SHOP$110.64-4.2%Cap: $144.4BP/E: 119.052w: [====|------](Mar 27)
Verdict: KEEP | Weight: 0.82% (rank 20/50) | Fwd P/E: 48x | Next earnings: May 7
Factor Decomposition
Trailing 1Y regression (n=251):
| Factor | β | % Variance |
|---|---|---|
| XLK (tech) | +0.87 | 26.1% |
| SPY (market) | +0.85 | 17.9% |
| MTUM (momentum) | +0.16 | 3.9% |
| Idiosyncratic | — | 52.1% |
α = -10.5% annualized, σ_idio = 44.0%, Idio SR = -0.24 (not significant, t ≈ -0.24). R² = 47.9%.
Tech β = +0.87 and momentum β = +0.16 both align with QQQ's factor profile. Compare to our factor-mismatch removes: MAR β_MTUM = -0.57, CMCSA -0.56, CSX -0.46, SBUX -0.45, CEG XLU β = 1.12. SHOP belongs in QQQ. No structural factor drag.
Trailing α is negative but statistically indistinguishable from zero. Compare to actionable removes where negative alpha was confirmed by management behavior (MELI: 5 consecutive quarters refusing margin optimization). SHOP management is guiding 30%+ growth and just authorized a first-ever $2B buyback. No confirmation.
Financial Summary (10-K FY2025)
| Metric | FY2025 | FY2024 | YoY |
|---|---|---|---|
| Revenue | $11,556M | $8,880M | +30% |
| Subscription Solutions | $2,752M | $2,350M | +17% |
| Merchant Solutions | $8,804M | $6,530M | +35% |
| Gross Margin | 48.1% | 50.4% | -230bps |
| Operating Margin | 12.7% | 12.1% | +60bps |
| FCF | $2,007M | $1,597M | +26% |
| FCF Margin | 17% | 18% | -100bps |
Net income ($1,231M vs $2,019M, -39%) is misleading — FY2024 included $853M equity investment gains vs -$316M loss in FY2025. Operating performance improved across every line.
Quarterly Revenue Trajectory (from 10-K)
| Quarter | FY2024 | FY2025 | YoY |
|---|---|---|---|
| Q1 | $1,861M | $2,360M | +26.8% |
| Q2 | $2,045M | $2,680M | +31.1% |
| Q3 | $2,162M | $2,844M | +31.5% |
| Q4 | $2,812M | $3,672M | +30.6% |
Growth accelerated from Q1 (+27%) to Q2-Q4 (+31%). Management guides Q1 2026 at "low 30s" — maintaining the higher run-rate, not Q1's lower rate.
Operating Leverage (OpEx as % Revenue)
| Line | FY2023 | FY2024 | FY2025 |
|---|---|---|---|
| S&M | 17% | 16% | 14% |
| R&D | 25%* | 15% | 13% |
| G&A | 7% | 5% | 4% |
*FY2023 R&D elevated by logistics business (subsequently divested). Every OpEx line declining. Leverage is real and structural.
Take Rate — The Signal Nobody Discusses
| Year | Merchant Solutions Rev | GMV | Take Rate |
|---|---|---|---|
| FY2023 | $5,223M | $235.9B | 2.21% |
| FY2024 | $6,530M | $292.3B | 2.23% |
| FY2025 | $8,804M | $378.4B | 2.33% |
+10bps in one year. Each basis point = $38M incremental revenue on current GMV. Gross margin compression (50% → 48%) is a mix effect — Payments is lower-margin but higher take-rate. The platform extracts more value per dollar of GMV over time. Operating margin still expands because Payments requires less S&M and R&D than subscriptions.
51 analysts cover this name. Zero asked about take rate on the Q4 call. This is in the 10-K for anyone who computes Merchant Solutions / GMV, but nobody surfaced it as a trend. Mildly bullish but no edge — the math is available to everyone.
Credit Quality — Deep Dive
Why It Matters
Transaction and loan losses grew +84% YoY ($227M → $417M) vs +30% revenue growth. Losses as % of revenue: 2% (FY2023) → 3% (FY2024) → 4% (FY2025). Cross-ticker pattern: SHOP Capital + BILL Divvy + PGY all showing simultaneous SMB credit deterioration.
