Cairn · Learning Loop

LAYERS OF THE MOVE:

1. Market context (1.28% S&P gain, VIX falling): The broad market moved up slightly and risk-off sentiment eased. This provided *no tailwind* to BAH and actually worked against it — when risk comes back, growth-oriented names should rally, but BAH did the opposite.

2. Company-specific catalyst (May 22, FY26 earnings release): BAH reported FY26 revenue of $11.2B (−6.4% YoY) and set cautious FY27 guidance of $11.2–$11.7B (0–4% growth, midpoint ~2%). The headline was *not* the earnings — it was CEO Rozanski stating "reacceleration of our business will take longer than we expected." This is exactly the crux we named: the FY26 gross-margin reclassification question turned out to mask *structural* revenue deceleration, not a benign accounting reset. Our thesis hinged on whether $87 or $70 was the right anchor; the guided slowdown pointed toward $70.

3. Sector/macro-narrative headwind (political uncertainty + DOGE headwind): Between June 7–21, defense contractors faced a two-layer sentiment reversal. First, <cite index="18-17,18-18">the DOGE cost-efficiency initiative aimed to reduce federal consulting and technology expenditures, resulting in slower procurement activity, contract reviews, and cancellation of several contracts within Booz Allen's civilian segment</cite>. Second, <cite index="28-2,28-3">political uncertainty over U.S. intelligence leadership, surveillance powers, and national security policy became a meaningful factor; when appointments stall and key laws such as Section 702 hang in the balance, contractors that rely on clear direction and timely approvals can see project pipelines and confidence put under pressure</cite>. BAH is 97–98% dependent on U.S. federal revenue and has heavy civilian exposure — a double hit.

4. Attribution breakdown: - Market: ~10% (broad market was flat/up; no support) - Sector: ~60% (political/DOGE pressure, guidance reset expectations in consulting/IT services) - Idiosyncratic: ~30% (the specific guidance miss and CEO tone that signaled structural slowdown, not just noise)

GRADE AGAINST OUR CALL:

We called BAH a Gem (Quality 1 / Value 23) on June 7 at $79.48. Our thesis was "Quality nibble now, scale on weakness." We explicitly named the FY26 gross-margin print as the crux — whether it held $87 fair value or signaled $70 anchor. The stock fell to $66.36 (−16.51%) in 14 days.

We were wrong on direction and magnitude: - We were right that the margin story mattered; we were wrong that "scaling on weakness" would work because the weakness *accelerated* and was *macro/sentiment-driven*, not company-specific misunderstanding. - The market *repriced* BAH on two frames: (a) revenues ARE slowing durably (not a one-time reset), and (b) political uncertainty + DOGE pressure + Section 702 turmoil create *option value losses* for a pure-play consulting services name with 97–98% federal exposure and slow organic growth.

WHAT WE MISSED:

Our Quality and Value lenses gave us a correct *bottom-up* read — the business is solid, the valuation was reasonable — but we had zero visibility into the top-down macro/sentiment force. In early June 2026: - Defense consulting was under re-rating pressure due to perceived DOGE-driven cost-cutting and intelligence leadership chaos. - Booz Allen specifically was marked for civilian segment pain (procurement delays, contract cancellations). - No positive catalyst was visible for a services/consulting name in a political friction environment.

We treated the May 22 earnings release as static data, when it was actually a *signal* that management's own forward guidance was tepid. We should have read "reacceleration will take longer" as "we are losing confidence," not "volatility = buying opportunity."

THE TIMING & SIZE: The move happened fast (11 trading days saw −16.5%), with the worst days on June 10–11 (broad selloff regime, BAH down −2.09% when S&P down −1.62%, so outpacing downside by 50 bps). The June 18 drop to $66.36 (−4.5% single day) came as the broad market rallied, confirming it was company-specific or sector-specific, not market-driven.

BOTTOM LINE: We nailed the *fundamentals* (the business is solid, margins are real, backlog is strong per-management). We whiffed on the *sentiment/regime*. BAH in June 2026 faced a narrative headwind (consulting = cost-cutting target; intelligence = political chaos; federal procurement = frozen) that our two-lens system cannot capture. We called it a Gem at the wrong time.

### Attribution of the +2.36% move over 14 days:

Market/Sector: ~1.0% (S&P was +1.28%, health insurance sector benefited from managed-care margin recovery narrative that started in April; broad market complacency + risk-on June 12-21 provided tailwinds).

Company-specific: ~1.4% — driven by two catalysts: 1. JPMorgan PT raise (June 8): From $169 to $191, a 13% upside call. This matched our analysis-date price exactly ($190.86), so the call was priced-in immediately with no material follow-through. 2. Illinois Medicaid contract win (June 10): A medium-term growth catalyst, but stock rose only $0.80 that day (0.4%), suggesting modest market impact.

Both catalysts were positive but incremental; the stock drifted sideways June 9-16 ($203 peak on June 9 faded to $195 by close June 21).

### How we fared against our thesis:

What we got right: - We called it a Gem tier at $191 (Quality 11 = Mixed, Value 22 = Modestly Cheap). JPMorgan's $191 target confirmed that valuation. - We named the crux precisely: 2026 state Medicaid rate updates restoring the medical-cost spread. This is exactly what the market is now pricing (margin trough thesis, recovery in 2027). - Our overall read ('get aggressive sub-$165') made sense — the stock stayed above $190, never giving a lower entry.

What we missed or under-weighted: - Membership loss severity: We did not flag that Molina burned through its *entire* 2026 disenrollment expectations in Q1 alone (2% drop). This is idiosyncratic to Molina and raises execution risk on the "margin recovery" story — *if Molina loses members faster than rates rise, the spread math breaks*. - Guidance conservatism: While peers (UNH, ELV) raised 2026 guidance post-Q1, Molina reaffirmed, citing "cost environment remains challenging." This is a yellow flag that execution uncertainty is *real*, not just noise. - Macro narrative blindness: We scored this as "Mixed Quality" and "Modestly Cheap," but the market *doesn't care* about the trough-year thesis if it doubts Molina's ability to *execute* the recovery. The 13% analyst upside from JPM signals the market *wants* to believe it, but the stock's refusal to run hard (peaked +7.38% on June 9, closed +2.36%) shows skepticism.

### Market and sentiment context: - The managed-care sector had a major tailwind: April 2026 saw Medicare Advantage rate guidance that was "better than feared," and the industry consensus shifted to "margin trough is here, recovery in 2027." - But MOH lagged its peers. On June 4, Wells Fargo stuck to a Hold; on June 8, JPM raised PT but kept Neutral. That's not a bullish chorus. - The Strait of Hormuz closure was live macro noise (risk-off early week), but by June 12 the market shifted risk-on. MOH rode the coattails but didn't outperform.

### The 14-day checkpoint: This is too early to kill the thesis, but it shows the thesis is contested, not confirmed. The stock had all the catalysts it needed (analyst upside, contract win, macro shift to "recovery") and still only gained 2.36% vs. the market's 1.28%. That's a weak delta. If the margin recovery story were compelling, MOH should have rallied 5-7% in a +1.3% market.

Market vs. Company:

BPOP +4.03% vs. S&P +1.28% in a 14-day window ending June 21. Market accounted for ~30% of the move; sector tailwind (rate stability, bank rotation) ~20%; idiosyncratic (BPOP fundamentals & analyst revisions) ~50%.

What Drove the Move:

Three factors fired together:

1. Margin recovery thesis lived. Q1 guidance (April 23) signaled NII tracking at high end of 5–7% guidance. By mid-June, analysts had incorporated this into revised estimates—Zacks Consensus up 5.8% over three months. This was the margin acceleration we scored as Q73 and assumed would compound slowly; instead, it announced loudly and analyst revision velocity accelerated.

2. Capital discipline reinforced. $155M buyback, tangible book value up $2.33/share, OpEx guidance cut. This validated the thesis.

3. No bad news. Evertec breach was old (May 15), explicitly immaterial. No other idiosyncratic hits. The coast was clear for bullish re-rating.

Where We Were Right: - Called the company GEM (Quality + Value coherent story); the fundamentals bore out. - Identified margin recovery as the crux; it delivered. - Risk-adjusted entry (starter 1/3 weight) was prudent.

Where We Missed: - Timing: We said 'open here, scale on dip to $145.' The market offered no dip. By the time analysts were revising up (early-mid June), momentum was live and the dip thesis was broken. - Signal gap: We didn't weight that consensus estimates were already in uptrend (+5.8% in 3 months). In a 14-day window, real-time revision momentum can override point-in-time valuation calls.

Magnitude Check:

Our prediction tier (GEM = buy starter, not pound the table) was too cautious on timing but roughly right on direction. The stock didn't hit our upside targets ($180–185 in 6 months), but it moved decisively higher in 2 weeks. A reader who followed our "starter here" call and added on no dip would have outperformed.

The Window: No major company news (Q1 earnings pre-date analysis; CMO/board hires post-date but were positive). Stock fell 7.22% vs. S&P +1.28%—a −8.5pp idiosyncratic miss.

The Real Driver: Structural—not news-driven. <cite index="37-1,37-8,37-12,37-13">TTD's Q1 2026 guidance projects 10% growth, continuing its downward trend. TTD has only experienced deceleration. TTD's downward trend kicked off in 2024. After averaging 26% growth through the first three quarters of 2024, TTD's Q4 growth slowed to 22%, resulting in missing guidance for the first time in 33 quarters</cite>. The market is re-rating this slow-motion train wreck.

Our Crux Was Right, Our Grade Was Wrong: We nailed the key question—does growth stay 20%+ or fall to mid-teens under Amazon pressure? The market is now saying mid-teens or worse. But we graded it GEM (buyer) on the assumption 20%+ holds. It doesn't. <cite index="40-1,40-2,40-3">Advertisers are shifting millions in CTV budgets from The Trade Desk to Amazon's. Key reasons: lower fees, better UX, measurement visibility. One auto brand moved $80M in annual spend by Q1</cite>. Sentiment is deteriorating faster than fundamentals are falling—a classic repricing.

Attribution: Market +1.28%, sector normalizing (not crashing), but TTD −8.5% idiosyncratic. The stock was down ~60% before our analysis; we were catching a knife. The 14-day move was the market continuing to catch it.

Attribution: The stock's flat performance (1.27%) tracks the S&P 500's move (1.28%) almost perfectly. No meaningful idiosyncratic driver: Q1 earnings occurred before the window (May 5), CEO's 10b5-1 share sales on June 15 are mechanical/pre-planned (immaterial signal), and the June 10 annual shareholder meeting was routine. The intra-window spike to $114.65 on June 15 coincided with a broad tech rally (risk-off environment easing, Strait of Hormuz repriced), not company-specific strength.

Analyst Repricing Dominates: The real headwind is not operational—QLYS delivered 10% YoY revenue growth and raised FY2026 guidance to 8–9%—but sentiment. Between May 5 and May 7, eight major firms (Morgan Stanley, Baird, Scotiabank, Piper Sandler, Canaccord, JPMorgan, Truist, Northland) cut price targets, citing 'uncertain long-term growth outlook' and 'growing debate around valuation.' This repricing happened *before* our analysis date (June 7) and persisted unchanged through the window.

Thesis Status: Our crux—'will competitive pressure compress growth toward zero?'—is *exactly* what the street grappled with post-earnings. Analysts downward-revised growth expectations (implicitly: if PANW/CRWD/MSFT pressure Qualys, growth may fall from 6–7% to 4–5%), justifying lower PE multiples. We named this risk and said the stock deserved $125–130 vs. $110 spot *if* growth holds. The market has now priced ~12–15% downside from our $125–130 FV by cutting analyst targets; this is a re-rating, not a refutation of fortress quality.

Macro/Narrative Overlay: The cybersecurity sector is rotating toward mega-cap platforms (PANW +45% YTD) and AI-security specialists. QLYS is solid remediation/vulnerability-focused, but that positioning isn't a growth narrative right now—it's 'steady state.' Even with tech rallying (June 15–19), QLYS didn't re-rate higher because the sentiment regime doesn't reward mid-cap, narrower-focus cybersecurity. This is a macro/sentiment headwind, not fundamentals.

### Market vs. Idiosyncratic

INTU fell −10% during a +1.28% S&P 500 and a recovering tech sector (June 16–19 showed 'tech_rally' regime). Attribution: ~75% idiosyncratic (DIY tax pricing loss + AI threat narrative), ~15% sector (software multiple reset), ~10% macro (small benefit from risk-on).

### The Timing Trap

Our analysis on 2026-06-07 was published *after* the May 20 earnings collapse (−$77, −20%) that revealed 'did not have the overall tax season we expected' and 2% TurboTax growth. By June 7, the stock was already at $297 from $384—we were buying the dip in a deteriorating stock, not recognizing the deterioration was structural.

### What Went Wrong

1. DIY Tax Margin Compression — We treated this as a tax-season cyclicality + IRS Direct File overhang. We missed that Intuit had *lost on price* to competitors and needed to restructure its low-end offering. This signals a durable, not temporary, revenue/margin hit.

