Union Pacific UNP

Reference table

1. What the business is, and the transaction the data omits

Chief executive Vincenzo James Vena; 28,716 employees; NYSE-listed; CIK 0000100885; reporting in USD.

Three commodity groups plus other revenue, and seg_prod ties exactly. The six fiscal-2025 lines sum to $24,510M against reported revenue of $24,510M — an exact match. Verified clean.

seg_geo is corrupt and REJECTED. The fiscal-2025 entry contains a single field named "M [X]" — a truncated country tag, evidently Mexico — with a value of $2,900M against revenue of $24,510M. No geographic figure appears anywhere in this dive.

The transaction, from Note 18 of the 10-Q filed 2026-07-23, in the company's own words:

> "On July 28, 2025, Union Pacific, Norfolk Southern, Ruby Merger Sub 1 Corporation, and Ruby Merger Sub 2 LLC, entered into an agreement and plan of merger..."

> "...each share of Norfolk Southern common stock issued and outstanding immediately prior to the first effective time... [receives] one validly issued, fully paid, and nonassessable share of Union Pacific common stock and $88.82 in cash, without interest. Assuming completion of the mergers, we expect approximately 225 million shares of common stock to be issued and approximately $20 billion of cash consideration to be paid. The cash consideration is expected to be funded through a combination of new debt and cash accumulated through cash provided by operating activities."

> "The combination of Norfolk Southern and Union Pacific would create America's first transcontinental railroad that spans over 50,000 miles across 43 states with access to 10 international interchanges and approximately 100 ports."

> "Completion of the mergers is conditioned on the receipt of Surface Transportation Board (STB) approval... On December 19, 2025, Union Pacific and Norfolk Southern [filed the application]."

> "Both Union Pacific and Norfolk Southern may be required to pay the other a termination fee of $2.5 billion if the merger agreement is terminated under certain circumstances."

> "As part of the pending acquisition of Norfolk Southern described in Note 18, we paused our share repurchases."

Four consequences, none of which appears in any vendor field:

1. Dilution of 37.9%. 225 million new shares against a June-quarter diluted count of 594.0 million.

2. Leverage. Approximately $20 billion of new debt on top of $31.814 billion existing takes gross debt toward $52 billion — roughly 4x combined EBITDA before synergies, against 2.39x today.

3. The buyback is gone. Union Pacific repurchased $2.679 billion in 2025, $1.505 billion in 2024 and $6.282 billion in 2022. It is now zero, and the diluted share count has stopped falling — 593.2M, 593.5M, 593.6M, 594.0M across the last four quarters, i.e. rising.

4. A $2.5 billion contingent obligation in either direction with no vendor field.

The regulatory timeline is unknowable from this file. The application went in on 2025-12-19 and the 10-Q filed seven months later reports no decision. We do not speculate on a date, and we treat the outcome as genuinely binary.

2. The standalone business — four flat years, and a 2026 acceleration

Every EPS figure in this table is computed by us. The vendor's epsdiluted field is NULL in all six annual rows and all nine quarterly rows. See Section 6, finding 1.

Read the revenue column. 2022: $24,875M. 2025: $24,510M. Three years, minus 1.5%. Earnings per share rose 7.3% over the same span — entirely because the diluted share count fell 4.9%, from 624.0M to 593.5M. Union Pacific's per-share earnings growth over the last three years has been a buyback, and the buyback is now paused.

2026 is genuinely different, and it is the strongest thing in this file:

The June quarter grew 11.5% and beat revenue consensus by 2.2% — after four years in which the top line went nowhere. First-half revenue is up 7.4%.

But the operating ratio moved the wrong way, and that is the number railroad investors trade on. The 10-Q: "The Company reported earnings of $3.36 per diluted share on net income of $2.0 billion and an operating ratio of 59.7% in the second quarter of 2026 compared to earnings of $3.15 per diluted share on net income of $1.9 billion and an operating ratio of 59.0% in the second quarter [of 2025]." A 70 basis point deterioration on 11.5% revenue growth means costs grew faster than revenue — the opposite of the operating leverage a volume recovery should produce.

3. Balance sheet and capital returns

Correction — netDebt omits $250M of short-term investments. The vendor nets total debt against cash and equivalents only. The vendor reports $30,539M; the correct figure is $30,289M; we use $30,289M. A 0.8% correction — small, and worth stating because the field fails the same way here as it does across this batch, just by less.

Rebuilt enterprise value:

> Market cap $176.054B (594.0M shares × $296.35)

> + total debt $31.814B − cash $1.275B − short-term investments $0.250B

> = corrected enterprise value $206.343B

The vendor reports $205.609B, understated by 0.4%. Trailing EBITDA is approximately $12.65B (trailing operating income of $10,146M plus roughly $2.5B of depreciation), giving EV/EBITDA of ~16.3x and net debt/EBITDA of 2.39x (vendor prints 15.44x and 2.219x). Interest coverage is 7.90x.

