CSX CSX

Industrials · Railroads · Synthos Deep Dive · 2026-08-04

$51.28

Hold

—

The Overview

CSX is a railroad. It owns about 20,000 miles of track across the eastern United States and moves heavy things — chemicals, cars, coal, grain, timber, metals and shipping containers — from one place to another. Nobody is going to build a competing rail network, so it is a genuinely irreplaceable asset.

For three years it was shrinking. Revenue fell from $14.7 billion in 2023 to $14.1 billion in 2025.

This year it turned around, and the numbers are good. In the three months to June, the amount of freight it hauled rose 6%, revenue rose 10%, and — the important part — costs rose only 6%. That gap is profit: operating profit rose 17% and earnings per share rose 23%. Over the first six months, costs actually fell while revenue rose 6%, and the profit margin improved by four full percentage points.

The shares have responded: up 46% in the past year, roughly double the market, and they now sit 4% below their highest price of the year.

Here is the problem. Analysts on average think CSX is worth $51.06. It closed at $51.03. On the street's own consensus, there is nothing left.

And there is something else in that distribution worth understanding. The most optimistic analyst says $60. The most pessimistic says $32 — 37% below today's price. That is not a normal spread of opinion about a railroad's earnings. It is two different scenarios.

The reason is one sentence in CSX's own quarterly filing. Explaining a small cost increase, the company mentions "advisory expenses related to potential industry consolidation." In plain terms: CSX is paying investment bankers to think about a merger. Nothing has been announced. But the American railroad industry has been consolidating for decades, there are very few independent networks left, and CSX brought in an outsider chief executive.

So a large part of what you are paying for at $51.03 is a deal that has not happened and might not. If it does, the shares are probably worth considerably more. If it does not, the $32 analyst is describing what happens when a railroad growing profits about 13% a year is valued like a railroad again.

Our estimate of fair value is $52, essentially the current price. That is a Hold — a genuinely good business, in a genuine recovery, at a price that has already paid for both, with a binary attached that we cannot handicap.

One honest note about our own research: the Synthos knowledge base contains 51,928 recorded expert claims and not a single one mentions CSX — not the company, not its recovery, not the consolidation wave its own filings say it is paying advisers to consider.

Downside Risk 7/10. A good asset priced for an unannounced event, 4% from its high.
Growth Quality 6/10. A real margin recovery on top of three years of revenue decline.
Exponential Potential 2/10. A railroad. The only non-linearity is corporate, not operational.

Putting a number on it: our fair-value estimate is $52 against a current price of $51.28 — real upside if our numbers are right.

Our summary metrics

Downside Risk (lower = safer)7/10High

"Rated 7 — a good asset in a genuine operating recovery, priced for an outcome that has not been announced. The supports: an irreplaceable eastern US rail network; a June-quarter operating margin of 38.3% and a six-month margin of 37.2%, up 400 basis points year on year; return on equity of 24.1% and return on capital employed of 12.2%; volume up 6% in the quarter and 4% for the half; and a share count reduced from 2,304 million in fiscal 2020 to 1,864 million in fiscal 2025, a 19% reduction. Against that, five risks. FIRST and largest, the entry: the stock is 4.1% below its 52-week high after a 45.8% twelve-month move and a 59.2% advance off its low, with a maximum drawdown over the year of only 4.1% — there has been no pullback to buy. SECOND, the price appears to embed a corporate event that does not exist in this archive: the 10-Q discloses 'advisory expenses related to potential industry consolidation' and nothing else, and the street's price-target distribution runs from a $60 high to a $32 LOW — 37.3% below spot — which is a two-outcome distribution wearing the clothes of a forecast. THIRD, leverage: total debt of $19,352 million against net debt of $18,682 million and roughly 2.74x EBITDA, in a business whose revenue declined in each of the last three fiscal years. FOURTH, the buyback has been cut hard — $4,731M in fiscal 2022 to $3,482M, $2,237M and $1,396M in fiscal 2025 — a 70% reduction over three years, which either signals capital being conserved or free cash flow under pressure. FIFTH, the chief accounting officer sold 30,000 shares at $53.29 on 2026-07-24, reducing her holding by 74%."

Growth Quality6/10High

"Rated 6 — a real and recent margin recovery on top of three years of revenue decline. The revenue record is poor: $14,853M (FY2022), $14,657M (FY2023), $14,540M (FY2024), $14,092M (FY2025) — down in each of the last three years. The 2026 inflection is genuine and filing-verified. From the June-quarter 10-Q: second-quarter volume of 1,676 thousand units against 1,580, UP 6%; revenue $3,935M against $3,574M, UP 10%; expense up 6%; operating income $1,506M against $1,283M, UP 17%; operating margin 38.3% against 35.9%, UP 240 basis points; and diluted EPS $0.54 against $0.44, UP 23%. For the six months the picture is stronger still: revenue $7,417M (+6%), expense DOWN $15M, operating income $2,759M (+19%), operating margin 37.2% against 33.2% — UP 400 basis points — and EPS $0.97 against $0.78, up 24%. Consensus wants $2.002 in FY2026 (16 analysts), $2.271 in FY2027 (16) and $2.512 in FY2028 (7) — 22.2%, 13.4%, 10.6% — on revenue rising from $14,092M actual to $15,180M and $15,952M. What holds this at 6: two-thirds of the improvement is cost and margin rather than volume, and margin recovery from a depressed base is finite. A railroad growing revenue 5% and earnings 13% is a good business, not a growth business."

Exponential Potential2/10Low

"Rated 2 — the lowest exponential score in this batch alongside Duke Energy's 3, and honestly so. CSX operates approximately 20,000 route miles of track across 26 states east of the Mississippi. Its two reportable segments are Rail ($3,709M of June-quarter revenue) and Trucking ($229M), and its business is moving chemicals, agricultural products, automobiles, coal, minerals, forest products, metals and intermodal containers from one place to another. The economics improve through operating discipline — the June quarter's 240-basis-point margin gain came from expenses growing 6% against revenue growing 10% — and through pricing, and they are bounded by industrial production and by competition from trucking. There is one genuinely non-linear possibility and it is corporate rather than operational: the 10-Q's disclosure of 'advisory expenses related to potential industry consolidation' points to a transaction that would, if it happened, change what this asset is worth in a single step. That is optionality, not compounding, and it is not the company's to control — any Class I combination requires Surface Transportation Board approval, which is a multi-year public process with a demonstrated history of conditions and rejections. A 2: a well-run, irreplaceable, entirely linear network with one binary corporate outcome attached."

Fair value$52 $37–$67

What does “fair value” mean?

Fair value is Synthos’s estimate of what one share is worth today, based on our model of the company’s future cash generation and the risks to it. It is not a price target or a prediction of where the stock trades next quarter — it’s the price where we think risk and reward are balanced. Above it, you’re paying for outcomes better than our base case; below it, the market is offering a margin of safety.

The model’s inputs and weightings are proprietary — they’re the product. What we publish is the output, plus an independent, public-math cross-check further down the page (“Check our number”) so you can judge it for yourself.

What do the 5 tiers mean? (Core · Tactical · Watch · Hold · Avoid)

Buy — CoreOwn it as a foundation — start or add now, size it for years, let dips be gifts.

Buy — TacticalGood price + confirmed trend + a defined exit — buy the setup, not a marriage.

WatchWe want the business, just not at this price/setup — act only when the listed trigger hits.

HoldFine to keep if you own it — no reason to buy more; new money does better elsewhere.

