Northrop Grumman NOC

Reference table

1. The business, and the backlog that defines it

Northrop reports four segments — Aeronautics Systems, Defense Systems, Mission Systems and Space Systems. Chief executive Kathy J. Warden; approximately 95,000 employees; NYSE.

From the 10-Q: "Company backlog as of June 30, 2026 was $104.7 billion. Of our June 30, 2026 backlog, we expect to recognize approximately 35 percent as revenue over the next 12 months."

That single sentence is the strongest fact in this file and it appears in no vendor field. $104.7 billion is 2.50 times fiscal-2025 revenue of $41,954 million. 35% conversion implies roughly $36.6 billion of the next twelve months' revenue is already contracted, against a consensus of $44,056 million for 2026 — i.e. approximately 83% of the current-year forecast is in hand. This is why the FY2027 consensus EPS range is only 7.7% wide, the second-narrowest in this batch: analysts are scheduling rather than forecasting.

Revenue, filed:

Two segments disclosed in the extracted quarterly text are growing well: sales of $3,519 million against $3,114 million (+13.0%) and $2,093 million against $1,991 million (+5.1%) in the June quarter, on the munitions and missile-defence cycle.

2. The June quarter — falling operating income, rising earnings per share, and the tax rate that explains it

Filed, from the 10-Q ($M):

The reconciliation is arithmetic and it should govern how this quarter is read. From the 10-Q: "The year to date 2026 ETR decreased to 10.4 percent from 17.5 percent primarily due to the remeasurement of UTPs described above, as well as higher research credits... The year to date 2026 ETR includes benefits of $244 million for research credits, inclusive of the UTP remeasurement, and $19 million for FDDEI, partially offset by $25 million of interest expense on unrecognized tax benefits. The year to date 2025 ETR included benefits of $103 million for research credits."

Holding the tax rate at 17.5%, first-half net earnings would have been approximately $1,813 million rather than $1,969 million — growth of 9.5% rather than 19.0%. Roughly half of the reported first-half earnings growth is the tax line. And the September-quarter consensus reflects it: $7.13 against $7.67 a year ago, a 7.0% DECLINE, on revenue up 6.7%. Analysts are modelling the normalisation. Anyone reading the June quarter's 12.6% beat as an operating result is not.

The estimate-at-completion adjustments — the recurring cost of fixed-price work

From the 10-Q, verbatim:

> 2026 · Stand-in Attack Weapon (SiAW) — "During the second quarter of 2026, the company recorded a $68 million unfavorable EAC adjustment on the SiAW program at Defense Systems driven by increases in the projected costs to support the design and qualification of system software and hardware architecture for the U.S. Air Force."

> 2026 · B-21 — "During the first quarter of 2026, the company reached an agreement with the U.S. Air Force to expand production capacity and increase the aircraft production rate for the B-21 program at Aeronautics Systems. We made no significant changes to the previously recognized loss on the program; however, impacts from the agreement and higher estimated production costs resulted in a $157 million net unfavorable EAC adjustment on the first four LRIP lots, which was offset by a net reduction in the loss contingency accrual on the remainder of the program."

> 2025 · B-21 — "During the first quarter of 2025, the company recorded a $226 million unfavorable EAC adjustment on the first and second LRIP lots of the B-21 program at Aeronautics Systems."

And the cash flow statement carries a B-21 loss provision of $477 million in the first half of 2026.

Read together: the B-21 low-rate initial production lots are fixed-price, the company has been absorbing the overrun for three years, and the March-2026 agreement to expand production capacity produced a further $157 million of unfavourable adjustment on the first four lots even as the loss accrual on the remainder was reduced. The programme is being expanded and it is still losing money on the contracted lots. This is the single most important operating fact in the file, and it is the direct refutation of one of the knowledge base's own claims (Section 6).

3. Balance sheet — and a debt figure the payload gets wrong by 25.8%

Filed, from the 10-Q ($M):

The payload reports totalDebt of $19,741 million for fiscal 2025. The filing says $15,696 million. The $4,045 million gap decomposes exactly:

Northrop's payload carries BOTH documented lease defects simultaneously: operating lease liabilities folded into the debt lines (the AMT signature) AND the same $1,857 million non-current lease liability counted a second time on a separate field (the HCA signature). The overstatement is 25.8% of filed debt.

Corrected: net debt at 2025-12-31 is $15,696M − $4,403M = $11,293M, against the payload's $15,338M. At 2026-06-30 it is $15,191M − $2,307M = $12,884M. Debt-to-equity on the filed figures is 0.85x, not the payload's 0.911x built on inflated debt against a differently-dated equity.

Two further balance-sheet facts matter.

Tangible equity is approximately nil. Goodwill of $17,439 million against total shareholders' equity of $17,884 million, before other intangibles. The vendor's own tangibleBookValuePerShareTTM of $3.13 and tangibleAssetValueTTM of $445 million both confirm it, and both are CORRECT rather than defective. A company earning a 26.6% return on equity on effectively no tangible capital is not levered in the ordinary sense — its capital is contracts and clearances — but it also has no asset backing whatever beneath the equity.

Cash fell $2,096 million in six months, from $4,403 million to $2,307 million, against first-half repurchases and dividends and the B-21 cash outflow.