2. AI Competitive Threat Underweighted — Goldman Sachs' June 2 downgrade (Neutral→Sell, $519→$276 PT) cited Perplexity Tax, Prime Meridian, Chime Tax as 'credible, scalable' competitors. Our Quality score (98) did not reflect this threat to TurboTax's 25% of revenue and operating income.

3. Mailchimp Softness Ignored — Supposed to hit double-digit growth by FY2026 end; posted YoY decline instead. 7% of revenue, but a signal of growth deceleration across the platform.

4. Workforce Restructuring as Negative Signal — 17% headcount reduction announced mid-window. This is a cost-cutting move that *should have flagged* we were wrong on margin trajectory.

### Market Context

Tech sold off early June (Broadcom caution, memory chip oversupply, geopolitical risk), then rallied June 16–19. INTU did not participate in the recovery—company-specific pain dominated. S&P +1.28% for the window; INTU −10%. Clear divergence.

### The Thesis Grading

We said: GEM, starter now, scale on weakness below $260. Direction: ✓ (it hit $267). Reasoning: ✗ (we assumed the weakness was an opportunity; it was a deterioration). Verdict: Partial. Right direction, wrong reasoning, and we arrived at the conclusion *after* the main event.

## The Move

EPAM fell 21.83% over 14 days to $76.64. The bulk (−11.6%) occurred on June 17–18, coinciding with Accenture's June 18 guidance cut (FY growth narrowed to 3–4% from 3–5%, consulting up only 1%, bookings down 2% YoY), which triggered an 18% crash for ACN and rippled across the entire IT services sector.

## Our Thesis vs. Reality

We published a GEM verdict (Quality 9, Value 10) on June 7, based on: - Mid-teens P/E on a fortress balance sheet with 25% of market cap in net cash - Clean accruals, disciplined buyback at ~1× sales - Crux: gross margin stabilization - Implied upside to $110–125 (12–25%)

Fundamentally, we were right about the fundamentals. Q1 2026 (published May 7) showed 7.6% YoY revenue growth, 14% YoY operating income growth, 18.7% EPS growth, and first gross margin improvement in years—exactly what we identified as the crux. The balance sheet remained fortress-like, the business was real, and the multiple was reasonable.

But the stock fell 22% anyway, ending at $76.64—well below our stated "real size in mid-$80s" entry point.

## Why: Contagion + Macro Narrative

The trigger was Accenture's June 18 guidance cut, which signaled that enterprise clients had become cautious on discretionary tech spending. This spoked the entire IT services complex, including EPAM. Simultaneously, a powerful macro narrative—that agentic AI could compress IT services work 10–15× and make headcount-heavy consulting models obsolete—reshaped sentiment around all legacy IT services firms, regardless of current financials.

EPAM was caught in a sector whipsaw. No EPAM-specific bad news occurred during the window. One analyst (Jefferies) downgraded from Buy to Hold, but only *after* the Accenture shock. The market simply repriced the entire sector on AI disruption fears.

## Attribution

• Market-wide: S&P +1.28%, VIX down 11.5%. Broad market was risk-on; EPAM moved *against* it. ~0–5% contribution.
• Sector (IT Services): Accenture guidance miss + AI disruption narrative. The entire consulting/engineering services sector sold off in sympathy. ~60–70% contribution.
• Company-specific: Q1 beat, margin improvement, fortress balance sheet, clean buyback. All positive; zero negative EPAM news in window. ~0–5% contribution.
• Macro/Narrative/Sentiment: The pervasive fear that AI is making IT services headcount models obsolete, supported by on-record voices from inside the tech tent (Microsoft). This was the second-order driver underneath the sector move. ~60–70% contribution (overlapping with sector).

## Where We Failed

We nailed Quality and Value. We correctly identified that the crux was margin stabilization, and margins *did* stabilize (first improvement in years). Our business analysis was sound.

But we were completely blind to the macro/narrative/sentiment re-rating in progress. Our two lenses (Quality and Value) cannot see top-down regime shifts—and June 2026 was a regime moment for IT services. A name could be high Quality AND high Value and still get crushed if a powerful macro narrative is pressing on it. EPAM was—and is—caught in that narrative.

We should have noted (a) the pre-existing AI disruption concerns flagged in our stored headlines (Morningstar: "No-Moat EPAM Systems Is Vulnerable to AI's Potential Disruption"), and (b) the heightened volatility in IT services sentiment. We called it a Gem, but we didn't hedge for the regime.

## Is the Thesis Still Intact?

Yes—but timing and price matter. If EPAM's fundamentals continue to execute (margin recovery, AI revenue ramp to $600M in FY2026, low P/E), the stock could recover as sentiment shifts. But a 22% two-week drop on *no bad news* shows our system had a blind spot to macro pressure.

Market vs. Company vs. Macro Decomposition:

1. Market (60% of move): S&P 500 rose 1.28% over the window; ZTS fell 0.92%. The initial 3-day selloff (June 8–11) was macro-driven (Strait of Hormuz geopolitical risk, VIX elevated at 18.92). ZTS underperformed the S&P by ~2.2%, which is modest for a stock fresh off a 21% earnings crash.

2. Sector (25%): Healthcare outperformed, gaining ~5.5% over the month. ZTS benefited from this rotation in the June 12–21 relief rally. Without sector tailwind, ZTS would have been worse.

3. Idiosyncratic (15%): Litigation filings (June 17–19) and persistent sentiment damage from the May 7 crash. No new operational catalysts, but re-weighting of downside risk kept the stock anchored despite sector strength.

The Real Issue:

We published on June 7, one month after Zoetis stock dropped 21.5%, closing at $87.31 per share on May 7, 2026, following Q1 2026 revenue of $2.3B, flat on an organic operational basis, against a backdrop of intensifying competition and weakening pet owner spending. Our thesis rested on the crux: whether the 2025 slowdown (8.4% → 2.3%) was a blip.

The May 7 earnings answered that question definitively: it was structural. The companion animal segment posted $1.5B in global revenue, down 4%, as key dermatology, parasiticides, and OA pain all declined simultaneously. Revenue has decelerated for four consecutive quarters. This is not cyclical pet-adoption normalization; it's structural competitive loss.

We acknowledged the crux correctly in our write-up but mis-calibrated the magnitude of the risk. We rated it Value 22 (modestly cheap) on the assumption that the margin of safety from EPV would protect us. But that assumes stable—not eroding—cash flow.

What We Got Right: - Quality score (100) is defensible: margins are holding, FCF is clean, balance sheet is strong. - The 'starter position, scale on weakness' playbook was sound in theory. - We didn't call it a slam-dunk; we said 'buy, but not a table-pound.'

What We Got Wrong: - We priced litigation risk at zero. By mid-window, a securities class action was filed against Zoetis for a class period of January 14, 2025 through May 6, 2026, adding tail-risk overhang. - We didn't weight the durability of post-earnings sentiment damage. The stock bounced to $82.20 on June 9 (intraday +4.46%), but litigation and analyst downgrades kept it anchored to $78–$80 for the rest of the window. - Our EPV-based 'downside protection' doesn't scale for *erosion* risk. If Companion Animal is structurally impaired, EPV math breaks.

### Market vs. Company vs. Sector Attribution

Market-wide context: S&P 500 +1.58% over the window, VIX fell from 21.51 to 16.78. Early window (June 5-11) showed broad risk-off regime; late window (June 15-19) saw broad rally, especially in tech. The Strait of Hormuz closure kept macro elevated early, but the early-window risk-off and subsequent rally are both market-wide forces, not idiosyncratic to CRM.

Sector pressure: <cite index="25-5,25-7">Investors focused on unit economics and whether AI-driven consumption could pay for itself, questioning whether AI agents could hit the core subscription model</cite>. This is sector-wide SaaS/AI anxiety, not CRM-specific. The broader AI-to-SaaS transition debate (seat compression, consumption-model pivot) loomed over all enterprise software.

Company-specific drivers—the real story:

1. May 18 Bank of America Underperform & $160 target: <cite index="21-2,21-4,21-5">BofA reinstated coverage with a sell-equivalent rating and $160 price target, citing concerns about slowdown in new customer growth, weakening pricing power, and underwhelming monetization of Agentforce</cite>. This hit *during* our analysis window's run-up (early June), creating headwind momentum.

2. Q1 FY27 earnings (May 27): Reported *after* our June 5 analysis but *within the window*. Results were strong on headline: <cite index="6-11,6-12">EPS of $3.88 beat estimates by 23.96%, revenue of $11.13B exceeded forecasts up 13% Y/Y, non-GAAP margin at 34.8% (up 250bps)</cite>. Stock rose modestly in after-hours (+0.42%) but the market's reaction betrayed deeper doubt: <cite index="28-22">current remaining performance obligation (cRPO) grew only 14%, a step down from the 16% pace late last year</cite>, and <cite index="28-19">more than 50% of bookings came from existing customers</cite>—not enough new logo growth. This planted a seed of deceleration that the market would water for the next two weeks.

3. June 15 Fin acquisition ($3.6B): <cite index="17-1,17-2">Salesforce announced acquisition of Fin for approximately $3.6 billion</cite>. The move was strategic (bolstering Agentforce), but the market read it as defensive—CRM is *spending* to acquire AI capability rather than organically growing it. <cite index="18-14">Analyst sentiment turned more cautious amid concerns about softer near-term demand, subdued enterprise renewals, and limited immediate AI revenue impact</cite>. The deal, intended to accelerate adoption, instead highlighted that organic Agentforce growth was slowing (already signaled by Q1's 50% upsell-focused bookings).

4. June 18 analyst moves: <cite index="25-20,25-21">Monness Crespi Hardt upgraded CRM to Buy from Neutral with a $200 target, citing margins, cash flow, buybacks, and gains from Agentforce and Data 360</cite>. But this late upgrade arrived *after* the stock had already fallen 18% from our June 5 entry—it was a bottom-fisher's signal, not a reversal driver.

### Why We Were Wrong

Our thesis rested on the crux: "Whether CRM's low-double-digit growth and margin expansion are durable through the AI transition." We answered 'yes' based on Q86 (strong business quality), the operating margin transformation to 21.5% GAAP, and FCF tripling to $14.4B. The market's answer was 'not yet durable.'

The key miss: We under-weighted the monetization risk in Agentforce.

• We saw $1.2B Agentforce ARR and 205% Y/Y growth and called that proof of the thesis.
• The market saw 50% of that growth coming from upsells (existing customers), not new logos, and smelled seat compression and pricing model disruption.
• <cite index="27-16,27-17">CRM faces persistent pricing pressures and uncertainty around transitioning from per-seat to consumption-based models; AI ARR growth to $1.2B is notable but remains small relative to CRM's total revenue</cite>.

Our DCF math assumed the margin expansion and growth deceleration we cited would hold in a "hybrid" model (seats + consumption). But the 14% cRPO growth—below prior 16% and below our low-double-digit assumption—signaled that the *math was breaking*. We did not have real-time Q1 cRPO data on June 5, but we had Q4 proxy signals. We missed the forward-looking deceleration signal embedded in the multiple (P/E 17.57 by late June, vs. our calculated mid-$260s fair value) before the market priced it in.

### The Crux Verdict

Our crux asked: "Is growth durable through the AI transition?" The window revealed: growth is *resuming* slightly (13% in Q1), but not durable at the rate and mix we assumed. Premium SKU adoption up 300% Q-o-Q is bullish; but if it's all upsells and not new logos, the long-term durability is in question. The market penalized the ambiguity by repricing from $186 to $151.78 (-18.25%).

### Attribution Breakdown

• Market (~25–30%): S&P rally late-window and early risk-off tone affected sector. Macro narrative (Strait closure, AI hype/skepticism swings) moved all enterprise software.
• Sector (~35–40%): SaaS/AI transition debate hit the entire group. The seat-compression, consumption-model pivot, and "show me the ROI" stance on Agentforce monetization are sector-wide anxieties that happened to hit CRM hardest given its exposure.
• Idiosyncratic (~30–35%): BofA downgrade (May 18), Q1 cRPO deceleration signal (May 27), Fin deal read as defensive (June 15), and the market's collective loss of conviction in the Agentforce ramp are all CRM-specific.

Market context: S&P 500 +1.58%, VIX eased from 21.5 to 16.78 (risk-on backdrop). CHKP fell -9.93% and underperformed the tech rally — entirely stock-specific.

Company-specific drivers: - Q1 earnings beat EPS but missed revenue by 0.68%; management cut FY2026 guidance due to firewall product weakness and go-to-market disruption not yet resolved. - Operating margins at 40% in Q1 (vs. our $30% stabilization thesis), still in decline and structurally pressured by 6% firewall growth vs. peers' double-digit expansion. - Multiple analyst downgrades: BMO ($210→$135), Citi ($190→$125), BofA ($260→$120). Street targets collapsed 30–50% in May–June. - June 8 VPN security vulnerability disclosure with confirmed active exploitation, triggering reputational/execution doubt and customer friction concerns at a time confidence was already shaken. - Insider selling signal (CEO Chelouche Form 4, June 2) at $139 added to negative sentiment.

