Cheap With a Grievance, or Cheap for a Reason? A Value-Trap Test From 1,013 Scorecards

What a documented grievance reads like

Wix (Growth Score 58, perception 12 — the only 12 in the library). The stock fell 27% in a single session on May 13, 2026, before any formal guidance change; the June 8 filing that then cut the outlook simultaneously raised the free-cash-flow target to about $420 million. The report's Q10 answer is a model of the genre: the market "compressed everything into one sentence" — AI destroys website builders — while the same disclosures showed ARR up 15% and new-cohort bookings up 46%. "A guide-down that raises the cash target is not the shape of structural collapse," the report argues; "the failure is one of horizon, not comprehension." It also named the pivot in advance: the Q2 release of August 4. That checkpoint has now passed, and the first reading landed on the bull side — revenue of $563.1 million, up 14.9% and ahead of estimates, ARR of $1.963 billion, and a first hard AI-margin disclosure (Base44 gross margin guided to roughly 60% in the second half). The shares, at 64.57, sit about 10% above the report's 58.55 snapshot. The report's own caveat stands: the gap is symmetric and should resolve within one or two more quarters.

Kuaishou (51, perception 10). Here the grievance is a mispriced option. At the report's HK$45.88, the company traded near 9.9x static earnings and 1.2x sales, which prices the main platform as a slow-growth utility and its Kling AI video business at almost nothing — the report reverse-engineers single-digit Hong Kong dollars per share of embedded Kling value, against HK$5–14 if one applies 6–15x to Kling's roughly $500 million ARR. The Q10 answer sorts the failure explicitly: a refusal to respect a maturing second-place platform, plus an unwillingness to look far on an unproven AI line — comprehension was never the problem. The pivot is specific: spin-off terms becoming contracts (press reports put a mooted valuation near $20 billion, with about 70% of Kling revenue from overseas), with downside bounded by over RMB 108 billion of cash-like assets.

CATL, the highest score in the grievance group at 59, needs one line: it anchored the June opportunity quadrant, scores 8 on the pair, and still sits a whisker above its own fair-buy gate.

What cheap for a reason reads like

Comtech (25, perception 4, rated Avoid). The trap signature is arithmetic that fails at the ceiling. The report's own optimistic intrinsic-value band tops out at $8 against the $27 a ten-year five-fold from its $5.40 scoring price would require; compound the report's most optimistic annualized return, 14%, for a full decade and you reach about 3.7x — the bull case, run flawlessly, misses the bar. Beneath it: trailing ROIC near 2.3% against a 14.7% cost of capital, and a $213 million preferred liquidation preference senior to the common. The Q10 answer inverts the usual complaint: far from being ignored, the stock had run +178% over the prior year on restructuring hopes — the market had noticed plenty and was discounting the balance sheet, correctly. Two months later the verdict arrived early: Comtech agreed to sell its satellite business to Gilat for $157.5 million — the restructuring event actually happened — yet quarterly sales fell 16.4%, the net loss reached $14.3 million, and the shares collapsed to $1.72, down about 68% from scoring. Today the stock trades below its own $3.50 fair-buy anchor, and the scorecard would still say no: a price gate answers "cheap enough?"; it never answers "should you?"

Sidus Space (22, perception 4, rated Avoid). The second trap signature is the double failure: an unproven company attached to an overpriced stock. At scoring in early June, revenue was $3.4 million and shrinking 27.6% a year, gross margin was minus 168%, and the shares — lifted by Russell-reconstitution passive buying and space-plus-AI theme flows — stood roughly 55% above the top of the report's reasonable value band. A stock can look cheap on screens and still be expensive against its own evidence. It has since roughly halved, to $2.07.

Two trap calls resolving downward inside two months is satisfying and proves little on its own — two anecdotes. The systematic test comes at the end of this piece.

