Pershing Square Challenge 2026 third place: Celsius $CELH

Summary

• The pitch’s core claim is a straightforward mispricing: Celsius trades at ~20x forward earnings with analyst-level earnings growth comparable to Monster, which trades at 34x, and below Coca-Cola’s ~24–25x despite growing far faster. Jas Tolani calls it “a clear arbitrage opportunity” — a below-market multiple on a company “growing way above market expectations” with expanding operating margins. The team attributes the discount to the market overlooking newly acquired Alani Nu, up about 60% YoY in Q1, while pricing in terminal-value risk from new entrants and Costco’s Kirkland launch.
• The team’s proprietary survey of 500+ consumers via the Prolific panel found Alani had “the highest repurchase score in the entire category,” Celsius scored in line with Monster and Red Bull, and newer entrants like Bloom and Ghost showed weak loyalty even among their own buyers. But 63% of respondents said they’d switch brands if their favorite were out of stock and another 8% would “buy whatever drink is available” — over 70% switching on the spot — which the team reads as evidence that “distribution and shelf presence are so important in this category.”
• Andrew Walker flipped the survey’s out-of-stock data into a bear case: if roughly nine in ten shoppers will still buy an energy drink when their brand is missing, that argues distribution beats brand — and a competitor could buy out the shelf. His analogy: refuse Diet Pepsi when the restaurant is out of Diet Coke and “it’s probably 2/3 saying no” — far stickier than the survey implies for energy. The team’s rebuttal: Pepsi’s network is among the most penetrated in North America, and a former Alani/current Monster sales rep described incumbents defending shelf space with promotions and price cuts.
• On the Kirkland scare that knocked the stock about 15% in February or March, the team argues private label works for price-sensitive staples like toilet paper, not brand-driven drinks — citing Kirkland sodas, where Coke and Pepsi perform well, and light beers, with private label holding about 0.5% of energy overall. Walker pushed back that Celsius’s Amazon-bulk-ordering, millennial-female consumer looks more Costco-vulnerable than the impulse convenience-store buyer; the team countered Costco is 10% of Celsius revenue versus a 5% industry average, still small, and Q1 growth hasn’t fallen.
• The Pepsi relationship is both the moat and the concentration risk: Pepsi distributes Celsius, holds an 11% equity stake, and had Rockstar, which Walker characterized as mismanaged before it was sold to Celsius — but Walker noted a 20x multiple underwrites 20+ years and “Pepsi might change their strategy 7 years from now,” with no fallback distributor since Coke has Monster and KDP has Ghost. Tolani’s reply leans on an ex-EVP call that innovation is difficult for legacy brands; Okada added that Monster has faced a comparable Coke risk without its multiple reflecting it.
• Against Walker’s protein-category analogy — Ensure and Muscle Milk disrupted by newcomers who were themselves disrupted, with Liquid Death, Bloom, C4, Bucked Up and influencer brands now flooding energy — the team argues both Celsius and Alani have reached about $1.5B in revenue, and “no energy drink company has gotten to that size and has failed.” Fads like Prime and Bang collapsed because their loyalty was short-lived; Tolani cited Celsius’s SG&A falling about 500 bps as a share of revenue as evidence the company is not throwing rising marketing spend at current sales.
• The team’s model underwrites 18% sales CAGR led by Alani, about 200 bps of operating-margin expansion, and about $2.60 of 2028 EPS; returning to the pre-Kirkland multiple produces roughly a 25% yield, with Monster’s international playbook — about 5–10% to 40% of sales in 10 years — as the longer-dated bull case. Walker’s framing of the tradeable angle: even conceding competition risk, it “plays double for Monster” — making the setup almost a relative long-Celsius/short-Monster arbitrage rather than a pounding-the-table absolute long.

Deep dive

1. Why two Columbia Business School students picked a contested stock they could actually test

• The team — Jas Tolani (private credit, venture debt and private equity experience in Mumbai, with a consumer focus) and Hideo Okada (Japanese commercial-bank credit analyst, company-sponsored at CBS) — took third in the Pershing Square Challenge with Celsius, chosen because “there was so much noise on both sides of the field in terms of it being a potential short as well as a potential long,” and a proprietary survey could help determine which side to take. Tolani is explicit: “had that come out to be negative, you could have easily gone short.”
• Okada’s disarming credential: “I’m probably one of the most experienced energy drink consumers” — matched by Walker, who auto-orders Celsius packets from Amazon (strawberry coconut) and opens by confessing energy drinks as a passion project (“do not tell my wife”).

2. The thesis: the market is paying for legacy Celsius and ignoring Alani

• Tolani’s setup: Celsius acquired Alani Nu, the market hasn’t valued its growth, and Q1 showed Alani up about 60% YoY — which, if the trajectory continues, supports the team’s projection of about 18% three-year-forward growth against a 20x forward multiple. “That’s below market multiple for a company that’s growing way above market expectations… as well as expanding operating margins.” Monster sits at 34x forward, Coca-Cola at about 24–25x, yet analysts believe Celsius earnings can grow as much as Monster’s: “a clear arbitrage opportunity from our perspective.”
• His two explanations for the discount: the market anchors on core Celsius, “sort of like a legacy brand now,” growing only about 6%, and a terminal-value problem from crowding — Bloom, KIMADE (the new Kim Kardashian brand), Ghost via the KDP tie-up, and the Kirkland private-label launch that “crashed the Celsius stock quite a bit.”