Book-equity metrics are distorted and partially rejected. priceToBookRatioTTM of 8.51x, returnOnEquityTTM of 38.6% and debtToEquityRatioTTM of 1.508x all rest on equity of $18,467M held down by more than a decade of buybacks — $6.282B in 2022 alone. returnOnInvestedCapitalTTM of 11.65% is the honest measure and is what we use. The vendor's own rating scores UNP 1 out of 5 on both debt-to-equity and price-to-book, dragging the composite to B / 3 — a rating artefact of the buyback, not a judgement about the railroad, and rejected as such.

Free cash flow is consistently around $5.5 billion on capital expenditure of roughly 15.5% of revenue — heavy, and unavoidable for a railroad. The dividend alone consumes $3.2 billion, or 59% of free cash flow. With the buyback paused, remaining free cash flow of roughly $2.3 billion a year is being accumulated toward the $20 billion of merger consideration, which is exactly what the 10-Q says: "funded through a combination of new debt and cash accumulated through cash provided by operating activities."

Dividend. The vendor reports lastDividend: 5.52 and dividendPerShareTTM: 5.52, yielding 1.86%. That $5.52 is the trailing-twelve-month total, not the last declared quarterly dividend — the same mislabelling seen across this batch. Dividends paid in 2025 were $3,236M on 593.5M shares, or $5.45 per share. Payout is 44.7% of trailing earnings.

Total shareholder yield is now 1.86% — the dividend alone. It was 3.4% in 2025 with the buyback. That 1.5-point reduction in cash return is a real, current cost of the transaction and no screen shows it.

4. Valuation — priced in or room?

At $296.35 (market cap $176.054B, corrected enterprise value $206.343B):

The estimate block is internally inconsistent from fiscal 2028 and those years are EXCLUDED — this is the most important valuation finding on this name.

Revenue of $30.523B (FY2028), $35.211B (FY2029) and $38.956B (FY2030) cannot be Union Pacific standalone. The company has produced $24.1-24.9B every year since 2022; $35.2B by 2029 would be +43.7% in four years, or 9.5% a year, from a business that grew 1.1% last year. Those lines consolidate Norfolk Southern, whose revenue base is roughly $12 billion.

But the EPS lines for the same years divide by a pre-merger share count. Dividing netIncomeAvg by epsAvg gives an implied diluted share count of 575.3M for FY2028 and 593.5M for FY2029 — against a post-merger count of roughly 819 million (594M + 225M). The revenue is consolidated; the shares are not. A reader computing a fiscal-2029 P/E of 16.2x is applying a merged revenue base to an unmerged share count, and the resulting earnings-per-share figure is roughly 38% too high.

We therefore anchor on the fiscal-2027 line (16 analysts on both revenue and EPS, and the last year plausibly free of the contamination) and exclude fiscal 2028 onward entirely.

est.ebitdaAvg and est.ebitAvg additionally carry the fixed-ratio fabrication signature. ebitdaAvg / revenueAvg = 51.79% and ebitAvg / revenueAvg = 41.94% in all eight years, to four significant figures. Actual fiscal-2025 EBITDA was 52.8% of revenue and operating margin 40.1% — close enough to look plausible and still constructed. Per the data contract, all forward valuation runs on epsAvg.

Peer context. No peer set is supplied in the payload for UNP — the peers array does not appear in the extracted fields. Union Pacific appears in Boeing's peer set in this batch, which is not a meaningful comparison. The genuine comparables are the other Class I railroads, none of which is in this file. No peer-multiple comparison is drawn.

4a. What today's price assumes (the inversion)

At $296.35 — 22.7x FY2026 consensus and 20.9x FY2027 — the price embeds:

• The Surface Transportation Board approves the Norfolk Southern merger. (Our framing; the application date is the filing's.) This is the most fragile assumption in the price and it is not a financial assumption at all. The stock has returned 35.1% over twelve months against 24.3% for the index on a business whose revenue grew 1.1% in 2025 — a large part of that eleven-point excess is merger anticipation. Refusal would remove it.
• Revenue reaches $26.753B in 2026 (+9.2%) and $27.920B in 2027. (Consensus; 16 analysts each.) First-half 2026 grew 7.4% and the June quarter 11.5%, so the fiscal-2026 figure is within reach — but it follows four years at zero, and the check is 2026-10-22 at $6.947B, +11.3%.
• The operating ratio returns below 59%. (Our number; the 59.7% and 59.0% are the filing's.) Consensus EPS growth of 8.5% in fiscal 2027 on 4.4% revenue growth requires margin expansion, and the most recent quarter delivered the opposite.
• The buyback resumes. (Our assumption.) The diluted share count has stopped falling and is now rising slightly. Roughly 2.5 points a year of the historical per-share earnings growth came from repurchases and it is currently zero.
• The market keeps paying 21-23x on the forward year. (Our number.) At 18x FY2027E the stock is $255; at 25x it is $354. Union Pacific's own history is 19-23x, so 22.7x on fiscal 2026 is at the upper end already.

