The TJX Companies TJX

> ### Earnings in 15 days — 2026-08-19

> TJX reports second-quarter fiscal 2027 results on Wednesday 2026-08-19, fifteen days after this dive. Consensus is $1.19 of diluted EPS on $15,183.3M of revenue — only +5.4% revenue growth after a first quarter that delivered +9.2% with comps accelerating from 3% to 6%. The verdict below is set with that timing as an explicit input: we are declining a full-priced entry two weeks ahead of a print, not declining the company.

Reference table

1. What the business is, and how the disclosure actually works

The 10-K: TJX operates approximately 5,200 stores through four segments. In the United States, Marmaxx (TJ Maxx, Marshalls, tjmaxx.com, marshalls.com, and Sierra) and HomeGoods (HomeGoods and Homesense). Outside, TJX Canada (Winners, HomeSense, Marshalls) and TJX International (TK Maxx, Homesense, tkmaxx.com, tkmaxx.de, tkmaxx.at in Europe; TK Maxx in Australia). Chief executive Ernie L. Herrman; approximately 377,000 employees; NYSE-listed since 1987.

Store base at the fiscal-2026 year-end (10-K)

The 10-K also publishes long-term store potential, and the gap is the growth runway: Marmaxx 3,000 against 2,603 today; HomeGoods 1,800 against 1,042. That is 1,155 US stores of headroom in the two largest concepts alone, at a rate of roughly 3% of the base per year.

Segment results — the vendor's product table is EXACT, and its geographic table is broken

Fiscal 2026 vs fiscal 2025 (10-K, US$M):

The seg_prod block sums to $60,372M for fiscal 2026 against reported net sales of $60,372M — an exact match — and reconciles in every prior year in the file. This is one of the cleanest vendor segment tables we have seen. We use it without adjustment.

Segment profit is defined by the company, and the definition matters: "pre-tax income or loss before general corporate expense and interest (income) expense, net, and certain separately disclosed unusual or infrequent items." The four segments total $8,089M against consolidated operating income of $7,178M; the $911M difference is general corporate expense. Note that TJX itself says segment profit excludes "certain separately disclosed unusual or infrequent items" — and yet the segment-margin commentary explicitly attributes part of the improvement in three of the four segments to the interchange settlement. We flag that in Section 2.

seg_geo is broken and unusable — and the break is large. The vendor supplies one year only (fiscal 2026) and two lines only: TJX Canada $5,629M and TJX International $7,986M, totalling $13,615M against $60,372M of revenue. The United States — 77.4% of the company — is absent entirely. No geographic figure in this dive comes from seg_geo; the international share of 22.6% is computed from the segment table instead.

2. The number that is not what it looks like: fiscal 2026 EPS

From the 10-K, verbatim:

> "Net income was $5.5 billion in fiscal 2026 compared to $4.9 billion in fiscal 2025. Diluted earnings per share in fiscal 2026 were $4.87 compared to $4.26 in fiscal 2025. The credit card interchange fees litigation settlement and related expenses resulted in a net benefit of $0.14 on diluted earnings per share in fiscal 2026. Foreign currency had a $0.01 negative impact on diluted earnings per share in fiscal 2026 compared to a $0.01 positive impact in fiscal 2025."

So the clean fiscal-2026 diluted EPS is $4.73, not $4.87, and the underlying growth rate was 11.0%, not 14.3%.

Two independent confirmations that this is right. First, the earnings calendar's reported actuals for the four fiscal-2026 quarters — $0.92, $1.10, $1.28 and $1.43 — sum to exactly $4.73, while the vendor's income statement quarters sum to $4.89. The gap is entirely in the fourth quarter, which is where the settlement landed. Second, the 10-K's segment commentary attributes margin improvement to it in three of four segments: Marmaxx's rise to 15.1% was "primarily driven by a net benefit from the credit card interchange fees litigation settlement and related expenses as well as favorable merchandise margin"; HomeGoods' rise to 12.2% likewise; while TJX Canada's margin FELL to 13.4% "primarily driven by expenses related to the credit card interchange fees litigation settlement" and TJX International's rise to 7.0% was "partially offset by expenses related to" it. The settlement was a net benefit in the US and a net cost abroad.

The vendor's income statement carries the gross $5,494M of net income and an eps of 4.89 with no flag whatsoever, and epsdiluted is NULL on every row. Every trailing multiple published on this company from that file is therefore struck on an inflated denominator.

The full annual record (10-K)

Note that fiscal 2024 was a 53-week year (ended 2024-02-03), which flatters its 8.6% growth by roughly two percentage points and understates fiscal 2025's 4.0%. The 10-K states it explicitly on the face of the comprehensive-income statement.

Also note the basic-EPS discrepancy. The 10-K reports basic EPS of $4.93 on 1,114M weighted-average basic shares and diluted EPS of $4.87 on 1,128M diluted shares. The vendor reports eps of 4.89 on weightedAverageShsOutDil of 1,125M — a share count that appears nowhere in the filing and an EPS figure that matches neither the basic nor the diluted number. We use the filing throughout.

3. Quarterly trajectory — and the acceleration

Revenue growth has accelerated in each of the last five quarters: +5.1%, +6.9%, +7.5%, +8.5%, +9.2%. Comparable-store sales went from +3% in the first quarter of fiscal 2026 to +6% in the first quarter of fiscal 2027, and the 10-Q is specific about the composition: "Comp sales for the first quarter of fiscal 2027 were driven by a higher average basket and an increase in customer transactions." Both components positive is the good version of a comp — traffic plus ticket, not price alone.

