance of $118,000 and $118,000 for 15-year and 30-year securities, respectively, and HARP pools had a weighted average original LTV of 126% and 136% for 15-year and 30-year securities, respectively.

As of September 30, 2020, lower balance specified pools had a weighted average original loan balance of $115,000 and $117,000 for 15-year and 30-year securities, respectively, and HARP pools had a weighted average original LTV of 126% and 137% for 15-year and 30-year securities, respectively.

WAC represents the weighted average coupon of the underlying collateral.

Portfolio yield incorporates a projected life CPR based on forward rate assumptions as of June 30, 2020.

Portfolio yield incorporates a projected life CPR based on forward rate assumptions as of September 30, 2020.

December 31, 2019

31 unchanged sentences

Portfolio yield incorporates a projected life CPR based on forward rate assumptions as of December 31, 2019.

As of June 30, 2020 and December 31, 2019, our investments in CRT and non-Agency securities had the following credit ratings:

June 30, 2020 December 31, 2019

As of September 30, 2020 and December 31, 2019, our investments in CRT and non-Agency securities had the following credit ratings:

September 30, 2020 December 31, 2019

CRT and Non-Agency Security Credit Ratings 1

28 unchanged sentences

The following selected financial data is derived from our interim consolidated financial statements and the notes thereto.

The tables below present our condensed consolidated balance sheets as of June 30, 2020 and December 31, 2019 and our condensed consolidated statements of comprehensive income and key statistics for the three and six months ended June 30, 2020 and 2019 (in millions, except per share amounts):

The tables below present our condensed consolidated balance sheets as of September 30, 2020 and December 31, 2019 and our condensed consolidated statements of comprehensive income and key statistics for the three and nine months ended September 30, 2020 and 2019 (in millions, except per share amounts):

($ in Millions, Except Per Share Amounts)

Balance Sheet Data June 30,

Balance Sheet Data September 30,

December 31, 2019

9 unchanged sentences

Three Months Ended

Six Months Ended

September 30,

Nine Months Ended

September 30,

Statement of Comprehensive Income Data (Unaudited) 2020 2019 2020 2019

20 unchanged sentences

Three Months Ended

Six Months Ended

September 30,

Nine Months Ended

September 30,

Other Data (Unaudited) * 2020 2019 2020 2019

32 unchanged sentences

Economic Interest Income and Asset Yields

The following table summarizes our economic interest income (a non-GAAP measure) for the three and six months ended June 30, 2020 and 2019, which includes the combination of interest income (a GAAP measure) on our holdings reported as investment securities on our consolidated balance sheets, adjusted to exclude estimated "catch-up" premium amortization adjustments for the cumulative effect from prior reporting periods of changes in our CPR forecast, and implied interest income on our TBA securities (dollars in millions):

Three Months Ended June 30,

Six Months Ended June 30,

The following table summarizes our economic interest income (a non-GAAP measure) for the three and nine months ended September 30, 2020 and 2019, which includes the combination of interest income (a GAAP measure) on our holdings reported as investment securities on our consolidated balance sheets, adjusted to exclude estimated "catch-up" premium amortization adjustments for the cumulative effect from prior reporting periods of changes in our CPR forecast, and implied interest income on our TBA securities (dollars in millions):

Three Months Ended September 30,

Nine Months Ended September 30,

2020 2019 2020 2019

26 unchanged sentences

The principal elements impacting our economic interest income are the size of our average investment portfolio and the yield (actual and implied) on our securities.

The following table includes a summary of the estimated impact of each of these elements on our economic interest income for the three and six months ended June 30, 2020 compared to the prior year period (in millions):

The following table includes a summary of the estimated impact of each of these elements on our economic interest income for the three and nine months ended September 30, 2020 compared to the prior year period (in millions):

Impact of Changes in the Principal Elements Impacting Economic Interest Income

Periods ended June 30, 2020 vs.

June 30, 2019

Periods ended September 30, 2020 vs.