Sector backdrop: Cybersecurity leaders (PANW +45.56% YTD) were surging on AI security narrative and platform consolidation. Check Point's firewall legacy, 72% of revenue growing at only 6%, was re-rated downward vs. growth-focused peers.

Attribution: Idiosyncratic ~70%, sector ~20%, market ~10%.

Verdict: We got the direction right (weakness was justified), but we underestimated the magnitude and failed to escalate the margin stabilization risk. Our crux was correct; our downside scenario under-specified what happens if margins don't hold — the street repriced it 30–50%, not 15–25%.

Price action: $73.56 → $73.44 (−0.16%), well within intra-window noise (+7.67% high, −6.89% low). S&P 500 −0.29%, no strong directional pressure.

Earnings/catalyst timing: Q2 earnings reported May 5 (before window); next catalyst is Q3 guidance (likely August). The 14-day window (6/9–6/23) was inter-earnings silence on SWKS fundamentals.

Sector/macro drivers: Market regimes shifted from risk-off (6/9–6/11) → rally (6/12–6/22) → RF/semiconductor pressure late (6/23 HBM slowdown news). <cite index="7-3">SWKS dropped 5.2% after the HBM4 report on 6/23</cite>, but this was post-close, almost entirely after our window.

Our crux (Apple content stability): <cite index="1-5">Company secured multi-generational Android OEM design win expected to generate $1B+ through 2030</cite> (disclosed May 5). This partially offsets the known iPhone 17 20–25% content loss but does NOT reset the Apple risk. No new Apple updates in June 9–23 window.

Attribution: ~50% market/sector macro, ~35% sector rotation (HBM), ~15% company-specific (largely dormant).

Market vs. stock attribution:

• Macro (55–65%): The window was dominated by geopolitical de-escalation (Strait of Hormuz peace deal, oil prices falling to <$74/bbl), which sparked a risk-on rally mid-window (June 12–21). S&P +1.28% masked a risk-off start (June 8–11). The late-window tech/complacency rally aligned perfectly with OPCH's strongest days (June 15–19), suggesting the stock rode the coattails of a sector rotation into defensives and a broader sentiment shift.

• Sector (25–30%): Healthcare benefited from a mid-June shift in investor positioning. The daily regime moved from "risk_off" (June 8–11) to "strong_up/tech_rally" (June 17–19), but healthcare (XLV) lagged tech. OPCH, as a healthcare services name, likely participated in dividend/defensive rotation as oil prices fell and geopolitical risk eased. No sector-specific catalyst in this window; rather, healthcare was a tactical beneficiary of broader macro relief.

• Idiosyncratic (5–15%): Minimal. No company news, earnings, guidance revision, or analyst action occurred during June 7–21. Goldman Sachs' Neutral initiation (early June) with $22 target and the prior week's downgrade cascade (late May) created a floor/consolidation sentiment, but no new fundamental catalyst tested our "gross margin stabilization" crux. The stock tested our predicted floor near EPV but didn't deliver clarity on whether margins would stabilize or bleed further.

Grade against our call:

Our thesis ("Starter position—scale in on weakness") was that the stock at $20.51 ≈ EPV floor ($19.79), so a buyer on dips. The stock did dip to $20.08 early (June 8), then rallied to $21.99 by end-of-window, handing us a +7.2% return.

• Positive: Direction was correct (up not down). The stock didn't break the $19.79 floor—our floor held.
• Negative: The move had almost nothing to do with our crux. We said "scale in on weakness" because *if margins stabilize, there's upside from a 0.7x multiple re-rate*. The market didn't test that thesis—it tested the stock via macro winds and sector rotation. The move is hollow: we got lucky on timing, not on thesis validation.
• The risk: If macro sentiment reverses (geopolitical re-escalation, oil spikes, tech corrects), the stock could re-test $20 or lower. Our BOUNCE call assumed we had an entry near EPV; we did. But the absence of company-specific good news means there's no new conviction in the margin story—the stock is still a *crux-dependent call*, and the crux is unresolved.

Macro pressure analysis: The prevailing force was the Iran/Hormuz resolution, which eased energy costs and geopolitical tail-risk, sparking a risk-on rotation (esp. June 12 onward). This lifted defensive, lower-volatility healthcare names like OPCH as an income/quality play while big tech rallied. The VIX drop from 18.92 to 16.78 signals risk appetite returning; OPCH's low beta (0.65 per Yahoo data) means it trails in absolute returns but benefits from "search for yield" when macro volatility declines.

Market vs. idiosyncratic: CAG +1.46% vs. SPX +1.28%. The outperformance is trivial and entirely driven by the 6/12 small-cap rally (stock hit $13.74); the rest of the window it drifted lower. By 6/21 it closed at $13.20, only 1.5% above entry, with max drawdown of -1.31%—i.e., no real move.

Our thesis vs. reality: We called BOUNCE tier on the logic that CAG is cheap (Value +48) but mediocre (Quality -22), and the crux was whether the ~$1.3B FCF stream stabilizes to deleverage the $8B debt. This was not a turnaround call—we explicitly said "don't need a turnaround, just 'not a disaster'." The market should have offered a bounce if FCF held.

What broke: (1) FCF is already declining. Nine months FY26 operating cash flow was $896M vs. $1.35B prior year, a 33% decline. (2) Demand is softening—Deutsche Bank on 6/18 explicitly cited "lackluster" consumption in scanned retail. (3) Analyst downgrades (Deutsche Bank $14→$12, Morgan Stanley $15→$13, BofA also down) did *not* create a panic selloff, meaning the market had already priced much of the pain; there was no capitulation bounce to buy into.

Attribution: ~60% market/sector (broad SPX strength carried CAG), ~35% company-specific (analyst cuts, FCF miss, demand weakness), ~5% macro narrative (consumer staples under pressure, but CAG didn't find a bid even as risk-on sentiment lifted peers).

Why we missed: We anchored on valuation (21% FCF yield, 11x EV/FCF) and insider buying as proof-of-life, but we under-weighted the forward deterioration in the *source* of that FCF (cash margin compression, volume loss). The thesis assumes the cash is "real and stable"; the data says it's already eroding. A 33% YoY cash flow decline in 9M is not a floor—it's a warning.

Market vs. Stock Move: - S&P 500 flat (−0.29%), Tech sector +4.52%, yet MNDY fell 15.88% - Even during the 'tech_rally' regime (Jun 17–19), MNDY kept declining - This is overwhelmingly idiosyncratic/sentiment-driven, not market-driven

What We Got Right: - Thesis called for downside: "wait for a dip, crux is whether we get a drawdown into the $60s" - Stock fell from $83.23 to $70.01, confirming directional risk - Our Quality score (100) was correct: Q1 2026 beat estimates with $351M revenue (+24% YoY), strong margins, FCF upside

What We Got Wrong: - We expected a pullback tied to execution risk or guidance miss - Instead, MNDY fell despite raising full-year guidance and beating Q1 EPS by 41.98% - The fall was driven entirely by macro narrative: "AI eats software" and "seat compression" fears are crushing the entire traditional SaaS category - Our Value score (−18 / fairly valued at $83) looked right on fundamentals, but sentiment doesn't care—the market is re-rating the entire category lower

Attribution: - Market: ~2% (S&P nearly flat) - Sector: ~3% (Tech rallied while MNDY fell) - Idiosyncratic/Sentiment: ~95% (pure re-rating on AI disruption fears)

Macro Pressure (Sentiment): - Direction: Negative (hard) - Strength: 80/100 (overwhelming) - Note: "AI eats software" narrative is repricing all SaaS work-management plays; MNDY hit by "seat compression" fears despite strong fundamentals

Market vs. Sector vs. Stock: - Market (S&P −0.29%, VIX flat): ~10% headwind. Strait of Hormuz closure early week created macro risk-off. - Sector (Retail): ~25% headwind. Q1 earnings strong but Q2 guidance downgraded; cautious tone on consumer spending. - BBY idiosyncratic: ~65% of move. Earnings beat (May 28, EPS $1.28 vs. est. $1.23), comp sales +2.0% (first turnaround in 8+ qtrs), Meta Lab partnership announced June 8, CFO Bilunas stepping down July 31 (initially −3% after-hours, recovered), analyst upgrades (JPM $84, Morgan Stanley $80, Telsey $90). Stock +5.84% peak (June 12, post-Meta/analyst surge), −3.47% trough (June 17), closed flat +0.04%.

Grade Against Our Thesis: - We said: Pass at $75.12; Quality +20 (disciplined, harvester), Value -11 (fairly priced). Work a bid at $62. Crux: revenue stabilization. - What happened: Revenue *did* stabilize — comps +2.0%, the inflection we named. But we called it a PASS / wait-for-dislocation, and the market repriced upward on the *first sign* of stabilization + new narratives (Meta Lab, CEO clarity). - Outcome: Stock flat, so no loss. But the window showed *upside sentiment* emerge that our system didn't flag. We were right on valuation (still fair), but too cautious on the re-rating velocity and catalyst visibility.

What We Missed: - Retail media upside (Meta Lab, Best Buy Ads). Market sees this as a real margin + revenue lever; we didn't weight it. Analysts repriced on earnings + partnership, not just hardware metrics. - CEO succession as positive (Bonfig is a 27-year e-commerce/supply-chain insider), not a risk. The CFO departure (announced June 22) was briefly negative but recovered—we didn't anticipate the positive leadership narrative. - Sentiment/narrative re-rating. Our Q/V system is bottom-up; it doesn't see "when a stock becomes a story." Meta Lab + comps stabilization + new CEO = a narrative re-rating force that we don't measure.

Market vs. Stock: S&P 500 was flat (-0.29%); NXT outperformed on sector tailwinds from Hormuz relief. Thesis vs. Reality: We correctly identified NXT as Quality (29) + Fairly Valued (-53) at $119, with a $110–$125 fair band. The stock dipped to $111.95 on Jun 10–11 (within our band, confirming our "wait for a dip" advice) but then rallied hard Jun 12–18 on: (a) Q4 FY2026 earnings beat on EBITDA/revenue (Jun 16, stock +15.5%), (b) Geopolitical relief as Hormuz tension eased, driving solar sector re-rating, (c) Zimmermann PV-Steel acquisition (Jun 22, $378M, adds €300M revenue/€45M EBITDA, announced at Intersolar). Stock peaked at $129.07 (Jun 22), then reversed -6.53% on Jun 23. Closed window at $120.64 (+1.15%), near our fair-value upper bound. Our miss: We didn't flag the timing risk of near-term catalysts (earnings on Jun 16, acquisition on Jun 22) that would temporarily push sentiment above fair value, nor did we weight the narrative upside of "platform expansion into energy storage/AI data center" (Prevalon deal + Zimmermann). Attribution: Geopolitical sentiment shift (Hormuz relief) + company-specific earnings beat + M&A narrative drove the spike; broad tech selloff on Jun 23 reversed it.

Market Context (Attribution): - S&P 500: -0.29% (nearly flat; no broad headwind) - VIX: 19.87 → 19.49 (risk-off eased slightly) - Regime: Started risk-off (6/9–6/11), shifted to strong risk-on by 6/12 onward (small-cap rally, tech rally, complacency) - Macro catalyst: Strait of Hormuz ceasefire/reopening (~6/15–6/18 window) de-risked energy but didn't crash FLS—it became a *confidence restorer* for industrial capex cycle

What Moved FLS: 1. Narrative inflection (50–60% of move): Starboard Value activism + management reaffirmation of margin/EPS targets in May created a *quality + operational execution* story. By early June, analysts and institutions rewiring their models around AI power/nuclear exposure. This was top-down sentiment shift, not bottom-up earnings surprise. 2. Hormuz ceasefire (20–25% of move): By mid-June, the Strait reopening removed a major tail risk (potential demand destruction, supply-chain chaos) that had weighed on industrial cyclicals. FLS beneficiary from both higher energy capex (from Hormuz premium) *and* reduced execution risk (certainty returned to project timelines). 3. Momentum/technicals (15–20% of move): Once the stock broke above $78, it held and ran; small-cap rally 6/12–6/18 swept FLS up. Analyst targets of $87–$91 (vs. our $69 6-month) were actively cited.

Our Analytical Failure: - Quality score was right (81/Strong): We correctly diagnosed a clean, margin-expanding operator. This held true and became THE reason it re-rated higher. - Value score was right at the math level (-58/Rich at $76): At $76 with 4.1× consensus 2026 EPS of $4.11, forward P/E ~18.5x — genuinely not cheap by historical standards. - But we missed the *narrative override*: In a regime where AI infrastructure is a dominant meta-theme (data centers → power generation → industrial equipment), quality *itself* gets re-rated upward. We priced in 3–5% annual growth; the market repriced for 8–10% on the back of AI capex cycles + nuclear renaissance narrative. - Timing error: We said "wait for a dip." The dip came (day 1: $72.86), but it was shallow and brief. We were looking for a 15–25% flush; the market never gave it because institutional money was already *leaning into* Flowserve on Starboard's push. By the time real-money investors realized the Starboard reaffirmation was credible, the dip window had closed. - Missed the activist catalyst timing: Starboard's late May letter and management's May 29 response landed RIGHT BEFORE our analysis date (6/9). We saw the headlines but didn't weight them heavily enough as a *sentiment inflection point* that would drive re-rating.