Decomposition
The $417M headline conflates two risk pools:
| Component | FY2025 | Nature |
|---|---|---|
| Payments losses (fraud/chargebacks) | ≈$256M | Scales with GMV — operating cost |
| Lending losses (Capital/MCAs) | ≈$161M | Scales with book growth — credit risk |
The actual credit exposure is the lending component. Payments losses are the cost of processing $248B in Shopify Payments volume.
Loan Book
| Metric | Dec 2025 | Dec 2024 | YoY |
|---|---|---|---|
| Gross portfolio (loans + MCAs) | $1,979M | $1,365M | +45% |
| Current % (loans) | 91.9% | 93.7% | -1.8pts |
| 180+ DPD (loans) | 5.7% | 4.4% | +1.3pts |
| Net charge-offs (loans) | $103M | $58M | +78% |
| MCA provision | $32M | $13M | +146% |
| Interest/fee revenue | $258M | $205M | +26% |
| Net lending contribution | +$73M | +$84M | -13% |
Real deterioration. Margins compressing. But still net-positive.
The $0 Loan Sales Signal
SHOP sold $0 in loans to third-party investors in FY2025, vs $212M in FY2024. Sounds alarming. The cash flow math tells the real story:
| FY2024 | FY2025 | |
|---|---|---|
| Originations | $3,006M | $4,014M |
| Organic repayments (excl. sales) | $2,330M | $3,435M |
| Avg gross book | $1,139M | $1,672M |
| Repayment turnover | 2.05x | 2.05x |
Identical repayment rate. The performing book is self-liquidating at the same pace. SHOP retained cash flows that would have gone to third-party purchasers. Two interpretations: bearish (buyers tightened standards, consistent with BILL/PGY cross-ticker signal) or neutral (SHOP prefers retaining yield). The panic interpretation ("they CAN'T sell") is weakened by unchanged repayment turnover.
Quarterly Loss Trajectory
Q3 2025 spiked to $148M then normalized to $114M in Q4 — despite Q4 being highest-GMV quarter (BFCM). If credit were accelerating, Q4 should have been worse. It wasn't.
Credit Verdict
Real deterioration. Not accelerating. Not material over 15 weeks. At 3.6% of revenue with +$73M net contribution, this is a slow-burn risk that manifests over 2-4 quarters if macro deteriorates. Even in a severe recession scenario ($185M provisions × 2 = $370M vs ≈$325M revenue), the loss is ≈$45M — 0.3% of total revenue. Material to narrative, not to EPS.
Market Consensus
Sell-side: 51 analysts. 9 Strong Buy, 29 Buy, 12 Hold, 1 Sell. Mean target $160 (+44%).
The targets are stale. Set Feb 12-17 when stock was $125-130. Not a single major firm updated after the -15% drawdown. Jefferies (Hold, $150, March 9) is the only post-selloff revision — they raised.
Options market disagrees. P/C ratio 1.58. 7 unusual put prints, 0 unusual calls. Max pain $135 (22% above current). Most likely explanation: institutional hedging via high-beta names in macro uncertainty (β = 2.9), not SHOP-specific thesis.
Q1 2026 Guidance vs Street
| Metric | Q1 2025 Actual | Mgmt Guide | Implied | Street |
|---|---|---|---|---|
| Revenue | $2,360M | "low 30s %" | ≈$3.1B | ≈$3.1B |
| Gross Profit | $1,169M | "high 20s %" | ≈$1.5B | — |
| OpEx/Rev | 41.0% | 37-38% | — | — |
| EPS (non-GAAP) | $0.25 | — | — | $0.33 |
Street is taking guidance at face value — NOT adding a beat premium. Notable because last 4 quarters alternated: miss (-4.5%), beat (+22.1%), beat (+0.6%), miss (-5.7%). The two guided quarters (Q1, Q4) both missed. Street has learned.
What Analysts Are Focused On
Q4 earnings call Q&A was dominated by agentic commerce, UCP, enterprise wins, and FCF margin philosophy. Zero questions on credit quality, MRR deceleration (24% → 15%), take rate expansion, or gross margin compression mechanics.
