Smart Money · Holder IQ
54.5 / 100
Above median holder quality (percentile 55 this quarter). Holder trend is not available.
AGNC
AGNC Investment Corp.NASDAQ
Real Estate • REIT - Mortgage
Factor tilt: Large-Cap Value
$8.93+1.71%
Close Nov 9, 2026
- Market cap
- $10.25B
- 1 month
- +3.0%
- 3 months
- -18.0%
- 1 year
- -12.5%
- 52-week range
- $8.44$12.17
Reading stored signals…
More market facts
- Dividend, last 12 mo
- $1.44 · 16.13% yield
- Last ex-date
- Oct 30, 2026 · $0.12
- Short interest
- 11.0% of float130M shares · settled Oct 15, 2026 · -0.9% since Sep 30, 2026 · FINRA
Fundamentals · trailing 12 months to Sep 30, 2026
- Revenue
- —
- Net income
- $59M
- Free cash flow
- $1.14B
- Book value
- $11.56B
- Total debt
- $96.72B
- P/E
- 173.7×
- Price / sales
- —
- Price / book
- 0.9×
- FCF yield
- +11.1%
- Net margin
- —
EODHD (TTM) · as of Nov 9, 2026
End-of-day close from stored market data, not a live quote.
About AGNC Investment Corp.
AGNC Investment Corp. provides private capital to housing market in the United States. It invests in residential mortgage pass-through securities and collateralized mortgage obligations for which the principal and interest payments are guaranteed by the United States government-sponsored enterprise or by the United States government agency.
Read more
The company qualifies as a real estate investment trust for federal income tax purposes. It generally would not be subject to federal or state corporate income taxes if it distributes at least 90% of its taxable income to its stockholders. The company was formerly known as American Capital Agency Corp. and changed its name to AGNC Investment Corp. in September 2016. AGNC Investment Corp. was incorporated in 2008 and is headquartered in Bethesda, Maryland.
- Website
- agnc.com
- Headquarters
- Bethesda, MD, United States
- Employees
- 54
- Listed
- May 13, 2008
- Fiscal year ends
- December
Company profile from EODHD, stored Oct 31, 2026.
Company Quality Grade
How steadily did the business compound its per-share book value with dividends over 40 quarters? This historical measure is not a stock-price return or a forecast.
As of · Window to
Computed, excluded from ranking
- Book with dividends · annual growth
- 3.4%
- Consistency of fitted trend · R²
- 0.053
- Recent / whole-window slope
- 14.91×
Basis: tangible common book (equity less preferred, goodwill and intangibles). Dividends are reinvested at book value, not at market prices. R² near 1 means a steadier fitted trend; it does not establish accounting quality.
No percentile or compounder designation is assigned because the admission rules were not met.
The compounder rule requires a top-decile composite within the tangible-book eligible pool, positive slope and no detected decline. Total-common-book fallback rows cannot qualify. A decline means the 90% interval for the recent 12-quarter / full-window slope ratio lies below 1; this row does not detect a decline. An absent decline is not proof that growth is accelerating.
- Share issuance, rather than earned profits, drove too much of the book-value increase.
issuance_driven_book - The filed share counts could not be reconciled consistently across the window.
share_basis_inconsistent
Link an account to see your stake, watchlist and past holdings. Use this to question the durability of the business behind your investment case, alongside its filings and valuation. A withheld grade is a data or method limit, not a sell signal. For a short position, it does not predict timing, borrowing costs or downside. It does not measure your account’s return.
Method, filing sources and publication receipt
Source: SEC first-filed XBRL. Midas Edge API’s published Green Compounders method uses only filings public by the as-of date, split-adjusted shares and the original filed values. Restatements do not replace them. Issuer pools and entity types reflect the SEC submissions captured for this run, not historical classifications at each as-of date.
The primary form fits log book value with reinvested dividends against time. The method checks the starting book and market-cap floors and excludes identified non-reliance/enforcement windows. Its cohort is fixed at capture time, so historical percentiles have capture and survival limits. It does not check audit-opinion going-concern text. This grade is separate from InsiderIQ, HolderIQ and NetCredibility.
Input filings (41 in this window): 0001423689-16-000091; 0001423689-16-000104; 0001423689-17-000011. The links show up to three of the input filings.
