Variant Perception
Where We Disagree With the Market
Our sharpest disagreement: the market is paying an Adjusted-EBITDA price for a GAAP-earnings business. The +147% rally from the April 2025 $16.54 low to $42.10 — $2 above the $40 ceiling of the announced Dutch tender — has compressed the discount-to-book setup from 0.46x P/B to roughly 1.2x. The implied valuation now requires $145M of through-cycle EBITDA, not the $15.8M FY25 GAAP operating income or the $55-60M three-year GAAP average. The market also reads Education Solutions as a cyclical ESSER-cliff bottom, while Stride — the strongest digital K-12 winner — does not name Scholastic as a competitor in its 10-K, a disclosure consistent with structural exit, not cycle recovery. Every disagreement below resolves inside a single filing window: the FY2026 10-K, expected late July 2026, which carries the FY26 final addback total, the 9 Story goodwill impairment test, the Q4 Children's-segment margin, and the post-tender share count.
Variant Strength (0-100)
Consensus Clarity (0-100)
Evidence Strength (0-100)
Time to Resolution (weeks)
The scorecard reads as a high-evidence, decision-relevant variant view with a single, dated resolution window. Variant strength is held below 80 because two of the four disagreements (Education structural exit, 9 Story impairment) require corroborating disclosure that does not yet exist. Consensus clarity is held at 62 because no syndicated sell-side coverage was staged — the consensus reads are inferred from price action, the tender mechanic, peer multiples, and management's own guide. The disclosures named are mandatory, not voluntary, which is what makes the short resolution window credible.
Variant Perception Scorecard
The scorecard isolates four variables where the market-implied assumption and the report evidence point in different directions. All four resolve inside an eight-week window starting late July 2026. That concentration of resolution is the single most important fact for sizing this disagreement: a PM is not waiting on multi-quarter trend confirmation, only on one filing.
Consensus Map
We agree with consensus on issue 6 (Children's-segment durability) — the 12-14% margin band held through schools physically closing, a stimulus boom, and a stimulus cliff. The variant view is not on the franchise itself but on the denominator being applied to it (issue 1), the tails that drag it down (issue 2), and the mechanical price-support that has carried the stock past where the discount-to-book trade rewards continued buybacks (issue 3).
No syndicated analyst coverage, target-price distribution, or short-seller report was retrieved for this run. Consensus signals above are inferred from price action versus the named tender cap, the peer-multiple gap (SCHL 5.2x vs WLY 8.4x and PSO 7.2x), management's own reaffirmed Adjusted EBITDA guide, and the qualitative tape that has tolerated each weak operating print since December 2025 without a re-rate.
The Disagreement Ledger
Disagreement 1 — the denominator question. Consensus would say that severance, real-estate exits, China reorganisation, and 9 Story integration are economically distinct from operating costs and belong out of the multiple base. The evidence disagrees because the categories repeat — paper-tariff drag, lease exits, M&A integration, and Asia/Canada restructuring all appear in both FY24 and FY25 and again in FY26 YTD at +148% YoY — and because the company's own bonus formula institutionalises the classification. If we are right, the market has to concede that 7-8x Adjusted EBITDA is mechanically the same as 18-22x GAAP, and the implied operating value falls from $1.2-1.4B to roughly $400M. The cleanest disconfirming signal is a single number on a single page in the FY26 10-K: total "one-time items" for the year. Under $15M would force us to drop this view; above $30M for a third consecutive year would mean management has stopped pretending these are transient.
Disagreement 2 — the Education cycle vs structure question. Consensus would say the YoY moderation through FY26 (Q1 -28%, Q2 -8%, Q3 -2%) is the textbook cyclical bottom — the ESSER federal-spend cliff is a one-year math headwind that stops being lapped after FY26, and Ready4Reading positions the segment for the next Science-of-Reading state adoption cycle. The evidence disagrees because Stride — the strongest digital K-12 winner of the ESSER era — does not name Scholastic as a competitor in its 10-K, because share has rotated permanently to phonics-native specialists in a category where Scholastic does not own the channel, and because three product repositionings in five years (PreK On My Way, Scholastic Literacy, the Q3 FY25 strategic review) is a pattern of misses, not a transition. If we are right, the consolidated multiple has no recovery leg to credit; the segment is either divested at a sensible multiple or quietly written down. Two consecutive quarters of flat-or-positive YoY in Q4 FY26 and Q1 FY27 would force us to drop this view; another -15% print on the easier comp would settle it the other way.