Market vs. Sector vs. Company Attribution: - Broad market: Essentially neutral (-0.29% S&P, but small-cap bias helped industrials) - Sector: Industrial equipment/flow-control benefited from Hormuz-driven energy confidence + AI infrastructure push - Company-specific: Management credibility (Starboard response), margin execution (Q1 beat on EBITDA despite revenue miss), strong analyst target consensus (most at $87–$91)

The Crux We Named vs. Reality: Our crux was: "Whether you get a cyclical or execution-driven pullback into the high $50s." We got neither. Instead, we got *negative catalysts removed* (Hormuz risk) + *positive narrative confirmation* (AI, nuclear, Starboard) that together re-rated the margin story higher, not lower. The stock found support at $72.86 and never looked back.

Market vs Stock Breakdown:

The S&P 500 fell -0.29% over the window while TOL rose +4.73%, so TOL outperformed the broad market by ~5% on an idiosyncratic basis. The Strait of Hormuz macro headwind was active but neutral/weak (VIX stable 19.87→19.49). Market regime was mixed (risk-off first two days, then tech/small-cap rally Jun 12–18). This move was heavily stock-specific and sector-positive.

What drove the move:

1. Keefe Bruyette upgrade (June 8–9, published at our analysis date): $158→$161 PT, +150bp rating lift from Market Perform to Outperform. This happened *on* our analysis date. The upgrade thesis centered on luxury demand resilience, land inflation hedge, and improved Q2 results.

2. Q2 earnings (May 19, pre-window): Beat consensus (EPS $2.72 vs $2.59), raised full-year guidance (deliveries narrowed to 10,400–10,700), and signed contracts up YoY (+7% units, +8% value). Management cited steady luxury demand and pricing power.

3. 21st Century ROAD Housing Act passage (Jun 22–23): Bill passed overwhelmingly (85-5 Senate, 358-32 House). Market repriced TOL up on regulatory relief (NEPA streamlining, modular home standards, CDBG incentives). TOL stock jumped ~7.5% on passage news as homebuilder sentiment inflated.

4. Trump cancellation threat (Jun 24, post-window close): Technically outside the window but the uncertainty did not hit prices during the 14-day period we assess (through Jun 23).

Attribution: - Idiosyncratic (company-specific): ~60–70%. Analyst upgrade + Q2 earnings beat + guidance raise + community openings (multiple announced Jun 17–23) drove base case. - Sector: ~25–30%. Housing bill passage lifted homebuilder sentiment broadly (LGI Homes +56% in 3 months, per headlines); TOL benefits from luxury-skew positioning vs entry-level peers (Lennar downgraded same day KB upgraded TOL). - Market (macro): ~5–10%. No macro tailwind; S&P flatish. Geopolitical risk (Hormuz) was offset by broader momentum into small/cyclical names during the 12–18 Jun rally.

Grade against our thesis:

We said WATCH (Quality 60 + Value -33) with the crux being: *'Is 2024 margin peak or new normal?'* and the call: 'wait for a dip.'

What happened: - TOL opened at $144.14, dipped to $138.90 on Jun 10 (-3.7%), then rallied +9.6% to $155.67 by Jun 18, closing the window at $150.96. - The stock *did* dip as we said ('wait for a dip'). Then it rallied *beyond* the range we implied as fair value ($115–$144). - But the rally wasn't on margin improvement — it was on analyst re-rating (KB), beat/raise, and sector tailwinds (housing bill). Our thesis was that margins were cyclically rolling over; the market is saying luxury demand + regulatory relief + strong order book = margin support longer than we thought.

What we got right: TOL's Quality fundamentals (21% share buyback, 4 Z-score, real FCF) are holding, and the luxury niche *is* more resilient than we gave it credit for. KB's logic (K-shaped market favors high-end) is sound.

What we underweighted: - Sentiment re-rating: The analyst upgrade and earnings surprise repriced the stock +11% of its 14-day move independently of fundamentals. We had a WATCH rating, implying wait-and-see; market had animal spirits. - Regulatory relief: We didn't model the housing bill as a real margin/growth tailwind for luxury builders. It's modest (NEPA streamlining, modular standards), but market priced it as relief from permitting drag, especially for TOL's big landbank. - Margin stability signal: Q2 guidance raise (10.4–10.7k units) and signed contracts up YoY suggested margins *aren't* rolling over as steeply as our analysis feared. We said 20% haircut from 27.9% GM; the market is saying maybe 5–10%.

Timing/sizing: We nailed direction (quality + wait = later = avoid downside). We missed magnitude (the upside surprise from analyst re-rating + sentiment shift + housing bill = +4.7%, not a dip-then-hold). A WATCH call is defensive; we got run past because the cohort (KB, BTIG, Citi all initiated/upgraded Jun 8–9) drove momentum.

Market & sector layer: The S&P 500 rallied 1.28% over the window (6/7–6/21), but TCBX declined 0.56%, tracing a sharp underperformance. The regime shifted from risk-off (6/8–6/11) to tech-led strength (6/17–6/21). Small-cap regional banks like TCBX have no AI upside narrative and underperformed during a mega-cap/growth rotation.

Company layer: No material catalysts during the window. Q1 earnings were reported 4/23 (outside window). The only 6/18 event was a preferred dividend declaration—mechanical, not price-moving. Keystone integration proceeds on plan (customer conversion deferred to summer 2026); no credit stress, no guidance miss.

Our call: We published a WATCH verdict at $39.21, saying the stock was fairly valued in the high-$30s/low-$40s range, with no edge—entry price, not quality, was the play. We flagged structural 19% CAGR dilution as a per-share compounder cap. Our thesis required a dip to entry; we got oscillation in the $38.99–$40.09 band instead. The stock *behaved exactly as we predicted*, but the call had zero catalyst in the window.

Grade: We were right on direction and magnitude (no surprise down, no surprise up), but the thesis was untested—the window was dominated by macro/sector forces that shifted small-caps lower regardless of TCBX's quality or value metrics.

Thesis accuracy: We pegged fair value at $40–42; stock hit $41.21 intra-window then recovered to $44.55 (+7.97%). Our 'wait for dip' call was sound—the dip occurred on June 10. Our Quality 70 was validated by strong earnings (net income +36%, Adj. EBITDA +23%). Value score −15 (fairly valued) proved accurate as a price anchor.

The miss: We named the crux correctly—ad take-rate expansion vs. slowing GTV—but we did not adequately weight management's May 6 earnings guidance: Q2 ads growth 11–14% and confidence in 4–5% LT take-rate. This directly addressed the crux and signaled the ad acceleration story was alive. Our playbook said 'fair value, not a buyer,' but failed to flag that a sentiment re-rating from "cash flow concern" (May 6) to "profitable growth at scale + buyback" (June) was plausible and likely.

Market vs. stock: S&P gained 1.28%; CART gained 7.97%. ~1.3% from market, ~0.5% from sector, ~6.2% from CART-specific sentiment recovery (post-earnings re-read, management guidance reset, new product announcements, capital allocation shift). The Strait of Hormuz geopolitical pressure was macro noise; retail/logistics was sector-neutral.

What went under-weighted: (1) Forward guidance resets more powerful than we priced; (2) Insider selling read as red flag when it was liquidity diversification; (3) Cash flow decline re-framed as buyback choice, not weakness. By June 14, the narrative had fully inverted from May 6 panic.

Market vs. Stock Context: - S&P 500 +1.28%, SIRI +3.78%; roughly 2.5× market move, so idiosyncratic strength. - Attribution: ~15% market, ~10% sector, ~75% idiosyncratic (index inclusion + Google partnership + narrative reframe).

What We Got Wrong: 1. Our Quality/Value model flagged SIRI as Fairly Valued at $27 and told us to wait for sub-$22. We underestimated the tailwind from mechanical index flows and narrative momentum. The stock's 5% after-hours pop on the June 8 announcement (index inclusion news) was forecastable but outside our scope. 2. We correctly noted podcast upside (+50% YoY growth) but modeled it as a *future* offset to satellite decay. The market immediately repriced it as a current growth driver, moving fair value to ~$28 consensus target. 3. Q1 2026 delivered 1% revenue growth (vs. 4 years of decline) + improved churn signal, but we didn't flag this as a *crux reversal*. A stabilizing subscriber base materially changes the FCF compounding math. 4. Google YouTube audio partnership (announced June 8) validated a new ad-tech moat. Combined with index inclusion, this created a re-rating that closed our entry point in one window.

Why We Missed the Macro Pressure: Our system doesn't scan for structural/mechanical catalysts (index inclusion, rebalance dates) or sentiment rotation speed. In late 2024–2026, both operate at elevated velocity. SIRI was a legitimate pass at $27 on fundamentals—but a 'wait for dip' call loses in a 2-week window where passive flows + positive narrative alignment override bottom-up valuation.

Timing vs. Direction: Our direction call (avoid / wait for dip) was structurally sound—the company IS a declining subscriber base with stable FCF and buybacks. But our *timing* was wrong by ~2 weeks because we didn't account for the index inclusion announcement and its mechanical flow impact. This is a timing miss, not a fundamental miss.

Market vs. Sector vs. Idiosyncratic Attribution: - Market-wide (S&P +1.28%, VIX compression): ~5–7 pts; broad risk-on mood lifted materials. - Sector (specialty chemicals, materials re-rating into risk-on, but selective on narrative quality): ~3–4 pts. - Company-specific (Q1 beat, raised full-year EBITDA guidance, CEO commentary on AI/datacenters, analyst target raises from $40.40 to $47.30): ~6–8 pts.

Our Thesis vs. Reality: Our call on 2026-06-07 was WATCH: Quality 20 (solid) + Value −91 (rich) = Pass at $39.84, revisit at mid-$20s. The crux was: "You can't pay elite-compounder prices for a solid-but-cyclical operator." We valued ESI at $13–15 DCF baseline and were right that it wasn't a traditional compounder. But the market was re-rating ESI as something different: a leveraged play on AI infrastructure capex, not a mature cyclical operator.

What Actually Happened: - Stock: +15.39% in 14 days (+17.27% max run-up, −2.06% max drawdown). S&P 500 only +1.28%. - Q1 2026 earnings came in late April/early May (outside our window but foundational to the move): Record results, 41% sales growth YoY (10% organic), Electronics segment +15% organic on AI/datacenters/high-performance electronics demand. - Full-year EBITDA guidance raised to $665–$685M; analyst consensus price target raised from $40.40 to $47.30. - Macro backdrop: Strait of Hormuz disruption (risk-on sentiment, energy tailwind), but ESI's move was largely company and narrative-specific.

The Gap: We saw "flat revenue, cyclical margins, levered balance sheet" and priced accordingly. Reality: The company was in the early innings of a structural demand acceleration in its highest-margin segment (Electronics) tied to AI infrastructure buildout. The revenue was not flat anymore; the business was inflecting into growth. And the leverage was being deployed to accelerate that growth (Micromax, EFC acquisitions), not a burden. The market re-rated the multiple from "mature cyclical" (~10–12x forward EBITDA) to "AI infrastructure play" (~18–20x forward), and we had no lens to see that re-rating happening. Our system is purely bottom-up (quality of business, price relative to DCF); it cannot price macro/narrative regime tailwinds that re-rate entire buckets of names regardless of intrinsic fundamentals.

Market and Macro Attribution:

• Market (S&P 500) +1.28% vs OGE −1.05%: OGE underperformed by ~230 bps, but in a narrow trading range ($46.62–$47.93). This is not catastrophic relative underperformance — it's consistent with a defensive utility lagging a broad risk-on market.

• Sector/Macro Pressure: The window captured a major narrative shift. Early window (6/8–11): broad risk-off, VIX elevated at 18.9. OGE held up well (+1.42% intraday on 6/10). But 6/12 onward: market rotated hard into risk-on (small-cap rally on 6/12, tech rally 6/15–19). By 6/19, Strait of Hormuz reopening narrative was in full relief-rally mode—falling oil, rising equity sentiment, energy sector outperformance.

• Utilities as contra-cyclical: Utilities typically underperform during commodity-down / risk-on rotations because they are defensive (low beta, high dividend yield). OGE's 3.59% yield became less attractive relative to a rising equity market. The falling oil/energy prices, while fundamentally positive for utility cost structure, were sentiment-negative because they signaled the macro macro-risk had abated and capital could rotate to higher-risk, higher-upside names.

Company-Specific Drivers (or lack thereof):

• No new catalyst in this window. The Google data center contracts, the $7.3B 5-year capex plan, the Q1 earnings miss (due to mild weather), and the reaffirmed 2026 guidance ($2.38–$2.48 EPS) were all announced on April 29, 2026. By June, this story was fully priced.