Method labels: going_concern_unchecked, intangibles_carried_forward, preferred_untagged_zero, red_finding_basis_d1_enforcement_naming_issuer, tbv_drop_gt15pct_unexplained, weighted_shares_fallback
- Parameters
- pg-g2-params.v17
- Plane
- sha256:841c1bb46826c45b96472bf9adff6a93c8ce5f5392636cff297045298e9bf517
- Grade row SHA-256
- sha256:5b043f46f4acc5b7a00a5add526effefaf521db5c1c8bc0025fe4c7b1104594b
- Manifest SHA-256
- sha256:c94c0931ffa7d9818ad0d34110b400f347a53da8a9c7d9aee1013095384d6a8a
- Engine SHA-256
- sha256:defa2e17f100b92a30576940d1632532b5cc4f37a2cb4bcc7fcc881856417811
- Hand-check receipt SHA-256
- sha256:03dd742b7b83e1130b4c01f9beaaf2671b20562c447a337dd047a41ad675ed3c
Short-interest history
FINRA reports open short positions for a settlement date. The publication date is later. These are reported positions, with a separate date for each observation.
Compare the reported short exposure around the dates you bought, sold or watched AGNC. A long position benefits from price rises; a short position benefits from falls. A change in these counts does not predict either move.
Settlement Oct 15, 2026
129,983,736 shares short
-1,201,551 shares (-0.9%) vs Sep 30, 2026
11% of the stored float. Published Oct 26, 2026.
Settlement Sep 30, 2026
131,185,287 shares short
+22,660,389 shares (+20.9%) vs Sep 15, 2026
11.1% of the stored float. Published Oct 9, 2026.
Settlement Sep 15, 2026
108,524,898 shares short
+2,806,000 shares (+2.7%) vs Aug 31, 2026
9.2% of the stored float. Published Sep 24, 2026.
Settlement Aug 31, 2026
105,718,898 shares short
+403,505 shares (+0.4%) vs Aug 14, 2026
Share of float unavailable. Published Sep 10, 2026.
Settlement Aug 14, 2026
105,315,393 shares short
+5,218,063 shares (+5.2%) vs Jul 31, 2026
Share of float unavailable. Published Aug 25, 2026.
Settlement Jul 31, 2026
100,097,330 shares short
-8,046,980 shares (-7.4%) vs Jul 15, 2026
Share of float unavailable. Published Aug 11, 2026.
Change = this settlement’s shares short − the previous recorded settlement’s shares short; percentage change divides by that earlier count. A zero earlier count has no percentage change. Split-spanning counts are left uncompared. Float percentages use the stored settlement observation, not today’s float.
Source: stored FINRA equity short-interest records, up to six settlements from the last 240 days. FINRA data and reporting dates · How to read short interest
Dividends and stock splits
See the recorded cash amounts and dates behind AGNC’s income, and the splits that change its share count. Dates below are stored event records, not confirmation of a payment to your account.
Cash-dividend records
USD 0.12 per share
Ex-dividend date: Oct 30, 2026
Recorded pay date: Nov 10, 2026 (today)
Cash dividend · Polygon event record
USD 0.12 per share
Ex-dividend date: Sep 30, 2026
Recorded pay date: Oct 9, 2026 (past date)
Cash dividend · Polygon event record
USD 0.12 per share
Ex-dividend date: Aug 31, 2026
Recorded pay date: Sep 10, 2026 (past date)
Cash dividend · Polygon event record
USD 0.12 per share
Ex-dividend date: Jul 31, 2026
Recorded pay date: Aug 11, 2026 (past date)
Cash dividend · Polygon event record
USD 0.12 per share
Ex-dividend date: Jun 30, 2026
Recorded pay date: Jul 10, 2026 (past date)
Cash dividend · Polygon event record
USD 0.12 per share
Ex-dividend date: May 29, 2026
Recorded pay date: Jun 9, 2026 (past date)
Cash dividend · Polygon event record
Up to six Polygon cash-dividend events, including recorded ex-dates within the next 120 days. Amounts retain their recorded currency and decimal precision. Historical amounts are not restated to a common split basis, annualized, or converted to dollars.