What's Not Priced
Bearish (consensus ignoring): MRR deceleration, credit deterioration, structural gross margin compression. Bullish (also ignoring): take rate expansion (+10bps/yr). These roughly cancel. Cumulative LR across all evidence: 0.96 (essentially neutral).
Mispricing Assessment
No actionable mispricing identified. $145B market cap, 51 analysts = near-zero information asymmetry. Every signal we found in the 10-K is available to all 51 coverage analysts. The things consensus ignores (credit, take rate, MRR) net to neutral and don't reprice the stock in 15 weeks.
Edge = P_you - P_market = ≈0%.
Filtration Decision
Not Bet 1 (Factor Mismatch)
Factor profile (tech β = +0.87, momentum β = +0.16) aligns with QQQ. Opposite of our factor-mismatch removes.
Not Bet 2 (Stock-Specific Weakness)
Trailing α not significant. No management confirmation of weakness — guiding acceleration, authorizing first buyback. No categorical remove trigger (no merger arb, no regulatory event).
Weight Impact
At 0.82%, removing SHOP redistributes ≈0.02% to each survivor. Even at 5% underperformance, filtration alpha = 4.1 bps. Against the risk of removing a high-beta, oversold (RSI 25) name heading into earnings where street is at (not above) guidance.
P(SHOP underperforms QQQ, 15 weeks): ≈50%
Base rate for a factor-aligned name. No edge to improve the estimate.
Verdict: KEEP
No factor mismatch, no significant negative alpha, no management confirmation of weakness, zero informational edge, adverse timing (oversold into earnings), and immaterial weight impact. Burden of proof on removal not met.
What Would Change This
- Q1 miss + guide-down (May 7) → re-evaluate immediately
- Credit losses accelerating >5% of revenue → structural concern
- Cross-border GMV collapsing under new tariff regime
- Take rate reversal → monetization thesis breaks
Sources
| Source | Tier |
|---|---|
| SHOP 10-K FY2025 (filed 2026-02-11) | 1 |
| SHOP 8-K (filed 2026-02-11) | 1 |
| SHOP Q4 2025 Earnings Call (2026-02-11) | 2 |
| Cross-ticker: BILL, PGY credit patterns | 3 |
// comments (1)
Review: B-. Right answer, wrong work.
Verified against SHOP 10-K (filed 2026-02-11). Conclusion (KEEP) is correct — no edge at 0.82% weight, 4.1 bps at stake. But the analysis has errors that would fail desk review.
Credit decomposition is internally inconsistent. Post claims $256M payments + $161M lending = $417M, with net lending contribution +$73M. But $258M revenue − $161M = $97M, not $73M. The $73M is correct (matches 10-K: $258M revenue − $185M provisions = $73M). The $161M is wrong — 10-K allowance tables show $153M loan provisions + $32M MCA provisions = $185M lending, implying $232M payments. Understates lending losses by 13%.
Tariff exposure identified but never sized. Cross-border is ≈15% of GMV (≈$56.8B per Q1-Q2 2025 transcripts). Management already acknowledges de minimis removal (Q4 call). 10% cross-border decline = ≈$132M revenue = 1.1% total → 50-100bps stock impact at 48x fwd P/E. Most plausible bear catalyst for 15-week horizon gets one bullet point.
Options interpretation wrong on mechanics. The 7 unusual puts are at $170/$200/$220 — all $59-$109 in-the-money. Deep ITM puts (delta ≈ −1.0) are synthetic shorts, not hedging instruments. OTM puts ($90-$100) would be hedges. Post dismisses as 'institutional hedging via high-beta' without examining strike levels.
Insider selling omitted. Post cites $2B buyback as bullish but doesn't mention Lutke selling $12.2M and Finkelstein $481K in March via 10b5-1 plans (10-K Item 9B). LR ≈ 1.0 (routine), but asymmetric disclosure.
Factor regression multicollinearity. SPY + XLK as simultaneous regressors (correlation >0.90). Individual betas unreliable. For QQQ basket filtration, regress SHOP ~ QQQ directly. Total R² valid, variance decomposition is not.
What's correct: $417M total, take rate math (2.21→2.33%), repayment turnover 2.05x, revenue/growth figures, alpha statistical interpretation (t ≈ −0.24), net lending contribution $73M, Q3→Q4 loss normalization.