Capital return, filed: fiscal 2025 repurchases $1,624 million and dividends $1,293 million, a combined $2,917 million against a $78,355 million market capitalisation — a 3.72% total shareholder yield. Fiscal 2024 repurchases were $2,514 million, so the pace has slowed 35%. Dividend $9.40 trailing, 1.70%.

4. Free cash flow — capex verified, and a clean vendor ratio

capexToDepreciationTTM of 0.9396 is plausible and internally consistent — fiscal-2025 capital expenditure of $1,450 million against depreciation and amortisation of $1,472 million is 0.985x, and the trailing figure moves with the first-half spend. Capital expenditure fell 17.9% year on year in fiscal 2025, which for a company expanding B-21 production capacity is worth watching rather than celebrating. No corroborating capital-expenditure figure was located in the extracted 10-Q text, so the vendor figures are used as reported and labelled as unverified against a filed line. The payload's freeCashFlowYieldTTM of 4.65% is recomputed here as 4.22% on the fiscal-2025 filed-basis figures; the difference is trailing-window composition and neither number is materially different.

5. Valuation — priced in or room?

At $551.55 (market capitalisation $78.35B; 142.4 million weighted-average diluted shares in the June quarter):

Estimate coverage is the best in this batch: 16 analysts on FY2027 EPS with a range of $29.50 to $31.79 — a 7.7% spread — and 18 on FY2028 revenue. That narrowness is a direct function of the $104.7 billion backlog, and it means the estimate base here is more trustworthy than anywhere else in this batch.

est.ebitdaAvg and est.ebitAvg are rejected. In every year from 2023 to 2030 ebitdaAvg is exactly 18.0170% of revenueAvg and ebitAvg exactly 14.5384%, to four decimal places. The implied 14.54% EBIT margin sits above the company's actual 10.1% June-quarter operating margin and above every full year since 2021. All forward valuation uses epsAvg only.

Enterprise value is rebuilt. The vendor prints enterpriseValueTTM of $92,336.8M, a $13,982.0M premium to market capitalisation. Rebuilding on the filed June-quarter figures gives $78,354.8M + $15,191M − $2,307M = $91,238.8M — 1.2% below the vendor's, because the vendor's inflated debt is partly offset by a different cash date. netDebtToEBITDATTM of 1.976x is REJECTED: on filed net debt of $12,884M and a rebuilt trailing EBITDA of roughly $7.2 billion, the correct figure is approximately 1.79x. minorityInterest is 0 and correctly so.

5a. What today's price assumes (the inversion)

At $551.55 — 19.1x the 2026 consensus and 18.2x 2027 — the price embeds roughly:

• EPS of $28.84 in 2026 and $30.39 in 2027. (Consensus; 14 and 16 analysts.) The first half delivered $13.82 ($6.14 plus $7.68), so 2026 needs $15.02 in the second half against a September consensus of $7.13 — itself a 7.0% decline.
• The effective tax rate normalises without breaking the earnings path. (Our derivation; the filing supplies the rates.) This is the most fragile assumption in the price. The year-to-date rate is 10.4% against 17.5%. A return to 17.5% removes approximately 8% from earnings, and consensus assumes revenue growth and margin recovery absorb it.
• Fixed-price programme losses stop. (Our number.) The B-21 has produced unfavourable estimate-at-completion adjustments in three of the last six quarters and carries a $477 million loss provision. Each $200 million of unfavourable adjustment is roughly $1.10 per share after tax.
• Backlog converts at approximately 35% a year and grows. (Company disclosure, half-weighted on the forward element.) $104.7 billion against consensus 2026 revenue of $44,056 million means roughly 83% of the current-year forecast is already contracted — the strongest support for any revenue line in this batch.
• The market keeps paying roughly 18x forward earnings and the capital return continues. (Our number.) At 16x FY2027E the stock is $486; at 21x it is $638. And the shareholder yield of 3.72% is explicitly in political scope — see Section 6.

5b. The return bridge (why the multiple moves)

Expected return over the next twelve months decomposes as EPS growth (+5.4%, from FY2026E $28.840 to FY2027E $30.388) + multiple drift + shareholder yield (+3.72%).

Our base assumes the multiple EXPANDS, from today's 18.2x on the 2027 year to 19.5x — a 7% re-rating — and that is the fragile leg, named as such. The mechanism is not a change of view about defence: it is the reversal of a de-rating that has already taken 28.2% off the price and pushed the stock 8.9% below its 200-day average. A company with $104.7 billion of backlog, contracted revenue growth of 6%, and a 3.7% shareholder yield trading below its own two-hundred-day average is priced for the programme losses to continue, and our base assumes they moderate rather than that anything improves.

At an unchanged 18.2x, fair value on the 2027 consensus is $552 — spot exactly — and the twelve-month return is 5.4% of earnings growth plus 3.7% of yield, roughly 9%. That is the floor case and it is not a bad one.

Sensitivity: on the FY2027E consensus mean of $30.388, 14.5x = $441, 16x = $486, 18.2x = $552 (spot), 19.5x = $593, 21x = $638, 24x = $729.

