Verdict: FILTER (65% confidence)

PEP is a consumer staples company in a tech-growth index. QQQ beta of 0.08 and MTUM beta of -0.44 make it a structural drag on relative performance in any momentum-up regime. The fundamentals — declining organic volume, PFNA margin compression, back-end loaded guidance — are real but fully priced by 23 analysts and a 16.73x forward multiple. There is no informational edge on this name. The edge is index construction arbitrage: QQQ must hold PEP at 1.15%; we don't have to. Expected alpha contribution: 2-5 bps, depending on QQQ direction. A process trade, not an insight trade.

Risk: QQQ is itself oversold (RSI 23.1). If momentum reverses and staples outperform tech over 15 weeks, this removal goes against us. The hedge partially neutralizes but doesn't eliminate this risk.

Factor Profile

Regression (250d daily, iev regress)

• Alpha vs SPY: +6.8% annualized — but t-stat = 0.31 (statistically zero; 95% CI: -37% to +51%). The R-squared is 5.6%. This regression explains almost nothing. The alpha estimate is noise.
• Idio vol: 21.8% annualized
• QQQ beta: 0.08 (from cross-sectional analysis). PEP participates in 8% of any QQQ move. The other 92% is dead weight in the basket.
• SPX beta: 0.38 (low — classic defensive staple)

The Factor Mismatch

PEP is NASDAQ-listed because Frito-Lay merged into PepsiCo when PepsiCo was already a NASDAQ company. That's it. There's no fundamental reason a consumer staples mega-cap should be in a tech-growth index.

Quantified:

• QQQ beta 0.08: For every 10% QQQ move (up or down), PEP moves 0.8%. The average QQQ constituent moves ≈9.5%. PEP contributes 1/12th the index-tracking of an average name, per unit weight.
• MTUM beta -0.44: QQQ is momentum-heavy by construction (cap-weighted tech = winners compound). PEP loads negatively on momentum. In momentum-up regimes, PEP is a headwind; in momentum-down regimes (like March 2026), PEP is neutral to helpful.
• Over 1 year: PEP +6.3% vs QQQ +17.2% = 10.9pp relative underperformance. The factor mismatch compounds over time.
• Over 1 month: PEP -8.7% vs QQQ -8.8%. Roughly flat — because the current regime is momentum-down (tech selling off). The mismatch only shows up in trending markets.

What This Means for 15 Weeks

The beta mismatch generates alpha only if QQQ trends up. In a flat or down QQQ, removing PEP is neutral to slightly negative. At 1.15% weight:

Expected case (QQQ drift +3-5% in 15 weeks at historical averages): +3 to +5 bps.

Fundamentals

FY2025 Financials (10-K, filed Feb 3, 2026)

Revenue: $93.9B (+2% reported, +2% organic). Growth is entirely pricing (+4% effective net pricing). Organic volume declined: -2% food, -3.5% NA beverages, -3% EMEA food. The only volume growth was Asia Pacific food (+5%) and flat in Latin America and international beverages.

The PFNA problem: Frito-Lay North America is the largest segment ($27.5B revenue) and the worst performer. Core operating profit fell 6% to $6.5B. Organic revenue -2%. Management closed 3 manufacturing plants, shut several production lines, and is cutting ≈20% of US SKUs. This is a business in restructuring, not growth.

Margins: Gross margin contracted 41 bps to 54.15% (commodity costs +5pp, tariffs 6pp in PBNA). Core operating margin contracted 12 bps to 15.88%. GAAP operating margin fell 179 bps to 12.24%, depressed by $1.99B in intangible impairments and $964M in restructuring.

Rockstar write-down: $1.86B impairment on the Rockstar energy brand — 48% of the $3.85B purchase price 5 years in. PEP now distributes CELSIUS and acquired poppi and Alani Nu, implicitly admitting the Rockstar thesis failed. Poor capital allocation signal.

EPS: Core EPS flat at $8.14 vs $8.16. GAAP EPS $6.00, down 14% from $6.95. The "flat core EPS on flat organic growth" is the headline — this is a no-growth mega-cap at current run rate.

Cash flow: FCF $8.2B (+9%), but the improvement was capex reduction (-17%), not operating improvement (OCF -3%). FCF conversion ≈73%, below the 80% target for FY2026.