Five consecutive beats, widening: +0.5%, +8.9%, +4.1%, +2.9%, +16.7%.

And the counterpoint, which is why this is a Watch. Consensus for full fiscal 2027 is $5.2348. The first quarter delivered $1.19. That leaves $4.04 for the remaining three quarters against $3.95 delivered in the equivalent period of fiscal 2026 — implying 2.4% growth, or about 6% once the fourth quarter's settlement benefit is stripped out of the base. Consensus is either markedly conservative or expecting the comp to decelerate sharply from 6%. The 2026-08-19 print will start to settle which, and $1.19 against a first quarter of $1.19 is not a demanding ask.

4. Balance sheet — where the vendor's headline is most misleading

As at 2026-05-02, from the 10-Q filed 2026-05-29 (US$M):

The vendor's netDebt of $7,259M is total debt of $13,489M less cash of $6,230M — and $10,620M of that "debt" is OPERATING lease liabilities carried in the capitalLeaseObligations field. TJX has no material finance leases; it leases its stores, and ASC 842 puts those liabilities on the balance sheet with a matching right-of-use asset.

The vendor reports net debt of $7,259M. On actual borrowings, the company held $2,878M of debt against $5,580M of cash at 2026-05-02 — $2,710M of NET FUNDED CASH. We use the funded figure, and we show the lease-inclusive figure alongside rather than hiding it.

Two independent confirmations that the funded figure is the economically right one. The 10-Q's debt note lists the actual instruments — including "1.600% senior unsecured notes, maturing May 15, 2031" ($500M) and "4.500% senior unsecured notes, maturing April 15, 2050" ($383M) — totalling $2,878M with $999M current. And interestCoverageRatioTTM reads 103.1x, which is what $74M of annual interest expense against $7.5B of operating income looks like; a company genuinely carrying $13.5B of debt could not produce that ratio.

Enterprise value — the vendor is RIGHT, and this is worth recording

> Market cap $174.046B (1,104,704,446 × $157.55)

> + current debt $0.999B + long-term debt $1.871B

> + current lease liabilities $1.714B + long-term lease liabilities $9.596B

> − cash $5.580B

> = $182.646B

The vendor reports enterpriseValueTTM of $182,646,185,467. The reconciliation is exact to the dollar-million, and it uses the 2026-05-02 balance sheet — not the stale annual one in bal_a. After the balance-sheet-staleness defects found across this programme, a vendor enterprise value that is both correct and current is a finding worth stating.

But evToEBITDATTM of 20.24x is wrong, and for a subtle reason. Trailing EBITDA of $9,026M (vendor, cross-checked: the four quarterly ebitda figures sum to exactly that) is struck after full rent expense, because operating-lease cost under ASC 842 runs through cost of sales and SG&A as a single lease cost rather than as depreciation plus interest. Putting the lease liability in the numerator while leaving rent in the denominator double-counts the leases. The internally consistent figures are:

We use 19.0x and note that no rent figure is available in the payload to construct the EBITDAR alternative.

Working capital is a genuine competitive asset and the numbers show it. Days of inventory outstanding 66.4, days of payables 42.0, days of sales outstanding 4.2 — a cash conversion cycle of 28.6 days. Inventory turns 5.5x. Return on invested capital 22.1%, return on capital employed 32.8%, return on equity 59.7%.

The one inventory flag. The 10-K: "Our consolidated average per store inventories... were up 10% at the end of fiscal 2026 as compared to the prior year." Against a 5% comp, that is inventory building faster than sales. The 10-K attributes part of it to Sierra stores entering the consolidated calculation in the first quarter of fiscal 2026 — a definitional change, not a demand signal — but the gap is worth watching in the 2026-08-19 print.

The vendor's composite rating is a lease artefact. rating gives TJX a B / overall score 3, with 5 out of 5 on return on equity and return on assets and 1 out of 5 on both debt-to-equity and price-to-book. The debt-to-equity score of 1 comes entirely from counting $10.6B of operating leases as debt against a $10.2B equity base that four decades of buybacks have kept small. Both sub-scores are rejected as lenses.

5. Cash flow and capital returns

The 10-K states it directly: "During fiscal 2026, we returned $4.3 billion to our shareholders through share repurchases and dividends." That is 87% of free cash flow and 2.5% of the current market capitalisation.

Capex is running at 3.2% of revenue and 1.65x depreciation (capexToDepreciationTTM 1.649), which is a company still building stores and distribution capacity rather than harvesting.

Buyback authority and the net-shrink problem. The 10-Q: "In February 2026, the Company announced that its Board of Directors had approved a new stock repurchase program that authorizes the repurchase of up to an additional $3.0 billion... TJX had approximately $3.5 billion available for repurchase as of May 2, 2026." That is 2.0% of market capitalisation in dry powder. In the first quarter of fiscal 2027 the company repurchased and retired 3.8 million shares for $604M. But shares outstanding went from 1,115,814,224 at 2025-05-03 to 1,105,051,999 at 2026-05-02 — a net reduction of only 0.96% across a full year in which roughly $2.4B was spent. Stock compensation and option exercises are absorbing a meaningful share of the buyback, and the gross-versus-net distinction matters when computing shareholder yield.