September 30, 2019

Due to Change in Average

7 unchanged sentences

Economic interest income, excluding "catch-up" amortization (non-GAAP measure) $ (284) $ (70) $ (214)

Six months ended:

Nine months ended:

Interest Income (GAAP measure) $ (790) $ (315) $ (475)

3 unchanged sentences

Economic interest income, excluding "catch-up" amortization (non-GAAP measure) $ (604) $ (141) $ (463)

Our average investment portfolio, inclusive of TBAs, decreased -16% and -5% (at cost) for the three and six months ended June 30, 2020, respectively, compared to the prior year periods primarily due to lower operating leverage.

The decrease in our average asset yield for the three and six months ended June 30, 2020 was due to the combination of changes in asset composition and higher premium amortization cost resulting from faster prepayment expectations.

Our average investment portfolio, inclusive of TBAs (at cost), decreased -11% and -7% for the three and nine months ended September 30, 2020, respectively, compared to the prior year periods primarily due to lower operating leverage.

The decrease in our average asset yield for the three and nine months ended September 30, 2020 was due to the combination of changes in asset composition and higher premium amortization cost resulting from faster prepayment expectations.

Our primary measure of leverage is our tangible net book value "at risk" leverage ratio, which is measured as the sum of our repurchase agreements and other debt used to fund our investment securities and net TBA position (at cost) (together referred to as "mortgage borrowings") and our net receivable/payable for unsettled investment securities, divided by our total stockholders' equity adjusted to exclude goodwill and other intangible assets.

8 unchanged sentences

and Other Debt 1

Net TBA Position Long/(Short) 2

Net TBA Position

Long/(Short) 2

Average Tangible Net Book Value

6 unchanged sentences

Amount Ending

September 30, 2020 $ 61,008 $ 69,628 $ 54,558 $ 27,785 $ 29,460 8.9:1 8.8:1

June 30, 2020 $ 69,552 $ 72,399 $ 69,370 $ 15,662 $ 20,413 8.8:1 9.2:1

12 unchanged sentences

Economic Interest Expense and Aggregate Cost of Funds

The following table summarizes our economic interest expense and aggregate cost of funds (non-GAAP measures) for the three and six months ended June 30, 2020 and 2019 (dollars in millions), which includes the combination of interest expense on Agency repurchase agreements and other debt (GAAP measure), implied financing cost (benefit) of our TBA securities and interest rate swap periodic interest (income) cost:

Three Months Ended June 30,

Six Months Ended June 30,

The following table summarizes our economic interest expense and aggregate cost of funds (non-GAAP measures) for the three and nine months ended September 30, 2020 and 2019 (dollars in millions), which includes the combination of interest expense on Agency repurchase agreements and other debt (GAAP measure), implied financing cost (benefit) of our TBA securities and interest rate swap periodic interest cost (income):

Three Months Ended September 30,

Nine Months Ended September 30,

2020 2019 2020 2019

4 unchanged sentences

(41) (0.58) % 52 2.00 % (13) (0.10) % 178 2.32 %

Economic interest expense - before interest rate swap periodic (income) costs, net 4

Economic interest expense - before interest rate swap periodic cost (income), net 4

21 0.09 % 609 2.43 % 609 0.87 % 1,846 2.55 %

Interest rate swap periodic interest (income) cost, net 2,5

Interest rate swap periodic interest cost (income), net 2,5

13 0.06 % (146) (0.58) % 41 0.06 % (317) (0.44) %

10 unchanged sentences

The principal elements impacting our economic interest expense are (i) the size of our average mortgage borrowings and interest rate swap portfolio outstanding during the period, (ii) the average interest rate (actual and implied) on our mortgage borrowings and (iii) the average net interest rate paid/received on our interest rate swaps.

The following table includes a summary of the estimated impact of these elements on our economic interest expense for the three and six months ended June 30, 2020 compared to the prior year period (in millions):

The following table includes a summary of the estimated impact of these elements on our economic interest expense for the three and nine months ended September 30, 2020 compared to the prior year period (in millions):

Impact of Changes in the Principal Elements of Economic Interest Expense

Periods ended June 30, 2020 vs.