For a long holding, compare the ex-date with when you held the shares and check the pay date against your broker’s cash history. Owning the stock today does not establish past entitlement. For a short holding, a dividend can create a payment obligation to the share lender.
Ex-dates and special dividend rules · Short-sale dividend obligations · Dividend data definitions
Split history
No qualified split events are stored. This does not establish that the stock has never split.
Up to six stored split dates through Nov 10, 2026. Equivalent vendor ratios are combined; conflicting ratios are left out. This history may be incomplete.
A split changes the share count and per-share price basis; it does not by itself add investment value. Put your holding’s earlier and later share counts on the same basis before comparing them. This applies to both long shares and shares owed on a short position.
How stock splits workPortfolio context
Link an account to see your stake, watchlist and past holdings.
We’ll show you if AGNC is in your portfolio and how much it contributes.
Sign inMarket context
Loading market dates
Smart money context
Holder IQ: MEDIUM
71 smart-money holders: institutions with a high Holder IQ
Q2 2026, filed Oct 29, 2026
Smart-money 13F trend: accumulating (light)
Unlock crowding heatmapUnlock filer discoveryInsider context
Sentiment: bearish
Last trade: sell 14 days ago
Read the insider context evidence
What it means for you
Sign in to read your holdings or watchlist context. No personal position or direction is inferred.
What happened
Stored recent-activity label: bearish. This context label is separate from the scored insider signal read.
Why it can matter
A count of reported activity describes a sample, not purchase conviction, an insider’s intent or a future price move. Neutral is not an all-clear, and absent direction is not neutral evidence.
How we got this
The current context rule reads at most 50 nonduplicate issuer/share-class records filed in the last 90 days. It counts a bullish direction or purchase code P toward bullish, and a bearish direction or sale code S toward bearish. A side must exceed the opposite count by more than two to set its label; otherwise nonzero directional counts produce neutral. Calculation unavailable for this saved label: its counts, exact records, as-of time and rule version are not supplied here. Do not substitute the separate scored read’s different window or denominator.
Where the data comes from
Stored SEC Forms 3/4/5 insider-signal records, ordered by filing date with amendment duplicates excluded. Filing date is not transaction date. The last-trade line locates recent reported activity; it does not date or reconstruct the complete count.
Red flags: 1
Read the insider context evidence
What it means for you
Sign in to read your holdings or watchlist context. No personal position or direction is inferred.
What happened
Stored recent-activity label: bearish. This context label is separate from the scored insider signal read.
Why it can matter
A count of reported activity describes a sample, not purchase conviction, an insider’s intent or a future price move. Neutral is not an all-clear, and absent direction is not neutral evidence.
How we got this
The current context rule reads at most 50 nonduplicate issuer/share-class records filed in the last 90 days. It counts a bullish direction or purchase code P toward bullish, and a bearish direction or sale code S toward bearish. A side must exceed the opposite count by more than two to set its label; otherwise nonzero directional counts produce neutral. Calculation unavailable for this saved label: its counts, exact records, as-of time and rule version are not supplied here. Do not substitute the separate scored read’s different window or denominator.
Where the data comes from
Stored SEC Forms 3/4/5 insider-signal records, ordered by filing date with amendment duplicates excluded. Filing date is not transaction date. The last-trade line locates recent reported activity; it does not date or reconstruct the complete count.
Largest 13-F holders · Q2 2026
709 filers reported a long position in AGNC for Q2 2026. 13-F holdings are filed up to 45 days after the quarter ends.