5c. Variant perception (where we differ, what would surprise)

• We are 7.6% below the street and the gap is the tax line. Our $593 against consensus $642, using the street's own EPS at a lower multiple. Where we genuinely differ is on how the June quarter should be read: consensus celebrated a 12.6% beat, and the same filing shows operating income down 23.1% with the difference in the tax rate. Watchable number: the effective tax rate in the 2026-10-20 print, against 10.4% year to date and 17.5% a year earlier.
• We grade one of the knowledge base's own claims as wrong, on the company's disclosure. stephen_wolfram 2025-08-21, conviction 45: "B-21 Raider stealth bomber is uniquely on-time and under-budget." The filings record a $226 million unfavourable adjustment in Q1 2025, a $157 million net unfavourable adjustment in Q1 2026 and a $477 million loss provision in the first half of 2026. We report the claim and grade it, rather than excluding it.
• We think the political risk to the shareholder yield is under-priced and it is the sharpest thing in the knowledge base. scott_bessent, speaker_role: policymaker, 2026-01-20, conviction 66: "Defense primes have failed, 5-7 years behind on deliveries while buying back stock instead of building facilities; good idea to bring them to heel by curbing dividends/buybacks." That is a named proposal to remove the 3.72% of expected return that this dive's base case relies on. Nothing in the consensus reflects it. Watchable event: any procurement-policy action linking contract awards to capital-return policy.
• We differ from the vendor data materially on leverage (Section 7): totalDebt is 25.8% too high because it carries both lease defects at once. Anyone screening this name on leverage is looking at a company with $4 billion of debt it does not have.
• Positive surprise that would force a re-rate: two clean quarters with no unfavourable estimate-at-completion adjustments; backlog growing above $110 billion; the solid-rocket-motor position given disclosed economics; or a repurchase acceleration back toward the fiscal-2024 pace of $2,514 million.
• Negative surprise that would break the thesis: another B-21 or Sentinel unfavourable adjustment above $200 million; the tax rate normalising above 18%; backlog declining; a policy action on defence-prime capital returns; or a cash balance falling below roughly $2 billion.

Synthos fair values

All three anchors are multiples of the FY2027 consensus EPS distribution — mean $30.38845, low $29.49996, high $31.78608, 16 analysts — cross-checked against the FY2026 and FY2028 rows and against the backlog.

• Bear ~$428 — 14.5x the FY2027 consensus LOW of $29.500. Cross-check: 15.3x FY2026E; 13.7% BELOW the 52-week low of $496.02. The scenario: fixed-price losses continue and widen, the tax rate normalises above 18%, capital returns are curbed politically, backlog conversion slows, and a defence prime with no tangible equity re-rates to the bottom of its historical band. −22.4%.
• Base ~$593 — 19.5x the FY2027 consensus MEAN of $30.388. Cross-check: 20.6x FY2026E; 18.0x FY2028E; 4.71x book. The scenario: programme losses moderate, the tax rate normalises but revenue growth of 6-7% absorbs it, backlog converts at 35%, the buyback continues at roughly $1.6 billion a year, and the multiple recovers 7% from a de-rated level. +7.5%.
• Bull ~$763 — 24.0x the FY2027 consensus HIGH of $31.786. Cross-check: 23.2x FY2028E; 0.7% below the 52-week high of $768.02; 2.8% below the street's high target of $785. The scenario: the fixed-price losses end, the munitions and missile-defence cycle accelerates backlog past $115 billion, the solid-rocket-motor choke-point converts into pricing, and the market pays a scarcity multiple for the only new bomber and the only new intercontinental ballistic missile. +38.3%.

Base is 7.5% above spot; asymmetry roughly 1.71:1 to the upside (22.4% down, 38.3% up), before a shareholder yield of 3.72%. That is the best payoff ratio in this batch and it is short of the Buy tier's threshold on the base return. The bear case sitting 13.7% BELOW the 52-week low is the honest measure of what would have to go wrong, and it is a genuinely severe scenario rather than a routine one — which is what a 1.71:1 ratio on a de-rated entry should look like.

6. Knowledge base — four name-level claims, one graded wrong, and a policymaker

Raw hits: 10. Exact case-sensitive entity matches: 4. Name-level claims on Northrop Grumman: 4. Bullish: 3. Bearish: 1. Text-only, of which 1 is highly material: 6.

The sweep ran entity terms NOC, Northrop, Northrop Grumman, B-21 and Sentinel, plus free text on "northrop", "b-21", "sentinel icbm", "defense prime" and "defence prime", across all 52,021 distilled claims.

The four name-level claims:

> 2026-03-16 · BULLISH · conviction 68 · thesis · entities: Northrop Grumman · channel: odd_lots · no named speaker

> "Northrop Grumman is the other dominant solid-rocket-motor producer; SRM is a scarce, choke-pointed capability central to the munitions ramp."

> 2026-04-22 · BULLISH · conviction 48 · thesis · entities: BA, RTX, NOC · channel: arthur_hayes · speaker: Arthur Hayes · skill 1.2

> "War means money printing and ramped bank lending to finance the merchants of death — Boeing, Raytheon, Northrop Grumman get the financing to expand weapons production."

> 2025-08-21 · BULLISH · conviction 45 · fact · entities: Northrop Grumman · channel: stephen_wolfram · no named speaker

> "B-21 Raider stealth bomber is uniquely on-time and under-budget, second unit already built, progressing faster than any prior military aircraft program."

> 2026-03-20 · BEARISH · conviction 68 · thesis · entities: Lockheed Martin, Northrop Grumman · channel: a16z · no named speaker

> "Post-Last-Supper consolidation bred conformity and financialization (dividends/buybacks over growth); primes can't harbor founder innovation."