Disagreement 3 — the per-share math at 1.1x P/B. Consensus would say the tender is unambiguously accretive — $200M removes ~5M shares (~20% of remaining float), and the residual $100M of authorization buys further per-share compounding at sub-1.2x book. The evidence disagrees because book value per share has been flat at $33-37 for eight years (buybacks have offset retained losses rather than created equity), because the discount-to-book that powered FY24-FY25 buybacks at $25 vs $34 book has compressed by 65% since April 2025, and because the controlling-holder participation in the tender is the most testable own-view signal on the file. If we are right, the buyback engine has to either pause, pivot to special dividends, or be deployed at multiples that destroy per-share value — meaning the Long-Term Thesis row 2 (per-share compounding via float retirement) has temporarily stopped working. The disconfirming signal is a token-or-zero controlling-holder tender plus a pause in the remaining $100M of open-market authorization; the confirming signal is the Estate tendering a material block at $40 and management resuming open-market buybacks immediately at $42+.
Disagreement 4 — 9 Story goodwill as a probable, not tail, binary. Consensus would say the $336.5M of carrying value reflects a deal price set only two years ago and that Q2 FY26 commentary on a turnaround in production greenlights is enough to defer concern. The evidence disagrees because the segment ran a $12.1M operating loss in its first full year, because the FY26 MD&A explicitly cites continued greenlight delays, and because the impairment-test discount rate applies the cash flow you have, not the cash flow you hope is coming. If we are right, the FY26 10-K shows a $50-100M write-down that arrives in the same filing window as the CEO contract decision — the worst possible moment for the controlling structure to be defending capital-allocation judgement. The cleanest disconfirming signal is a clean impairment test paired with named series and dollar-quantified streaming commitments; the confirming signal is any impairment line or a widening segment loss.
Evidence That Changes the Odds
How This Gets Resolved
Every signal above is observable in a single SEC filing or 8-K disclosure window. There is no "watch how the business performs over the next year" signal in the list — that is the deliberate point. The FY26 10-K alone resolves four of the seven; the Schedule 13E-4 and CEO contract decision land in the same window; only the Q1 FY27 segment print extends beyond the same eight-week resolution band.
The single highest-conviction disagreement is #1 — through-cycle earning power. It is the only variant view that re-prices the entire equity on a denominator change rather than a single segment outcome, and it has the cleanest disconfirming signal: one number on one page of the FY26 10-K reconciliation table.
What Would Make Us Wrong
The clearest way we are wrong is if FY26 turns out to be the actual cycle-bottom year and the addback line falls toward zero in FY27 of its own accord. The FY26 partial-year sale-leaseback rent drag of $14M is genuinely cycle-specific and will not repeat. If no new restructuring is initiated, if 9 Story integration costs anniversary out cleanly, and if Education Solutions doesn't trigger a fourth "repositioning" charge, the FY27 addback could legitimately compress to $10-15M. In that world, the three-year addback recurrence reads as a transition cost of a managed-down trough, not as institutionalised cost classification, and the SOTP framework that applies 7-8x EV/EBITDA to the Children's segment becomes the right way to value the equity. The "Move to Lean Long" trigger Stan named — addback line below ~$15M, no Entertainment goodwill impairment, Education Solutions flat or better — is exactly the package that would close out our variant view in a single filing.
A second way we are wrong is on Education Solutions. The Stride non-naming is a powerful piece of evidence but it is a single competitor's 10-K language, not a market-share dataset. State Science-of-Reading adoption cycles are large and lumpy, and a single Ready4Reading win in a state like Texas or Florida in FY27 would force a re-read of the structural-exit thesis. Magazines+ classroom-magazine subscriptions could also reset higher if classroom-print teacher demand stabilises post-pandemic. Two consecutive quarters of flat-or-positive YoY in Q4 FY26 and Q1 FY27 would mean the moderation is genuinely a cycle bottom, not a stall before further decline — and would dissolve disagreement 2.
A third way we are wrong is on the per-share math. If FY26 Adjusted EBITDA prints near the top of the $146-$156M guide, FY27 steps up cleanly, and the tariff pass-through fully holds the Children's margin band, book value per share could climb materially over the next two years — making continued buybacks at $42 less dilutive than the static 1.1x P/B implies. The judgement on disagreement 3 is partly a forward-book-value call, and if operating margins recover faster than the current data suggests, we will look like we under-credited the compounding mechanism for too long.
A fourth way we are wrong is on 9 Story. The impairment-test framework relies on assumptions about future cash flows, terminal value, and discount rate that management controls within auditor oversight. If Brown Bag Films can demonstrate dollar-quantified production commitments from major streaming buyers — beyond the qualitative "three premium animated series in production" language — the test can legitimately clear even on a segment that posted a FY25 operating loss. We are explicitly assigning Medium (not High) confidence to disagreement 4 for this reason.
The first thing to watch is the FY26 10-K "one-time items" reconciliation total, expected in late July 2026.