June 30, 2019

Periods ended September 30, 2020 vs.

September 30, 2019

Due to Change in Average

5 unchanged sentences

Total change in economic interest expense $ (429) $ (11) $ (418)

Six months ended:

Nine months ended:

Repurchase agreements and other debt interest expense $ (1,046) $ (210) $ (836)

2 unchanged sentences

Total change in economic interest expense $ (879) $ (51) $ (828)

Our average mortgage borrowings, inclusive of TBAs, decreased by -13% and -1% for the three and six months ended June 30, 2020, respectively, due to lower operating leverage.

The decline in our average interest rate (actual and implied) on our mortgage borrowings for the three and six month periods was due to lower short-term interest rates.

Additionally, during the three months ended June 30, 2020, we terminated $3.7 billion of repurchase agreements with a weighted average interest rate of 2.11% and a weighted average remaining maturity of 2.2 years.

The terminated agreements were replaced with shorter duration repurchase agreements at lower prevailing market rates.

We recognized losses on debt extinguishment of $146 million in other gain (loss), net for the three and six months ended June 30, 2020 associated with the terminated repurchase agreements.

The increase in our interest rate swap periodic cost for the three and six months ended June 30, 2020 was due to declines in the average rate received on our interest rate swaps as the receive leg of our pay-fixed interest rate swaps reset to lower prevailing rates, which were partially offset by a decline in our average pay-rate.

The following table presents a summary of the ratio of our average interest rates swaps outstanding, excluding forward starting swaps, to our average mortgage borrowings and the weighted average pay-fixed / receive-floating rates on our interest rate swaps for the three and six months ended June 30, 2020 and 2019 (dollars in millions):

Our average mortgage borrowings, inclusive of TBAs, decreased by -9% and -4% for the three and nine months ended September 30, 2020, respectively, largely due to lower operating leverage.

The decline in our average interest rate (actual and implied) on our mortgage borrowings these periods was due to lower short-term interest rates.

Additionally, during the nine months ended September 30, 2020, we terminated $3.7 billion of longer-dated repurchase agreements and replaced them with shorter duration repurchase agreements at lower prevailing market rates.

We recognized losses on debt extinguishment totaling $146 million in other gain (loss) during the period associated with the terminated agreements.

We did not terminate any repurchase agreements during the three months ended September 30, 2020 or the prior year periods.

The increase in our interest rate swap periodic cost for the three and nine months ended September 30, 2020 was due to declines in the average rate received on our interest rate swaps as the receive leg of our pay-fixed interest rate swaps reset to lower prevailing rates, which were partially offset by a decline in our average pay-rate.

The following table presents a summary of the ratio of our average interest rates swaps outstanding, excluding forward starting swaps, to our average mortgage borrowings and the weighted average pay-fixed / receive-floating rates on our interest rate swaps for the three and nine months ended September 30, 2020 and 2019 (dollars in millions):

Three Months Ended

Six Months Ended

September 30,

Nine Months Ended

September 30,

Average Ratio of Interest Rate Swaps (Excluding Forward Starting Swaps) to Mortgage Borrowings Outstanding 2020 2019 2020 2019

14 unchanged sentences

0.13 % (0.74) % 0.10 % (0.73) %

For the three and six months ended June 30, 2020, we had an average forward starting swap balance of $0.9 billion and $1.2 billion, respectively.

For the three and six months ended June 30, 2019, we had an average forward starting swap balance of $3.5 billion and $4.3 billion, respectively.

For the three and nine months ended September 30, 2020, we had an average forward starting swap balance of $0.7 billion and $1.0 billion, respectively.

For the three and nine months ended September 30, 2019, we had an average forward starting swap balance of $2.4 billion and $3.7 billion, respectively.

Forward starting interest rate swaps do not impact our economic interest expense and aggregate cost of funds until they commence accruing net interest settlements on their forward start dates.

Including forward starting swaps, our average ratio of interest rate swaps outstanding to our average mortgage borrowings for the three and six months ended June 30, 2020 was 53% and 63%, respectively, and 57% and 56%, respectively, for the corresponding prior year periods.