FY2026 Guidance (Dec 8, 2025 8-K)

• Organic revenue +2-4% ("expect high end of range during second half")
• Core constant currency EPS +4-6% (implies $8.47-$8.63)
• Core operating margin expansion of 100+ bps over 3 years (2026-2028)
• FCF conversion >= 80%

Nuance on timing: Total company organic acceleration is back-end loaded (H2) — driven by acquisition math (poppi, Alani Nu become organic). But CEO Laguarta specifically said PFNA growth comes "early in the year" — double-digit Frito-Lay space gains in March-April retail resets, affordability investments active now. These are different statements about different segments, and the research initially conflated them. PFNA could beat in Q1 while total company organic remains soft.

8-K Highlights (Last 6 Months)

• Dec 8, 2025: Elliott engagement confirmed. Elliott is "supportive" of PEP's investment/cost-cutting plan. Multi-year thesis — benefits extend well beyond our 15-week window. The NA go-to-market review update is promised for "late 2026."
• Dec 15, 2025: Steven Williams (CEO North America) moved to EVP/Vice Chairman/Global Chief Commercial Officer. Org chart reshuffling under Elliott pressure.
• Oct 9, 2025: New CFO Steve Schmitt from Walmart US. First full quarterly report is April 16, 2026. Cost-cutting DNA aligns with Elliott mandate but adds execution uncertainty.
• Feb 11, 2026: EUR 2.5B senior notes offering. Routine refinancing, not a red flag.

Consensus

Sell-Side

14 Hold / 8 Buy / 1 Sell. Mean target $171 (+12%). Median $173. Range $130-$191.

Recent upgrades (Citi $182, JPM $176, Piper Sandler $181) reflect the post-Elliott engagement narrative: activist pressure + new CFO + cost cuts = margin expansion. The 14 Holds reflect appropriate skepticism about timing. Barclays ($144) is the lone bear.

The Street's consensus narrative: "Quality compounder in a soft patch. Elliott accelerates the fix. Buy the dip for dividend + margin expansion over 2-3 years."

Options Market (April 17 expiry, captures April 16 earnings)

• Implied earnings move: ±5.6% (derived from term structure: Apr 17 IV 33.3% vs Apr 10 IV 28.5%). Large for a staples name.
• ATM IV: 33.3% — 113th percentile of 52-week range (14.4%-31.1%). Most expensive options in a year.
• IV vs HV: 33.3% implied vs 22.2% realized = 50% premium. Options are overpriced.
• P/C ratio: 0.73 OI (bullish), 0.63 volume (more bullish). Market leans toward a bounce.
• Put skew: OTM puts +45.5% vs ATM. Meaningful downside protection bid despite the bullish lean.
• Unusual activity: Deep ITM $130 calls saw 3,330 volume vs 5 OI (666x ratio) — institutional synthetic long or position roll, not a speculative punt.
• Max pain: $150 (2% below current).
• OI concentration: $160 call has 8,354 OI — the largest single position, marking the upside target the market watches.

Translation: The options market expects a bounce (bullish positioning), acknowledges downside risk (elevated put skew), and is pricing more event volatility than PEP typically delivers (50% IV premium over realized). Someone institutional is building or rolling a bullish position.

Positioning

• Short interest: 1.6% of float, 3.4 days to cover. Nobody is aggressively short.
• RSI: 27.6 (deeply oversold). The March -8.7% decline was rotation-driven (staples → growth), not conviction-driven.
• Insider selling: CEO Laguarta sold $4.7M on March 2. Officer Willemsen sold $1.1M on March 4. February 27 transactions were annual compensation grants. Net discretionary activity is mildly negative. LR ≈1.5 — weak signal, could be tax/estate planning.

Relative Valuation (Staples Peers)

PEP is middle-of-pack. Tracking worse than KO (+6.3% vs +10.1% 1Y) — the PEP-KO spread says "same sector, but PEP has the snack problem and KO doesn't." Not in the distress camp (GIS, MKC). The 3.77% yield vs KO's 2.76% shows the market charging PEP a discount for execution risk.

Edge Assessment

Where the edge is

Index construction arbitrage. QQQ must hold PEP at 1.15%. We have discretion to remove it. The counterparty is the index methodology, not an informed participant. No one on the other side of this trade is expressing a view on PEP's fundamentals — QQQ holds PEP because it's a top-100 non-financial NASDAQ stock. Period.

This is the same category as CEG (utility proxy), CMCSA (cable/media), CSX (railroad), CTAS (uniforms), and MAR (hotels). Factor outliers that QQQ holds mechanically and we can filter systematically.