4b. The return bridge (why the multiple moves)

Expected return to the end of fiscal 2027 decomposes as: adjusted EPS growth (+8.5%, from FY2026E $13.046 to FY2027E $14.153) + multiple drift (COMPRESSION, from 22.7x to 21.5x, −5%) + dividend yield (+1.86% a year) ≈ +5% to +7% over the period, or roughly +4% annualised.

This is the weakest return bridge in this batch and the reason is structural rather than a view. The three historical sources of Union Pacific shareholder return were volume growth (zero for four years), margin expansion (reversing — the operating ratio deteriorated 70 basis points last quarter) and the buyback (paused). What remains is a 1.86% dividend and whatever the merger delivers.

The merger arithmetic, stated plainly and simply. Issuing 225 million shares and paying $20 billion of cash for Norfolk Southern's roughly $12 billion of revenue and roughly $3 billion of net income is accretive to earnings per share only if synergies are large. Post-deal share count of roughly 819 million against a combined net income of roughly $10.5 billion (UNP $7.5B + NS $3.0B) before synergies and before the interest on $20 billion of new debt — perhaps $900 million after tax — gives roughly $11.70 of pro forma EPS against $13.05 standalone in fiscal 2026. On the arithmetic available in this file the transaction is dilutive to earnings per share before synergies, and the case for it is the synergies and the single-line network, neither of which the filings quantify. We flag this as our own computation from filing inputs, not as a company or consensus figure, and it is the single most important number a reader of this name should carry.

If the multiple simply HELD at 22.7x on FY2027E the price would be $321 (+8%); if it compressed to 18x it would be $255 (−14%). The bull case at $355 is 25x the FY2027 consensus HIGH of $14.50; the bear case at $229 is 17x the FY2027 consensus LOW of $13.49 and sits 6.6% above the 52-week low.

4c. Variant perception (where we differ, what would surprise)

• We differ from the consensus estimate block on its own internal consistency, and this is a finding rather than an opinion. Fiscal-2028 to fiscal-2030 revenue consolidates Norfolk Southern; the EPS lines for those years use a 575-594 million share count against a post-merger ~819 million. Anyone valuing UNP on a fiscal-2029 P/E of 16.2x is combining merged revenue with unmerged shares.
• We compute the transaction as dilutive to earnings per share before synergies, on filing inputs: ~819 million pro forma shares, ~$10.5 billion of combined pre-synergy net income, less roughly $900 million of after-tax interest on $20 billion of new debt, gives roughly $11.70 against $13.05 standalone. The filings do not quantify synergies and the payload contains nothing about them. This is our arithmetic and we label it as such — but the burden of proof sits with the transaction, and the price does not appear to reflect that.
• We think the paused buyback is materially under-appreciated. Union Pacific returned $5,915 million in 2025 — $3,236M of dividend and $2,679M of repurchase, a 3.4% total yield. It is now 1.86%, the dividend alone, and the diluted share count has stopped falling. Roughly 2.5 points of the company's historical per-share growth rate has been switched off and the switch stays off until the deal closes or dies.
• We differ on the standalone growth picture. Consensus fiscal-2026 revenue growth of +9.2% follows four years of nothing (2022 $24.875B, 2025 $24.510B). The June quarter's +11.5% is genuine and encouraging. Watchable number: revenue in the 2026-10-22 print against $6.947B, and the operating ratio, which must return below 59% for the fiscal-2027 margin assumption to hold.
• Positive surprise that would force a re-rate: a Surface Transportation Board approval with limited conditions; a quantified synergy figure from management; the operating ratio back below 58%; or a resumption of the buyback, which would signal confidence in the financing.
• Negative surprise that would break the thesis: a Board refusal or an approval loaded with trackage-rights and rate conditions; the operating ratio above 60% for two consecutive quarters; a slowdown in the 2026 revenue acceleration; or a credit-rating action on the $20 billion of prospective financing. At the 88th percentile of its annual range and 22.7x forward, none of these is absorbed by the price.

Synthos fair values

All three anchors are multiples of the fiscal-2027 consensus adjusted-EPS distribution (mean $14.153, low $13.49, high $14.50, 16 analysts — the last uncontaminated forward year), and none uses fiscal 2028 or beyond, for the reasons in Section 4.

• Bear ~$229 — 17x the FY2027 consensus LOW of $13.49. Cross-check: 18.3x FY2026E; 6.6% above the 52-week low of $214.91; 8.9% below the 200-day average of $251.38. The scenario: the Surface Transportation Board refuses or conditions the merger heavily, the buyback stays paused through a termination process costing up to $2.5 billion, the 2026 revenue acceleration proves to be a comparison effect, and the eleven points of twelve-month outperformance unwind. −22.7%.
• Base ~$305 — 21.6x the FY2027 consensus MEAN of $14.153. Cross-check: 23.4x FY2026E; corrected EV of $215.3B on FY2027E revenue of $27.920B is 7.71x sales. Sensitivity, stated openly: 20x FY2027E gives $283, 4.5% BELOW spot, and 23x gives $326, close to the street's mean. The whole answer lives in a 20-23x band. The scenario: the merger completes on roughly the announced terms, the pre-synergy dilution is offset over time by cost savings the filings do not quantify, and the multiple holds near where it is. +2.9%, plus 1.86% a year of dividend.
• Bull ~$355 — 25x the FY2027 consensus HIGH of $14.50. Cross-check: 27.2x FY2026E; below the street's $363 high. The scenario: the Board approves cleanly, management quantifies large synergies, the single-line transcontinental network takes share from long-haul trucking as the knowledge base's cost claim implies it should, the buyback resumes, and Union Pacific re-rates as a growth industrial. +19.8%.