• June 12 credit facility upgrade: Positive signal (larger revolving lines, extended to 2031), but routine corporate finance — not a moved-the-needle event. Stock was at $47.80 on 6/7, at $47.93 on 6/15 (best of window), then declined to $47.30 by 6/18. The credit facility (signed 6/12) was contemporaneous with the flat-to-down move.

• Regulatory/valuation: Analyst price targets are in the $47–$50 range (post-April), fairly valued to modestly cheap. No analyst downgrades or upgrades surfaced in this window.

Why our thesis was untested:

Our call was "Pass at current price; bid in the low-$40s." We said the stock was a coin flip at $47.80 ($46.77 composite FV) and lacked margin of safety for a utility. The stock finished at $47.30, exactly where our thesis would place it — barely moved, still in the "no position" zone. The move (down 1.05%) was within noise and almost entirely driven by market-wide rotation into cyclicals away from defensives. Our bottom-up Quality×Value thesis was never tested — no earnings surprise, no regulatory shock, no capital-allocation miss, no data center contract win/loss in this window. The macro regime change is what moved the stock, not the company.

Market Context: S&P +1.28% over window; BSY −10.78%. Stock underperformed by ~12%, driven by sector rotation out of SaaS amid AI-agent fears, not by Bentley-specific deterioration.

What We Got Right: Our verdict "wait for a dip, entry is the only lever" was directionally sound. The stock did dip from $32.93 to $30.57 (6.4% down), and recovered toward $29.38 end. We correctly scored Bentley as high-Quality (99), fairly valued (−48 Value), and not a buy at FV $31.48 vs $32.93 price.

What We Missed: (1) We couldn't see the macro narrative—enterprise software sentiment collapse due to fears that AI agents would compress seat-based licensing. (2) Analyst downgrades (Mizuho: Outperform → Neutral on valuation; targets cut $60→$50) happened near our analysis date and created momentum selling independent of company news. (3) We said "dip to mid-$20s" but didn't explain why the market might not *allow* a full dip in a sector-wide panic; sentiment overshoot can prevent fair value realization.

The Attribution: Market/sector accounted for ~80% of the move (broad risk-off early, then software sector contagion from AI fears). Company-specific (analyst moves, modest ARR growth reset) accounts for ~20%. Bentley's fundamentals held; the stock fell on *re-rating of software multiples and narrative risk*, not on deteriorating business.

Market Context: S&P 500 +1.28%, VIX down 11% (risk-on pivot). Early window (Jun 8-12) was risk-off, but Jun 16-19 was strong up/tech rally/complacency. Macro backdrop shifted favorably.

Company Catalyst: On June 18, WMS hosted Investor Day unveiling FY30 targets (>$4B revenue, >8% organic growth, >30% EBITDA margins) and showcasing pure-play water/infrastructure narrative. Analyst consensus price targets raised from ~$139 to $155 post-event. Stock closed at $147.12 by Jun 18 close, a +13% move from $130.15.

Our Error: We scored Quality=58 (strong, correct) and Value=-79 (rich, correct on absolute math). Our thesis was "wait for a dip" anchored on: (a) FY24 peak earnings vs. FY26 rollover, and (b) $1.4B net debt. But we failed to account for: (1) Investor Day as a near-term catalyst that would reframe the narrative, (2) the shift in analyst positioning from "earnings compression" to "long-term secular growth in water/infrastructure," and (3) the macro sentiment toward infrastructure/resilience names during a risk-on environment.

Attribution Breakdown: (1) Market/broad risk-on: ~15%; (2) Sector (water/infrastructure/construction materials beneficiary of IIJA + climate narrative): ~35%; (3) Idiosyncratic (WMS-specific Investor Day, analyst repricing, 2030 vision reframe): ~50%.

What We Missed: Our Quality/Value lenses capture *current* business health and *current* price-to-cash-flow. They do NOT capture (a) narrative inflection points, (b) catalyst timing, or (c) sentiment-driven re-rating events. The stock was rich on FY26 earnings, but investors repriced it upward on the promise of a secular story (pure-play water, infrastructure, climate resilience) backed by near-term IIJA spending and management's explicit 2030 roadmap. That inflection happened in this window and overwhelmed the valuation warning.

Market vs. Company Attribution:

The 14-day move splits cleanly: ~-11% was market headwind (S&P +1.28 vs. ENSG -9.78), leaving ~-10.8% as idiosyncratic. Of that idiosyncratic loss, ~-8.2% came from Hunterbrook's June 8 short report, another ~-3% from Muddy Waters on June 11, and then a recovery of ~+3% mid-window (June 16-19) as sentiment recovered and the buyback announcement on June 15 provided some support. Net: the two short reports account for >85% of the move.

Our Call vs. Reality:

1. Direction & Entry: We said "wait for a dip" and pegged fair value at $147–153 (vs. $170.30 at analysis). ENSG hit $147.13 on June 11—exactly our target. *This is a tactical win.* But the dip was triggered by a company-specific fraud allegation, not the valuation-driven pullback we implied.

2. The Crux Problem: Our crux was "Whether you get a pullback into the $140s; the quality is settled, the entry isn't." We treated pullback as a macro/vol event, not a company-specific risk event. The Hunterbrook report alleged that ENSG inflated its CMS star ratings—a metric CEO Barry Port cited on the Q1 2026 earnings call stating "85% of all of our operations are at 4- or 5-star quality measures"—and that staffing levels and care quality at ENSG facilities were materially worse than publicly represented. This directly contradicts the Quality = 73 (Strong) score we gave, which rested on clean cash conversion, strong operations, and acquisition integration. If the allegations are true, our Quality score was built on fraudulent metrics.

3. Fair Value Vindication & Danger: The fact that ENSG fell *exactly* to our fair-value band is NOT a win—it's a red flag. The market didn't reprrice the stock based on our DCF math; it repriced it because two credible short-sellers made allegations that, if substantiated, would *crater* the DCF (worse quality, worse margins, regulatory risk, litigation).

Why We Missed the Risk:

Our thesis was pure bottom-up: "Is Ensign a good business (Quality) at a good price (Value)?" We scored Quality at +73 based on revenue growth (18.4% YoY in Q1), EPS beats (+21.7%), occupancy records, acquisition discipline, and strong OCF conversion. *None of our metrics flagged that the quality metrics themselves might be gamed.* This is a data/assumption gap: we used management's reported CMS star ratings and occupancy figures as truth, when the entire short case hinges on those being fabricated or misleading.

Market Context (Macro/Sentiment Pressure):

The broader market (S&P +1.28%, VIX 18.92 → 16.78) was in a strong up / complacency regime for most of the window (especially June 16-19 tech rally). The Strait of Hormuz closure was a minor geopolitical overhang early in the window, but healthcare was not a sector under systematic pressure. Medical Care Facilities as a sub-industry showed -0.59% performance within the broader healthcare sector, suggesting the sector itself was flat. ENSG's underperformance is idiosyncratic, not macro-driven.

Analyst & Buyback Context:

According to 5 analysts, the average rating for ENSG stock is "Buy," with a 12-month stock price target of $220.4, representing a 42.41% increase from the latest price. Wall Street remained constructive despite the short reports, and the June 15 $100M buyback expansion signaled management confidence. The stock partially recovered (from -16.86% low to -9.78% at window close), suggesting some conviction that the allegations are contestable or overblown.

Missed Opportunity & System Gap:

We had a clean entry thesis (buy on dip to $140s) but no scandal/litigation/fraud monitoring. We did not screen for: (i) existing short-seller scrutiny, (ii) regulatory compliance flags, (iii) prior allegations of quality gaming in the SNF sector. This is a data gap—our system scores business quality without interrogating the *verifiability* of that quality.

Market context: S&P 500 +1.28%, VIX down 18.92→16.78; risk sentiment recovered strongly from June 12 onward, favoring small-caps. FCFS underperformed broad market by ~57 bps despite favorable macro.

Stock-specific events in window: (1) Q1 earnings beat + raised FY26 guidance announced April 23, pre-window, but may have created lingering positive momentum. (2) Annual meeting on June 9 approved Delaware→Texas reincorporation; finalized June 18. This was a governance event, not a fundamental catalyst.

Critical fact: The Ramsdens acquisition (174 UK pawn stores, £206M / $273M deal) was announced June 23, 2026—two trading days *after* the June 21 window close. It did not drive the observed price action.

Price path: $225.44 (June 7) → max drawdown -8.29% (June 10, during Strait of Hormuz risk-off) → recovered and ended $227.05 (June 21, +0.71%). The stock traded noise within a range and slightly underperformed a rising market.

Attribution: Market sentiment recovery (65%), sector tailwind (20%), idiosyncratic events (15% from reincorporation, minimal impact). Our company thesis was untested and remains valid: FCFS is high Quality, fairly valued, levered; entry requires either a dip or a major capital event. That event occurred June 23, outside the window.

Market Attribution: S&P 500 +1.28% over the window; RRC -6.93% = -8.21% relative. Energy sector underperformed due to Hormuz reopening narrative pushing oil/gas lower. Our thesis was long-dated and company-specific (Quality vs Value), not commodity-price-sensitive.

Sector Context: Natural gas remained flat/weak despite EIA forecasts for $3.34/MMBtu in 2H26 (not bullish). Oil fell on peace deal hopes—the window coincided with a whipsaw: Pakistan announced MoU on June 18 (Strait reopening), Iran re-closed on June 20, then markets priced a reopening narrative anyway. This collapsed energy sentiment across E&P.

Company Specifics: Q1 2026 results (published before window) showed strong revenue (+20.6% YoY), record NGL/gas premiums, and best-in-company balance sheet. RRC had a dividend ex-date on June 12 ($0.10), mechanically neutral. No earnings surprise, no guidance cut, no company news. The 6.93% decline was pure sector beta, not a revaluation of RRC's fundamentals.

Our Thesis Assessment: We said 'wait for a dip'—it dipped. Our Quality (+22) and Value (-11) lenses both held; we were right that $39.10 lacked margin of safety. BUT we got the timing and catalyst wrong. We expected a patient, fundamental reassessment (earnings miss, commodity cycle shift). Instead, a 14-day macro panic sold the entire sector on Strait-reopening hope. The dip is real but not because of company-specific bad news—RRC's business quality is unchanged.

Follow-Through Risk: If oil prices stabilize and Hormuz uncertainty ebbs, the sector could rebound. Our thesis is untested by actual company deterioration.

Market-wide: S&P gained 1.28%; VIX fell from 18.92 to 16.78. Early window (6/8–6/12) was risk-off/geopolitical shock (Strait of Hormuz); MKC rallied 5.9% as defensive hedge despite broad weakness. Mid-window (6/15+) reversed to tech-led complacency; MKC sold off as flows rotated out of staples into growth.

Sector: Consumer Staples is structurally challenged—weak revenue growth (2% CAGR), margin-constrained, and ignored in favor of AI. MKC is a poster child: 2% organic growth, flat NI, but high leverage ($3.9B net debt). The geopolitical bid for defensives was temporary; the regime reasserts staples as a lagging, crowded-out sector.

Company: No catalysts during the window. Q2 earnings scheduled for 6/25 (after window). Analyst revisions trivial: JPM −$1 (still Overweight), UBS −$2 (Neutral). EPS consensus down 0.4% over 30 days—immaterial. Unilever deal and Mexico acquisition are ongoing narratives, not window-specific. Activist Toms Capital stake was known news; no new activism.

Our thesis: We said WATCH, wait for a dip to $40s. Stock rallied to $49.69, then drifted to $46.64 (−1.27% net). We wanted the dip but got noise. Direction was correct (down), magnitude tiny and unconvincing. No dip to $40s materialized, and the move was as much macro rotation (risk-off → risk-on) as company-specific.

Verdict: We were not *wrong*, but not *right* either. The thesis stood untested because MKC had no company catalyst to validate or invalidate the Quality/Value scores. The move was driven entirely by macro/sector flows, which our system doesn't price.

Market vs. Company Attribution:

• Market (60%): S&P 500 +1.28% over the window; CLX +1.76% is only +48 bps outperformance. Early risk-off (Jun 8-11) weighed on CLX; mid-week Fed news and Strait of Hormuz breakthrough sparked risk-on rotation Jun 12+. VIX fell from 18.92 to 16.78, a sign of de-risking and shift into defensive positioning.

• Sector (30%): Consumer staples saw a rotation inflow as equity markets broadened away from mega-cap tech into defensive plays. This is a structural macro/narrative play—not CLX-specific. Staples' low correlation to AI trade became an asset.

• Company (10%): Jun 17 reorganization (Hyder → COO, Barton → Chief Growth Officer) signaled internal efforts to streamline execution. Healthcare wipes launch (Jun 2, pre-window) and CEO search (orderly succession, no shock) were incremental. CFRA had already downcut guidance pre-window.

The Thesis Test:

Our crux was simple: "the recovery is already in the tape; no edge without a high-$70s dip." The stock never dipped below $94.97. It hit +7% intraday Jun 9, then consolidated. By Jun 21, it closed $95.80—only $1.66 above analysis price. We were anchored to waiting for drawdown into $76–$79 range (high $70s). We did not get tested on that thesis—the market didn't offer us the price we said was necessary to move the needle. This is not a miss; it's a thesis that remains unproven because market conditions never triggered our signal.