AvoidDon't own it — the problem is the business or the expectations, so a cheaper price won't fix it.

The Road Ahead

What we expect to matter in each window, and the evidence that would prove us wrong.

Short term 0-6 months

Neutral

Driver: "The strongest chart in this batch and the smallest expected return in it. CSX closed 2026-08-04 at $51.03, up 2.41%, which is 4.1% below its 52-week high of $53.23 and 59.2% ABOVE its low of $32.05, 5.6% above a 50-day average of $48.31 and 23.1% above a 200-day average of $41.45. RSI is 57.4 and MACD is +0.73. The maximum drawdown from peak over the trailing year is 4.1% — the shallowest in this batch after Duke Energy's 6.9% — so a 46% advance has been given back essentially nothing. Relative performance is the second-best in the batch: 3-month +14.1% against SPY's +7.6%, 6-month +32.9% against +11.1%, 12-month +45.8% against +24.3%. And the fundamental support is real: the June quarter, reported thirteen days ago, showed volume up 6%, revenue up 10%, operating income up 17% and the operating margin up 240 basis points to 38.3%. What makes the stance neutral rather than a tailwind is that the vendor's consensus price target is $51.06 against a $51.03 close — the street's average sees essentially zero upside — and the same distribution contains a $32 low, 37.3% below spot. Buying a railroad 4% below its high after a 46% year, when the average analyst target equals the price, is not a setup with a margin of safety."
What we’re watching: "The 2026-10-15 print against consensus diluted EPS of $0.54 and revenue of $3,929M — a forecast of NO sequential growth on either line, which is itself informative about how much of the recovery the street thinks is already banked. Within it: the operating margin, 38.3% in the June quarter against 35.9%, and whether the 400-basis-point six-month improvement is sustained or gives back; volume, up 6% in the quarter and 4% for the half, and whether the intermodal strength continues; and locomotive fuel expense, which rose from $226M to $379M year on year in the June quarter — a 68% increase that partly offsets the labour and purchased-services discipline. Separately and more importantly: any development on the 'advisory expenses related to potential industry consolidation' disclosed in the same filing. That single line is the only evidence in this archive that a corporate transaction is under consideration, and its resolution in either direction is the dominant short-term variable."
Confidence: Low

Medium term 6-24 months

Neutral

Driver: "The medium term is a good railroad compounding earnings at low-double digits, priced as though something more is coming. Consensus has diluted EPS at $2.002 in FY2026, $2.271 in FY2027 and $2.512 in FY2028 — 22.2%, 13.4% and 10.6% — on revenue rising from $14,092M actual to $15,180M, $15,952M and $16,499M, growth of 7.7%, 5.1% and 3.4%. The FY2026 step of 22.2% is largely the margin recovery already visible in the first half; from there the growth rate settles into the low double digits, which is what a well-run Class I earns through pricing above cost inflation plus modest volume. Capital return is a smaller contributor than it was: the dividend of $0.54 is a 1.06% yield on a 31.2% payout, and share repurchases have fallen from $4,731M in fiscal 2022 to $3,482M, $2,237M and $1,396M — the fiscal-2025 figure is 1.5% of market capitalisation against 5.0% three years earlier. Total shareholder yield is therefore roughly 2.5%, not the 6% this company used to return. Whether that reduction reflects free cash flow constrained by a rising capital programme ($2,133M in fiscal 2022 rising to $2,902M in fiscal 2025) or capital deliberately conserved for a transaction is the question the filings do not answer."
What we’re watching: "Whether a Class I combination is announced, and on what terms — the 10-Q's 'advisory expenses related to potential industry consolidation' is the only signal in this archive and it is a cost line, not a disclosure. If a transaction comes, the Surface Transportation Board approval process is multi-year, public and historically hostile to conditions-free approvals. Whether the operating margin holds above 37%, which is what the FY2027 consensus implicitly requires. Whether volume growth persists — 6% in the June quarter and 4% for the half is well above the industrial-production trend and may reflect share recapture from trucking rather than a durable demand shift. Whether the buyback recovers from $1,396M or falls further; on 1,852 million shares, the difference between a $1.4 billion and a $3.5 billion annual repurchase is roughly two points of annual EPS growth. Whether the capital programme, now $2,902M against $1,680M of depreciation, moderates. And whether locomotive fuel cost, up 68% year on year in the quarter, is a pass-through or a margin headwind."
Confidence: Low

Long term 2+ years

Neutral

Driver: "Long-run, CSX owns an asset that cannot be rebuilt: roughly 20,000 route miles of rail across the eastern United States, assembled over a century and a half, in a country that will not grant the rights-of-way again. That is the entire long-term case and it is a good one — railroads are the lowest-cost mover of heavy freight over land, they are structurally advantaged on fuel efficiency, and their pricing power derives from the physical impossibility of replication. The long-run return on such an asset is roughly the rate at which it can raise price above cost inflation, plus whatever volume the industrial economy provides, plus the share count reduction. Historically that has been a high-single to low-double-digit compound and it is what the FY2027 and FY2028 consensus implies. The structural question overlaying it is consolidation. North American Class I railroading has been consolidating for forty years and there are now very few independent networks left; a transcontinental combination changes the competitive map for everyone remaining. CSX's own filings list 'consolidation or financial distress within the transportation industry' as a risk factor and separately disclose that it is paying advisers about it. Stephen F. Angel — an outsider, previously chief executive of an industrial gases company — became President and Chief Executive Officer, which is the kind of appointment a board makes when it wants either a cost transformation or a transaction, and possibly both."
What we’re watching: "Whether the eastern network remains independent, is acquired, or acquires — and if any of those, what the Surface Transportation Board permits. Whether precision-scheduled-railroading margins, now at 38.3%, prove to be a plateau or a ceiling; the industry has repeatedly discovered that service quality degrades when operating ratios are pushed too hard, and CSX's own operating statistics on train velocity, dwell and trip-plan performance are the leading indicators. Whether truck competition intensifies as autonomous long-haul technology matures — the single most plausible long-term threat to rail's cost advantage, and one this file does not address. Whether coal, at approximately $1,900 million of fiscal-2025 revenue and declining, stabilises or continues to erode. Whether intermodal, the growth line, is a share gain from trucking or merely cyclical. And whether the capital intensity of maintaining a 20,000-mile network — $2,902 million against $1,680 million of depreciation in fiscal 2025 — is a temporary catch-up or the new normal, because the difference is the whole free-cash-flow yield."
Confidence: Low

Exponential Potential

Exponential Potential2/10Low

What could take this further than the base case — and where the market may be underpricing it. Full forward-growth and acceleration math in Deeper analysis, §4 below.

Deeper analysis

Technicals, fundamentals, valuation, and the full expert-claim evidence panel — the detail behind the numbers above.

Check our number Market-implied growth ≈ 15%/yr This isn’t how we calculate fair value — it’s public math you can verify yourself. To justify today’s $51, earnings would have to compound roughly 15% a year for 10 years (9% discount rate). Analysts forecast ~7%/yr, so the market is pricing in MORE than what the Street expects.