Base is 2.9% above spot; asymmetry roughly 0.87:1 — NEGATIVE (22.7% down, 19.8% up). Together with Boeing, this is one of only two names in this batch where the asymmetry runs against the buyer, and it is the central reason for the verdict. The business is excellent. The entry, after a 35% year and with the outcome in a regulator's hands, is not.

5. Knowledge base — two claims name the company, neither mentions the merger

Raw hits: 25 (2 entity, 23 text). Used: 6. Discarded: 19. Searches covered Union Pacific, Norfolk Southern, railroad and railroads.

The dominant discard class, and it is instructive. TWELVE of the twenty-three text hits use "railroads" as a historical analogy for a technology bubble — "Investors reliably overbid new technologies (railroads, radio, talkies, dot-com) into bubbles that crash and burn"; "New tech follows a boom-bust life cycle (like 19th-century railroads)"; "Like railroads and internet telecom, the bubble builds real infrastructure that ends up amazing"; "analogous to the 1990s and the late-1800s railroads"; "US 1850-1970 railroads forced machine tools then interchangeable parts then autos". Not one is about the freight-rail industry as an operating business. Railroads are among the most frequently used metaphors in investment conversation, and any keyword sweep on this name will be dominated by them — this is the semantic equivalent of the Visa/"visa" homograph problem and it is the single most important collision class for UNP. Seven further text hits are macro or policy claims where railroads appear incidentally.

The two entity claims:

> 2022-07-18 · neutral · conviction 60 · fact · channel: invest_like_the_best · no named speaker · entity: Union Pacific

> "Precision scheduled railroading cut Union Pacific from 42,000 employees to 30,000 in five years, then left it critically short-staffed post-COVID."

> ADJUDICATED against the current filing and the trend has continued, not reversed: profile.fullTimeEmployees is 28,716 today — BELOW the 30,000 the claim describes as already critically short. Union Pacific has removed a further 1,284 positions since. Whether that is efficiency or under-resourcing is exactly the question the 70 basis point operating-ratio deterioration in the June quarter raises, and the claim has aged into relevance rather than out of it.

> 2026-02-06 · neutral · conviction 60 · principle · channel: compound_and_friends · no named speaker · entities: Union Pacific, Prologis

> "Rails and industrial storage/REITs (Union Pacific, Prologis) are valuation/risk peers because freight must be stored — the market prices them symbiotically."

> A valuation-framework observation rather than a view. Noted as colour.

Four category-level text claims are used, because they are genuinely about the business rather than the metaphor:

> 2022-07-18 · bullish · conviction 65 · principle · channel: invest_like_the_best

> "Railroads have durable advantages: ~4x cheaper than trucks, no new lines being built, stable growers with pricing power."

> The core structural case, independently stated, and the strongest support for the moat.

> 2024-10-17 · neutral · conviction 60 · principle · channel: no_priors

> "Physical-moat assets like railroads are fully AI-robust — AI doesn't help or threaten them; durability comes from irreproducible footprint, not tech."

> 2024-07-18 · bullish · conviction 58 · principle · channel: no_priors

> "Own AI-durable businesses (railroads, certain SaaS) where AI adds little — that indifference itself signals competitive durability."

> 2022-10-24 · neutral · conviction 78 · principle · channel: we_study_billionaires

> "Favor stable industries (railroads, energy, insurance) with consistent operating history; severe change and exceptional returns rarely mix."

> Worth pausing on: this claim's own logic argues AGAINST the current setup. Union Pacific is attempting the most severe change in its history.

The finding that matters most. Not one claim in 51,928 mentions the Norfolk Southern merger — the transaction that determines this security's value, announced 2025-07-28 and pending before a regulator since 2025-12-19. The Synthos knowledge base has a good structural view of railroads and no view whatsoever on the single event that will decide this stock. Conviction Very Low, breadth 6, and none of the six claims carries a named speaker. No management voice and no promoter-class voice appears.

6. Data integrity — what we rejected from the vendor file and why

Eight findings. UNP's file is clean on segments and defective on the earnings-per-share field, the geographic block and the forward estimates.

1. epsdiluted is NULL in every income-statement row — annual and quarterly — and every EPS figure in this dive is computed. All six annual rows and all nine quarterly rows return None. This is the zeroed-field defect class: a null that hides nothing sinister but forces every per-share figure to be derived. We compute EPS as net income divided by weightedAverageShsOutDil, which validates against both the vendor's own ratio and the filing: our trailing figure of $12.35 against priceToEarningsDilutedRatioTTM of 23.996 implies $12.35 exactly, and our June-quarter figure of $3.36 matches the 10-Q's stated "earnings of $3.36 per diluted share" to the cent. The derivation is sound; the field is missing; every EPS number here is labelled as computed.