Dividend, and the standing labelling trap. The vendor reports dividendPerShareTTM of $1.755 and a 1.11% yield; profile.lastDiv is null. That $1.755 is the trailing-twelve-month total. The filings give the declared rate: the 10-K records "A dividend of $0.425 per share was declared in the fourth quarter of fiscal 2026 and paid in March 2026", and the 10-Q's per-share table shows cash dividends declared per share of $0.480 in the first quarter of fiscal 2027 against $0.425 a year earlier — a 12.9% increase. The forward run-rate is $1.92 per share, a 1.22% yield, not 1.11%. Payout is 37.4% of clean trailing diluted EPS.

Total shareholder yield is therefore approximately 2.6% — 1.22% dividend plus roughly 1.4% of net buyback (using the actual net share reduction rather than the gross spend, because the gross figure overstates the return by about 40 basis points).

6. Valuation — priced in or room?

At $157.55, market cap $174.046B, enterprise value $182.646B (or $171.344B excluding lease liabilities):

Estimate coverage is good on fiscal 2027 and 2028 — 15 analysts on revenue and 10 on EPS for each. The fiscal-2029 EPS row rests on 5 analysts and the fiscal-2030 row on ONE; the latter is excluded from every conclusion.

est.ebitdaAvg, est.ebitAvg and est.sgaExpenseAvg are fabricated in this file and we prove it. Across every year from fiscal 2026 through fiscal 2030, ebitdaAvg is exactly 12.600% of revenueAvg, ebitAvg is exactly 10.727% and sgaExpenseAvg is exactly 18.877% — identical to five significant figures in five consecutive years. Independently, the fiscal-2027 ebitAvg of $6,881M is 4.1% below the operating income implied by consensus EPS and known share counts. All forward valuation in this dive runs on epsAvg and revenueAvg.

Peer context. The vendor peer set contains one true comparable — Ross Stores ($80.6B) — plus Burlington ($23.2B) and Boot Barn ($4.9B) as adjacent off-price and specialty names, and then a set that is not relevant here (Booking Holdings, Lowe's, McDonald's, MercadoLibre, Nike, O'Reilly, PDD). No peer multiples are supplied in the file, so no peer-multiple comparison is drawn. The structural note that can be made: TJX at $174.0B of market capitalisation is larger than Ross and Burlington combined by a factor of 1.7, and its 15.1% Marmaxx segment margin is the benchmark the rest of the category is measured against.

6a. What today's price assumes (the inversion)

At $157.55 — 30.1x the fiscal-2027 consensus and 27.3x fiscal 2028 — the price embeds the following falsifiable claims, each with a number and a date:

• Diluted EPS reaches $5.23 in the year to January 2027 and $5.76 by January 2028. (Consensus-derived; 10 analysts each.) Note what that requires of the remaining three quarters of fiscal 2027: $4.04 against $3.95 delivered a year earlier, or about 2.4% growth on the reported base and 6% on the clean base. After a first quarter that grew 29.3%. This is the one place the price looks conservative rather than stretched, and the 2026-08-19 print against $1.19 is the immediate test.
• Comparable-store sales stay at 4-6%. (Our number, derived from the revenue estimates net of approximately 3% square-footage growth.) Fiscal 2026 delivered 5% and the first quarter of fiscal 2027 delivered 6%. A drop back to 2-3% breaks the revenue line before it touches the margin line.
• Pre-tax margin holds at or above 12%. (Filing-confirmed base of 12.1% for fiscal 2026; our assumption on the hold.) This is where the interchange settlement bites: $0.14 of diluted EPS, roughly 25 basis points of pre-tax margin, does not repeat, so holding 12.1% in fiscal 2027 requires genuine operating improvement just to stand still.
• No tariff pass-through problem. (Our assumption; the 10-K discusses trade and sourcing risk but quantifies nothing.) TJX imports a large share of its merchandise and earns 22.6% of revenue outside the United States. The off-price model is a net beneficiary of supply-chain dislocation — stranded inventory is its raw material — but that is a hypothesis, not a disclosure.
• The market keeps paying roughly 30x the forward year. (Our number.) This is the most fragile assumption in the price. At 25x fiscal 2028E the stock is $144; at 34x it is $196. Nothing in the business changes between those two numbers. The entire dispersion in the outcome sits in a multiple we do not control and, given 53 analysts already positive and none negative, have no informational edge on.

6b. The return bridge (why the multiple moves)

Expected return over the next twelve months decomposes as: diluted EPS growth (+10.1%, from FY2027E $5.2348 to FY2028E $5.7627) + multiple drift (COMPRESSION, from 30.1x to roughly 28x on the respective forward year, −7.0%) + shareholder yield (+2.6%) ≈ +5% to +6%.

Our base case assumes modest multiple COMPRESSION and we should be explicit that this is a judgement, not an artefact. Today the market pays 30.1x on the next fiscal year. Our $162 is 28.1x the then-next fiscal year. That is a genuine 7% de-rating, not the mechanical effect of the calendar advancing. We assume it because the clean growth rate is 10-11% and 30x is roughly 3x PEG — a level TJX has earned in periods of comp acceleration and given back in periods of normalisation, and consensus itself models deceleration from 9.2% revenue growth in the first quarter to 5.8% in fiscal 2028.

Almost the entire base-case return is earnings growth and the multiple is working against it. If the multiple simply HELD at 30.1x on fiscal 2028E, the price would be $173.5 (+10.1%); if it compressed to 25x, $144.1 (−8.5%); if it expanded to 31.7x — which is what the street's $182.75 implies — $182.7 (+16.0%). The street's target is a multiple-expansion call, and it should be labelled as one.