Why Stock Held Mid-$90s:

Staples were bid into mid-June as both a hedge against Fed rate-hike risk (Warsh signaled Q4 2026 hike possibility) and as a rotation out of crowded tech. CLX's 5.5% dividend yield and defensive profile made it sticky. However, the stock did *not* re-rate on fundamentals: CFRA kept its Hold, margin guidance stayed weak, and the CEO search is unresolved. Buyers were trading staples as a sector play, not CLX-on-its-own merits.

Macro/Narrative Pressure:

The Hormuz closure-to-reopening arc (Jun 15–19) reduced energy risk and shifted sentiment from "geopolitical shock, buy bonds" to "risk-on, buy value and staples as cyclical stabilizer." This benefited CLX passively, not actively.

Macro context: The window spanned two major shocks: (1) early-June Strait of Hormuz geopolitical risk (broad risk-off June 5–11), and (2) June 17 FOMC meeting where new Chair Warsh signaled potential 2026 rate hikes (dot plot moved from 3.4% to 3.8% year-end), killing the easing-cycle narrative that had powered software valuations. The 2-year Treasury yield rose 11 bps in the session.

Company fundamentals held: PAYC beat Q1 earnings (EPS $3.15 vs $2.99 est., revenue $572M vs $565M), reaffirmed 2026 guidance (6–7% revenue growth, 44% EBITDA margin at midpoint), and management reported strong Beti/IWant adoption. No company-specific miss or guidance cut.

The move was sector-driven: Software as a whole repriced down as higher interest rates compress long-duration DCF valuations. PAYC fell 4.1% on June 18 (FOMC day) in line with the broader HR-software cohort (Adobe also −4.1%). The headline was: "Software companies are priced on earnings 5–10 years out; every basis point rate increase cuts present value." This is not a company problem; it's a regime problem.

Attribution breakdown: Market/geopolitical (~25%), Sector/software rate re-rating (~60%), Idiosyncratic/company-specific (~15%). PAYC's thesis (Quality starter, crux is margin durability vs. transition) was never tested—it was overrun by macro.

What we got right: Calling it WATCH, cautious on timing, no firm buy signal. We didn't walk into a landmine.

What we missed entirely: The macro regime. Our Quality and Value scores are bottom-up and blind to Fed dots, sector re-ratings, and narrative shifts. A stock can be high Quality, modestly cheap, and still fall 9% in two weeks if the discount rate applied to its 5–10-year cash flows jumps 30+ bps due to Fed action. We had zero signal for that.

Market vs. Company: S&P 500 up 1.58% and VIX down 31% (risk-on), yet OMDA down 7% from $19 IPO price (−3 ppts idiosyncratic underperformance). This is entirely company/sentiment-driven, not macro. Digital health sector was favorable (Hinge held its pop); OMDA lost the peer battle.

Our Call vs. Reality: We issued a WATCH verdict on June 5 at $17.93 (pre-IPO). Stock IPO'd June 6 at $19 (+5.9%), then declined to $17.64 by June 19 (−7% from IPO, −1.6% from our stated price). The pullback matches our "scale on pullback" language, but the driver was not business news—it was analyst repricing.

Key Catalyst in Window: Evercore ISI cut OMDA price target from $20 to $15 (25% cut) on valuation and scale concerns. This repricing occurred post-IPO and anchored the stock down. We did not anticipate this level of analyst skepticism on an IPO that had strong initial demand.

Operational Thesis: Still intact. Op margin remains −4.6%, gross margin 65.7%, membership growth 51% YoY. No deterioration; the inflection story has not been tested yet (too early in 2026). Revenue growth on track for 25–30%+.

Attribution: Market 10% (small tailwind offset by growth repricing), sector 20% (Hinge stronger), idiosyncratic 70% (IPO pop reversal + analyst downgrade + valuation repricing from $19 to justified $15–18 range).

## Layers

Market (June 5–10): Broad risk-off pressure (VIX 21.5, S&P weak) drove DELL down ~6.5% on June 5–9. This was market-wide, not company-specific, and validated our "dip" thesis. No error here.

Macro shift (June 11–19): On June 17, Trump and Iran signed a peace memorandum; June 18 saw Iran pledge to reopen Strait of Hormuz. This triggered a risk-on reversal, a tech/cyclical re-rating, and capex narrative lift. DELL rallied hardest June 17–18 (+10%+ intraday). Our system has no third lens for this macro/narrative momentum; it was invisible to us.

Company-specific (throughout): We analyzed DELL on June 5, seven days *after* May 28 earnings. That Q1 reported 88% YoY revenue growth, 757% YoY AI server revenue growth, 81% ASP uplift, and a $51.3B backlog. Our playbook said "I like the company more than consensus," but we under-weighted the *confirmation* that earnings provided. The blowout reduced tail risk on the reversion crux; we still treated reversion as the base case. The stock closed +3.84%; our call to "wait for sub-$230" now looks overly pessimistic given the earnings-driven confidence boost and pricing-power visibility.

## What We Got Right

• Correctly flagged overvaluation (DCF $166–179 vs. $394 tape).
• Correctly identified Quality (clean cash, share buybacks, low SBC).
• Correctly predicted a dip (it came June 5–10).

## What We Missed

1. Post-earnings momentum blindness. Earnings data materially reduced reversion risk; we did not fully reprieve our base-case assumptions. A 757% YoY AI server revenue jump is a *structural signal*, not noise. 2. Macro narrative re-rate (Hormuz). The June 17 peace deal and June 18 Strait reopening pledge triggered a capex/growth re-rating that lifted cyclical/tech names. Our system cannot see this. 3. ISG margin stickiness. The 81% ASP uplift and strong gross-margin data suggested AI server margins are *not reverting* like prior hardware cycles. Our crux assumed they would. Evidence now suggests we were wrong on the crux.

Market vs. Stock Attribution:

• S&P 500: +1.58% (positive backdrop)
• BSX: -6.71% (underperformance vs. market = idiosyncratic)
• Medical sector context: BSX lost 17.41% in prior month while Medical sector gained 4.11%, confirming stock-specific damage.

The Crux Broke in April, Not During Our Window:

<cite index="20-3,20-4">Q1 2026 earnings on April 22 showed solid results but included a lowered full-year guidance—organic growth cut to 6.5–8% from prior 10–11%, and adjusted EPS to $3.34–$3.41, citing unanticipated headwinds in EP, WATCHMAN, and Urology segments.</cite> This directly invalidated the thesis's central claim: that "the EP/Farapulse + Watchman product cycle keeps driving the 2025 margin step-up."

<cite index="9-1,9-2">Watchman sales in Q1 totaled $506 million (+19.2% YoY), below expectations, with management guiding only mid-teens growth for 2026.</cite> <cite index="9-14,9-17">Boston Scientific lost more market share than expected in EP despite solid results overall, as competitors like Medtronic, Johnson & Johnson and Abbott expanded their presence.</cite>

The June 5–19 Window:

By our June 5 analysis date, the thesis had been visibly broken for 6 weeks. Our "WATCH" + "starter position" call came *after* the April guidance slash and *after* the street had begun downgrading. The June window captured continued analyst target cuts and sentiment cascade; the stock underperformed a positive-regime S&P 500, confirming that sentiment (not macro) was the pressure.

What We Missed:

We assessed Quality (83) and Value (-15) correctly as *business fundamentals today*, but failed to recognize that the market had already repriced the *forward earnings trajectory*. The "margin inflection" we anchored to was no longer credible to consensus after April guidance. A 6.5–8% organic growth guidance is not a margin-inflection story; it's a low-single-digit compounder story.

Attribution estimate: Market 5%, Sector 15%, Idiosyncratic 80%.

The move: BILL -9.73% in 14 days vs. S&P +1.58%; stock fell from $35.97 to $32.47, hitting $32.39 on June 10 (nearly the 'high-$20s' buy level we named). The early-window price action was driven by a broad market risk-off (June 5–11) overlaid with JPMorgan's June 5 PT cut ($60→$50) and a surgical June 10 Truist downgrade (Buy→Hold, citing slowing revenue growth and rising competition). Leadership departures (Payments GM exiting June 30) and insider selling also signaled management awareness of pressure.

Attribution: Roughly 65% idiosyncratic (downgrade, growth reassessment, exec departure), 25% market risk-off early-window, 10% sentiment/sector rotation lag. The broad market recovered late-window (tech_rally June 17–18), but BILL did not re-rate upward, confirming company-specific weakness.

Our thesis performance: We said WATCH—not a buy at $35.97, buy only at high-$20s. The stock dipped to $32.39, so we got direction right. However, we predicted the dip would be driven by valuation compression and SBC dilution concerns. Instead, it was driven by a *fundamental growth slowdown reassessment*—a change in consensus earnings assumptions, not multiple re-rating. The Truist downgrade (mid-teens vs. prior mid-high-teens growth) was the real driver.

What we missed: (1) JPMorgan's PT cut on June 5 was a leading signal of consensus drift that our Q/V scores didn't capture. (2) The May 26 leadership churn (Payments GM departure June 30) was a harbinger for take-rate/volume risk that we labeled as 'known' but didn't weight into the downgrade risk. (3) Insider selling + downgrade combo is a high-confidence negative signal we didn't monitor in the prior 4 weeks of analyst actions. Our thesis was not *wrong*—the price did dip—but we didn't predict or explain the driver.

The Crux Was Right, the Timing and Grading Were Wrong

• Our core thesis—that MLR is the deciding factor for a health insurer—was correct. Q1 2026 proved the MLR did recover to 70.5% from 75.4% YoY, driven by disciplined pricing and favorable mix/reserve development. That was exactly the 'one-off vs. structural' question we posed.
• But we analyzed it too late and miss-timed the catalyst. Oscar's Q1 earnings were released May 6, 2026. Our analysis date was June 9. By the time we published, Wells Fargo had already upgraded (June 4), the Goldman Sachs conference was imminent (June 8), and the repricing was underway. We had the data; we didn't act on it urgently.
• Valuation vs. narrative. We graded it -71 Value (Rich) and -4 Quality (Mixed), concluding "no asymmetry at $27." But the stock was repricing *because* the MLR inflection de-risked the narrative. By June 8–10, Barclays upgraded to Overweight with a $35 PT, citing "direct leverage to a multi-year multiple re-rating" of the ACA market. The market was saying: this stock is now a 13–15x earnings story, not a cheap cyclical. We were still pricing it as if it needed a dip; it didn't.

Attribution: Primarily idiosyncratic, some sector lift

• S&P 500: -0.29% (flat). VIX stable. No tailwind here.
• Sector (Health Insurance): Not specifically tracked in the data, but health insurance broadly was benefiting from lower policy-risk sentiment and better margin visibility in 2026. However, Oscar was a *single-name beat*, not a sector trade.
• Company: +9.96%. This was driven by (1) Q1 earnings beat and MLR inflection, (2) conference confidence reaffirmation, (3) analyst upgrades *during the window*, and (4) repositioning from value/distressed to growth/inflection narrative.

Market regime: S&P up/neutral through the window; VIX slightly declining. This was mildly favorable sentiment but *not* the reason for the move. OSCR was driven by company specifics, not market breadth.

What we under-weighted

1. Timing urgency: We knew the MLR question was settled (Q1 data was public). We should have graded the *urgency* of the catalyst window, not just the direction. By June 9, the analyst consensus was shifting; we analyzed post facto. 2. Multiple expansion basis: We were fixated on valuation on 2025 trailing earnings (-$443M GAAP loss). We didn't adequately account for the *repricing from loss-making to profit* that Q1 signaled. A company inflecting from $400M loss to $250–450M operating earnings guidance is a multiple-expansion event, not a cheap stock looking for a dip. 3. Narrative momentum: Our brief mentioned "~10%/yr dilution silently raising per-share hurdle," but didn't acknowledge that if MLR and margins inflect, EPS accretion can overwhelm dilution for several years, validating a higher multiple *now*.

Why we were wrong on direction: We said Avoid; it went up 10%. The crux insight was right (MLR was the key); the verdict was wrong (we should have upgraded or been neutral, not AVOID) because we didn't weight the inflection enough or act on the timing.

Market vs. Company: S&P 500 returned −0.29% over the window; REXR fell 3.74%, a 3.45% spread. Sector (industrial REITs) was mixed; the move is ~80% idiosyncratic to REXR. The Catalyst: JPMorgan downgraded REXR to Underweight (June 16) on forward-looking concern: CFO per share growth will be "muted or potentially negative" in 2027–2028. Stock crashed −5.62% on June 17 (the day after the downgrade, in a strong tech rally—so pure company pain). Scotiabank upgraded it the next day (June 18) on "valuation aligned" narrative, reversing some losses, but the mood stayed cautious.\n\nOur Verdict: Right Direction, Incomplete Diagnosis. We said AVOID; the stock fell. We named the crux (dilution hollows per-share economics) correctly—JPMorgan independently reached the same conclusion, just forward-looking. But we did not flag the risk that Q1 2026 earnings (released April 23) would confirm the per-share decline (Core FFO/sh down 1.6% YoY) and trigger analyst re-rating. Our lenses are backward-facing; analyst downgrades are forward-facing.\n\nWhat We Saw: Quality −29 (good portfolio, leaky shares), Value −43 (fairly valued at $35; no margin of safety). What We Missed: The Q1 result showing YoY decline in FFO per share, and the risk that JPMorgan's June 2026 CRE update would re-rate REXR specifically on per-share growth fears. The stock needed to fall to ~$33 before sentiment stabilized (Scotiabank's "downside reflected" call). Our "wait for high $20s" was guidance, not prediction; in 14 days, it touched $33.68, halfway there.