2. The consensus estimate block is internally inconsistent from fiscal 2028 and those years are EXCLUDED. Revenue of $30.523B (FY2028), $35.211B (FY2029) and $38.956B (FY2030) cannot be Union Pacific standalone — the company has produced $24.1-24.9B every year since 2022, and $35.2B by 2029 is +43.7% in four years. Those lines consolidate Norfolk Southern. But dividing netIncomeAvg by epsAvg gives implied share counts of 575.3M (FY2028) and 593.5M (FY2029) against a post-merger count of roughly 819 million. The revenue reflects the merger and the share count does not, so the EPS lines are roughly 38% too high. All three fair-value anchors use fiscal 2027 (16 analysts on both lines) and nothing beyond it.

3. est.ebitdaAvg and est.ebitAvg carry the fixed-ratio fabrication signature — REJECTED. ebitdaAvg / revenueAvg = 51.79% and ebitAvg / revenueAvg = 41.94% in all eight years, to four significant figures. Actual fiscal-2025 figures were 52.8% and 40.1% — close enough to pass a glance and still constructed. All forward valuation runs on epsAvg.

4. seg_geo is corrupt and REJECTED in full. The fiscal-2025 entry contains a single field named "M [X]" — a truncated country tag — valued at $2,900M against revenue of $24,510M. No geographic figure appears anywhere in this dive.

5. netDebt omits $250M of short-term investments — corrected. The vendor reports $30,539M; netting the short-term investments gives $30,289M. A 0.8% correction, small by the standards of this batch, and recorded because the field fails the same way here as everywhere else. Consequential: enterpriseValueTTM $205.609B → $206.343B (the vendor's EV is separately understated relative to its own net-debt figure); netDebtToEBITDATTM 2.219x → 2.39x on our recomputed EBITDA; evToEBITDATTM 15.44x → ~16.3x.

6. The transaction is in no vendor field at all — added entirely from the filing. Approximately 225 million shares to be issued (37.9% dilution), approximately $20 billion of cash consideration funded with new debt, a $2.5 billion termination fee each way, share repurchases paused, and a Surface Transportation Board application filed 2025-12-19 with no ruling. None of bal_a, cf_a, km_ttm, ratios_ttm or est contains any indication that this is happening, and the share count, leverage and shareholder yield in every screen are therefore describing a company that is about to cease to exist in its current form.

7. lastDividend: 5.52 is the trailing-twelve-month total, not the last declared dividend. Dividends paid in fiscal 2025 were $3,236M on 593.5M shares, or $5.45 per share. The field is the TTM figure, consistent with the mislabelling seen throughout this batch.

8. Book-equity metrics are distorted by a decade of buybacks — partially rejected. priceToBookRatioTTM of 8.51x, returnOnEquityTTM of 38.6% and debtToEquityRatioTTM of 1.508x rest on equity of $18,467M against $69,698M of assets, held down by repurchases including $6,282M in 2022 alone. The vendor's rating scores UNP 1 out of 5 on both debt-to-equity and price-to-book, dragging the composite to B / 3. Both sub-scores are artefacts of the buyback and are rejected as lenses. returnOnInvestedCapitalTTM of 11.65% is the honest measure.

seg_prod ties exactly — a clean check. The six fiscal-2025 lines sum to $24,510M against reported revenue of $24,510M. Verified clean across the years used.

Share count — checked and consistent. Market cap of $176,054,273,832 divided by $296.35 gives 594,072,000, against a June-quarter diluted weighted average of 594.0M — agreement to 0.01%. No share-class error. Note that the count has RISEN across the last four quarters (593.2M, 593.5M, 593.6M, 594.0M), which is the buyback pause becoming visible in the data.

quote.yearHigh/yearLow ($315.99 / $210.84) disagree with tech.hi52/lo52 ($307.32 / $214.91) by 2.8% and 1.9% — the largest such discrepancy in this batch. We use tech throughout, per house rule.

No peers data is available for this name, so no peer comparison of any kind is drawn.

Non-equity tripwire — checked and passed. UNP is NYSE-listed common stock. Price $296.35 is not par-like; beta 0.965; the dividend is variable with a long record of increases; volume 2.35M shares (~$696M of turnover); the 52-week band of $214.91 to $307.32 is a 43% range. This is common equity.

7. Technicals

• Price $296.35. −3.6% from the 52-week high of $307.32; +37.9% above the low of $214.91. Position within the annual range: 88th percentile — the second-highest in this batch.
• Moving averages, both well below and both rising: +6.4% above a 50-day average of $278.48 and +17.9% above a 200-day of $251.38. A clean uptrend.
• RSI 56.3 — neutral. MACD +4.65 — positive.
• Maximum drawdown from peak over the trailing year: −3.6%. This stock has essentially not corrected.
• Relative performance, strong across every horizon: 3-month +12.5% vs SPY +7.6%; 6-month +26.0% vs SPY +11.1% and QQQ +15.7%; 12-month +35.1% vs SPY +24.3% and QQQ +30.8%. An eleven-point twelve-month outperformance of the index by a railroad whose revenue grew 1.1% last year — that gap is the merger being priced in.
• Sentiment: 1 strong buy, 27 buy, 18 hold, 1 sell across 47 analysts. Consensus target $330.50 (+11.5%), median $334, high $363 (+22.5%), low $294 — 0.8% below spot.