Reference table

Street consensus	$51.06 (+0.06%) — the consensus target equals the price to within three cents · median $53 (+3.9%) · high $60 (+17.6%) · low $32 — 37.3% BELOW spot · 0 strong buy / 27 buy / 17 hold / 2 sell across 46 analysts
Valuation	28.4x trailing adjusted EPS ($1.80) · 25.5x FY2026E $2.002 · 22.5x FY2027E $2.271 · 20.3x FY2028E $2.512 · 16.8x EV/EBITDA · 6.72x book
The June quarter — filing-verified	Volume 1,676k units, +6% · revenue $3,935M, +10% · expense +6% · operating income $1,506M, +17% · operating margin 38.3%, +240bp · diluted EPS $0.54, +23%
The six months — stronger	Revenue $7,417M, +6% · expense DOWN $15M · operating income $2,759M, +19% · operating margin 37.2% vs 33.2%, +400bp · EPS $0.97 vs $0.78, +24%
The sentence that explains the price	From the same 10-Q: "All other net costs increased $11 million, primarily due to intermodal volume increases and advisory expenses related to potential industry consolidation." No transaction is announced anywhere in this archive
The revenue record before 2026	$14,853M (FY2022) → $14,657M (FY2023) → $14,540M (FY2024) → $14,092M (FY2025) — down three years running
Capital return, and it has been cut	Dividend $0.54, 1.06% yield, 31.2% payout · buyback $4,731M → $3,482M → $2,237M → $1,396M across FY2022-FY2025, a 70% reduction. Total shareholder yield ≈ 2.5%, against roughly 6% three years ago
Conviction	NONE. 0 raw KB hits — zero entity matches AND zero free-text hits. The only completely null sweep in this batch alongside JCI. No expert in the knowledge base has expressed any recorded view on this company
Technicals	−4.1% from the 52-week high of $53.23; +59.2% above the low of $32.05; +5.6% above the 50-DMA ($48.31) and +23.1% above the 200-DMA ($41.45); RSI 57.4; MACD +0.73; max drawdown −4.1%; 12-month +45.8% vs SPY +24.3%

What the experts actually said

No independent expert claims in the Synthos knowledge base yet for CSX — this dive is fundamentals- and technicals-driven, not panel-driven.

Price & moving averages 12 months · 50 & 200-day averages · 52-week range

Solid line = price · dashed line = 50-day average · dotted line = 200-day average · the two thin horizontal lines mark the 52-week high and low. Price above both averages is an uptrend.

Bollinger Bands 20-day average ± 2 standard deviations

The shaded band widens when the stock gets more volatile. Riding the upper edge = strong momentum (sometimes stretched); the lower edge = weak / potentially oversold.

RSI (14) momentum gauge · 0–100

Above 70 (overbought zone, shaded) = overbought, below 30 (oversold zone, shaded) = oversold. Currently 57.

MACD 12 / 26 / 9 · trend & momentum

The MACD line crossing above the signal line (bars flip to the up color) = momentum turning up; crossing below (bars flip to the down color) = turning down. Bar height = the size of that gap.

Relative performance vs S&P 500 & its sector (XLI (sector)), set to 100 a year ago

Solid = CSX · dashed = S&P 500 · dotted = XLI (sector). A rising line means it is beating that benchmark — the sector line shows whether it is a leader or laggard within its own group.

Forward revenue & earnings actual → estimate · "FY" = fiscal year, "E" = estimate

Darker bars = actual results, brighter = analyst estimates. Taller bars to the right = expected growth.

Key stats an RIA wants

Price$51.28

Market cap$95B

P/E trailing30×

P/E FY26E / FY27E26× / 23×

EV / Sales7.8×

EV / EBITDA17.0×

Gross margin54.9%

Net margin22.2%

Dividend yield1.05%

Beta1.211

52-wk range$32 – $53

RSI(14)67

50 / 200-DMA$50 / $43

12-mo return+58% (SPY +19%)

Street target$51 ($32–$60)

Analyst grades27 Buy · 17 Hold · 2 Sell

FMP ratingB

Next earnings2026-10-15 (Q3 2026 earnings, 72 days away; vendor consensus diluted EPS $0.54 and revenue $3,929M, implying flat EPS and a 0.2% revenue decline against the June quarter's $0.54 and $3,935M). Second-quarter results were released 2026-07-22, thirteen days before this dive, and beat consensus EPS by 4.2% ($0.54 against $0.518) on revenue of $3,935M against $3,894M estimated.

1. The business and the June quarter

CSX Corporation is a Virginia-incorporated railroad holding company, Nasdaq-listed, with approximately 23,000 employees and Stephen F. Angel as President and Chief Executive Officer. Two reportable segments, Rail and Trucking, and the June-quarter segment table reconciles exactly:

($M, Q2 2026)	Q2 2026	Q2 2025
Rail revenue	$3,709	$3,363
Trucking revenue	$229	$219
Elimination of intersegment revenues	(3)	(8)
Total consolidated revenue	$3,935	$3,574
Labour and fringe	782	736
Purchased services and other	535	605
Depreciation and amortisation	396	410
Locomotive fuel	379	226
Non-locomotive fuel	34	23
Equipment and other rents	92	90
Gain on property disposition	(2)	(4)
Rail segment operating income	$1,493	$1,277
Trucking expenses	216	213
Total consolidated operating income	$1,506	$1,283
Interest expense	(211)	(212)
Other income, net	25	22
Earnings before income taxes	$1,320	$1,093

The headline summary, quoted directly from the 10-Q filed 2026-07-22:

	Q2 2026	Q2 2025	change	H1 2026	H1 2025	change
Volume (thousands)	1,676	1,580	+6%	3,235	3,098	+4%
Revenue	$3,935M	$3,574M	+10%	$7,417M	$6,997M	+6%
Expense	$2,429M	$2,291M	+6%	$4,658M	$4,673M	−$15M
Operating income	$1,506M	$1,283M	+17%	$2,759M	$2,324M	+19%
Operating margin	38.3%	35.9%	+240 bp	37.2%	33.2%	+400 bp
Diluted EPS	$0.54	$0.44	+23%	$0.97	$0.78	+24%

Read the six-month expense line, because it is the whole recovery. Revenue rose 6% and expenses FELL $15 million. That is a 400-basis-point margin improvement generated by holding cost flat against growing volume — the classic railroad operating-leverage result, and it is what a new chief executive is hired to produce.

Two components deserve separate attention. Purchased services and other fell from $605M to $535M — down 11.6% — which is where the cost discipline is concentrated. And locomotive fuel rose from $226M to $379M, up 68%, which is a large adverse move that the rest of the cost base absorbed. A 68% fuel increase inside a 240-basis-point margin expansion means the underlying operating improvement is larger than the headline suggests, and equally that a reversal in fuel prices would flatter future comparisons.

The revenue history before this recovery is not good and must be stated:

Fiscal year	Revenue	change	Operating income	Diluted EPS	Diluted shares
FY2021	$12,522M	—	$5,159M	$1.68	2,255M
FY2022	$14,853M	+18.6%	$5,954M	$1.95	2,141M
FY2023	$14,657M	−1.3%	$5,467M	$1.83	2,013M
FY2024	$14,540M	−0.8%	$5,366M	$1.79	1,943M
FY2025	$14,092M	−3.1%	$4,521M	$1.54	1,864M

Revenue declined in each of the last three fiscal years and operating income fell 24% from its FY2022 peak. Earnings per share held up better than operating income only because the share count fell 13% over the same period. That is the base from which the 2026 recovery is being measured, and it is why a 22.2% consensus EPS increase for FY2026 is less impressive than it sounds — it takes EPS to $2.00 against $1.95 four years earlier.