The bull case at $197 requires re-rating and says so. The bear case at $118 is 21x the fiscal-2028 consensus low of $5.600 — a de-rating to the multiple TJX carried during the 2022 comp trough.

6c. Variant perception (where we differ, what would surprise)

• On the business we do not differ from the street at all, and that is the finding. 53 analysts, 47 of them positive, one sell, and a lowest target of $160 that sits 1.6% ABOVE the current price. There is no analyst left to upgrade and no dispersion to arbitrage. Our base of $162 is 11.4% BELOW consensus and 2.8% above spot. When our own arithmetic lands below the crowd's on a name where the knowledge base contributes literally nothing, the honest conclusion is that we have no edge, and no-edge names default toward Watch.
• We differ on the reported earnings base, and it is a real correction rather than a quibble. Fiscal 2026's $4.87 of diluted EPS includes a disclosed $0.14 net benefit from the credit-card interchange fee litigation settlement. The clean figure is $4.73, the clean growth rate was 11.0% and not 14.3%, and the trailing P/E is 31.5x rather than 30.6x. The vendor's income statement carries the gross number unflagged, so every screen currently shows TJX's fiscal-2026 growth about three points faster than it was. This cuts against the stock, and we report it for that reason.
• We differ materially on the balance sheet. Every screen shows $7,259M of net debt and a 0.95x net-debt-to-EBITDA ratio. The company has $2,710M of net funded CASH; the difference is $10.6-11.3B of operating lease liabilities coded as debt. That moves TJX from "modestly levered" to "net lender", and it is why the risk score is a 4 driven by valuation rather than by the balance sheet. It also means the vendor's evToEBITDATTM of 20.24x double-counts the leases; the consistent figure is 19.0x.
• Positive surprise that would force a re-rate: second-quarter comps at or above 6% on 2026-08-19 with EPS materially above $1.19 — which, given consensus asks for only 2.4% EPS growth across the back three quarters, is a live possibility rather than a stretch. A second, cheaper surprise: TJX International's segment margin continuing to converge; it went 5.9% to 7.0% and carries $7,986M of revenue earning less than half the Marmaxx return, so every 100 basis points there is roughly $80M of pre-tax profit.
• Negative surprise that would break the thesis: comps back to 2-3%; per-store inventories still up 10% or more against a decelerating comp, which would foreshadow markdowns; any quantified tariff cost in the cost-of-sales ratio; or simply a de-rating toward 25x, which costs 17% and requires no news at all.

Synthos fair values

All three anchors are multiples of the fiscal-2028 consensus diluted EPS distribution (mean $5.7627, low $5.600, high $5.844, 10 analysts), cross-checked against fiscal 2027E of $5.2348.

• Bear ~$118 — 21.1x the FY2028 consensus LOW of $5.600. Cross-check: 22.5x FY2027E; roughly 8% below the 52-week low of $128.77. The scenario: comps decelerate to 2-3% as the consumer softens, per-store inventories force markdowns, the interchange benefit's absence shows up as a margin decline, and a 30x multiple on a decelerating retailer becomes a 21x multiple. −25.1%.
• Base ~$162 — 28.1x the FY2028 consensus MEAN of $5.7627. Cross-check: 30.9x FY2027E; enterprise value excluding leases of $178.2B on FY2028E revenue of $67.891B is 2.62x sales. Sensitivity, stated openly: 27x gives $156 — below spot — and 30x gives $173. The whole answer lives in a 27-30x band and we make no claim to finer resolution than that. The scenario: comps normalise to 4-5%, margin holds, unit growth continues at 3%, and the multiple compresses modestly as the growth rate settles into the low double digits. +2.8%, plus about 2.6% of shareholder yield.
• Bull ~$197 — 33.7x the FY2028 consensus HIGH of $5.844. Cross-check: 31.0x FY2029E of $6.3628; equal to the street's high target of $197. This bull case requires multiple EXPANSION as well as an earnings beat, and we say so rather than disguising it — comps hold at 6%, TJX International's margin converges toward the domestic segments, and the market decides a 5,214-store off-price compounder with net cash deserves a mid-thirties multiple. +25.0%.

Base is 2.8% above spot; asymmetry approximately 1.00:1 (25.1% down, 25.0% up). A one-to-one payoff with a base case at the current price, fifteen days before a print, on a name where the knowledge base is empty, is a Watch by construction.

7. Knowledge base — one claim, and it does not name the company

Raw hits: 1. Entity matches: 1. Text matches: 0. Case-sensitive bare-ticker matches: 0. Used as name-level conviction: 0. Discarded for collision: 0.

The primary sweep covered the entities TJX, TJ Maxx, TJMaxx, Marshalls, HomeGoods and The TJX Companies across all 51,928 distilled claims. A supplementary text sweep on off-price, TJ Maxx, Marshalls, HomeGoods and treasure hunt returned zero additional claims. A case-sensitive word-boundary sweep on the bare ticker TJX in claim text returned zero. There is no homograph problem, no metaphor collision, and nothing to run a concentration sensitivity test on, because one channel supplying one claim is not a concentration — it is an absence.

The single claim, verbatim:

> 2023-05-01 · bullish · conviction 70 · skill 1.0 · horizon: principle · channel: business_breakdowns · no named speaker

> "Best-in-class retailers win by maniacal customer/employee focus over unit expansion; service and store experience protect the moat as merchandise commoditizes."