Today's move and the entry

UNP closed 2026-08-04 at $296.35, up 1.76% or $5.12 from $291.23, on 2.35M shares. It opened at $289.34 and closed within 0.2% of the day's high of $296.81 — a 3.1% intraday advance and a strong close. No company-specific news appears in this file for 2026-08-04; the last event was the 2026-07-23 second-quarter release twelve days earlier.

The honest read. The chart is among the best in this batch and that is precisely the problem: at the 88th percentile of its range, 17.9% above the 200-day, with a twelve-month outperformance of eleven points earned on 1.1% revenue growth, the stock has already priced a regulatory approval that has not been granted.

What would change the answer:

• A price near $250 — roughly 17.7x FY2027E, at the 200-day average of $251.38 — which would put our $305 base at +22% and invert the asymmetry to roughly 2:1. That is 15.6% below today's close.
• Or a Surface Transportation Board approval with limited conditions and a quantified synergy figure, which would convert a regulatory bet into an arithmetic one. At that point paying 22x is defensible because the binary has resolved.

8. Insiders

The insider file contains only routine equity-plan mechanics. The largest entry is an A-Award of 4.879 shares to an executive vice president at a reference price of $286.96 — a dividend-equivalent accrual on an existing 61,133-share position, not a transaction of any economic significance.

There are no open-market purchases and no open-market sales in the file. That is a neutral reading. With the stock up 35% over twelve months and within 4% of its high, the absence of insider selling is a mild positive; the absence of buying is unremarkable in a pending-merger blackout period, which is likely the operative constraint.

9. Verdict, kill-criteria and flip conditions

Watch.

What is genuinely excellent: an irreplaceable western rail network with an operating margin of 40.1% and an operating ratio just under 60; $5.5 billion of annual free cash flow; 11.65% return on invested capital; independent structural corroboration that railroads are "~4x cheaper than trucks, no new lines being built, stable growers with pricing power" and "fully AI-robust"; a June quarter that grew 11.5% and beat on both lines after four flat years; a 0.965 beta; a segment table that ties to the dollar; and a prospective 50,000-mile transcontinental network across 43 states that would be a genuine improvement in the product rather than a cost synergy.

What we are declining to pay for: approximately 225 million new shares — 37.9% dilution — and approximately $20 billion of cash funded with new debt, on a transaction that on our own arithmetic from filing inputs is dilutive to earnings per share before unquantified synergies; a paused buyback that has taken total shareholder yield from 3.4% to 1.86% and stopped the share count falling; a $2.5 billion termination fee in either direction; four years of no revenue growth (2022 $24.875B, 2025 $24.510B); an operating ratio that deteriorated to 59.7% from 59.0% on 11.5% revenue growth; a consensus estimate block that consolidates Norfolk Southern's revenue while dividing by an unmerged share count; a stock at the 88th percentile of its range after eleven points of twelve-month outperformance; negative asymmetry of 0.87:1; and a base fair value of $305 that is 2.9% above spot and 7.7% below the street's $330.50.

The distinction that matters. This is not a judgement about the railroad, which is one of the finest assets in the listed market. It is that the equity has been re-rated 35% in a year on the anticipation of a regulatory decision nobody in this file has made, that the terms of the transaction dilute existing holders by 38% and lever the company toward 4x, and that the cash return that historically carried the stock has been switched off in the meantime. In the Synthos frame, a name where our fair value is at spot, below consensus, with negative asymmetry and no name-level conviction, is a Watch by construction — and the wanted price is 15% lower.

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

• A Surface Transportation Board refusal, or an approval conditioned on trackage rights, rate protections or service commitments material enough to erode the synergy case.
• The operating ratio above 60% for two consecutive quarters. It was 59.7% in June against 59.0%.
• The 2026 revenue acceleration reversing — third-quarter revenue below roughly $6.6 billion against the $6.947 billion consensus.
• A credit-rating action on the prospective $20 billion of merger financing, which would raise the cost of a deal already dilutive before synergies.
• The buyback remaining paused into 2028 without a completed transaction, which would mean shareholders had foregone roughly $5 billion of repurchases for nothing.

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

• Price at or below approximately $250 (roughly 17.7x FY2027E, at the 200-day average). This is the most likely path to a Buy and it is 15.6% below today's close. At $250 the base case is +22% and the asymmetry inverts to roughly 2:1.
• A Surface Transportation Board approval with limited conditions, which converts the dominant uncertainty into arithmetic.
• A quantified synergy figure from management sufficient to make the transaction accretive on the pro forma share count — the filings currently quantify nothing.
• The operating ratio back below 58% with revenue growth sustained above 5%.
• Resumption of the buyback, which would signal that the financing is secured and restore 1.5 points of shareholder yield.