The sentence that explains the share price

From the same 10-Q, in the discussion of second-quarter expenses:

> "All other net costs increased $11 million, primarily due to intermodal volume increases and advisory expenses related to potential industry consolidation. These increases were partially offset by an insurance recovery related to 2024 property damage on the Blue Ridge subdivision."

That is the only reference to a corporate transaction anywhere in this archive, and it is a cost line rather than a disclosure. The 10-K's risk factors mention "the impact of industry competition, conditions, performance and consolidation" and "competition and consolidation or financial distress within the transportation industry generally" in the standard forward-looking-statements language, which every railroad carries.

What we can say with confidence: CSX is incurring advisory fees in connection with potential industry consolidation, and it has disclosed nothing further. What we cannot say: whether CSX is a buyer, a seller, a defender, or merely prepared. What the market appears to be saying: the stock is up 45.8% in twelve months and 32.9% in six against a business whose revenue fell 3.1% in the most recent full year, and the street's price-target distribution runs from $32 to $60.

Two structural facts frame the possibility without settling it. Any combination of Class I railroads requires Surface Transportation Board approval — a multi-year, public, contested process. And Stephen F. Angel, an outsider from outside the rail industry, is President and Chief Executive Officer — the kind of appointment a board makes when it wants a cost transformation, a transaction, or both.

2. Valuation — priced in or room?

At $51.03 (market capitalisation $94.53B, implied 1,852.3M shares):

	Trailing	FY2026E	FY2027E	FY2028E
Diluted EPS	$1.80	$2.002 (16)	$2.271 (16)	$2.512 (7)
Estimate range	—	$1.921–$2.073 (7.9%)	$2.148–$2.404 (11.9%)	$2.363–$2.664 (12.7%)
Growth	—	+22.2%	+13.4%	+10.6%
P/E	28.4x	25.5x	22.5x	20.3x
Revenue	$14,092M (FY2025)	$15,180M (15)	$15,952M (16)	$16,499M (12)
Revenue growth	−3.1%	+7.7%	+5.1%	+3.4%
EV / EBITDA	16.8x	—	—	—
Net debt / EBITDA	2.74x	—	—	—
Price / book	6.72x	—	—	—
Return on equity	24.1%	—	—	—
Dividend yield	1.06%	—	—	—

Estimate coverage is excellent on the near years and thin beyond. Sixteen analysts on both FY2026 and FY2027 EPS — the most consistent coverage in this batch — falling to 7 on FY2028 and 2 on FY2029. The FY2029 row is excluded from every conclusion. The FY2027 range of $2.148 to $2.404 is 11.9% wide.

est.ebitdaAvg and est.ebitAvg carry the fixed-ratio fabrication signature — REJECTED. ebitdaAvg is exactly 49.528% of revenueAvg and ebitAvg exactly 38.874% in FY2026, FY2027, FY2028, FY2029 and FY2030 alike. Five consecutive years to three decimal places. Note the implied 38.9% EBIT margin is close to the 38.3% the company actually achieved in the June quarter, which makes the fabrication harder to spot and no less fabricated. All forward valuation uses epsAvg, validated against the four earn_cal actuals summing to $1.80 and the 10-Q's reported $0.54 and $0.97.

Enterprise value is clean. $112,882M = market capitalisation plus net debt of approximately $18.4 billion. Noncontrolling interests are $5 million — immaterial, and the smallest NCI in this batch by three orders of magnitude. Finance lease obligations of $479M ARE included in total debt of $19,352M — no lease double-count of the T or TJX class.

Peer context is mixed but contains the two comparables that matter. The vendor set is Cummins, Canadian National, Canadian Pacific Kansas City, FedEx, Illinois Tool Works, Norfolk Southern, Quanta Services, Republic Services, Thomson Reuters and United Rentals. Norfolk Southern ($76.7B), Canadian National ($77.0B) and Canadian Pacific Kansas City ($79.7B) are the three genuine Class I railroad comparables, and Norfolk Southern is the closest — the other eastern network, and the obvious counterparty in any eastern consolidation. Illinois Tool Works, Quanta, Republic Services, Thomson Reuters and United Rentals are unrelated. No cross-multiple is drawn, because this file contains no forward-EPS data for the peers.

2a. What today's price assumes (the inversion)

At $51.03 — 25.5x FY2026 consensus, 22.5x FY2027, 16.8x EV/EBITDA — the price embeds:

Diluted EPS reaches $2.002 in 2026 and $2.271 in 2027. (Consensus; 16 analysts each.) The first half delivered $0.97, so FY2026 needs $1.03 in the second half — and the September guide is $0.54, implying $0.49 in the December quarter. That is a modest ask and the first-half performance supports it.
The operating margin holds above 37%. (Filing-verified 38.3% in Q2 and 37.2% for the half.) The FY2027 consensus of 13.4% growth on 5.1% revenue growth requires continued margin expansion beyond the 400 basis points already captured. This is the most fragile operating assumption in the price — margin recovery from a depressed base is finite, and the June quarter absorbed a 68% locomotive fuel increase that will not repeat as a tailwind.
Something happens on industry consolidation. (Filing-disclosed advisory expense; no transaction announced.) This is the assumption that actually explains the multiple. A railroad growing earnings 13% a year does not obviously deserve 22.5x its second forward year when its revenue declined in each of the three prior years. The street's $32 low target — 37.3% below spot — is a direct measurement of what the shares are worth without it.
The buyback does not fall further. (Our number, from cf_a.) Repurchases have gone $4,731M → $3,482M → $2,237M → $1,396M. At the current pace the buyback contributes roughly 1.5 percentage points to per-share growth against 5.0 points three years ago. A further cut removes a meaningful slice of the consensus EPS path.
The market keeps paying 20-25x forward earnings. (Our number.) At 18x FY2027E the stock is $41; at 26x it is $59. CSX has traded across most of that band in the last five years, and the current 22.5x is toward the upper end for a business whose revenue has been flat to down.

2b. The return bridge (why the multiple moves)

Expected return over the next twelve months decomposes as: diluted EPS growth (+13.4%, from FY2026E $2.002 to FY2027E $2.271) + multiple COMPRESSION (25.5x on the forward year moving to roughly 22.9x on the then-forward year, −10%) + shareholder yield (+2.5%: a 1.06% dividend and roughly 1.5% of buyback) ≈ +2% to +5%.

A caution the addendum requires: the rolldown from 25.5x FY2026 to 22.5x FY2027 at a constant price is mechanical and is not counted as compression. What our base assumes is that the market pays roughly 23x the FY2027 number a year from now, against 25.5x on FY2026 today — modest genuine de-rating, on the ground that a 13% earnings grower with declining revenue history does not sustain a mid-twenties multiple absent a corporate event.

Note what this makes the return: earnings growth almost entirely offset by de-rating, leaving a 2.5% yield. That is the arithmetic of a good business bought after a 46% move, and it is why this is a Hold.

If the multiple fell to 18x FY2027E the price is $41 (−19.7%). If it held at 25.5x, $58 (+13.5%).