Including forward starting swaps, our average ratio of interest rate swaps outstanding to our average mortgage borrowings for the three and nine months ended September 30, 2020 was 47% and 58%, respectively, and 82% and 65%, respectively, for the corresponding prior year periods.

Net Interest Spread

The following table presents a summary of our net interest spread (including the impact of TBA dollar roll income, interest rate swaps and excluding "catch-up" premium amortization) for the three and six months ended June 30, 2020 and 2019:

Three Months Ended June 30,

Six Months Ended

Three Months Ended September 30,

Nine Months Ended

September 30,

Investment and TBA Securities - Net Interest Spread 2020 2019 2020 2019

3 unchanged sentences

Net Spread and Dollar Roll Income

The following table presents a summary of our net spread and dollar roll income, excluding estimated "catch-up" premium amortization, per diluted common share (a non-GAAP financial measure) and a reconciliation to our net interest income (the most comparable GAAP financial measure) for the three and six months ended June 30, 2020 and 2019 (dollars in millions):

Three Months Ended June 30,

Six Months Ended

The following table presents a summary of our net spread and dollar roll income, excluding estimated "catch-up" premium amortization, per diluted common share (a non-GAAP financial measure) and a reconciliation to our net interest income (the most comparable GAAP financial measure) for the three and nine months ended September 30, 2020 and 2019 (dollars in millions):

Three Months Ended September 30,

Nine Months Ended

September 30,

2020 2019 2020 2019

1 unchanged sentence

TBA dollar roll income, net 1

Interest rate swap periodic interest income (cost), net 1

155 29 249 70

Interest rate swap periodic interest (cost) income, net 1

(13) 146 (41) 317

Other interest and dividend income 1

15 unchanged sentences

Gain (Loss) on Investment Securities, Net

The following table is a summary of our net gain (loss) on investment securities for the three and six months ended June 30, 2020 and 2019 (in millions):

Three Months Ended June 30,

Six Months Ended

The following table is a summary of our net gain (loss) on investment securities for the three and nine months ended September 30, 2020 and 2019 (in millions):

Three Months Ended September 30,

Nine Months Ended

September 30,

Gain (Loss) on Investment Securities, Net 1

9 unchanged sentences

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

The following table is a summary of our gain (loss) on derivative instruments and other securities, net for the three and six months ended June 30, 2020 and 2019 (in millions):

Three Months Ended June 30,

Six Months Ended

The following table is a summary of our gain (loss) on derivative instruments and other securities, net for the three and nine months ended September 30, 2020 and 2019 (in millions):

Three Months Ended September 30,

Nine Months Ended

September 30,

2020 2019 2020 2019

2 unchanged sentences

TBA securities - dollar roll income, net

155 29 249 70

TBA securities - mark-to-market net gain (loss)

10 unchanged sentences

Losses on debt extinguishment

(146) — (146) —

Total realized gain (loss) on derivative instruments and other securities, net 402 (767) (2,787) (2,920)

7 unchanged sentences

Treasury securities - long position

(17) (1) (2) 4

Treasury securities - short position

2 unchanged sentences

(3) 54 (20) 61

(3) (2) (7) 1

Total unrealized gain (loss) on derivative instruments and other securities, net 11 73 (311) (383)

2 unchanged sentences

Estimated Taxable Income

For the three months ended June 30, 2020 and 2019, we had estimated taxable income (loss) available (attributable) to common stockholders of $(4) million and $171 million, or $(0.01) and $0.32 per diluted common share, respectively.

For the six months ended June 30, 2020 and 2019, we had estimated taxable income available to common stockholders of $107 million and $367 million, or $0.19 and $0.68 per diluted common share, respectively.

For the three months ended September 30, 2020 and 2019, we had estimated taxable income (loss) available (attributable) to common stockholders of $508 million and $156 million, or $0.92 and $0.29 per diluted common share, respectively.

For the nine months ended September 30, 2020 and 2019, we had estimated taxable income available to common stockholders of $615 million and $523 million, or $1.11 and $0.97 per diluted common share, respectively.