Two boundary cases that break the shortcut

TSMC (Growth Score 57, perception 4). The mechanical cut files TSMC in the reason group, and the reason is instructive: quality without a grievance. At NT$2,425 the stock sat in the lower fifth of a very wide fair-value band — and the Q10 answer opens by demolishing the premise: "This is the most closely watched manufacturing franchise on earth." Trailing P/E of 28.1x ran 48% above its ten-year average; the plausible misjudgments the report identifies are symmetric (anchoring on 2025's 59.9% gross margin versus extrapolating Q2 2026's 67.7%), which describes disagreement around a fair price, and the price "sits near the midpoint of the conservative blended fair-value range, which is what a market that has done the work looks like." A low band position with no documentable grievance is a fair price wearing a cheap costume. TSMC is by far the strongest business in the reason basket, and we leave it there because the rule is the rule; its next dated checkpoint is the October 8 earnings call, with the report's gross-margin thresholds at roughly 66% and 62%.

Stellantis (35, perception 7). The mirror image: a real grievance attached to a fragile business. The mispricing mechanism is concrete — a structural buyer's strike. With about $10.8 billion of industrial net cash inside a $16.8 billion market cap, the industrial business trades near 0.10x sales, and the report notes the stock is temporarily unownable for entire mandate categories: no dividend, a trailing loss, no growth story. But the quality axis vetoes conviction: 35 points, a six-condition chain to any five-fold outcome, and even flawless execution of the 2030 plan computes to roughly 4x. The dated checkpoint came on July 30, and the release was the second consecutive improving quarter — revenue up 13% to €43.5 billion, industrial free cash flow back above zero, guidance reaffirmed — and the shares still drifted from $5.81 to $5.63. Buyer's strikes lift slowly. A grievance without quality is an option, and options get option-sized positions.

A five-check discriminator

What separates the Wix file from the Comtech file generalizes into five checks, each of which a report (or a diligent reader) can run:

1. Name the failure. A real grievance specifies which of the three failures is operating — comprehension, respect, or horizon — and cites the disclosure the market misread. Wix: horizon, with the June 8 filing as evidence. "The market is wrong" with no mechanism is a mood, and moods score 5.
2. Date the pivot. Grievances worth owning come with a calendar entry and a threshold: Wix's August 4 print, Kuaishou's spin-off terms, TSMC's 66%/62% margin lines. A thesis that can never be falsified by a specific release will drift for years.
3. Compound the bull case. Run the report's own optimistic scenario to its horizon. Comtech's best case compounds to 3.7x against the 5x bar — when even the bull case misses, cheapness is a ceiling story, and the discount is the market reading the ceiling correctly.
4. Price the price side separately. Sidus screened as a beaten-down micro cap while trading 55% above its own reasonable value band. Cheap-looking and overpriced can be true simultaneously; check the band, never the chart.
5. Demand both axes. Quality without a grievance is fairly priced excellence (TSMC — wait for the price, or accept fair returns). A grievance without quality is optionality (Stellantis — size it like one). The June quadrant already showed high score × low band position is rare; this list shows why the perception pair must confirm it.

Two baskets, on the record

Method, so the future can grade us: membership is mechanical from the 2026-08-06 snapshot (valuation position ≤ 0.33; perception ≥ 7 versus ≤ 4), which is why TSMC sits in the reason basket over our own commentary. Several members — Comtech, Sidus, Legend Biotech, ACM Research among them — entered the cheap band by falling, so this is a fresh-start forward test from today's prices, with the crash already behind them and no credit claimed for it. Prices are our library's live feed (refreshed roughly every 8 hours); an asterisk marks names where no live price passes our currency gate and the report snapshot price stands in. The bet: equal-weight, local-currency price return, dividends excluded, graded on 2027-08-06 alongside our quarterly index. We put 70% on the grievance basket finishing ahead of the reason basket — our subjective probability, and if quality-plus-documented-grievance carries no forward information, this is where it fails in public.

The grievance basket — cheap with a documented case:

The reason basket — cheap, with the discount explained:

The live ranking behind all of this is on the Growth Board, machine-sorted for opportunity with ?sort=val; the scoring standards are on the methodology page. Everything above is a quantified reading of research-report signals with stated time stamps, and report views age; none of it is investment advice.