2c. Variant perception (where we differ, what would surprise)

We are 1.8% above consensus and that is not the disagreement. Our $52 against a mean of $51.06 — which is the closing price to within three cents. The disagreement is not about the number; it is that a consensus target equal to spot, sitting inside a $32-to-$60 distribution, is not a forecast at all. It is a probability-weighted average of two outcomes, and publishing its midpoint as a fair value conceals that. We publish $52 because the format requires a point estimate and we flag that the honest answer here is bimodal.
We think the operating recovery is being conflated with the consolidation speculation, and only one of them is verifiable. The margin improvement is filing-verified and large: 240 basis points in the quarter, 400 in the half, on expenses that fell in absolute terms. The consolidation is a line item in an expense discussion. Watchable event: any 8-K under Item 1.01 or 8.01, or any Surface Transportation Board filing. Until then the two should be valued separately and only one can be.
We think the buyback reduction is under-examined. Repurchases fell from $4,731 million in fiscal 2022 to $1,396 million in fiscal 2025 — a 70% cut — while capital expenditure rose from $2,133 million to $2,902 million. Watchable number: the fiscal-2026 repurchase run-rate. Capital being conserved is consistent with a transaction; capital being consumed by the network is consistent with a lower structural free-cash-flow yield. The filings do not distinguish them and the difference is worth several points of annual return.
We differ from the vendor data on the cash-flow position and the difference is large. The payload reports a free-cash-flow yield of 5.31% and price-to-free-cash-flow of 18.83x, on a trailing capital expenditure figure of roughly $300 million — against $2,902 million in the company's own fiscal-2025 cash-flow statement, a 90% understatement. The corrected free-cash-flow yield is nearer 2.5%. Section 3 sets out the arithmetic.
Positive surprise that would force a re-rate: an announced combination on terms favourable to CSX shareholders; a September quarter with the operating margin above 39%; volume growth sustained above 5%; or a restored buyback above $2.5 billion annualised.
Negative surprise that would break the thesis: confirmation that no transaction is in prospect, which on the street's own low target is worth 37% of the share price; an operating margin retreating below 36%; volume growth stalling as the truck market loosens; a further buyback cut; or a Surface Transportation Board decision — in any Class I proceeding — that makes eastern consolidation materially harder.

Synthos fair values

All three anchors are multiples of the FY2027 consensus diluted EPS distribution (mean $2.271, low $2.148, high $2.404, 16 analysts).

Bear ~$37 — 17.2x the FY2027 consensus LOW of $2.148, cross-checked at 18.5x FY2026E. 15.4% above the 52-week low of $32.05, and 15.6% above the street's own low target of $32. The scenario: no transaction materialises, the margin recovery plateaus, revenue growth reverts toward the flat-to-down trend of 2023-2025, the buyback stays cut, and a Class I railroad is valued as one. −27.5%.
Base ~$52 — 22.9x the FY2027 consensus MEAN of $2.271, cross-checked at 26.0x FY2026E and 20.7x FY2028E, and approximately 6.85x book. Sensitivity, stated openly: 22.5x — the current multiple held flat — gives $51.10, which is spot; 25x gives $57. The scenario: the operating recovery holds, consensus delivers, the buyback stays near $1.4 billion, the consolidation question stays open, and the multiple gives back a little. +1.9%.
Bull ~$67 — 27.9x the FY2027 consensus HIGH of $2.404, cross-checked at 26.7x FY2028E. 25.9% above the 52-week high of $53.23 and 11.7% above the street's own high target of $60. The scenario: a combination is announced at a control premium, or the margin recovery extends well beyond 39% while volumes keep growing. +31.3%.

Base is 1.9% above spot; asymmetry roughly 1.14:1 (27.5% down, 31.3% up), plus a shareholder yield near 2.5%. A base case inside 2% of the price, a payoff ratio of roughly one-to-one, and a consensus target three cents above spot is as complete a definition of fairly valued as this programme has produced. This is a Hold.

3. Data integrity — what we rejected and why

Seven findings. One would materially mislead a screen, one segment block is catastrophically double-counted in its history, and one block is empty.

1. Trailing capital expenditure is roughly 90% understated, so the free-cash-flow yield is roughly double the truth — REJECTED. The arithmetic is visible in three fields:

Field	Value	Implied TTM total
`operatingCashFlowPerShareTTM`	$2.8690	~$5,321M
`capexPerShareTTM`	$0.16226	~$301M
`freeCashFlowPerShareTTM`	$2.7067	~$5,021M

The vendor's own cf_a reports fiscal-2025 capital expenditure of $2,902 million, and $2,529M, $2,257M and $2,133M in the three prior years. capexToDepreciationTTM of 0.182 is the visible tell: a railroad does not spend 18% of its depreciation on capital, and the 10-Q reports $396 million of depreciation in the June quarter alone (roughly $1.58 billion annualised). Correcting it: trailing operating cash flow of approximately $5,321 million less roughly $2,900 million of capital expenditure gives free cash flow near $2,420 million and a free-cash-flow yield of approximately 2.56%, not the reported 5.31%. priceToFreeCashFlowRatioTTM of 18.83x should be nearer 39x; evToFreeCashFlowTTM of 22.48x nearer 47x; capexToOperatingCashFlowTTM of 0.0566 should be approximately 0.55; capitalExpenditureCoverageRatioTTM of 17.68 and dividendPaidAndCapexCoverageRatioTTM of 4.08 are both void. This is the same defect family as FCX (capex set to zero) and DUK (a sign error on an 84% understatement) elsewhere in this batch. No cash-flow-based measure from this file is used anywhere in this dive.

2. seg_prod DOUBLE-COUNTS to 159.7% of revenue in fiscal 2023 and earlier — REJECTED for those years. The FY2023 entry contains both a "Total Merchandise" line of $8,700M and the seven individual merchandise components (Agricultural and Food Products $1,657M, Automotive $1,219M, Chemicals $2,599M, Fertilizers $516M, Forest Products $1,012M, Metals and Equipment $917M, Minerals $733M — summing to $8,653M). Summing every line in the block gives $23,413M against reported revenue of $14,657M — 159.7%. The same structure appears in FY2022, FY2021, FY2020, FY2019 and FY2018. The FY2025 and FY2024 entries drop the components and are usable: FY2025 sums to $13,616M against $14,092M (96.6%, the residual being the Other Services line the block omits) and FY2024 to $13,944M against $14,540M (95.9%). A reader summing this block across years would conclude CSX's revenue grew from $23.4 billion to $13.6 billion. Only the FY2025 and FY2024 entries are used, and only with the 3-4% shortfall stated.

3. seg_geo is EMPTY — zero entries, for any year. There is no geographic revenue disclosure for CSX in the payload at all. We report it as empty rather than substituting anything. The company is a domestic US railroad operating east of the Mississippi, so the analytical cost is low, and we say that rather than treating the gap as material.

4. earn_cal contains one row whose revenueActual is wrong by 24.3%, manufacturing a fake 24.9% revenue beat. The 2025-10-16 row reports revenueActual of $4,460,000,000 against revenueEstimated of $3,572,206,117 — an apparent +24.9% beat. The inc_q figure for the same quarter is $3,587,000,000, and the four fiscal-2025 quarterly revenues sum to $14,092M, exactly matching inc_a. So inc_q is right and this one earn_cal row is wrong by $873 million. The other five rows agree with inc_q to the dollar. This is a milder relative of the SPGI mixed-basis defect — one corrupted row rather than a systematic basis change — and it is recorded because a screen reading revenue surprises would register a 25% beat that did not occur. The EPS column is unaffected and is used.

5. inc_q.grossProfit is variously equal to revenue or equal to operating income — the field is broken. The June 2026 quarter reports grossProfit of $3,935M, identical to revenue — a 100% gross margin. The June 2025, March 2025, December 2025, September 2024 and June 2024 quarters report grossProfit identical to operatingIncome. The March 2026 and December 2025 quarters report a third, different value. Three inconsistent definitions across nine quarters. ratios_ttm.grossProfitMarginTTM of 54.9% inherits it. The field is not used; all margin analysis in this dive uses the 10-Q's operating margin of 38.3% and 37.2%.