Income determined under GAAP differs from income determined under U.S.

2 unchanged sentences

Furthermore, our estimated taxable income is subject to potential adjustments up to the time of filing our appropriate tax returns, which occurs after the end of our fiscal year.

The following is a reconciliation of our GAAP net income to our estimated taxable income for the three and six months ended June 30, 2020 and 2019 (dollars in millions, except per share amounts):

Three Months Ended June 30,

Six Months Ended

The following is a reconciliation of our GAAP net income to our estimated taxable income for the three and nine months ended September 30, 2020 and 2019 (dollars in millions, except per share amounts):

Three Months Ended September 30,

Nine Months Ended

September 30,

2020 2019 2020 2019

27 unchanged sentences

Our leverage will vary depending on market conditions and our assessment of relative risks and returns, but we generally would expect our leverage to be between six and twelve times the amount of our tangible stockholders' equity, measured as the sum of our total mortgage borrowings and net payable / (receivable) for unsettled investment securities, divided by the sum of our total stockholders' equity adjusted to exclude goodwill.

Our tangible net book value "at risk" leverage ratio was 9.2x and 9.4x as of June 30, 2020 and December 31, 2019, respectively.

The following table includes a summary of our mortgage borrowings outstanding as of June 30, 2020 and December 31, 2019 (dollars in millions).

Our tangible net book value "at risk" leverage ratio was 8.8x and 9.4x as of September 30, 2020 and December 31, 2019, respectively.

The following table includes a summary of our mortgage borrowings outstanding as of September 30, 2020 and December 31, 2019 (dollars in millions).

For additional details of our mortgage borrowings refer to Notes 3, 5 and 6 to our Consolidated Financial Statements in this Form 10-Q.

June 30, 2020 December 31, 2019

September 30, 2020 December 31, 2019

Mortgage Borrowings Amount % Amount %

7 unchanged sentences

Amount excludes $200 million and $97 million of repurchase agreements used to fund purchases of U.S.

Treasury securities as of June 30, 2020 and December 31, 2019, respectively.

Treasury securities as of September 30, 2020 and December 31, 2019, respectively.

Repurchase arrangements involve the sale and a simultaneous agreement to repurchase the transferred assets at a future date and are accounted for as collateralized financing transactions.

5 unchanged sentences

None of our repo counterparties are obligated to renew or otherwise enter into new borrowings at the conclusion of our existing borrowings.

The use of TBA dollar roll transactions increases our funding diversification, expands our available pool of assets, and increases our overall liquidity position, as TBA contracts have lower implied haircuts relative to Agency RMBS pools held on balance sheet and funded with repo financing.

However, if it were to become uneconomical to roll our TBA contracts into future months it may be necessary to take physical delivery of the underlying securities and fund those assets with cash or other financing sources, which could reduce our liquidity position.

The collateral requirements under our TBA contracts are governed by the Mortgage-Backed Securities Division ("MBSD") of the FICC and by our brokerage agreements, which may establish margin levels in excess of the MBSD.

To help manage the adverse impact of interest rate changes on the value of our investment portfolio as well as our cash flows, we utilize an interest rate risk management strategy under which we use derivative financial instruments.

To help manage the adverse impact of interest rate changes on the value of our investment portfolio as well as our cash flows, we utilize an interest rate risk management strategy involving the use of derivative financial instruments.

In particular, we attempt to mitigate the risk of the cost of our variable rate liabilities increasing at a faster rate than the earnings of our long-term fixed rate assets during a period of rising interest rates.

7 unchanged sentences

Haircuts under repo agreements are typically determined on an individual transaction basis and reflect our counterparties' assessment of the underlying risk associated with the specific collateral.

Our derivative agreements also require that we maintain a minimum collateral balance regardless of the value of the derivative instrument based on our counterparties' assessment of risk specific to the derivative instrument and clearing exchange rules.

Haircut levels and minimum margin requirements can change overtime and could be expected to increase during periods of elevated market volatility.