Where UNP fits in the Synthos Framework Portfolio. The transport and physical-infrastructure sleeve, at 0% today with a 1.5-2% target on a fill near $250 or a clean regulatory approval. On batch overlap: UNP sits with LIN and NEE as the contracted-or-regulated cash-flow assets here. Among the three it has the best asset, the worst growth record and the only binary regulatory event. It is also, notably, the only name in this batch whose share count is currently RISING. Logged as a tracked Synthos call (Watch) as of 2026-08-04 at $296.35.

Single biggest risk: the Surface Transportation Board. The application was filed 2025-12-19 and there is no ruling in this file. The stock has outperformed the index by eleven points over twelve months on 1.1% revenue growth, and a large part of that is anticipation. A refusal returns Union Pacific to being a flat-revenue railroad at 22.7x forward earnings, with a termination fee of up to $2.5 billion, a buyback that has been paused for a year and a half, and a share count that has stopped shrinking. There is no valuation cushion under that: our bear case of $229 is 22.7% below the current price and it assumes only a return to 17x on a slightly disappointing number.

Most fragile assumption in the price: that the merger creates value for existing Union Pacific holders. Every other assumption is a volume or margin assumption, observable quarterly. This one is arithmetic we can do and the filings cannot answer. Issuing 225 million shares onto a 594 million base and paying $20 billion of cash funded with debt for a business with roughly $12 billion of revenue and roughly $3 billion of net income gives, on our computation from filing inputs, approximately $11.70 of pro forma earnings per share against $13.05 standalone in fiscal 2026 — before synergies and after roughly $900 million of after-tax interest on the new debt. The transaction is dilutive on the numbers available and the entire case rests on synergies that no document in this archive quantifies. That is not a reason to conclude it fails. It is a reason not to pay 22.7x for it in advance.