What this is, stated without inflation. It is a general principle about retail, tagged to TJX as an entity by the distiller, that does not name the company anywhere in its text. It is more than three years old, carries no named speaker, and is attributed only to a channel — the weakest sourcing the 4-lane attribution policy admits. It is directionally consistent with the qualitative case in Section 1 and it is emphatically not a view on the security.

Conclusion, stated as a finding rather than dressed up. The Synthos knowledge base has no view on The TJX Companies. Not a thin view, not a mixed view — none. Conviction rating: Very low. Breadth: 1. Net conviction: none. The absence is an explicit input to the Watch verdict: on a name where the price is full, the street is unanimous and our own research library is silent, there is no basis for a differentiated position.

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

Nine findings. TJX is a mixed file with one exact reconciliation, one exact verification, and one segment table that is missing three-quarters of the company.

1. netDebt counts $10,620M of OPERATING lease liabilities as debt — REJECTED and rebuilt. The capitalLeaseObligations field carries $10,620M (FY2026), $9,912M (FY2025) and $9,680M (FY2024), which are the operating lease liabilities from the balance sheet, and totalDebt of $13,489M includes them. The vendor reports net debt of $7,259M; the 10-Q shows total borrowings of $2,878M principal against $5,580M of cash at 2026-05-02, i.e. $2,710M of NET FUNDED CASH; we use the funded figure and disclose the lease-inclusive figure alongside. The 10-Q's debt note lists the actual instruments, and interestCoverageRatioTTM of 103.1x independently confirms that only ~$2.9B of interest-bearing debt exists.

2. Fiscal-2026 diluted EPS is inflated by a disclosed one-off and the vendor does not flag it — corrected from the 10-K. "The credit card interchange fees litigation settlement and related expenses resulted in a net benefit of $0.14 on diluted earnings per share in fiscal 2026." Clean diluted EPS $4.73, not $4.87. Clean growth 11.0%, not 14.3%. Clean trailing P/E 31.5x, not 30.6x. Independently confirmed by the earnings-calendar actuals ($0.92 + $1.10 + $1.28 + $1.43 = $4.73) and by the 10-K's segment commentary, which attributes margin improvement to the settlement in Marmaxx and HomeGoods and margin DETERIORATION to it in TJX Canada.

3. epsdiluted is NULL on every row and eps matches neither the filing's basic nor its diluted figure — REJECTED. The 10-K reports fiscal-2026 basic EPS of $4.93 on 1,114M shares and diluted EPS of $4.87 on 1,128M shares. The vendor reports eps of 4.89 on weightedAverageShsOutDil of 1,125M — a share count that appears nowhere in the filing. The same pattern holds for fiscal 2025 (vendor 4.31 against filed basic $4.31 and diluted $4.26) and fiscal 2024 (vendor 3.90 against filed basic $3.90 and diluted $3.86). The vendor's eps is the BASIC figure in FY2025 and FY2024 and an unexplained hybrid in FY2026. All EPS figures in this dive are filing-sourced.

4. seg_geo is missing 77.4% of the company — REJECTED in full. The block supplies one year (fiscal 2026) and two lines: TJX Canada $5,629M and TJX International $7,986M, totalling $13,615M against $60,372M of revenue. The United States is absent entirely. No geographic figure in this dive comes from seg_geo; the international revenue share of 22.6% is derived from the segment table instead.

5. est.ebitdaAvg, est.ebitAvg and est.sgaExpenseAvg are fabricated — NOT USED. From fiscal 2026 through fiscal 2030, ebitdaAvg is exactly 12.600% of revenueAvg, ebitAvg is exactly 10.727% and sgaExpenseAvg is exactly 18.877% — identical across five consecutive years to five significant figures. This is the fixed-ratio fabrication signature. All forward valuation runs on epsAvg and revenueAvg.

6. evToEBITDATTM of 20.24x double-counts operating leases — recomputed. Trailing EBITDA of $9,026M is struck after full lease expense, because ASC 842 operating-lease cost flows through cost of sales and SG&A as a single line rather than as depreciation plus interest. Placing $11,310M of lease liabilities in the numerator while leaving rent in the denominator is inconsistent. On enterprise value excluding lease liabilities ($171.344B) against the same EBITDA, the multiple is 19.0x, and that is what we use.

7. lastDividend is absent and dividendPerShareTTM of $1.755 is the trailing total — corrected from the filings. profile.lastDiv is null. The 10-K records a $0.425 per share dividend declared in the fourth quarter of fiscal 2026; the 10-Q's per-share table shows $0.480 declared in the first quarter of fiscal 2027 against $0.425 a year earlier, a 12.9% increase. The forward run-rate is $1.92 — a 1.22% yield, not the vendor's 1.11%.

8. quote.yearHigh/yearLow ($170.00 / $128.48) disagree with tech.hi52/lo52 ($168.41 / $128.77) — we use tech. The discrepancies are 0.9% and 0.2%. The tech block is computed from the same close series that produces the moving averages and relative returns, and mixing sources would make the percentages inconsistent.

9. bal_a is annual only and therefore three months stale — stated, and worked around. The most recent balance sheet in the payload is 2026-01-31. The 10-Q filed 2026-05-29 supplies the 2026-05-02 position and every balance-sheet figure in this dive uses it. Notably, the changes are material: cash fell from $6,230M to $5,580M and long-term lease liabilities rose from $8,894M to $9,596M in the quarter.