4 unchanged sentences

Our agreements generally provide that in the event of a margin call, collateral must be posted on the same business day, subject to notice requirements.

As of June 30, 2020, we had met all our margin requirements.

As of September 30, 2020, we had met all our margin requirements.

The value of Agency RMBS is reduced by principal pay-downs on the mortgage pools underlying them.

5 unchanged sentences

We attempt to manage the liquidity risk associated with principal pay-downs by monitoring conditions impacting prepayment rates and through asset selection.

As of June 30, 2020, given the current market conditions and elevated prepayment risk associated with historically low interest rates, our portfolio largely consisted of higher coupon specified pool securities, which have a lower risk of prepayment compared to generic Agency RMBS, and TBA securities, which do not expose us to margin calls due to prepayments.

As of September 30, 2020, given the current market conditions and elevated prepayment risk associated with historically low interest rates, our portfolio largely consisted of higher coupon specified pool securities, which have a lower risk of prepayment compared to generic Agency RMBS, and TBA securities, which do not expose us to margin calls due to prepayments.

Haircut levels and minimum margin requirements imposed by counterparties reduce the amount of our unencumbered assets and limit the amount we can borrow against our investment securities.

For the six months ended June 30, 2020, haircuts on Agency RMBS collateral remained stable.

Haircuts and funding levels for our less liquid, credit-oriented securities, however, were adversely impacted by the dislocation in the financial markets in the first quarter, but improved to more normal levels during the second quarter.

As of June 30, 2020, the weighted average haircut on our repurchase agreements was approximately 4.4% of the value of our collateral, inclusive of collateral funded through BES, unchanged from December 31, 2019.

As of June 30, 2020, our unencumbered assets totaled 54% of our tangible net equity, unchanged from December 31, 2019.

For the nine months ended September 30, 2020, haircuts on Agency RMBS collateral remained stable.

As of September 30, 2020, the weighted average haircut on our repurchase agreements was approximately 4.4% of the value of our collateral, inclusive of collateral funded through BES, largely unchanged from December 31, 2019.

As of September 30, 2020, our unencumbered assets totaled 60% of our tangible net equity, compared to 54% as of December 31, 2019.

The majority of our liquidity is held at AGNC, but we also maintain capital and excess liquidity at BES for regulatory standards, risk management purposes, and to meet expectations of its counterparties, clearing banks and clearing organizations.

As of June 30, 2020, we had cash and unencumbered Agency RMBS totaling $4.5 billion, which excludes unencumbered CRT and non-Agency securities and assets held at BES.

As of September 30, 2020, we had cash and unencumbered Agency RMBS totaling $5.2 billion, which excludes unencumbered CRT securities, non-Agency securities and assets held at BES.

Collateral haircuts and minimum margin requirements imposed by counterparties subject us to counterparty credit risk.

1 unchanged sentence

We also diversify our funding across multiple counterparties and by region.

As of June 30, 2020, we had master repurchase agreements with 47 financial institutions located throughout North America, Europe and Asia, including repo counterparties accessed through BES.

BES' direct access to the General Collateral Finance Repo service offered by the FICC and other triparty and bi-lateral repo funding provides us greater depth and diversity of funding at favorable terms relative to traditional bilateral repurchase agreement funding.

As of June 30, 2020, $33.3 billion of our repurchase agreement funding was sourced through BES.

The table below includes a summary of our Agency RMBS repurchase agreement funding by number of repo counterparties and counterparty region as of June 30, 2020.

June 30, 2020

Counter-Party Region Number of Counter-Parties Percent of Repurchase Agreement Funding 1

North America:

Total North America 28 87%

Europe 14 10%

Total 47 100%

________________________________

Percent of repurchase agreement funding includes U.S.

Treasury repurchase agreements.

As of June 30, 2020, our maximum amount at risk (or the excess value of collateral pledged over our repurchase liabilities) with any of our repurchase agreement counterparties, excluding the FICC, was less than 3% of our tangible stockholders' equity, with our top five repo counterparties, excluding the FICC, representing less than 9% of our tangible stockholders' equity.