Provenance & disclosures

• Traceability: 2 tagged knowledge-base entity claims name Union Pacific (raw hits 25: 2 entity, 23 text; 6 used, 19 discarded; breadth 6, net conviction positive-negligible). The dominant discard class is METAPHOR: twelve of the twenty-three text hits use "railroads" as a historical analogy for a technology bubble ("Investors reliably overbid new technologies (railroads, radio, talkies, dot-com)"; "like 19th-century railroads"; "Like railroads and internet telecom, the bubble builds real infrastructure"; "analogous to the 1990s and the late-1800s railroads"; and eight others), and not one is about the freight-rail industry as an operating business. This is the semantic equivalent of the Visa/"visa" homograph problem and it is the single most important collision class for this name. Seven further text hits are macro or policy claims where railroads appear incidentally. Four category-level claims ARE used because they are genuinely about the business: railroads are "~4x cheaper than trucks, no new lines being built, stable growers with pricing power" (bullish 65); they are "fully AI-robust — AI doesn't help or threaten them" (neutral 60) and "AI-durable... that indifference itself signals competitive durability" (bullish 58); and "favor stable industries (railroads, energy, insurance)... severe change and exceptional returns rarely mix" (neutral 78) — the last of which argues against the current setup on its own logic. One entity claim is ADJUDICATED: the 2022-07-18 observation that precision scheduled railroading "cut Union Pacific from 42,000 employees to 30,000 in five years, then left it critically short-staffed post-COVID" is confirmed and extended — profile.fullTimeEmployees is 28,716 today, BELOW the level the claim already called critically short. NONE of the six claims carries a named speaker, and NOT ONE CLAIM in 51,928 mentions the Norfolk Southern merger — the single event that determines this security — which we report as an absence rather than padding around it. No management voice and no promoter-class voice appears. Searches: Union Pacific, Norfolk Southern, railroad, railroads. All quotes are verbatim from the stored claim text.
• Data as-of: fundamentals — annual statements to 2025-12-31 from the vendor payload; quarterly income statement to 2026-06-30; most recent filing-verified financials 2026-06-30, from the 10-Q filed 2026-07-23 · estimates 2026-08-04 · prices 2026-08-04, quote timestamp 1785873602 = 2026-08-04T20:00:02Z ($296.35, +1.76%; 50-DMA $278.48; 200-DMA $251.38; RSI 56.3; MACD +4.65) · knowledge-base claims 2026-08-04. Fiscal note: Union Pacific's fiscal year is the calendar year. All figures come from the Synthos vendor data file for UNP or from the SEC filings in the UNP archive; no figure comes from memory, recall or external retrieval.
• Filing archive contents and its limitation: 10-K filed 2026-02-06 (fiscal 2025); 10-Q filed 2026-04-23 (March 2026 quarter); 10-Q filed 2026-07-23 (June 2026 quarter); 8-K filed 2026-04-23; 8-K filed 2026-05-18; 8-K filed 2026-07-23 (second-quarter results). The UNP archive is flagged tables: false in the filings index — financial-statement tables were NOT preserved — so every filing figure in this dive comes from narrative prose (the notes and MD&A) rather than from a statement table. That is why the merger terms, the operating ratio and the buyback pause are quoted at length while no June-quarter balance sheet is reproduced: the balance-sheet figures in this dive are the vendor's 2025-12-31 statement, and we state explicitly that no filing-verified balance sheet as at 2026-06-30 was available to us. The three 8-Ks are cover documents with no extracted financials.
• Where the filings contradicted or corrected the vendor (detailed in Section 6): the entire Norfolk Southern transaction added from the filing, because NO vendor field contains it — approximately 225 million shares to be issued (37.9% dilution on 594.0M), approximately $20 billion of cash consideration "funded through a combination of new debt and cash accumulated through cash provided by operating activities", a $2.5 billion termination fee each way, share repurchases paused, and a Surface Transportation Board application filed 2025-12-19 with no ruling; the consensus estimate block shown to be internally inconsistent from fiscal 2028, with revenue lines of $30.523B, $35.211B and $38.956B that consolidate Norfolk Southern against implied share counts of 575.3M and 593.5M derived from netIncomeAvg ÷ epsAvg — versus a post-merger count of roughly 819 million — so fiscal 2028 onward is excluded entirely and all three fair-value anchors use fiscal 2027; epsdiluted found NULL in all fifteen income-statement rows, with every EPS figure computed and validated against both priceToEarningsDilutedRatioTTM (our $12.35 trailing implies exactly 23.996x) and the 10-Q's stated "earnings of $3.36 per diluted share" for the June quarter; the operating ratio taken from the filing (59.7% against 59.0%), a metric with no vendor field; est.ebitdaAvg/ebitAvg rejected on a fixed-ratio fabrication signature (51.79% and 41.94% of revenue in all eight years); seg_geo rejected in full for containing a single truncated field "M [X]" at $2,900M against $24,510M of revenue; netDebt corrected from $30,539M to $30,289M for the omission of $250M of short-term investments; lastDividend of $5.52 identified as the TTM total; and book-equity metrics (P/B 8.51x, ROE 38.6%, D/E 1.508x) rejected as buyback artefacts along with the vendor's B/3 composite rating. Where vendor and filing AGREED — worth recording: the June-quarter diluted EPS (our computed $3.36 against the 10-Q's stated $3.36); the share count (implied 594,072,000 against a 594.0M diluted weighted average); seg_prod summing exactly to reported revenue; and the quarterly revenue series against the earnings calendar.
• Basis note: est.epsAvg is an adjusted (non-GAAP) EPS measure. The earn_cal actuals ($3.08, $2.86, $2.93, $3.41) differ from our computed GAAP diluted EPS ($3.01, $3.11, $2.87, $3.36) in every quarter, and the June-quarter gap of $0.05 corresponds to the acquisition-related expense the 10-Q discloses. The gap is small — roughly 1.5% on the trailing year — so the basis choice moves the multiple by about 0.4 turns and no conclusion turns on it. Forward multiples use consensus adjusted EPS; the trailing multiple of 24.0x is computed on our derived GAAP figure of $12.35.
• Estimate coverage: 14 analysts on FY2026 EPS and 16 on revenue; 16 and 16 on FY2027 — the anchor year; 6 and 12 on FY2028; 4 and 11 on FY2029; 2 and 6 on FY2030. Fiscal 2028 through 2030 are excluded entirely, not for thin coverage but for the internal inconsistency described above.
• Peer note: no peers array is available for this name in the payload, so no peer comparison of any kind is drawn. The genuine comparables are the other Class I railroads, none of which appears. Union Pacific itself appears in Boeing's peer set elsewhere in this batch, which is not a meaningful comparison.
• Fair-value caveat: the $229 / $305 / $355 anchors are multiples of the fiscal-2027 consensus adjusted-EPS distribution — 17x the low of $13.49, 21.6x the mean of $14.153, and 25x the high of $14.50, on 16 analysts. Stated arithmetic, not a discounted cash flow, and deliberately confined to the last forward year the estimate block treats consistently. The base is sensitivity-disclosed: 20x gives $283, 4.5% BELOW spot; 23x gives $326, close to the street's mean. The whole answer lives in a 20-23x band. Note that all three anchors are STANDALONE valuations. We do not attempt a pro forma fair value for the merged entity, because the filings quantify no synergies and any such figure would be invented. Our pro forma earnings-per-share computation in Section 4b (roughly $11.70 against $13.05 standalone, before synergies) is our own arithmetic from filing inputs and is labelled as such throughout — it is not a company figure, not a consensus figure, and not used as a valuation anchor.
• Timing: second-quarter 2026 results were released 2026-07-23, twelve days before this dive, beating on both lines (adjusted EPS $3.41 against $3.26; revenue $6,864M against $6,715M) while the operating ratio deteriorated to 59.7% from 59.0%. The next scheduled print is 2026-10-22, 79 days away. The dominant catalyst — a Surface Transportation Board ruling on the merger — is unscheduled and no date appears in any filing in this archive. 2026-08-04 was a firm session; UNP's +1.76% carried no company-specific news in this file.
• 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.