Not defects — and two of them are verifications worth recording:

A. The share count reconciles to five decimal places. The 10-Q cover page states 1,104,704,446 shares outstanding as of 2026-05-22. The vendor's market capitalisation of $174,046,185,467 divided by $157.55 implies 1,104,704,764 shares. Agreement to 0.00003%. No share-class error, no partial-unit error, no split problem.

B. enterpriseValueTTM is correct AND current. $174.046B + $999M + $1,871M + $1,714M + $9,596M − $5,580M = $182,646M against the vendor's $182,646,185,467. Exact, and struck on the 2026-05-02 balance sheet rather than the stale annual one. Given how often the enterprise value in these files is wrong, a correct and current one is a finding.

C. seg_prod sums exactly to reported revenue in every year available — fiscal 2026 totals $60,372M against reported net sales of $60,372M; fiscal 2025 totals $56,360M against $56,360M. No double-counting, no missing segment.

Also correctly reported: revenuePerShareTTM of $54.986 × 1,120M diluted shares gives $61,584M, matching the filing-derived trailing revenue exactly; incomeQualityTTM of 1.312 correctly reflects $6,874M of operating cash flow against $5,494M of net income; interestCoverageRatioTTM of 103.1x is consistent with the debt and income figures; effectiveTaxRateTTM of 24.55% matches the filing's provision; and dividendPayoutRatioTTM of 32.7% is arithmetically correct on the trailing figures used.

Non-equity tripwire — checked and passed. TJX is common stock, NYSE-listed, $1 par value, 1,800,000,000 shares authorised and 1,105,051,999 issued and outstanding per the 10-Q. Beta is 0.623 — low, but nowhere near the sub-0.3 typical of a fixed-income-like instrument; the dividend is variable and has been raised annually, most recently by 12.9%; the 52-week band of $128.77 to $168.41 is a 30.8% range; volume was 5.12 million shares (roughly $807M of turnover). This is common equity.

9. Technicals

• Price $157.55. −6.4% from the 52-week high of $168.41; +22.3% above the 52-week low of $128.77. Position within the annual range: 73rd percentile.
• The stock is sitting on its 50-day moving average to the cent: $157.55 against $157.534 — a 0.01% difference. It is +1.9% above the 200-day at $154.576. There is no trend here in either direction, and that is a genuinely unusual configuration to write a dive against.
• RSI 65.7 — warm but short of overbought. MACD +0.61 — barely positive, effectively flat.
• Maximum drawdown from peak over the trailing year: −6.4%. This stock does not have corrections; it has pauses.
• Relative performance has rolled over, and the shape matters: 3-month +1.9% vs SPY +7.6% and QQQ +7.7% — 5.7 points behind; 6-month +4.9% vs SPY +11.1% and QQQ +15.7%; 12-month +25.0% vs SPY +24.3% and QQQ +30.8% — level with the index. The full twelve-month outperformance was earned in the first six months and has been given back since. A stock that is level with the index over a year, below it over six months and well below it over three, while its comparable-store sales accelerate, is a stock whose multiple is doing the work in the wrong direction.
• Sentiment: 1 strong buy, 46 buy, 5 hold, 1 sell across 53 analysts. Consensus target $182.75 (+16.0%), median $187.50, high $197 (+25.0%), low $160 — 1.6% ABOVE spot. Every published target is above the price. We read that as agreement rather than as information: there is no analyst positioned to become more constructive, and a consensus with no dispersion offers nothing to trade against.

Today's move

TJX closed 2026-08-04 at $157.55, up $0.05 or 0.03% from $157.50. It opened at $157.13, traded $155.30 to $158.49, and closed mid-range on 5,120,832 shares — roughly $807M of turnover. An entirely uneventful session inside a market whose energy was elsewhere. There is no gap to pay for and no news in this file dated 2026-08-04. The entry price is simply the price.

10. Insiders — routine board compensation and one trivial sale

Eight transactions, all filed 2026-06-11 or 2026-06-12, all involving directors:

The reading, and it is short. There is not a single executive transaction in this file — no chief executive, no chief financial officer, no operating officer. Every entry is a non-employee director's annual deferred-stock-unit award or the conversion of a matured award into common stock. The only cash transaction in the file is Jackwyn Nemerov's sale of 957 shares at $168.595 — approximately $161,000, and executed at a price 7.0% above the current one, near the 52-week high.

What the file does NOT contain is any open-market purchase, and it does not contain any executive selling either. That is a neutral signal and it is reported as neutral. The absence of executive transactions of any kind across a two-month window is the least informative insider table in this batch, and no weight is placed on it.

11. Verdict, kill-criteria and flip conditions

Watch.

What is genuinely excellent, and essentially none of it is in dispute: comparable-store sales of +5% in fiscal 2026 accelerating to +6% in the first quarter of fiscal 2027, driven by both basket and transactions; all four segments expanding margin, with Marmaxx at 15.1%, HomeGoods at 12.2% and TJX International closing from 5.9% to 7.0%; pre-tax margin from 11.5% to 12.1% with the cost-of-sales ratio down 40 basis points and SG&A down 30; return on invested capital 22.1% and return on equity 59.7%; a 28.6-day cash conversion cycle and 5.5x inventory turns; $2,710M of net funded cash and 103x interest coverage; $4,917M of free cash flow and $4.3B returned to shareholders; 1,155 US stores of stated headroom in the two largest concepts; and five consecutive EPS beats.