As of June 30, 2020, approximately 11% of our tangible stockholder's equity was at risk with the FICC.

Excluding central clearing exchanges, as of June 30, 2020, our amount at risk with any counterparty to our derivative agreements was less than 1% of our stockholders' equity.

Additionally, BES' direct access to the General Collateral Finance Repo service offered by the FICC and other triparty and bi-lateral repo funding provides us greater depth and diversity of funding at favorable terms relative to traditional bilateral repurchase agreement funding.

As of September 30, 2020, $26.2 billion of our repurchase agreement funding was sourced through BES.

As of September 30, 2020, our maximum amount at risk (or the excess value of collateral pledged over our repurchase liabilities) with any of our repurchase agreement counterparties, excluding the FICC, was less than 2% of our tangible stockholders' equity, with our top five repo counterparties, excluding the FICC, representing less than 6% of our tangible stockholders' equity.

As of September 30, 2020, approximately 10% of our tangible stockholder's equity was at risk with the FICC.

Excluding central clearing exchanges, as of September 30, 2020, our amount at risk with any counterparty to our derivative agreements was less than 1% of our stockholders' equity.

Agency RMBS securities are among the most liquid fixed income securities, and the TBA market is the second most liquid market (after the U.S.

2 unchanged sentences

Under certain market conditions, however, we may be unable to realize the full "pay-up" value, or premium relative to generic Agency RMBS, of our specified pool securities;

consequently, we seek to operate with a minimum level of securities that trade at or near TBA values that in our estimation enhances portfolio liquidity across a wide range of market conditions.

consequently, we seek to operate with a minimum level of securities that trade at or near TBA values that in our estimation enhances our portfolio liquidity across a wide range of market conditions.

Capital Markets

3 unchanged sentences

There can be no assurance that we will be able to raise additional equity capital at any particular time or on any particular terms.

Furthermore, when the trading price of our common stock is less than our estimate of our current tangible net book value per common share,

among other conditions, we may repurchase shares of our common stock.

Please refer to Note 10 of our Consolidated Financial Statements in this Form 10-Q for further details regarding our recent equity capital transactions.

Furthermore, when the trading price of our common stock is less than our estimate of our current tangible net book value per common share, among other conditions, we may repurchase shares of our common stock.

Please refer to Notes 10 and 11 of our Consolidated Financial Statements in this Form 10-Q for further details regarding our recent equity capital transactions and our stock repurchase plan.

OFF-BALANCE SHEET ARRANGEMENTS

As of June 30, 2020, we did not maintain relationships with unconsolidated entities or financial partnerships, such as entities often referred to as structured finance, or special purpose or variable interest entities, established to facilitate off-balance sheet arrangements or other contractually narrow or limited purposes.

Additionally, as of June 30, 2020, we had not guaranteed obligations of unconsolidated entities or entered into a commitment or intent to provide funding to such entities.

As of September 30, 2020, we did not maintain relationships with unconsolidated entities or financial partnerships, such as entities often referred to as structured finance, or special purpose or variable interest entities, established to facilitate off-balance sheet arrangements or other contractually narrow or limited purposes.

Additionally, as of September 30, 2020, we had not guaranteed obligations of unconsolidated entities or entered into a commitment or intent to provide funding to such entities.

FORWARD-LOOKING STATEMENTS

4 unchanged sentences

Factors that could cause actual results to vary from our forward-looking statements include, but are not limited to, the following:

• the impact of the Pandemic and of measures taken in response to the Pandemic by various governmental authorities, businesses and other third parties;

• the impact of the COVID-19 pandemic and of measures taken in response to the COVID-19 pandemic by various governmental authorities, businesses and other third parties;

• actions by the federal, state, or local governments to stabilize the economy, the housing sector or financial markets;

14 unchanged sentences

Compared sentence by sentence after normalising whitespace, quotation marks, case and digits, so re-formatting and restated figures do not read as changed language. Wording changes appear as one removal and one addition. The current filing and the prior one are authoritative.

What changed 10-Q

Item 2. Management’s Discussion and Analysis