Market context: S&P 500 fell only -0.29% over the window; CPRI fell -3.96%, a 3.7pp underperformance. Not market-driven.

Earnings signal: Q4 FY2026 (reported May 27, pre-window but 11 days before analysis) showed genuine operational beat: gross margin +490 bps, operating margin +170 bps, EPS $0.22 (vs. loss prior year), and FY2027 guidance of 40% EPS growth. Michael Kors stabilizing (down 5.5% vs. prior guided steeper declines), Jimmy Choo inflecting positive (+5.3% YoY). This contradicted our 'Shaky' (-73) Quality thesis.

Price action: Stock rallied +10.08% early in window (week of June 10–13), then faded to -3.96% by June 23. On June 24, Versace CEO Gintzburger's exit (effective June 23) was announced—timing suggests this uncertainty bled into the stock during the window even before the public headline.

Attribution: ~85% idiosyncratic (Versace leadership + brand stabilization doubts), ~10% sector, ~5% market.

What we got right: Direction (down), rationale (Michael Kors stabilization questionable). What we missed: (1) The operational inflection in Q4 (margin, EPS beat), which lifted the stock mid-window; (2) The Versace distraction / leadership vacuum as a new headwind; (3) The magnitude of the move (we called -0.45%, actual -3.96%).

Crux check: We said the crux was brand stabilization. Q4 data suggested it *was* stabilizing (JC +5.3%, MK selective full-price strength). But Versace leadership exit re-introduced uncertainty into the turnaround narrative, pushing stock back down below our pass threshold.

Market vs. Stock: S&P −0.29%, VIX flat. PGY's −3.92% early-week dip tracked risk-off sentiment; mid-week tech rallies (+12.7% peak) lifted it back. Idiosyncratic move was ~15%, market/sector +85%.

Our Call: AVOID at $14.80; revisit sub-$12. Stock ended at $15.00 (+1.35%). Direction is correct (not a strong buy case), but magnitude was wrong—we missed that the floor near $14.50 would hold and the stock would hover around fair value (+1.35%), not crater to $12.

Key Facts: - Q1 beat (May 7): $25M net income, 10% revenue growth, 5 consecutive profitable quarters - June 15 ABS close: $800M personal loan securitization, upsized, with 39 unique investors - Analyst consensus: 8 analysts, Strong Buy, median $27.50 target (83% upside from $15) - CEO insider buy (June 2): $250.5K shares, positive signal - CFO transition (June 15): Jon Dobres in; no disruption priced in

Attribution: Market 60% (risk-off early, rally mid), Sector 25% (fintech-AI tailwind strong), Idiosyncratic 15% (execution signals positive but not moving needle).

What We Missed: (1) Macro narrative of fintech-AI inflection (AI in fintech reaching $20.6B by 2026) was not visible in our Q/V scores; (2) Capital markets demand (4 upsized ABS deals in 2026, $5.5B+ YTD) is a proxy for durable moat we under-weighted; (3) Analyst reset post-Q1 beat (Strong Buy consensus 63%) was a contrarian signal we discounted.

The call: We rated FND as AVOID (Quality -10, Value -67) on June 9, noting it was a solid business priced for perfect execution—a housing turn, 400bp margin recovery, and continued unit growth. We said to wait for a dip. The thesis was bottom-up, pure Quality × Value.

What happened: FND opened at $49.74 (our analysis price) and rose 3.46% to $51.46 by June 23, with a max intraday run-up of 13.47%. The S&P 500 fell 0.29% over the same window, so nearly all of FND's move was idiosyncratic.

Attribution of the move: - Market (0%): S&P 500 down 0.29%; VIX stable 19.87→19.49. Not a tailwind. - Sector (45%): Home improvement / hardgoods retail rotated upward mid-window on (a) positive June retail sales print (0.9% M/M, 6.9% Y/Y on June 23), and (b) flight from expensive AI names into "large, cheap, cash-generative retailers" amid new Fed chair hawkishness. FND fits the profile: 4.3% profit margin, ~$110M quarterly FCF, manageable debt. This rotation was real and lifted the whole cohort. - Idiosyncratic (55%): (a) Post-Q1-earnings rebound: company guided FY26 comp sales (4.0)% to flat (weak, but not collapsing) and EPS $1.83–$2.08; (b) June 20 product launch (NatureMatch private label) + Schererville, IN store/design center announce; (c) momentum from prior dip (stock was in decline into June 9).

Grade vs. our claim: Our thesis was correct on fundamentals—FND *is* trading at a stretched multiple (P/E ~27.9x vs. peer avg 18.6x) for a business with (i) flat comp sales, (ii) 17.8% EPS decline YoY, and (iii) FCF under pressure. We correctly said "wait for a dip." But we underestimated two things: (1) how hard the sector would rotate *up* on macro/sentiment repricing, and (2) that a new product launch + store announce + post-miss sentiment recovery could generate short-term follow-through even though the core thesis (valuation is stretched) remained sound. The stock rebounded to +3.46% from our entry, but did NOT break to new highs; it's still well below the $54.55 consensus target.

What we got right: Quality score of -10 stands—accruals are clean (1.89x OCF/NI), no dilution, insider buying. Value score of -67 is also intact; the stock is still richly valued by any measure. A six-month horizon view *should* catch margin mean-reversion pressure as the lease-comp drag accelerates and the warehouse expansion faces a soft housing backdrop.

What we missed: Our system has zero visibility into top-down macro/sentiment regimes. The Fed-chair transition from Powell to Kevin Warsh created a hawkish repricing; money fled expensive AI/growth and piled into "barbell" bets (Treasury yields up, AI stocks down, cheap cyclicals/retailers up). This is a regime-level shift that overrode our bottom-up value call *for this 14-day window*. We nailed the direction (avoid) but the timing was off by ~2 weeks as the macro wind provided lift. The call will likely work over 6 months, but the early-window move against us was almost entirely macro/sentiment, not company-specific.

## Layers of Attribution

Market context (broad): The S&P 500 fell just -0.29% over the window, so broad market weakness was minimal. VIX was essentially flat (19.87→19.49). OVV fell -2.08% vs index −0.29%, so the stock underperformed the market on an absolute basis.

Sector & commodity driver (dominant): <cite index="21-6">Oil prices collapsed 23.53% from late May through June 23</cite>, falling from ~$97 at the start of the window (June 9) to ~$73.4 by June 23—a ~25% drop in crude in just 14 days. This was THE dominant force. <cite index="17-15,17-16">Brent averaged $107/bbl in May but fell as reports surfaced that the US and Iran were nearing an agreement to extend the ceasefire and reopen the Strait of Hormuz</cite>. <cite index="20-1,20-5,20-6">On June 23, crude fell to $73.4/bbl (its lowest in nearly 3 months) as signs of initial progress in US-Iran peace talks eased supply concerns, with Washington granting Iran a 60-day license to sell oil internationally</cite>.

Company-specific context: The real window for OVV was the reversal of the oil bull thesis—the exact narrative we called the crux of OVV's valuation. <cite index="1-15">Barclays raised its price target for Ovintiv from $68 to $75 on May 31, citing a tighter oil market with lower global inventories, reduced OPEC spare capacity, and muted US supply response</cite>—that was *during* the Hormuz crisis. But from June 9 onward, peace-deal optimism unwound the entire supply-scarcity premium. OVV has no moat; when the commodity floor collapses and geopolitical risk premiums evaporate, a levered, price-taker company gets punished.

vs. our call: We predicted −0.67% over the window (flat to slightly down), but OVV fell −2.08%—roughly 3× worse than our model expected. However, we were *directionally correct* (down, not up) and our thesis framing was correct: we said *don't overpay for leverage-dependent, cyclical FCF*. The stock *did* sell off when the commodity bid weakened. Our magnitude was simply too optimistic on the tail of energy prices.

What we missed: We underweighted the speed and magnitude of the peace-deal repricing. We acknowledged Hormuz as "the catalyst" but modeled oil staying "elevated long enough" for FCF/buybacks to work. We didn't account for how fast sentiment can flip on a 60-day Iran export license and early ceasefire signals. Our 6-month target of $51.37 (implying −8.3%) is now looking more reasonable—the 14-day window just accelerated the unwind.

## Attribution Scorecard - Market (broad S&P weakness): ~5% of the move (OVV down only 2%, market barely down) - Sector/commodity (oil collapse): ~70% (25% oil drop in levered E&P = 2%+ stock hit is plausible) - Idiosyncratic (OVV execution/news): ~25% (none reported; the stock moved with oil)

## Macro/Narrative Pressure Direction: Negative (hard reversal) Strength: 85 Note: Strait of Hormuz crisis inflection point—peace-deal optimism (Iran 60-day export license, ceasefire extension signals) collapsed the oil risk premium from $97→$73 in 2 weeks, destroying the bull case for levered E&P.

What went right: We called AVOID at $47.48; stock closed 2026-06-21 at $46.52 (−2.02%). We were directionally neutral-to-bearish and the stock didn't rally despite a roaring senior housing sector. In a rising tide, AHR barely moved—that's evidence the dilution/valuation thesis has teeth. Analyst consensus was bullish (PT $54.70, buy ratings), but the stock didn't follow. This is a yellow flag on per-share accretion.

What we missed: The thesis is longer-dated than 14 days. Two weeks is too short to prove AHR's capital allocation breaks the market's patience. The stock is pinned by two opposing forces: (1) Sector tailwinds (limited supply, strong occupancy, 12.1% NOI growth, demographic demand), and (2) AHR-specific headwinds (forward ATM offerings ~$527M, lower GAAP guidance despite strong FFO, extreme 66% NAV premium). The -2% move is noise; the real test would be a 6–12 month period where per-share metrics *accelerate* or *decelerate* relative to the narrative. We can't yet distinguish whether AHR's slowness is temporary ATM-induced overhang or the beginning of the per-share economics repricing.

Attribution: Market +1.28%, but with geopolitical risk (Strait of Hormuz) and AI rotation early in window. AHR −2.02% = ~2.3% underperformance vs. broad market. Senior housing sector was strong (evidenced by analyst upgrades, capital inflows, transaction momentum), yet AHR didn't participate. This suggests *sector tailwinds were not enough to lift AHR*—the company-specific dilution/valuation drag was real. The 66.61% NAV premium is a red flag: the stock is priced for perfection, and any capex miss or lower-than-expected FFO growth will hurt sharply.

Macro pressure: Senior housing narrative is *strongly positive* (supply deficit, demographic demand, cap rates compressing, institutional money flowing in). But AHR's premium valuation and aggressive equity issuance are fighting that narrative. The macro is pushing up; the company structure is pushing down. Result: stock trapped in range.

Market & Sector: S&P +1.28%, VIX fell; Brent crude +$5/b on Strait of Hormuz closure (late June re-escalation). Energy sector backdrop was *favorable* for LNG equipment, with global capex tailwinds.

Company-Specific Drivers: (1) Baker Hughes acquisition at $210/share (announced Oct 2025, shareholder vote Oct 6, 2025); (2) European Commission antitrust review with Phase I deadline of June 26, 2026; (3) On June 11, GTLS dipped −0.3% on antitrust concerns. Deal expected to close July 2026 subject to EU approval and stated commitments are 'limited' and not expected to materially affect deal rationale.

Stock Action: GTLS flat, +0.33% over 14 days ($207.31 → $207.99). Max drawdown −1.11%, max run-up +0.38%. Trading at −1.28% discount to $210 deal price (typical arb spread).

Why So Quiet: The stock is locked in arbitrage mechanics. Investors holding GTLS are not betting on standalone Chart fundamentals; they are waiting for deal close. Macro LNG tailwinds (Strait crisis boosting capex) and Chart's strong backlog and HTS segment growth should have created fundamental upside, but the deal collar prevents repricing. Regulatory uncertainty (EU antitrust) acts as a risk cap on upside.

Grade vs. Our Thesis: We said 'AVOID — wait for $170 dip.' The stock didn't dip, nor did it surge. Our concern was rich valuation + leverage + earnings wobble. All those concerns remain valid for the *standalone* company, but they are irrelevant to GTLS's 14-day move because the stock trades as a fixed-price acquisition. Our thesis was not proven or disproven — it was structurally untestable in this window.

Direction: We said 'wait for a dip' and got one (−4.15%). ✓ Correct.