What we are declining to pay for: 30.1x the fiscal-2027 consensus and 27.3x fiscal 2028 for an earnings line whose clean growth rate was 11.0%, not the 14.3% the reported figure implies — because $0.14 of last year's $4.87 came from a credit-card interchange litigation settlement that will not repeat. A base fair value of $162 that is 2.8% above spot and 11.4% below the street's $182.75. A consensus with 47 positive analysts, one negative, and a lowest target already above the price — no dispersion, no informational edge. A knowledge-base lane of exactly one claim that does not name the company. And a second-quarter print in fifteen days.

The distinction that matters. This is not a criticism of TJX. It is the observation that everything good about this company is known, published, and in the multiple — and that our own arithmetic, run independently, lands below the crowd's on a name where our research library has contributed nothing to differentiate the view. In the Synthos frame, a name where the base case sits at spot, the fair value sits below consensus, and the conviction lane is empty is a Watch by construction, not by hesitation. We want this business. We want it at a price that has a return in it.

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

• Comparable-store sales below 3% in any quarter. That is the number the entire multiple rests on, and it is reported every quarter.
• Per-store inventories still up 10% or more against a decelerating comp at the 2026-08-19 print — markdowns follow, and gross margin follows markdowns.
• Pre-tax margin below 11.5%, which would mean the interchange benefit's absence is not being offset by operating improvement.
• Any quantified tariff or landed-cost impact appearing in the cost-of-sales ratio, which the 10-K currently discusses only as a risk factor.
• A break below the 200-day moving average at $154.58 on company-specific news rather than market beta.
• A material slowdown in unit growth below 2%, which would remove the mechanical half of the growth algorithm.

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

• A price below approximately $138 (roughly 24x FY2028E, 26x FY2027E), which would put the base case at +17% and make the risk/reward genuinely asymmetric. This is the single most likely path to a Buy and it is 12.4% below today's close — well inside the range this stock traversed between last October and this February.
• The 2026-08-19 print delivering comps at or above 6% with EPS materially above $1.19 — which, given consensus asks for only 2.4% EPS growth across the remaining three quarters after a 29% first quarter, is a realistic outcome. A beat of that shape would validate the fiscal-2027 line and justify revisiting the multiple assumption in 6b.
• TJX International's segment margin reaching 9% or better, which on $8.0B of revenue is roughly $160M of incremental pre-tax profit and is the largest single identified source of upside to the base case.
• A genuine name-level knowledge-base claim appearing. At present the lane contains one three-year-old generic retail principle with no named speaker, and that alone is worth roughly one tier of caution.

Where TJX fits in the Synthos Framework Portfolio. The consumer / defensive-retail sleeve, at 0% today with a 2-3% target on a fill near $138 or a confirmed 6%-plus comp on 2026-08-19. On batch overlap: TJX is the lowest-beta name in this batch (0.623) and the only one whose model is countercyclical in a specific, mechanical way — a weak apparel cycle is a supply opportunity, not a demand problem. That is a genuine portfolio-construction argument for owning it at some price. It is not an argument for owning it at 30x. Logged as a tracked Synthos call (Watch) as of 2026-08-04 at $157.55.

Single biggest risk: the multiple, and there is nothing under it. At 30.1x forward earnings and 2.78x sales for a business growing clean EPS at 11%, TJX carries no valuation cushion whatsoever. The balance sheet, excellent as it is, cannot supply one — $2.7B of net cash against a $174.0B market capitalisation is 1.6% of the price. A de-rating to 25x — which is where this stock traded as recently as its fiscal-2023 comp trough — costs 17% with no change in the business at all. That is the whole bear case, and the reason a 25% downside case sits against a 25% upside case on a name where almost everything is going right.

