The Setup

Vertex, Inc. is a tax compliance SaaS company. Indirect tax determination, e-invoicing, compliance reporting. 45+ years in the business, majority of the Fortune 500 as customers, 4,867 direct accounts, $671M ARR. The kind of boring, mission-critical compliance software that should trade at 25x forward earnings and never make headlines.

Instead it's at $13.26, down 69% from its 52-week high of $44.59. Forward P/E of 15x. RSI 35. Sitting at 4% of its 52-week range.

The headline story is simple: NRR decelerated from 109% to 105%, the market panicked, and VERX got caught in the broader Feb 2026 SaaS selloff (TEAM -73%, TRI -50%, INTA -53%, NOW -50%). Growth decelerating + sector de-rating = 70% drawdown.

The interesting story is in the Form 4s.

What The 10-K Actually Says

The 2025 10-K (filed Feb 24) is a mixed bag. Revenue $748.4M (+12.2%), first year of GAAP profitability ($7.2M net income), $314M cash, no debt drawn on the revolver. The subscription business grew 12.8%. Cloud mix hit 55% (up from 49%). International revenue grew from 8% to 10%. All fine.

The problems are real but specific:

NRR decelerated 400bps to 105%. Management's explanation has two parts. First, "customers' annual growth has slowed, keeping them within current bands of usage" — VERX prices on customer revenue, so when their customers slow down, VERX's usage-based pricing mechanically produces lower expansion. This is pro-cyclical, not structural. Second, "slightly higher customer attrition" concentrated in small accounts under $50K ARR (vs $138K average). The whales are staying. The minnows are leaving.

Customer count actually declined by 48 net. But new logos were up 20%. The math: they're adding bigger customers and losing smaller ones. AARPC rose 12.4% to $137,867. This is upmarket migration, not customer flight.

A one-time true-up headwind masked the real growth. This is the detail the market hasn't fully processed. CFO Schwab on the Q4 call: true-up revenue in 2025 was approximately $10M lower than 2024. That alone reduced full-year revenue growth by ≈2 percentage points and Q4 growth by ≈4 percentage points. Underlying subscription growth was ≈14%, not the 8.9% that Q4 headline showed. The true-up cliff doesn't recur in 2026 — it was customers staying within their contracted entitlements rather than exceeding them. One-time reset.

New CEO as of November 10. David DeStefano (longtime CEO) retired. Christopher Young (ex-Microsoft AI partnerships) appointed. The 10-K explicitly flags: "We have recently experienced workplace changes, including a new Chief Executive Officer, and we expect changes to continue for the foreseeable future." Severance spiked to $6.8M from $3.0M. Young has been in the chair for 3.5 months. His strategy: AI-first transformation, accelerate e-invoicing, expand customer success coverage using AI tools. Directionally right but execution is TBD.

Services margin compressed hard. Services gross margin dropped from 34.7% to 27.4% — COGS grew 21.4% on 9.2% revenue growth. Managed services (outsourced tax filing, lower margin) growing faster than implementation services. Mix shift that may continue.

Avalara litigation escalating. VERX sued Avalara (now Thomson Reuters subsidiary) in January 2022 for trade secret misappropriation, intentional interference, unfair competition. Still in discovery. Legal costs: $10.754M in 2025 vs $2.032M in 2024. New risk factor: "We have recently experienced efforts by a significant competitor to solicit our employees." Avalara is actively poaching VERX talent while the lawsuit proceeds. Asymmetric outcome — potential large damages recovery if VERX wins, but ongoing cash drain regardless.

What The Market Is Pricing

At $13.26 and 15x forward P/E, the market implies roughly:

• 50% probability of the bear case (NRR continues to ≈100%, growth stalls to <8%, stock goes to $9-10)
• 40% probability of base case (guidance met, NRR stabilizes at 103-105%, stock drifts to $15-16)
• 10% probability of bull case (e-invoicing + AI re-accelerate growth, stock re-rates to $22+)

In other words: coin flip the business is structurally impaired.

Is that reasonable? GRR is 94%, within the company's own 94-96% target. The base isn't eroding. New logos are up 20%. The true-up headwind was mechanical and won't recur. Four consecutive quarterly beats. Management guided margin EXPANSION for 2026 (EBITDA 23% vs 21.6%). The CFO explicitly said the guidance philosophy is conservative: "took thoughtful approach, took everything into consideration."

50% bear probability on this evidence set is fear, not analysis.

The Insider Signal

Jeffrey Westphal is the founder and a 10%+ beneficial owner. Here's what his family trusts did over the past 8 months:

June 2025: His wife's irrevocable trust sold 405,000 shares at $35-37. Total proceeds: $14.8M. Stock was near the 52-week high.

February 2026: His generation-skipping trust (different entity, different beneficiaries) bought 397,740 shares at $12.53-$13.08. Total cost: $5.12M. Stock was near the 52-week low.

Different trusts. Same brain. He sold the top through one vehicle and bought the bottom through another. The buy price is 65% below the sell price.

The timing is precise. Q4 earnings were February 11. The stock cratered on NRR. Westphal waited two days, then bought 247,740 shares on February 13 at $13.08. Four days later he bought another 150,000 shares at $12.53 — a LOWER price. He averaged down. He saw the Q4 numbers, saw the market's reaction, and stepped in with $5.1 million.

Three days after Westphal's first buy, director Eric Andersen purchased 40,000 shares at $12.91 for $516K — a 25% increase to his direct position. Open-market director purchases are rare. Two insiders buying independently within a week at multi-year lows is a pattern, not a coincidence.