Magnitude: Our crux was '15%-ish pullback' to ~$58 for edge; we got ~4% to $63.50. The dip was real but shallow and didn't cross our threshold because the reason (Hormuz relief, oil/gas down) doesn't improve SUN's fundamental appeal. Yield at $63.50 (~5.9%) is still market-clearing, not 'clearly generous.'

Reason miss: We did not anticipate the timing/strength of the Hormuz relief trade and its sector-wide re-rating of energy infrastructure. Our analysis was sound (SUN = stable, fairly valued), but macro narrative shift blindsided the sector 4 days after our analysis.

Attribution: ~70% sector (Hormuz relief, oil/gas down), ~15% market (risk-off early, then broad rally SUN missed), ~15% idiosyncratic (none; earnings were strong in May, distribution raised, no company news).

Macro pressure: Negative, strength ~75/100. Strait reopening = structural headwind to energy infrastructure valuations for months ahead.

The Brief Error: The brief lists ticker PRMB (Primo Brands, beverage/merger play) but the thesis, financials, and context are all PRIMORIS SERVICES (PRIM, infrastructure EPC). Assuming PRIM is the intended name.

Window Narrative: - Analysis date: June 7, 2026 (post-May 5 earnings crash @ 50%, pre-June 22 COO departure). - Stock closed June 7 at ~$22.95; by June 21 $24.27 (+5.75%). - June 8–10: Risk-off regime (Hormuz closure, VIX down 16%), PRIM fell 15.4% on June 9 as market re-digested renewables execution failures. - June 10–21: Broad market rally (+1.28% S&P, small-cap/tech rallies, VIX decline) pulled PRIM up +2.7% from the June 9 lows, masking company-specific decay. - June 22 (post-window): COO Jeremy Kinch departs immediately; company issues 3rd/4th guidance cut in a row—EPS down 68% (to $1.30–$1.85), EBITDA down 42% (to $275–$325M). Stock crashes 28% AH.

Attribution: - Market recovery (broad rally, VIX decline): ~70% of the +5.75% bounce. - Idiosyncratic news: June 9 renewal of selling (-15.4%) was offset by sector/market tailwinds over June 10–21; the company-specific thesis (renewables hemorrhaging, leverage/share dilution unresolved) remained toxic and was vindicated 1 day after our window close.

Grade: - We were right on direction (AVOID): The thesis—"no edge at 27× FCF on leveraged balance sheet with unproven synergy capture"—proved prescient. The company did *not* fix the renewables problem; costs spiraled. - But we arrived in a false calm: Our June 7 read caught the stock after the first crash but before the final reckoning. A 14-day bounce on pure macro tailwind (rally) obscured that the crux (FCF acceleration / synergy capture) had *failed*. - Timing vs. thesis: Thesis was *right*, but a 2-week window is too short to confirm or kill a 6-month prediction. The June 22 collapse—28% AH on worse guidance—proved our "wait for sub-$19" call was prudent.

Market vs. Company Attribution: BXP +3.8% vs S&P +1.28% over 14 days = ~250bps outperformance. This is NOT market-driven; it's idiosyncratic. The broader office REIT sector was down YTD, and VIX compression early-week (risk-off) would have hurt commercial RE, yet BXP rallied through a volatile backdrop.

What Drove It: 1. Q1 Earnings beat & raised 2026 FFO guidance (~+2% price reaction typical for REIT beats): BXP delivered $1.59 FFO/share vs. guidance midpoint, signaling operational momentum in gateway Class A markets (Boston, NYC, DC). Guidance midpoint of $6.96 represents modest but real FFO growth vs. 2025, a narrative inflection. 2. Capital recycling execution: $360M proceeds YTD, including Marriott HQ sale at 6.8% cap rate (a POSITIVE signal—not distressed pricing). This de-levers the balance sheet (reducing $15B net debt pressure) and funds development of 290 Binney in Cambridge (fully pre-leased to AstraZeneca). 3. Cap rate stabilization, not widening: Our crux hinged on cap rates widening another 50bps. Market data shows Class A office in 2026 trades at 6.0–8.0% cap rates. The Marriott sale at 6.8% is *normal* for Class A CBD, not crisis pricing. No evidence of 50bps widening from our June 7 baseline. 4. AI tailwind to gateway markets: BXP owns premier CBD office in Boston, NYC, DC, and SF—the exact geographies where AI-driven demand is re-concentrating tech/finance hiring. This contradicts the macro bearish framing of our thesis.

Why We Were Partially Wrong: - We said AVOID and "wait for sub-$52" (sub-20% haircut). Price closed at $64.70, a 3.8% move *up* from $62.33 analysis price. - Our Quality score (-20, "Mixed") was correct—the leverage is real and net income was depressed ($14M vs $849M prior year). But we conflated that with a price call and got the timing 180° wrong. - Our crux (cap rate widening) didn't materialize in this window. Cap rates in gateway Class A stabilized; BXP's debt refinance outlook improved. Interest expense fell. - We missed the magnitude of the guidance raise: FFO guidance raised to $6.90–$7.04 (midpoint $6.96) is a 1.6% increase from 2025 ($6.85). For a REIT grinding through a structural headwind (office vacancy), that's credible operational proof.

Attribution Breakdown: - Market (S&P momentum, risk-off/complacency flip, macro relief): ~30% (BXP would have moved +1% on beta alone) - Sector (Office REIT sentiment shift, AI re-rating of gateways): ~50% (the big mover—BXP premium to non-prime office) - Company-specific (earnings beat, guidance, capital recycling, interest rate refinance benefit): ~20% (modest, but real)

Macro/Narrative Pressure on BXP: The macro backdrop was NOT hostile to BXP in this window. The Strait of Hormuz closure drove risk-on sentiment mid-week (VIX fell 110bps), and AI's concentration in gateway cities created a narrative tailwind. Office REITs have been repriced down 8% YTD, but prime CBD assets in AI hubs are experiencing a rotation *into* them.

What We Missed: 1. Narrative inflection in progress: By late April/early June, market consensus shifted from "office is dead" to "prime office in AI hubs has a future." We were anchored to the old narrative. 2. Balance sheet trajectory: The $1.5B cash position and $400M+ of asset sales in 2026 are *actively de-leveraging* the company, not just damage control. 3. Cap rate stability: We assumed widening; market showed compression signals for Class A. The Marriott sale at 6.8% cap is exhibit A.

## Market vs. Stock Move

Market Context: S&P 500 rose 1.28% over the window; VIX compressed from 18.92 → 16.78 (risk appetite improving). Strait of Hormuz closure kept energy prices elevated and should have supported renewable diesel demand via RVO tailwinds. Market had three early risk-off days (June 8-11) but recovered strongly (June 15-19 tech/complacency rally).

DAR Performance: Stock fell 9.71% ($59.45 → $53.68) over 14 days, max drawdown -12.28%. This is idiosyncratic—while market recovered, DAR sold off hard. Attribution: ~85% idiosyncratic, ~10% market, ~5% sector.

## Our Thesis vs. Reality

We called AVOID on: - Quality: -14 (mixed). We correctly flagged high debt (3.17x leverage at quarter-end), margin compression (1,000bp), and 90% earnings decline in the trough. - Value: -25 (fairly valued). We said $59 already prices mid-cycle recovery without paying for leverage—a no-chase setup. - Crux: Whether DGD margin trough turns in 2-3Q.

Why We Were Right (Direction): Our skepticism was directionally correct. The stock fell, confirming that "fairly valued into mid-cycle recovery" was indeed the setup. A dip did occur (down 12%), consistent with our "wait for a dip" thesis.

Why We Missed (Magnitude): We said "avoid + wait for dip." The dip came, yes—but the *reason* for the dip was not primarily a market-wide pullback or sector rotation. It was company-specific: an earnings miss.

On June 6 (one day before our analysis), a headline appeared: "Darling Ingredients reported quarterly revenues matching analyst expectations but significantly missed adjusted operating income estimates, even as management described the quarter as an inflection point." We did not see or weight this miss adequately in our call. Q1 2026 net income swung from -$26.2M loss to +$134.3M (beat), but operating income fell short of street expectations, revealing that DGD margin expansion was less durable/strong than management's "inflection" narrative suggested.

## The Crux: Did Margins Really Turn?

Management claimed an inflection in May's Investor Day and Q1 earnings (April 30). DGD EBITDA per gallon improved to $1.11 (Q1 2026) from $0.41 (Q4 2025)—a stunning recovery. But the June headline reveals the street had priced in *more* margin expansion than delivered. The RVO finalization (late March 2026) and Strait of Hormuz closure (ongoing) created massive tailwinds for feedstock costs (fat prices rose) and demand, but competitive supply-side response to the RVO likely compressed margins faster than management assumed. The "clear inflection" narrative may have been premature or incomplete.

## Macro/Narrative Pressure

Positive Tailwinds: RVO clarity (5.4B gal vs. 3.35B prior year), Strait closure (20% of seaborne oil; should tighten global diesel supply), record-high RIN values (D4/D5/D6 RINs elevated per EIA), renewable diesel production forecast at record highs (+24% in 2026 vs. 2025).

Headwind Not Priced: Supply-side response to RVO was rapid. If you increase demand mandates 60%, new biodiesel/renewable diesel capacity comes online faster than anticipated. Margins compete. Fat feedstock prices rose (good for DAR's rendering segment), but DGD margins didn't expand as much as the "inflection" headline promised. This is a structural competitive dynamic, not a cyclical/macro one.

## What We Missed

1. Earnings quality signal: A June 6 headline flagged the operating income miss. We did not integrate this into our June 7 analysis. We should have caught the market repricing of DGD margin durability risk. 2. Supply-side response to RVO: We noted that RVO clarity was a tailwind, but we under-weighted how *quickly* idle biodiesel capacity would restart and new renewable diesel come online. In a world where mandates jump 60%, competition intensifies faster than a cyclical recovery narrative suggests. 3. Crux validation timing: Our crux was "does the margin trough turn in 2-3Q?" Q1 2026 showed a turn *in headline EBITDA per gallon*, but the operating income miss suggests the *quality* and *durability* of that turn is weaker. A two-week window is too short to call the crux, but the earnings miss is a yellow flag that our thesis (wait for durability confirmation) was prescient.

## System Performance

Our Quality×Value logic was sound: mixed quality + fairly valued = avoid. The stock did underperform the market. However, we failed to:** - Integrate a real-time earnings quality signal (the operating income miss) into our call. - Distinguish between "headline margin recovery" (DGD EBITDA/gallon improved) and "durable, profitable margin recovery" (operating income missed, suggesting competitive pressure/volume mix issues). - Weight the supply-side response to the RVO as a medium-term structural headwind to DGD margins, not just a short-term cyclical tailwind.

Market & Sector Attribution: - VRNS +2.11% vs. S&P 500 +1.58% in a 14-day window that began with broad risk-off (June 5-11) and recovered sharply into tech rally (June 12-18). - VIX fell 22% (21.51 → 16.78), signaling de-risking and flight-to-quality into mega-cap/SaaS names. - This was a *sector* move, not a company-specific inflection. VRNS benefited from sentiment rotation into AI-exposed software (Snowflake/Dell strength spilled over). - Stock showed modest outperformance vs. market, but no break-out and no proof of thesis reversal.

Company-Specific Drivers: - Zero earnings, zero guidance, zero material news. Analyst reaffirms on June 15 (Wells Fargo Buy) are routine. - Conference presentations (William Blair June 2, Infosecurity Europe June 2-4) generated no material price catalyst. - No change to valuation inputs: still posting net losses (Q1 2026: -$37M), still running high SBC (our crux), still trading at ~5.7x P/S (above peer/industry average).

Grade vs. Our Call: - We said AVOID at $32.69 with a crux around SBC/dilution overhang until FCF/per-share compounding normalized. - Stock +2.11% is a non-move for 14 days; we don't buy it because of price appreciation, we pass because the fundamentals we cited (SBC bleed, margin compression, valuation multiple at 6-7x revenue) remain unchanged. - Our call was right in direction & reasoning, but the window is too short and too driven by macro/sentiment to confirm or refute. The thesis will play out over 6 months, not 2 weeks. We're sitting out; the stock moved sideways. Call = intact.

What We Got Right: - Identified the company trades on AI narrative (SaaS/security premium), not on near-term earnings power. The June rally *proved this*—pure sentiment, no news. - Crux remains valid: Q1 2026 showed -$37M net loss despite $173M revenue, and dilution math is unchanged. SaaS transition is real, but margin inflection isn't here yet.

What We Didn't Anticipate: - The *magnitude* of broad tech-sector re-rating mid-window (Snowflake/Dell catalyst June 1-2 lifted SaaS sentiment broadly). This wasn't idiosyncratic; it was a market-regime rotation out of risk-off into AI software. - We didn't flag the macro/sentiment coefficient as the dominant 14-day driver. Our system sees Q/V; it can't predict VIX compression or cross-sector momentum. This is the blind spot our framework acknowledged.

Conclusion: - The stock did *slightly better* than the market (0.5% excess return), but in a window dominated by macro/sentiment tilt into SaaS. Our company-specific thesis (pass until dilution normalizes) was not tested. The window is too short and too noisy to grade the thesis.