Provenance & disclosures

• Traceability: 1 tagged knowledge-base claim touches TJX across 51,928 distilled claims (raw hits 1, entity matches 1, text matches 0, case-sensitive bare-ticker matches 0, discarded for collision 0; breadth 1, net conviction none). Source: business_breakdowns (2023-05-01, bullish, conviction 70, skill 1.0, horizon: principle, no named speaker) — a general claim that best-in-class retailers win on customer and employee focus rather than unit expansion, which does not name TJX anywhere in its thesis text and is tagged to the company by entity extraction alone. Searches: the primary entity sweep on TJX, TJ Maxx, TJMaxx, Marshalls, HomeGoods and The TJX Companies (1 hit); a supplementary text sweep on off-price, TJ Maxx, Marshalls, HomeGoods and treasure hunt (0 hits); and a case-sensitive word-boundary sweep on the bare ticker TJX in claim text (0 hits). No concentration sensitivity test was run because a single claim from a single channel is an absence, not a concentration. No claim carries a management speaker_role. The quote is verbatim from the stored claim text.
• Data as-of: fundamentals — first-quarter fiscal 2027 income statement and balance sheet as at 2026-05-02, from the 10-Q filed 2026-05-29; annual figures for fiscal 2026 (ended 2026-01-31) from the 10-K filed 2026-03-31 · estimates 2026-08-04 · prices 2026-08-04, quote timestamp 1785873728 = 2026-08-04T20:02:08Z ($157.55, +0.03%; 50-DMA $157.534; 200-DMA $154.576; RSI 65.7; MACD +0.61) · knowledge-base claims 2026-08-04. Fiscal-year note: TJX's fiscal year ends in late January or early February. "Fiscal 2026" means the 52 weeks to 2026-01-31; "fiscal 2027" means the year ending 2027-01-30. Fiscal 2024, ended 2024-02-03, was a 53-week year and the 10-K flags it. All figures come from the Synthos vendor data file for TJX or from the SEC filings in the TJX archive; no figure comes from memory, recall or external retrieval.
• Filing archive contents and its gaps: 10-K filed 2026-03-31 (fiscal 2026, 335,885 characters, financial-statement tables preserved); 10-Q filed 2025-12-02 (third quarter fiscal 2026); 10-Q filed 2026-05-29 (first quarter fiscal 2027); and three 8-Ks (2026-02-25, 2026-05-20, 2026-06-11). All three 8-Ks are present as cover documents only, 3,512 to 4,375 characters each, with no exhibits extracted — the fourth-quarter and first-quarter earnings releases they announce are NOT in the archive. The quarterly figures used here therefore come from the 10-K and 10-Q financial statements and the vendor's earn_cal block, and the fiscal-2026 quarterly split is reconstructed from the vendor's inc_q series cross-checked against the earnings-calendar actuals.
• Where the filings contradicted or corrected the vendor (detailed in Section 8): net debt rebuilt from $7,259M to $2,710M of NET FUNDED CASH after identifying $10,620M of operating lease liabilities coded into capitalLeaseObligations, with the 10-Q's debt note confirming total borrowings of $2,878M principal; fiscal-2026 diluted EPS corrected from the vendor's implicit $4.89 to the filed $4.87, and then to a clean $4.73 after the 10-K's disclosure that "the credit card interchange fees litigation settlement and related expenses resulted in a net benefit of $0.14 on diluted earnings per share" — the vendor carries the gross figure unflagged; epsdiluted NULL on every row and eps matching neither the filed basic nor the filed diluted figure, on a share count of 1,125M that appears in no filing; seg_geo rejected in full for omitting the United States, 77.4% of revenue; est.ebitdaAvg/ebitAvg/sgaExpenseAvg rejected as fabricated on fixed revenue ratios of 12.600%/10.727%/18.877% across five consecutive years; evToEBITDATTM recomputed from 20.24x to 19.0x to stop double-counting lease liabilities against a rent-inclusive EBITDA; dividendPerShareTTM of $1.755 identified as the trailing total, with the filings giving a declared rate of $0.480 per quarter, a forward run-rate of $1.92 and a 1.22% yield; quote.yearHigh/yearLow rejected in favour of tech.hi52/lo52; and bal_a noted as annual-only and three months stale, with every balance-sheet figure taken from the 2026-05-02 10-Q instead. Where vendor and filing AGREED — worth recording: the share count (10-Q cover 1,104,704,446 at 2026-05-22 against 1,104,704,764 implied by market capitalisation — agreement to 0.00003%); enterpriseValueTTM of $182.646B, which reconciles EXACTLY to the 2026-05-02 balance sheet and is current rather than stale; seg_prod, which sums exactly to reported net sales in every year available; revenuePerShareTTM; incomeQualityTTM; and interestCoverageRatioTTM.
• Basis note: TJX reports on a GAAP basis only and publishes no non-GAAP earnings measure; analyst estimates are struck on GAAP diluted EPS, and the earnings-calendar actuals match the company's reported diluted EPS except in the fourth quarter of fiscal 2026, where the $0.14 interchange settlement benefit was excluded from the comparison figure. No GAAP-to-non-GAAP bridge is required and none is drawn. Trailing multiples are shown on both the reported ($5.14) and clean ($5.00) bases.
• Estimate coverage: 15 analysts on fiscal-2027 revenue and 10 on EPS; 15 and 10 on fiscal 2028; 7 and 5 on fiscal 2029; 5 and ONE on fiscal 2030. The fiscal-2029 EPS row is used only as a bull-case cross-check and the fiscal-2030 row is excluded from every conclusion. The fair value is struck on the fiscal-2028 distribution, which carries 10 analysts.
• Peer note: the vendor peer set contains one true comparable — Ross Stores ($80.6B) — plus Burlington ($23.2B) and Boot Barn ($4.9B) as adjacent names, alongside a set that is not relevant (Booking Holdings $150.5B, Lowe's $122.3B, McDonald's $190.7B, MercadoLibre $95.7B, Nike $61.5B, O'Reilly $76.1B, PDD $32.4B). No peer multiples are supplied in the file, so no peer-multiple comparison is drawn.
• Fair-value caveat: the $118 / $162 / $197 anchors are multiples of the fiscal-2028 consensus diluted EPS distribution — 21.1x the low of $5.600, 28.1x the mean of $5.7627, and 33.7x the high of $5.844. Stated arithmetic, not a discounted cash flow. The base is sensitivity-disclosed: 27x gives $156, below spot, and 30x gives $173. The base assumes modest multiple compression from 30.1x to 28.1x and explains why; the bull case requires expansion and says so; the bear assumes both an earnings shortfall and a de-rating to the multiple TJX carried at its last comp trough.
• Timing: first-quarter fiscal 2027 results were released 2026-05-20 and the 10-Q filed 2026-05-29. The next print is 2026-08-19, FIFTEEN DAYS after this dive — consensus $1.19 of diluted EPS on $15,183.3M of revenue. The proximity of that print is an explicit input to the Watch verdict: with a base case 2.8% above spot, there is no argument for taking full-priced exposure two weeks ahead of an event that will move the stock several percent in either direction. 2026-08-04 was an uneventful session for this name (+0.03% on 5.12 million shares).
• 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.