The odd_lots claim is the best in the lane and it is a specific structural argument rather than a directional opinion. Solid rocket motors are a duopoly capability, they are the binding constraint on the munitions replenishment cycle, and Northrop holds one of the two positions. Nothing in the filings quantifies the segment's economics separately, so the claim is unfalsifiable from this archive — but it is the correct thing to have identified and it is corroborated by a second, unrelated claim in the text-only lane (natalie_brunell 2025-06-28, conviction 60: defence primes "can't produce rocket motors domestically — a manufacturing/supply-chain constraint").

The stephen_wolfram claim is WRONG on the company's own disclosure and we grade it rather than exclude it. It states the B-21 is "uniquely on-time and under-budget." The filings record: a $226 million unfavourable estimate-at-completion adjustment on the first and second low-rate initial production lots in the first quarter of 2025; a $157 million net unfavourable adjustment on the first four lots in the first quarter of 2026; and a $477 million B-21 loss provision in the first-half 2026 cash flow statement, on top of the $1.56 billion loss recognised in 2023. The programme may well be ahead of schedule — the claim's second half about a second unit being built is not contradicted — but "under-budget" is the opposite of what a fixed-price loss provision means. A knowledge base that only surfaces claims which aged well is not a knowledge base, and this one is reported and marked wrong.

And the most consequential item in the lane is text-only, because it names no company:

> 2026-01-20 · BEARISH · conviction 66 · thesis · channel: scott_bessent · speaker: Scott Bessent · role: POLICYMAKER

> "Defense primes have failed, 5-7 years behind on deliveries while buying back stock instead of building facilities; good idea to bring them to heel by curbing dividends/buybacks."

This is a named sitting policymaker proposing to curb the dividends and share repurchases of defence primes. Northrop returned $2,917 million to shareholders in fiscal 2025 — 3.72% of its market capitalisation — and that yield is one of the three components of this dive's expected return. The speaker_role field is documented as unreliable in this store; here the content matches the named speaker and the role, so it is accepted on content rather than on the tag. It enters no conviction number, because it names no company. It is an explicit input to the verdict, and it is the reason a defence prime at 1.71:1 asymmetry is a Watch rather than a Buy.

One further text-only claim is directly competitive:

> 2026-01-15 · bullish · conviction 80 · thesis · entities: SpaceX, Anduril, Palantir · channel: no_priors

> "New-entrant defense primes (SpaceX, Anduril, Palantir) once had to sue for contracts; DoW now actively courting lower-cost, faster, more sophisticated startup options."

Four remaining text-only claims concern European defence — niall_ferguson 2025-10-06 on Rheinmetall and Ukrainian drone makers, mikkel_rosenvold 2025-03-25 on European primes' order books being full, natalie_brunell 2025-05-06 arguing a trillion-dollar defence budget should be halved, and the rocket-motor constraint claim already cited. Only the last is relevant to Northrop and it is supportive.

Attribution note. Two of the four name-level claims carry no named speaker and are channel-attributed only. Two of the most consequential items — arthur_hayes and scott_bessent — carry named speakers, and one carries a policymaker role.

Conclusion. Breadth 5, claim count 4, net conviction mixed. The Synthos knowledge base holds one strong structural bull case on Northrop (the solid-rocket-motor choke-point), one macro bull case, one factual claim that the filings contradict, one structural bear case on the prime model, and — outside the entity lane — a named policymaker proposing to curb the exact capital return this dive is counting on. Genuinely two-sided, genuinely relevant, and a direct input to the verdict rather than to a score.

7. Data integrity — what we rejected and why

Five findings. Northrop's payload carries the most precisely-decomposable debt error in this batch — both documented lease defects in the same field — against otherwise sound data.

1. totalDebt is overstated by $4,045 million, 25.8%, by BOTH lease defects at once — REJECTED and rebuilt. Filed at 2025-12-31: long-term debt $15,162M plus current portion $534M = $15,696M. Payload totalDebt $19,741M. The decomposition is exact:

• longTermDebt of $17,019M = filed $15,162M plus the $1,857M non-current operating lease liability shown separately on the balance sheet. (The AMT signature: operating leases folded into debt.)
• shortTermDebt of $865M = filed current portion $534M plus $331M of current operating lease liabilities.
• capitalLeaseObligations of $1,857M is then added AGAIN on top. (The HCA signature: the same lease counted twice.)

Total: $1,857M + $331M + $1,857M = $4,045M. Consequential rejections: netDebt $15,338M (correct at 2025-12-31: $11,293M; at 2026-06-30: $12,884M), netDebtToEBITDATTM 1.976x (correct approximately 1.79x), debtToEquityRatioTTM 0.911x (correct on filed figures: 0.85x), debtToAssetsRatioTTM 0.321 and enterpriseValueTTM $92,336.8M (rebuilt: $91,238.8M). This is the first name in this programme to exhibit both lease defects in a single payload.

2. inc_q depreciation for the June 2026 quarter reads ZERO — REJECTED. depreciationAndAmortization is 0 for Q2 2026 against $372M in Q1 2026, $406M in Q4 2025 and comparable figures throughout. The neighbouring fields in the same row — revenue $10,876M, operating income $1,096M, net income $1,094M and diluted EPS $7.68 — all reconcile exactly to the 10-Q. A single field in an otherwise correct row: check fields individually, not rows.