Where the edge is NOT

There is no informational edge. $209B market cap. 23 analysts. Full coverage. Our 10-K and transcript work confirmed what the sell-side already knows. The volume declines, margin compression, tariff exposure, back-end loaded guidance, management transition, and Rockstar write-down are all fully priced in the 16.73x forward multiple and 14 Hold ratings.

The alpha from regression (+6.8% vs SPY) is statistical noise: t-stat 0.31, R-squared 5.6%. We cannot distinguish PEP's true alpha from zero. This is not a name where deep research uncovers what 23 analysts missed.

Honest sizing of the edge

At 1.15% weight, with QQQ expected drift of +3-5% in 15 weeks (historical averages, though current oversold condition adds uncertainty), the expected alpha contribution from removing PEP is 2-5 basis points. This is correct as a systematic process trade and not worth more than its weight implies.

Catalyst Calendar

Q1 Earnings Scenario

• Beat (40%): PFNA volume turns positive on space gains. Street models small beat. Stock bounces 3-5% from oversold. But QQQ likely bounces harder in the same environment (higher beta). Net relative impact: neutral to slightly negative for our filter thesis.
• In-line (35%): Mixed quarter. PFNA volume flat-to-slightly-positive, total organic soft. Stock range-bound. Factor mismatch continues to drag vs QQQ in any up-tape. Net: mildly positive for filter thesis.
• Miss (25%): PFNA affordability investments hurt margins without driving volume. New CFO conservative on guide. Stock drops 4-7%. Net: positive for filter thesis, PEP underperforms QQQ significantly.

Signals

Combined LR Calculation

Independent bearish signals: QQQ beta (0.75) × volume decline (0.80) = 0.60 Correlated signals (MTUM, PFNA OP, tariffs): take strongest, not product. Already captured in QQQ beta and volume. Bullish offset: RSI mean-reversion (1.3) × Elliott (1.1) = 1.43 Partially correlated with options bullish lean — don't multiply again.

Net LR: 0.60 × 1.43 = 0.86 → rounds to ≈0.7 accounting for timing uncertainty.

Prior P(underperform QQQ) = 50% (base rate). Posterior: odds = 0.5/(1-0.5) × (1/0.7) = 1.43:1 → P(underperform) = 59%.

The 65% prediction in worldview was slightly aggressive. Honest math says 59-65% depending on how you weight the factor mismatch mechanical argument (which isn't really a probabilistic signal — QQQ beta 0.08 is a measurement, not a forecast).

What Would Change the Verdict

To KEEP:

• QQQ enters sustained correction (>15% drawdown over 15 weeks). PEP's low beta becomes protective and our hedge can't fully neutralize the intra-basket tracking error.
• PFNA reports a genuine volume inflection in Q1 (positive organic volume, not just positive revenue via pricing). This would signal the turnaround is real and front-loaded, making PEP's absolute return potentially competitive.
• Elliott announces a strategic transaction (spinoff, divestiture) that re-rates the stock 15-20%. Low probability in window but non-zero.

To higher-conviction FILTER:

• QQQ recovers from oversold and resumes uptrend. The factor mismatch becomes the dominant signal as PEP fails to participate.
• Q1 misses on PFNA volume. Space gains don't translate. Affordability investments hurt margins without driving traffic.
• Momentum factor re-asserts after March correction. PEP's MTUM beta of -0.44 becomes an active headwind.

Self-Critique

Strongest part of the thesis: The QQQ beta of 0.08 is a measurement, not an opinion. It doesn't depend on any assumption about PEP's fundamentals, management quality, or competitive position. PEP is in the wrong index. Full stop.

Weakest part of the thesis: The magnitude is small. At 1.15% weight, even a 10% relative underperformance generates only 11.5 bps. And the direction depends on QQQ going up — which is uncertain given QQQ's own oversold condition (RSI 23.1). We are implicitly long momentum by removing anti-momentum names, and momentum is currently in drawdown.

What I got wrong initially: The initial research conflated PFNA-specific guidance ("early in the year") with total company guidance ("second half acceleration"). These are different statements about different segments. Corrected in this memo. Also: the alpha estimate of +6.8% was initially presented without the t-stat context. It's noise, not signal.

Correlation with other removes: PEP joins 5 other factor mismatch removes (CEG, CMCSA, CSX, CTAS, MAR). These are effectively one bet on momentum continuing and low-beta names lagging. If momentum reverses, all six go against us simultaneously. The combined factor mismatch bucket is 6 positions with zero IC and correlated risk — acknowledged in basket notes, proceeding anyway.