6. est.ebitdaAvg and est.ebitAvg carry the fixed-ratio fabrication signature — REJECTED. ebitdaAvg is exactly 49.528% of revenueAvg and ebitAvg exactly 38.874% in FY2026, FY2027, FY2028, FY2029 and FY2030 alike. The implied 38.9% EBIT margin sits within a percentage point of the 38.3% actually achieved in the June quarter, which makes this instance unusually plausible-looking and no less fabricated. All forward valuation uses epsAvg.

7. inc_a FY2025 depreciation of $1,276M contradicts cf_a FY2025 depreciation of $1,680M. The two blocks disagree by 24%. The 10-Q reports $396 million of rail-segment depreciation in the June quarter — roughly $1.58 billion annualised — so the cash-flow figure is the credible one. inc_a FY2025 ebitda of $5,866M is built on the smaller number and is not used; EBITDA in this dive is derived from inc_q (four quarters summing to approximately $6.70 billion).

Also checked and recorded. quote.yearHigh/yearLow ($53.60 / $31.80) against tech.hi52/lo52 ($53.23 / $32.05) — a 0.7% and 0.8% discrepancy; tech is used throughout. inc_q passed the MO/APP two-basis test — all nine quarters on one revenue basis, the four fiscal-2025 quarters summing to $14,092M exactly matching inc_a, and the June-quarter revenue of $3,935M and operating income of $1,506M matching the 10-Q to the dollar. The share count is consistent: market capitalisation ÷ price gives 1,852.3M against revenuePerShareTTM implying approximately 1,855M, a 0.15% difference. effectiveTaxRateTTM of 23.3% is consistent with incomeTaxExpense of $880M on incomeBeforeTax of $3,769M (23.3%) — an exact match. dividendYieldTTM of 1.058% is consistent with dividendPerShareTTM of $0.54 at $51.03. Noncontrolling interests are $5 million, so the enterprise value omission that materially distorts KKR (35.6%) and FCX (11.5%) elsewhere in this batch is immaterial here.

returnOnEquityTTM — source stated. Present in km_ttm at 24.06% and absent from ratios_ttm. Used as reported.

Vendor composite rating — noted, low weight. B− / 3 overall, with 5 out of 5 on return on equity, 4 on return on assets, and 1 out of 5 on both debt-to-equity and price-to-book, and 2 out of 5 on discounted cash flow. The discounted-cash-flow sub-score rests on the free-cash-flow figure destroyed in finding 1. The debt-to-equity score of 1 penalises a capital-intensive network business for the leverage its asset base supports.

Non-equity tripwire — checked and passed. CSX is common stock, Nasdaq-listed. Price of $51.03 is not par-like; beta is 1.211; the dividend is variable ($0.54, 1.06%); volume was 13.78M shares (~$703M of turnover); the 52-week band of $32.05 to $53.23 is a 66% range. This is common equity.

4. Knowledge base — a completely empty lane

Raw hits: 0. Entity matches: 0. Free-text hits: 0. Used: 0. Discarded: 0.

The sweep ran the case-sensitive entity tokens CSX and CSX Corporation, plus a case-sensitive word-bounded free-text pattern on CSX, across all 51,928 distilled claims. It returned nothing at all.

This is the only completely null sweep in this batch alongside JCI, and it differs from the USB result in an informative way. USB returned three hits and all three were the Universal Serial Bus — a homograph problem. CSX returned zero hits of any kind, which means the null is a genuine measurement rather than a filtering artefact: "CSX" is a distinctive three-letter string with low collision potential, and there is simply nothing there.

What that means, stated plainly. No expert in the Synthos knowledge base has expressed a recorded view on CSX Corporation, on the eastern US rail franchise, on the margin recovery documented in Section 1, or on the North American railroad consolidation wave that CSX's own filings say it is paying advisers to consider. Given that consolidation is the single largest swing factor in this security's value, the absence of any commentary on it is a real limitation of the qualitative input available here, and it is stated rather than papered over.

We do not pad the lane. Adjacent macro claims about freight volumes, tariffs, industrial production or supply chains exist in the store and none of them names this company, so none is admitted. Breadth 0, claim count 0, net conviction none. This dive is built entirely on the filings and the vendor payload.

5. Technicals

Price $51.03. −4.1% from the 52-week high of $53.23; +59.2% above the 52-week low of $32.05. Position within the annual range: 90th percentile.
Above both moving averages by wide margins: +5.6% above a 50-day average of $48.31; +23.1% above a 200-day average of $41.45.
RSI 57.4 — unstretched for a stock at the 90th percentile of its range. MACD +0.73.
Maximum drawdown from peak over the trailing year: −4.1% — the second-shallowest in this batch after Duke Energy's 6.9%. A 59% advance off the low with a maximum 4% give-back is an exceptionally orderly uptrend and a statement about momentum, not about value.
Relative performance, the second-strongest in this batch after Freeport: 3-month +14.1% against SPY +7.6% and QQQ +7.7%; 6-month +32.9% against SPY +11.1%; 12-month +45.8% against SPY +24.3% and QQQ +30.8% — a 21-point outperformance against the S&P.
Sentiment, and the shape is the story: 0 strong buy, 27 buy, 17 hold, 2 sell across 46 analysts; consensus target $51.06 — three cents above the closing price; median $53 (+3.9%); high $60 (+17.6%); low $32 — 37.3% BELOW spot. A $32-to-$60 range on a Class I railroad is an 88% spread, and it is the second-widest in this batch after KKR's. That is not disagreement about a growth rate; it is two scenarios, and the consensus mean sitting exactly at the market price is the arithmetic consequence of averaging them.

Today's move and what it does to the entry

CSX closed 2026-08-04 at $51.03, up 2.41% or $1.20 from $49.83. It opened at $49.73, traded $49.42 to $51.08, and closed within 0.1% of the day high on 13.78 million shares. No company-specific filing is dated 2026-08-04; the last events were the 2026-07-22 results and the 2026-07-27 insider filings.

The honest read: there is no entry here, only a decision about whether to keep holding. The stock closed at the day high, 4% below its 52-week high, 23% above its 200-day average, having risen 59% off its low with a maximum drawdown of 4%. A Hold verdict on a stock in this configuration is not a bearish call; it is a refusal to pay a full price for a good business plus an unannounced transaction, when the same analyst community's most pessimistic member thinks the shares are worth 37% less.

6. Insiders — one large outright sale at the high

Date	Person	Role	Type	Shares	Price	Held after
2026-06-03	Kevin S. Boone	EVP & Chief Financial Officer	M-Exempt (option exercise)	123,253	$23.48	345,330
2026-06-03	Kevin S. Boone	EVP & Chief Financial Officer	M-Exempt (option exercise)	13,455	$22.70	222,077
2026-06-15	J. Steven Whisler	Director	A-Award	791	$47.39	97,326
2026-07-01	Stephen F. Angel	Director, President & CEO	A-Award (phantom stock)	187	$0	2,021
2026-07-24	Angela C. Williams	VP & Chief Accounting Officer	S-Sale	30,000	$53.29	10,437
2026-07-24	Michael S. Burns	SVP, Chief Legal Officer & Corporate Secretary	M-Exempt (option exercise)	13,000	$26.50	72,643
2026-07-24	Michael S. Burns	SVP, Chief Legal Officer & Corporate Secretary	S-Sale	13,000	$52.68	59,643
2026-07-24	Michael S. Burns	SVP, Chief Legal Officer & Corporate Secretary	M-Exempt (option disposition)	13,000	$26.50	10,457

Eight transactions, two open-market sales, no purchases.