3. est.ebitdaAvg and est.ebitAvg carry the fixed-ratio fabrication signature and exceed the actual margin — REJECTED. In all eight years from 2023 to 2030, ebitdaAvg is exactly 18.0170% of revenueAvg and ebitAvg exactly 14.5384%. The implied 14.54% EBIT margin is above Northrop's actual June-quarter operating margin of 10.1% and above every full year since 2021 (10.2%, 10.6%, 6.5%). All forward valuation uses epsAvg only.

4. researchAndDevelopementToRevenueTTM reads ZERO while inc_a carries R&D of $1,100 million. Internally inconsistent within the same file. Not used.

5. Backlog — the single most important operating metric for this company — is absent from every vendor field. $104.7 billion at 2026-06-30, 2.50x annual revenue, with approximately 35% expected to convert within twelve months. This is not a defect in the ordinary sense; it is a coverage gap, and it means any screen of this name is missing the fact that roughly 83% of the current-year revenue consensus is already contracted. Recorded because it changes the interpretation of every growth estimate in the file.

Verified clean — recorded:

• inc_q for the June 2026 quarter is correct in every field except depreciation — revenue $10,876M, operating income $1,096M, net earnings $1,094M and diluted EPS $7.68 all reconcile to the 10-Q exactly, as do the prior-year comparatives ($10,351M, $1,425M, $1,174M).
• bal_a FY2025 goodwill of $17,437M matches the filed balance sheet exactly, as does cashAndCashEquivalents of $4,403M and totalAssets of $51,377M.
• minorityInterest is 0 and that is CORRECT — Northrop has no material non-controlling interests. The NCI-omission defect class was checked and does not apply.
• tangibleBookValuePerShareTTM of $3.13 and tangibleAssetValueTTM of $445 million are CORRECT rather than defective — goodwill of $17,439M genuinely leaves almost no tangible equity beneath $17,884M of total equity. Properly DISTINCT from bookValuePerShareTTM of $125.86; the identical-book defect was checked and does not apply.
• capexToDepreciationTTM of 0.9396 is plausible and internally consistent with fiscal-2025 capital expenditure of $1,450M against depreciation and amortisation of $1,472M (0.985x). No filed capital-expenditure line was located in the extracted text, so this is recorded as consistent rather than as verified.
• effectiveTaxRateTTM of 14.54% sits between the filed year-to-date 2026 rate of 10.4% and the 2025 rate of 17.5%, which is exactly where a trailing figure should be. Reconciles.

Peer set — REJECTED as half-irrelevant. The vendor peers are Emerson Electric, General Dynamics, Howmet Aerospace, Illinois Tool Works, L3Harris, Lockheed Martin, 3M, TransDigm, United Parcel Service and Waste Management. Four are genuine aerospace-and-defence comparables (General Dynamics, Howmet, L3Harris, Lockheed); the rest are two diversified industrials, a components manufacturer, a parcel carrier and a waste company. RTX and Boeing — both of which appear alongside Northrop in a knowledge-base claim — are absent. No peer-multiple comparison is drawn.

Vendor composite rating — noted. rating gives A− / 4 overall, with 2 on debt-to-equity. That sub-score is computed on the $19,741M debt figure rejected in finding 1; on the filed $15,696M the ratio is 0.85x rather than 0.911x, and the sub-score should be higher. This is a rare case where the vendor's own defect makes its rating too harsh rather than too generous.

Non-equity tripwire — checked and passed. NOC is common stock, NYSE-listed. Beta −0.102 — the only negative beta in this batch, which for a defence prime in a rising market is plausible rather than anomalous; volume 721 thousand shares (~$398M of turnover); the 52-week band of $496.02 to $768.02 is a 55% range. This is common equity.

8. Technicals

• Price $551.55, +0.56%. 28.2% BELOW the 52-week high of $768.02; 11.2% above the low of $496.02. Position within the annual range: 20th percentile — by far the lowest in this batch.
• +3.0% above the 50-day of $535.47 but 8.9% BELOW the 200-day of $605.67, with the 50-day well below the 200-day. NOC is the only name in this batch trading below its two-hundred-day average.
• RSI 64.4 — firm but not overbought, and the mildest reading among the names here that have risen off a low. MACD +3.13.
• On max_dd_from_peak. The field is misnamed: it computes (last − 6-year peak) / peak, i.e. the CURRENT distance from the multi-year high, not the deepest drawdown suffered over the window. A name that fell hard and fully recovered reads 0.00%. It coincides with pct_from_hi whenever the 6-year peak falls inside the last 252 days. Verified across 3,668 payloads: the ordering never violates (it is always equal to or deeper than pct_from_hi), so the values are internally consistent — only the label misleads.

• Relative performance: 3-month −2.7% against SPY +7.6%; 6-month −19.5% against +11.1%; 12-month −5.9% against SPY +24.3% and QQQ +30.8% — a thirty-point twelve-month deficit and the worst absolute return in this batch.
• Sentiment: 20 buy, 14 hold, 1 sell across 35 analysts; consensus target $642 (+16.4%), median $620 (+12.4%), high $785 (+42.3%), low $500 (−9.3%). The street is constructive and has been wrong for a year, which is worth stating: the same community carried similar targets while the stock fell 28% from its high.