The top 10 hold
58.2%
of the 475.8M shares 13-F filers reported
Moves vs Q1 2026
9 added0 trimmed
1 not comparable
Net shares held
+7.7%
across the 9 holders with a comparable Q1 2026 book
| Rank | Holder | Value | Shares | Shares vs Q1 2026 |
|---|---|---|---|---|
| 1 | BlackRock, Inc. | $704.1M · 64.6M shares | 64.6M | +6.9% |
| 2 | Vanguard Capital Management LLC | $562.3M · 51.6M shares | 51.6M | +4.4% |
| 3 | Vanguard Portfolio Management LLC | $502.7M · 46.1M shares | 46.1M | +2.7% |
| 4 | UBS Group AG | $297.7M · 27.3M shares | 27.3M | +18.3% |
| 5 | Geode Capital Management, LLC | $267M · 24.5M shares | 24.5M | +17.2% |
| 6 | State Street Corp | $227.5M · 20.6M shares | 20.6M | +8.2% |
| 7 | Norges Bank | $168.7M · 15.5M shares | 15.5M | — |
| 8 | Equity Investment Corp | $106.1M · 9.7M shares | 9.7M | +0.4% |
| 9 | Victory Capital Management Inc | $95.7M · 8.8M shares | 8.8M | +17.5% |
| 10 | Charles Schwab Investment Management Inc | $88.8M · 8.1M shares | 8.1M | +5.6% |
Sign in to see each holder’s Holder IQ, factor-adjusted Holder IQ and rank, and compare these holders with all scored filers. Available on every plan.
Sign in5% holder filings (13D / 13G)
Holders of more than 5% of AGNC file with the SEC within days. A 13D holder may seek to influence the company; a 13G holder is passive. Each holder’s latest filing from the last three years.
| Filed | Holder | % of class |
|---|---|---|
| Oct 27, 2026 | BlackRock, Inc.13G · passive · amended | 9.4% |
| Apr 29, 2026 | Vanguard Capital Management LLC13G · passive | 5.14% |
| Mar 26, 2026 | Vanguard Group Inc13G · passive · amended | 0%under 5% |
Moves the factor model did not explain
How much of AGNC’s move accompanied the market and its measured style leanings, and how much was left? The residual is unexplained by this fitted model. It does not establish a cause, skill or future alpha.
Return split unavailable. A complete, reconciled model for every market session in this window is unavailable. Missing dates are not filled.
Link an account with dated transactions to see what share of your own holding’s move since its purchase-date close was unexplained, with long and short positions distinguished. Your private holdings never enter this public chart.
Dated model observations (20)
Daily percentage returns, not window contributions. Source: the same dated warehouse rows and model v1.2.0; observed 2026-11-10T15:16:15.910Z. Absent sessions are never interpolated; missing or failed inputs stay unavailable.
| Date | Stock | Market | Styles | Residual |
|---|---|---|---|---|
| Unavailable | Unavailable | Unavailable | Unavailable | |
| Unavailable | Unavailable | Unavailable | Unavailable | |
| Unavailable | Unavailable | Unavailable | Unavailable | |
| Unavailable | Unavailable | Unavailable | Unavailable | |
| Unavailable | Unavailable | Unavailable | Unavailable | |
| Unavailable | Unavailable | Unavailable | Unavailable | |
| Unavailable | Unavailable | Unavailable | Unavailable | |
| Unavailable | Unavailable | Unavailable | Unavailable | |
| Unavailable | Unavailable | Unavailable | Unavailable | |
| Unavailable | Unavailable | Unavailable | Unavailable | |
| Unavailable | Unavailable | Unavailable | Unavailable | |
| Unavailable | Unavailable | Unavailable | Unavailable | |
| Unavailable | Unavailable | Unavailable | Unavailable | |
| Unavailable | Unavailable | Unavailable | Unavailable | |
| Unavailable | Unavailable | Unavailable | Unavailable | |
| Unavailable | Unavailable | Unavailable | Unavailable | |
| Unavailable | Unavailable | Unavailable | Unavailable | |
| Unavailable | Unavailable | Unavailable | Unavailable | |
| Unavailable | Unavailable | Unavailable | Unavailable | |
| Unavailable | Unavailable | Unavailable | Unavailable |
How this split is calculated
Each day uses that day’s fitted market intercept and style exposures multiplied by that day’s published factor returns. Their sum plus the independently stored residual must reconcile to the stock return. Window effects are weighted by stock wealth at each day’s start and added; the stock return compounds. Every expected US market session must be present and pass source, identity, model and numerical checks. This retrospective fit is an explanation of measured returns, not evidence of causation or an investable forecast. The window ends at the latest published model session, which can lag the market; today’s unfinished session is excluded.
What insiders reported owning
Reading dated filing evidence…
Get the Full Analysis for AGNC
Unlock detailed insider transactions, institutional holder changes, factor exposures, and personalized alerts. Start free with your own portfolio.