The Burns sequence on 2026-07-24 is mechanical: exercise 13,000 options at $26.50, sell 13,000 shares at $52.68, retain 59,643. That is an exercise-and-sell to fund the strike and the tax, and it carries little signal.

The one transaction that does carry signal is Angela Williams's. The Vice President and Chief Accounting Officer sold 30,000 shares at $53.29 on 2026-07-24 — two days after the second-quarter results and at a price 4.4% ABOVE today's — retaining 10,437 shares. That is a 74.2% reduction, the largest proportional insider disposal in this batch, executed at the top of the range by the officer responsible for the company's accounting. It is not, on its own, evidence of anything specific. It is the opposite of what a person who believed a control premium was imminent would do, and it is reported as such.

Kevin Boone, the chief financial officer, exercised 136,708 options in June at $22.70 and $23.48 and — on the evidence of this eight-transaction window — did not sell. His holding is recorded at 345,330 shares. That is the most constructive item in the file and it is the mirror image of the Williams disposal.

What the file does not contain: any open-market purchase by anyone. Stephen Angel's only transaction is a 187-share phantom-stock accrual.

7. Verdict, kill-criteria and flip conditions

Hold.

The case for owning it. CSX operates an irreplaceable eastern US rail network and has just produced the best operating quarter in several years: volume up 6%, revenue up 10% to $3,935 million, expenses up only 6%, operating income up 17%, an operating margin of 38.3% up 240 basis points, and diluted EPS up 23% to $0.54. For the six months, expenses FELL $15 million while revenue rose 6%, lifting the margin 400 basis points to 37.2% and EPS 24% to $0.97. Return on equity is 24.1%. The share count has fallen 19% in five years. And a transaction may be coming: the 10-Q discloses "advisory expenses related to potential industry consolidation."

The case against buying it here, which is the operative one. Our base fair value of $52 is 1.9% above spot and the vendor's consensus target of $51.06 is three cents above the close. Asymmetry is 1.14:1. The stock is 4.1% below its 52-week high after a 45.8% twelve-month move, at the 90th percentile of its annual range, having drawn down at most 4.1% all year. Revenue declined in each of the three fiscal years before this one, and FY2026 consensus EPS of $2.00 merely returns the company to its FY2022 level of $1.95. The buyback has been cut 70% in three years and total shareholder yield is roughly 2.5% against 6% three years ago. The chief accounting officer sold 74% of her holding at $53.29 two days after the results. The Synthos knowledge base contains not one claim about this company. And the largest single component of the current price — a corporate transaction — is evidenced by nothing more than an advisory-fee line item, with the street's own low target at $32, 37.3% below spot, measuring what the shares are worth without it.

We are not going to manufacture a Buy out of a 1.9% base case on a one-to-one payoff. Nor is there any case for Avoid: the operating recovery is genuine, filing-verified and large, the asset is irreplaceable, and the bear case requires the consolidation premium to unwind rather than the business to fail. Hold is the honest tier, and $40 is where the arithmetic changes.

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

Explicit confirmation that no industry transaction is in prospect, or the abandonment of advisory work — on the street's own distribution that is worth up to 37% of the share price.
An operating margin below 36% in any quarter, reversing the 400-basis-point six-month gain.
Volume growth below 2% for two consecutive quarters, indicating the 6% June-quarter figure was cyclical share recapture rather than durable.
A further buyback cut below $1.0 billion annualised, which would confirm free cash flow is constrained rather than conserved.
Capital expenditure guided above $3.2 billion, against $2,902M in fiscal 2025 and roughly $1,680M of depreciation.
A Surface Transportation Board decision in any Class I proceeding that materially raises the bar for eastern consolidation.
A break below the 200-day moving average of $41.45 on company-specific news.

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

A price at or below $40, which is roughly 17.6x FY2027 consensus. At that level the same forecast produces a 30% base-case return and the verdict changes on arithmetic alone.
An announced combination on terms favourable to CSX shareholders, with a stated Surface Transportation Board timetable.
Two more quarters with the operating margin above 38.5% and volume growth above 4%, which would establish the recovery as structural rather than cyclical.
A restored buyback above $2.5 billion annualised, which would add roughly a point and a half to annual shareholder return.
Insider buying, on a file whose only outright sale was a 74% reduction by the chief accounting officer at the high.

Where CSX fits in the Synthos Framework Portfolio. The industrials sleeve, held at existing weight if already owned, no new capital today. Sizing note: the operative discipline here is to separate the two things being bought. The operating recovery is verifiable and worth roughly what the price says; the consolidation premium is not verifiable at all. A position sized for the first and not the second is the correct exposure, and at $51.03 the market is charging for both. Ours is $40. Logged as a tracked Synthos call (Hold) as of 2026-08-04 at $51.03, with the fair-value anchors, kill criteria and the $40 upgrade trigger all gradeable.

Single biggest risk: that the price contains a transaction that does not happen. The only evidence of industry consolidation anywhere in this archive is one clause in a quarterly expense discussion — "advisory expenses related to potential industry consolidation" — and the stock has risen 45.8% in twelve months against a business whose revenue fell 3.1% in the most recent full year. The street's price-target distribution runs from $60 to $32; the mean of $51.06 equals the market price; and the $32 low is a direct measurement of what a Class I railroad growing earnings 13% a year is worth without a control premium — 37.3% below spot. The operating recovery is real and it justifies a good part of the move; nothing in this file allows us to handicap the rest. A verdict of Hold on a stock at the 90th percentile of its range, with the average analyst target equal to the price, is not caution — it is the arithmetic, and the kill criteria above are set on the two things that would resolve it.