9. Insiders — one transaction, and it is eight shares

The entire insider file for Northrop Grumman contains one transaction: a director's sale of EIGHT shares at $544.53 on 2026-08-03, for approximately $4,356 — 0.18% of his 4,488-share position.

This is a null file and it is reported as one. There is no open-market purchase, no executive-officer transaction, no option exercise and no tax withholding in the window. On a stock 28% below its high, the complete absence of insider buying by any of the company's officers is mildly notable, and eight shares is not evidence of anything at all. Compared with the other names in this batch — where executives were selling into strength — the absence here is at least not a negative.

10. Verdict, kill-criteria and flip conditions

Watch.

Why this is the most constructive name in the batch. Northrop holds $104.7 billion of backlog — 2.50 times annual revenue, with approximately 35% converting within twelve months, meaning roughly 83% of the 2026 revenue consensus is already contracted. It is the sole prime on the United States' only new strategic bomber and its intercontinental ballistic missile replacement, and one of two dominant solid-rocket-motor producers — a position the knowledge base independently identifies at conviction 68 as "a scarce, choke-pointed capability." Consensus revenue grows 5.0%, 6.9% and 6.2% over three years and the FY2027 earnings estimate range is only 7.7% wide, the second-narrowest in this batch. It returns 3.72% a year to shareholders. And unlike every other name here, the price has already fallen: 28.2% below its high, 8.9% below its 200-day average, down 5.9% over twelve months against an index up 24.3%. Base fair value $593, +7.5%, with 1.71:1 asymmetry — the best payoff ratio in this batch.

Why it is not a Buy — three specific reasons.

First, the current year's earnings are tax, not operations. Second-quarter operating income FELL 23.1% to $1,096 million while earnings per share rose 8.0% and beat by 12.6%. The year-to-date effective tax rate is 10.4% against 17.5%, including $244 million of research credits and an unrecognised-tax-position remeasurement. Holding the rate constant, first-half earnings grew 9.5% rather than 19.0%. The September consensus already embeds the reversal: $7.13 against $7.67, a 7.0% decline on 6.7% higher revenue.

Second, the fixed-price programme losses are live. A $477 million B-21 loss provision in the first-half cash flow; a $157 million net unfavourable estimate-at-completion adjustment in March 2026; $226 million in March 2025; $68 million on the Stand-in Attack Weapon programme in June. A March-2026 agreement to expand B-21 production capacity produced a further unfavourable adjustment on the contracted lots. The company is being asked to build more of something it loses money on.

Third, the shareholder yield is in political scope. The knowledge base carries a named sitting policymaker, at conviction 66, proposing to "bring them to heel by curbing dividends/buybacks" at defence primes. Northrop returned $2,917 million in fiscal 2025 — 3.72% of market capitalisation and roughly a third of this dive's expected return. Nothing in the consensus reflects that risk.

And one structural note that is not a reason for Watch but belongs in the record: tangible equity is approximately $250 million against $17,439 million of goodwill. That is normal for a defence prime and it means there is no asset backing beneath the equity if the contracts ever stop.

Pre-registered UPGRADE conditions — what would take this to Buy — Tactical:

• A price below approximately $510, which is 16.8x the FY2027 consensus mean and would take asymmetry above 2.3:1 without any change to the estimates. That is 2.8% above the 52-week low.
• Two consecutive quarters with NO unfavourable estimate-at-completion adjustment, ending the pattern that has run through three of the last six quarters.
• Backlog above $110 billion, confirming the munitions and missile-defence cycle is still adding rather than converting.
• The September quarter delivering at or above $7.13 with an effective tax rate above 15%, proving the earnings survive normalisation.
• A repurchase acceleration back toward the fiscal-2024 pace of $2,514 million.

Pre-registered KILL criteria — what would take this to Hold or Avoid:

• Another B-21 or Sentinel unfavourable adjustment above $200 million.
• The effective tax rate normalising above 18% without offsetting operating improvement.
• Backlog declining below $100 billion.
• Any procurement-policy action linking contract awards to capital-return policy, which is the specific risk the knowledge base's policymaker claim describes.
• Cash falling below approximately $2 billion, from $2,307 million at 2026-06-30 and $4,403 million six months earlier.
• A break below the 52-week low of $496.02.

Where NOC fits in the Synthos Framework Portfolio. Not entered today. Held on the watch list with a trigger price of ~$510 and an event trigger of a clean September quarter with a normalised tax rate. Sizing note: this is the only de-rated entry in this batch and the only one where the bear case sits below the 52-week low rather than above it — which is what a genuinely asymmetric setup looks like. The correct response to 1.71:1 with a base return of 7.5% is a limit order 7% lower, where the same estimates give 2.3:1. Logged as a tracked Synthos call (Watch) as of 2026-08-04 at $551.55, with the fair-value anchors, kill criteria and upgrade conditions all gradeable.