Westphal sees Q1 pipeline data. He sees the e-invoicing deal flow. He sees what the new CEO is doing. He has more information than any analyst covering this stock. And he just put $5.1 million into the generation-skipping trust — money he's positioning for his grandchildren. That's not a trade. That's a conviction about the next decade.

The notable absence: CEO Christopher Young owns zero open-market shares. His 1,300,728 RSU hiring grant ($17.2M notional) provides alignment but not a personal conviction signal. He's been in the chair for 3.5 months. Not penalizing, but it's the one missing piece.

The C-suite "acquisitions" (Schwab, Rowland, Leib) are all RSU vestings with automatic tax withholding sales. Code M and Code F. Zero signal.

The Latent Factor

E-invoicing is the structural catalyst the market is ignoring. Governments globally are mandating "clearance-model" electronic invoicing — every invoice must be validated by a government platform before it's legally valid. Belgium's mandate went live January 2026. Poland, France, Germany are in pipeline for H2 2026 and beyond.

VERX's e-invoicing platform operates in 38 countries. They acquired ecosio in 2024 for EU capabilities and partnered with Brinta in 2025 for Latin America. Their existing Fortune 500 customers — who already trust VERX for tax determination — are natural buyers of the e-invoicing solution as each mandate goes live.

From the Q4 call, specific deals: a global payments company adopted e-invoicing for Belgium, Poland, and France. A consumer electronics company for Italy, Belgium, Poland, and Denmark. A consumer products company for Germany, Belgium, Poland. CEO Young stated: "E-invoicing cross-sell increased ARR for these customers on average over 20%."

Twenty percent ARR uplift per customer. That's the number. If even 500 of VERX's 4,867 customers adopt e-invoicing in 2026, that's roughly $14M of incremental ARR. Enough to add 2 percentage points to NRR by itself.

This is a Paleologo latent factor — a force that will drive returns but doesn't exist in the regression yet. The market's model assigns zero weight to e-invoicing because it hasn't shown up in quarterly revenue at scale. Our model assigns positive weight because the mandates are public record with specific dates and the per-customer economics are stated by management.

When the mandates go live and the revenue appears in H2 2026 numbers, this factor materializes. The market re-classifies e-invoicing from "narrative" to "revenue contributor." But by then the alpha is gone.

Factor Decomposition

Statistical (backward, 1Y regression): 91% idiosyncratic, 9% XLK, 1% SPY. R-squared 9.1%. Factors explain almost nothing — this stock moves entirely on its own.

Forward (where returns come from):

Edge-weighted factors: ≈44.5% of forward variance. At 91% idio, this is a high-purity situation.

Alpha Calculation

Scenario-weighted target (post-insider Bayesian update):

• Bull (21%): $23.10 — NRR stabilizes, e-invoicing lands, CEO executes, sector re-rates → 22x × $1.05 EPS
• Base (57%): $15.84 — guidance met, NRR 103-105%, modest e-invoicing contribution → 18x × $0.88 EPS
• Bear (22%): $9.10 — NRR continues to 100%, macro worsens, growth stalls → 14x × $0.65 EPS

Probability-weighted target: $15.89 (+20% from $13.26).

After stripping factor returns (SPY +8%, XLK +10%, MTUM +5%): r_idio = ≈14%.

Alpha = 14.0% × 0.70 conviction × 0.445 edge = 4.3% annualized idiosyncratic alpha.

This is a 2.5-3% starter position. Not a high-conviction concentrated bet — the NRR trend and CEO unknowns cap conviction. But the math works and the risk-reward is asymmetric (15-20% downside to $10-11 vs 35-65% upside to $18-22).

What Kills It

• NRR prints below 100% at Q1 earnings (May 6). If the base is shrinking, not just slowing, the thesis is dead. Exit.
• GRR drops below 93%. GRR is the floor — revenue retained before expansion. Currently 94%. If it falls, customer base is eroding. Kill switch.
• CEO announces dilutive acquisition or strategic pivot. New CEO risk is real.

What Upgrades It

• Q1 NRR prints 103%+. The founder was right. Scale to 5%+.
• E-invoicing revenue becomes visible in Q2/Q3 numbers. Latent factor materializes. The 20% ARR uplift per customer is no longer narrative — it's revenue.
• CEO announces specific AI product roadmap with customer commitments. Resolves the 15% "CEO execution" factor.

The Bottom Line

The 10-K by itself is a nothing-burger — steady SaaS company, some deceleration, some tailwinds, new CEO. The filing doesn't scream buy or sell.

The situation around the 10-K is what matters. A 69% crash that prices 50% probability of structural impairment. A founder buying $5.1 million at the trough through his family's generational wealth vehicle. A one-time true-up headwind that masked 14% underlying growth. A latent e-invoicing catalyst with specific mandate dates and quantifiable per-customer economics. A forward P/E of 15x for a company with 94% gross retention, $314M cash, and majority of the Fortune 500 as customers.

The market has over-extrapolated the NRR deceleration trend. Three years of 112→109→105 looks like a straight line to 100%. But the composition tells a different story: small account attrition (not enterprise), mechanical true-up headwind (one-time), and macro-driven deal delays (cyclical). The base — 94% GRR, 60%+ of the Fortune 500, mission-critical compliance infrastructure — is intact.

May 6 is the falsification date. If Q1 NRR prints 103%+, the bear case collapses and this re-rates. If it prints below 100%, the founder was wrong and so was I. The position is sized for surviving that outcome.

Westphal isn't guessing. He sold the top and bought the bottom. He sees Q1 data we won't see for 10 weeks. He put $5.1 million where his grandchildren are.

That's not a vibes trade. That's informed conviction by the person who knows this business better than anyone alive.