Provenance & disclosures

Traceability: ZERO knowledge-base claims name CSX out of 51,928 distilled claims (raw hits 0, entity matches 0, free-text hits 0, used 0, discarded 0; breadth 0, net conviction none). This is the only completely null sweep in this batch alongside JCI, and it differs informatively from the USB result: USB returned three hits and all three were the Universal Serial Bus, a homograph problem, whereas CSX returned NOTHING of any kind. Because "CSX" is a distinctive three-letter string with low collision potential, the null is a genuine measurement rather than a filtering artefact — no expert in the Synthos knowledge base has expressed a recorded view on this company, on the eastern US rail franchise, on the margin recovery documented above, or on the North American railroad consolidation wave that CSX's own filings say it is paying advisers to consider. Given that consolidation is the largest single swing factor in this security's value, the absence of any commentary on it is a real limitation of the qualitative input, and it is stated rather than papered over. The lane is reported as empty rather than padded with adjacent macro claims about freight, tariffs or industrial production, none of which name this company.
Data as-of: fundamentals — segment revenue and expense detail, volume, operating margin, earnings per share and the industry-consolidation advisory disclosure through 2026-06-30, from the 10-Q filed 2026-07-22 · estimates 2026-08-04 · prices 2026-08-04, quote timestamp 1785873601 = 2026-08-04T20:00:01Z ($51.03, +2.41%; 50-DMA $48.31; 200-DMA $41.45; RSI 57.4; MACD +0.73; max drawdown −4.1%) · knowledge-base claims 2026-08-04. CSX's fiscal year is the calendar year. All figures come from the Synthos vendor data file for CSX or from the SEC filings in the CSX archive; no figure comes from memory, recall or external retrieval.
Filing archive contents: 10-K filed 2026-02-12 (fiscal 2025, 93 preserved [TABLE] blocks); 10-Q filed 2026-04-22 (March 2026 quarter, 31 tables); 10-Q filed 2026-07-22 (June 2026 quarter, 32 tables — the primary source, and the origin of the volume, revenue, expense, operating-margin, earnings-per-share and consolidation-advisory disclosures); 8-K filed 2026-04-22 (first-quarter results, Item 2.02 — furnished, with the press release and Quarterly Financial Report as Exhibits 99.1 and 99.2, neither in the extracted text); 8-K filed 2026-05-14 (separation of Stephen Fortune, EVP and Chief Digital & Technology Officer, effective immediately; plus 2026 annual meeting voting results — twelve directors elected); 8-K filed 2026-07-22 (second-quarter results, Item 2.02, same furnished structure). Statement tables ARE preserved for this name and were used directly.
Where the filings contradicted or corrected the vendor: capital expenditure — capexPerShareTTM of $0.16226 implies roughly $301 million of trailing capital expenditure against $2,902 million in the vendor's own cf_a for fiscal 2025 and $396 million of rail-segment depreciation in the June quarter alone, a roughly 90% understatement, so the corrected free-cash-flow yield is approximately 2.56% rather than the reported 5.31% and price-to-free-cash-flow is nearer 39x than 18.83x; the revenue history — one earn_cal row (2025-10-16) reports revenueActual of $4,460M where inc_q reports $3,587M for the same quarter, a $873 million / 24.3% error that manufactures an apparent 24.9% revenue beat, with the four fiscal-2025 quarterly revenues summing to $14,092M exactly matching inc_a and confirming inc_q is correct; depreciation — inc_a FY2025 reports $1,276M against cf_a's $1,680M, a 24% discrepancy resolved in favour of the cash-flow figure by the 10-Q's $396 million quarterly rail depreciation; and the corporate context — the disclosure of "advisory expenses related to potential industry consolidation" and the 2026-05-14 separation of the Chief Digital & Technology Officer, neither of which appears in any vendor field. Where vendor and filing AGREED — recorded: inc_q June-quarter revenue of $3,935M and operating income of $1,506M match the 10-Q to the dollar; the four fiscal-2025 quarterly revenues sum exactly to inc_a's $14,092M; effectiveTaxRateTTM of 23.3% matches incomeTaxExpense ÷ incomeBeforeTax exactly; and the implied share count (1,852.3M) is within 0.15% of revenuePerShareTTM's implication.
Rejected as a class: the entire cash-flow block (freeCashFlowYieldTTM 5.31%, priceToFreeCashFlowRatioTTM 18.83x, evToFreeCashFlowTTM 22.48x, capexToOperatingCashFlowTTM 0.0566, capexToDepreciationTTM 0.182, capexToRevenueTTM 2.07%, capitalExpenditureCoverageRatioTTM 17.68, dividendPaidAndCapexCoverageRatioTTM 4.08, freeCashFlowToEquityTTM $3,983M) for the capital-expenditure understatement, together with the vendor composite rating's discounted-cash-flow sub-score of 2 out of 5, which rests on the same figure; seg_prod for fiscal 2023 and earlier, where the block contains both a "Total Merchandise" aggregate of $8,700M and its seven individual components summing to $8,653M, so that summing every line gives $23,413M against $14,657M of revenue — 159.7% (the FY2025 and FY2024 entries drop the components and reconcile to 96.6% and 95.9%, and only those are used); inc_q.grossProfit, which is identical to revenue in the June 2026 quarter (a 100% gross margin), identical to operating income in five other quarters, and a third value in two more — three inconsistent definitions across nine quarters, with grossProfitMarginTTM of 54.9% inheriting it; est.ebitdaAvg and est.ebitAvg for the fixed-ratio signature (exactly 49.528% and 38.874% of revenueAvg in five consecutive forward years, an instance made unusually plausible-looking by the 38.9% implied EBIT margin sitting within a point of the 38.3% actually achieved); inc_a FY2025 ebitda of $5,866M, built on the understated depreciation figure; and the FY2029 estimate row, which rests on 2 analysts.
seg_geo is EMPTY — zero entries for any year. Reported as empty; the analytical cost is low for a domestic railroad operating east of the Mississippi, and that is stated rather than the gap being treated as material.
returnOnEquityTTM — source stated: present in km_ttm at 24.06% and absent from ratios_ttm. Used as reported.
Enterprise value — clean. $112,882M, with noncontrolling interests of $5 million, immaterial and the smallest in this batch. Finance lease obligations of $479M are correctly included in total debt of $19,352M; no lease double-count of the T or TJX class.
Estimate coverage: 16 analysts on both FY2026 and FY2027 diluted EPS — the most consistent near-year coverage in this batch — the anchor for all three fair values, with an FY2027 range of $2.148 to $2.404 (11.9%); 7 on FY2028; 2 on FY2029 and 5 on FY2030, both excluded from every conclusion.
Peer note: the vendor peer set is Cummins, Canadian National, Canadian Pacific Kansas City, FedEx, Illinois Tool Works, Norfolk Southern, Quanta Services, Republic Services, Thomson Reuters and United Rentals. Three are genuine Class I railroads — Norfolk Southern ($76.7B), Canadian National ($77.0B) and Canadian Pacific Kansas City ($79.7B) — and Norfolk Southern is the closest, being the other eastern network and the obvious counterparty in any eastern consolidation. Illinois Tool Works, Quanta, Republic Services, Thomson Reuters and United Rentals are unrelated. No cross-multiple is drawn.
Fair-value caveat: the $37 / $52 / $67 anchors are multiples of the FY2027 consensus diluted EPS distribution — 17.2x the low of $2.148, 22.9x the mean of $2.271, and 27.9x the high of $2.404 — cross-checked against FY2026 and FY2028 and against book value of approximately $7.59 per share. Stated arithmetic, not a discounted cash flow. No cash-flow cross-check is applied, because the vendor's capital-expenditure figure is roughly 90% understated. The base is sensitivity-disclosed: 22.5x — the current FY2027 multiple held flat — gives $51.10, which is spot; 25x gives $57. The base assumes modest genuine de-rating from the 25.5x currently paid on FY2026; the mechanical rolldown from 25.5x to 22.5x at a constant price is NOT counted as compression. A structural caveat applies to all three: the street's own price-target distribution runs from $32 to $60 with a mean of $51.06 — three cents above the market price — which is a bimodal outcome averaged into a point estimate, and any single fair value for this security inherits that ambiguity.
Timing: second-quarter 2026 results were released 2026-07-22, thirteen days before this dive, beating consensus EPS by 4.2% ($0.54 against $0.518) on revenue of $3,935M against $3,894M estimated. The next print is 2026-10-15, 72 days away, and consensus of $0.54 EPS and $3,929M of revenue implies NO sequential growth on either line. The most recent insider filings are dated 2026-07-27 and include a 30,000-share sale at $53.29 by the Vice President and Chief Accounting Officer, a 74.2% reduction of her holding, executed two days after the results at a price 4.4% above the current one. 2026-08-04 carried no company-specific filing, and the stock closed at the day high, up 2.41%.
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.