Single biggest risk: that fixed-price development losses continue, because the terms have not changed. The B-21 low-rate initial production lots are fixed-price. Northrop recognised a $1.56 billion loss on them in 2023, a $226 million unfavourable adjustment in the first quarter of 2025, a $157 million net unfavourable adjustment in the first quarter of 2026, and carries a $477 million loss provision in the first half of 2026. In the same quarter it agreed with the customer to expand production capacity and increase the production rate — and the agreement itself produced the unfavourable adjustment. That is the structural problem the knowledge base's a16z claim describes from a different angle: a consolidated prime with monopoly capability that has nonetheless accepted terms on which its most important programme loses money. Each $200 million of further adjustment is roughly $1.10 per share, and the market has already taken 28% off the price in anticipation of more. The bull case requires them to stop; the bear case requires only that they continue at the current rate.

Provenance & disclosures

• Traceability: 4 name-level knowledge-base claims name Northrop Grumman out of 52,021 distilled claims — 3 bullish, 1 bearish, net conviction MIXED (raw hits 10, exact case-sensitive entity matches 4, text-only 6, breadth 5). The strongest is odd_lots 2026-03-16 (bullish, conviction 68) — "Northrop Grumman is the other dominant solid-rocket-motor producer; SRM is a scarce, choke-pointed capability central to the munitions ramp" — a specific structural moat argument, independently corroborated by natalie_brunell 2025-06-28 (conviction 60) on the domestic rocket-motor constraint. arthur_hayes 2026-04-22 (bullish 48, skill 1.2, speaker-attributed) is a macro financing claim rather than a company one. stephen_wolfram 2025-08-21 (bullish 45) states the "B-21 Raider stealth bomber is uniquely on-time and under-budget" — and we grade it WRONG on the company's own disclosure: a $226 million unfavourable estimate-at-completion adjustment in Q1 2025, a $157 million net unfavourable adjustment in Q1 2026, a $477 million B-21 loss provision in the first-half 2026 cash flow, and a $1.56 billion loss recognised in 2023. Reported and graded rather than excluded. a16z 2026-03-20 (bearish 68) argues consolidation "bred conformity and financialization (dividends/buybacks over growth)". The most consequential item in the lane is text-only and enters no score: scott_bessent 2026-01-20, bearish, conviction 66, speaker_role POLICYMAKER — "Defense primes have failed, 5-7 years behind on deliveries while buying back stock instead of building facilities; good idea to bring them to heel by curbing dividends/buybacks." That is a named proposal to curb the 3.72% shareholder yield on which a third of this dive's expected return depends, it appears in no consensus estimate, and it is an explicit input to the Watch rather than Buy designation. no_priors 2026-01-15 (bullish 80) on new-entrant primes being actively courted is the competitive complement. Four further text-only claims concern European defence and the defence budget and are not used. The speaker_role field is documented as unreliable in this store; the Bessent claim is accepted on CONTENT matching the named speaker, not on the tag. All quotes are verbatim from the stored claim text.
• Data as-of: income statement, balance sheet, backlog, estimate-at-completion adjustments, the effective tax rate and the B-21 loss provision through 2026-06-30, all filing-verified from the 10-Q filed 2026-07-21; annual figures from the 10-K filed 2026-01-27 · estimates 2026-08-04 · prices 2026-08-04, quote timestamp 1785873602 = 2026-08-04T20:00:02Z ($551.55, +0.56%; 50-DMA $535.47; 200-DMA $605.67; RSI 64.4; MACD +3.13) · knowledge-base claims 2026-08-04. Northrop's fiscal year is the calendar year. All figures come from the Synthos vendor data file for NOC or from the SEC filings in the NOC archive; no figure comes from memory, recall or external retrieval.
• Filing archive contents: 10-K filed 2026-01-27 (fiscal 2025); 10-Q filed 2026-04-21 (March 2026 quarter); 10-Q filed 2026-07-21 (June 2026 quarter — the source of the backlog, the estimate-at-completion adjustments, the tax-rate disclosure and the balance sheet); 8-K filed 2026-04-21 (Item 2.02, first-quarter results — the earnings release is Exhibit 99 and was NOT extracted); 8-K filed 2026-05-21 (Item 5.07, annual meeting results — directors elected, auditor ratified, and a shareholder proposal for an independent board chair DEFEATED, 23,610,695 for against 88,773,053 against); 8-K filed 2026-07-21 (Item 2.02, second-quarter results — exhibit not extracted). All three 8-Ks were read in full. There is NO corporate action in this archive — no acquisition, divestiture, financing, restructuring or guidance revision. All filings carry preserved [TABLE] statement data.
• Where the filings contradicted or corrected the vendor (detailed in Section 7): totalDebt of $19,741M against a filed $15,696M — a $4,045 million, 25.8% overstatement carrying BOTH documented lease defects in a single payload: operating lease liabilities of $1,857M folded into longTermDebt and $331M into shortTermDebt (the AMT signature), and the same $1,857M non-current lease liability then added AGAIN as capitalLeaseObligations (the HCA signature) — invalidating netDebt ($15,338M against a correct $11,293M at 2025-12-31 and $12,884M at 2026-06-30), netDebtToEBITDATTM (1.976x against approximately 1.79x), debtToEquityRatioTTM (0.911x against 0.85x on filed figures), debtToAssetsRatioTTM and enterpriseValueTTM ($92,336.8M against a rebuilt $91,238.8M); inc_q depreciation for the June 2026 quarter reading ZERO while every neighbouring field in the same row reconciles exactly to the 10-Q; est.ebitdaAvg/ebitAvg, at exactly 18.0170% and 14.5384% of revenueAvg in all eight years from 2023 to 2030, with the implied 14.54% EBIT margin above the company's actual 10.1% June-quarter operating margin and above every full year since 2021; and researchAndDevelopementToRevenueTTM reading ZERO while inc_a carries $1,100 million of R&D. Separately, and not a defect but a coverage gap that changes every interpretation: BACKLOG — $104.7 billion at 2026-06-30, 2.50x annual revenue, with approximately 35% expected to convert within twelve months — appears in NO vendor field, which means roughly 83% of the 2026 revenue consensus is contracted and invisible to any screen. Where vendor and filing AGREED — recorded: the entire June-2026 inc_q row except depreciation — revenue $10,876M, operating income $1,096M, net earnings $1,094M, diluted EPS $7.68, and the prior-year comparatives $10,351M, $1,425M and $1,174M — all reconcile exactly; bal_a FY2025 goodwill of $17,437M, cash of $4,403M and total assets of $51,377M all match the filed balance sheet; minorityInterest of 0 is CORRECT and the NCI-omission class does not apply; tangibleBookValuePerShareTTM of $3.13 and tangibleAssetValueTTM of $445 million are CORRECT rather than defective and properly distinct from bookValuePerShareTTM of $125.86, so the identical-book defect was checked and does not apply; capexToDepreciationTTM of 0.9396 is internally consistent with fiscal-2025 capital expenditure of $1,450M against $1,472M of depreciation and amortisation, though no filed capital-expenditure line was located in the extracted text, so it is recorded as consistent rather than verified; and effectiveTaxRateTTM of 14.54% sits correctly between the filed 2026 year-to-date rate of 10.4% and the 2025 rate of 17.5%.
• Estimate coverage: 14 analysts on FY2026 EPS, 16 on FY2027 — the anchor for all three fair values, range $29.49996 to $31.78608, a 7.7% spread, the second-narrowest in this batch and a direct consequence of the $104.7 billion backlog — 15 on FY2028 and 18 on FY2028 revenue. FY2029 rests on 10 analysts and FY2030 on 6; both are used only as directional cross-checks. est.ebitdaAvg and est.ebitAvg are rejected outright and all forward valuation uses epsAvg only.
• Peer note: the vendor peer set — Emerson Electric, General Dynamics, Howmet Aerospace, Illinois Tool Works, L3Harris, Lockheed Martin, 3M, TransDigm, United Parcel Service, Waste Management — contains four genuine aerospace-and-defence comparables out of ten, alongside two diversified industrials, a parcel carrier and a waste-management company. RTX and Boeing are absent. No peer-multiple comparison is drawn.
• Fair-value caveat: the $428 / $593 / $763 anchors are multiples of the FY2027 consensus EPS distribution — 14.5x the low of $29.49996, 19.5x the mean of $30.38845 and 24.0x the high of $31.78608 — cross-checked against the FY2026 and FY2028 rows and against the backlog. Stated arithmetic, not a discounted cash flow. The base is sensitivity-disclosed: 14.5x gives $441, 16x gives $486, 18.2x gives $552 — spot exactly — and 21x gives $638. The base assumes a 7% multiple EXPANSION from a level that already reflects a 28.2% de-rating, which is the fragile leg and is named as such; at an unchanged multiple the twelve-month return is 5.4% of earnings growth plus 3.72% of shareholder yield, roughly 9%. The bear case of $428 sits 13.7% BELOW the 52-week low of $496.02 — a genuinely severe scenario rather than a routine one, which is what a 1.71:1 payoff ratio on an already-corrected entry should look like.
• Basis note: analyst estimates are on an adjusted basis that runs BELOW reported GAAP earnings in this file — the fiscal-2025 consensus of $26.06 against reported GAAP diluted EPS of $29.08, and the December-2025 quarter's $7.23 against a reported $9.99 — which is the reverse of the usual direction and reflects the exclusion of gains rather than of charges. All forward multiples use the adjusted consensus; the trailing 19.2x is computed on the sum of four filed adjusted quarters ($7.67, $7.23, $6.14, $7.68 = $28.72).
• Guidance note: the earnings releases carrying management's outlook are Exhibit 99 to the 2026-04-21 and 2026-07-21 8-Ks and were not extracted, so no management guidance is available to this dive and none is quoted. The only forward statement used is the 10-Q's backlog-conversion disclosure — "we expect to recognize approximately 35 percent as revenue over the next 12 months" — which is half-weighted per house rule. Earnings-call transcripts are not available on the current data plan.
• Timing: second-quarter 2026 results were released 2026-07-21, fourteen days before this dive, and beat consensus EPS by 12.6% ($7.68 against $6.82) on revenue of $10,876M against $10,801M estimated — while operating income fell 23.1%. The next print is 2026-10-20, 77 days away, at a consensus of $7.13 against $7.67 a year ago — a 7.0% DECLINE on 6.7% higher revenue. The only insider filing is dated 2026-08-03 and is a director's sale of eight shares. 2026-08-04 carried no company-specific filing.
• Accessibility note: no information in this dive is conveyed by colour. All emphasis is carried by bold text, table structure and explicit labelling.
• Not investment advice. Independent research, educational and informational only, never personalised. No recommendation to buy, sell or hold any security is made to any person.
• Version: 2026-08-04-full.