3. Kirkland is a headline risk, not a thesis risk — with one honest caveat

• The rebuttal has two legs: roughly 70% of energy-drink purchases in convenience stores are impulse buys (“you ain’t going to go to Costco and buy this in bulk unless you’re a die-hard fan”), and private label works better for price-sensitive staples than branded drinks — Coke and Pepsi perform well against Kirkland sodas, with a similar point for light beers, while private label holds about 0.5% of energy overall. Tolani’s kicker: “if they decide to copy Celsius now and not Monster and Red Bull… why would you copy something that’s not working?”
• Walker’s pushback is specific to Celsius’s consumer: management itself touts over-indexing to Amazon, with millennial working women buying 12-packs for a daily morning ritual — “when I hear order from Amazon, I am hearing we order in bulk, and that Costco is a risk.” Tolani concedes the skew (Costco is 10% of Celsius revenue versus a 5% industry average) but calls it small, notes both brands are still sold at Costco, and points to Q1 growth holding, with Q2 as “a great indicator for that thesis.”
• Walker’s supporting history: even identical-tasting store brands have struggled in drinks because the brand still matters, unlike paper towels where the brand is largely invisible.

4. The survey — and the host’s inversion of its best slide

• Tolani’s three findings from the Prolific-panel survey: Alani had “the highest repurchase score in the entire category”; Celsius scored broadly in line with Monster and Red Bull (“not a fad brand”); and Bloom and Ghost buyers rated established brands higher than their own — “entrants in the female segment are struggling to build real loyalty, while Alani already has it.” On out-of-stock behavior: Tolani reported that 63% switch brands, while 8% buy whatever is available — “more than 70% of buyers are willing to switch on the spot rather than walk out empty-handed.”
• Walker’s contrarian read of slide 10: he rounded the categories to roughly nine in ten still buying an energy drink, while 12% would go to another store and 4% would skip the purchase entirely — so a Ghost or Monster could buy out a 7-Eleven’s shelf knowing “nine out of 10 people are going to drink it and still buy the Monster.” His counter-anecdote: offered Diet Pepsi when the restaurant lacks Diet Coke, “it’s probably 2/3 saying no” — real brand loyalty looks stickier than the survey’s switching data.
• Tolani’s answer: brand loyalty matters when all options are on the shelf, and Pepsi’s network — one of the most penetrated in North America — keeps Celsius and Alani there. A sales representative who moved from Alani to Monster described the incumbent playbook: “when they want more shelf space, they simply lower prices, put out promos, and get that done.” Tolani also argued that if competitors’ customers are more likely to choose Alani, that strengthens Alani’s standing. Okada’s summation: “Celsius has both the brand and the distribution network… somebody who has both is pretty rare. We thought this is the moat.”

5. Everything runs through Pepsi — moat and single point of failure

• Walker’s structural worry: at 20x “you kind of need 20 years plus of earnings to just get your money back,” Pepsi could change strategy in seven, and there is no alternate distributor — Coke is locked to Monster, while KDP has Ghost and, in Walker’s view, less effective distribution. “It feels like all the power’s over there.”
• The team’s two-part reply: an ex-EVP told them “innovation is typically very tough to come by with existing legacy brands” — Monster has tried to compete with Celsius but has not captured its market, and Coke’s last decade has been acquisition-led rather than focused on organic launches; and Pepsi’s 11% equity stake means starting its own energy brand “would be for it to devalue its own equity stake.” Okada also pointed out that Pepsi had Rockstar and could have grown an energy brand through it, but did not; Walker had characterized Rockstar as mismanaged before its sale to Celsius.
• Walker then pointed to the Monster-Coke precedent: the companies went “to the brink” once or twice and always came back to the table, while Okada argued that Monster’s multiple has never reflected the comparable risk — making it unfair to penalize Celsius solely for that concern.

6. Is Alani a sugar high? The protein-aisle analogy vs. escape velocity

• Walker channels a consumer analyst from a trade call: energy today looks like protein a few years ago — Ensure and Muscle Milk were disrupted by newcomers, then “the newcomers got disrupted by the new newcomers.” He can’t recognize 95% of protein brands at Vitamin World 20 years after working there, and lists the incoming wave: Liquid Death (“if they can sell freaking bottled water in a can, what can they not sell”), Bloom, C4, Form, Bucked Up, Bomb Energy and influencer brands.
• Tolani’s counter-data: both brands are about $1.5B in revenue today and “no energy drink company has gotten to that size and has failed.” Prime — which the speakers identify as Jake Paul’s brand — briefly ranked above Monster in Google trends, then fell equally fast because customers tried it once and did not keep buying. Walker adds Bang’s bikini-Instagram virality as the same pattern. The differentiator is “continuous repeat customer purchase without spending marketing dollars”; Tolani cited SG&A falling about 500 bps as a percentage of revenue this quarter as evidence that the current sales are not being supported by rising marketing spend.
• Alani’s community supplies flavor ideas: the cotton-candy flavor, one of its top-selling SKUs, came from TikTok comments. Tolani argued that community-building, scale and Pepsi distribution make the brands less likely to be fads.
• On the “buying growth because the core stalled” critique: Tolani reframes the deal as $1.8B for a brand that did $1.3B in sales that year, filling a Gen Z-women hole Celsius wasn’t winning — “pretty efficient capital allocation” — and creating a portfolio-of-brands approach. International expansion, including an Aston Martin F1 sponsorship and Suntory distribution, is deliberately cautious and represents only about 5% of revenue, with a three-to-five-year horizon.

7. Valuation: modest assumptions, a re-rating, and the relative-value angle

• The model: about 18% sales CAGR led by Alani, about 200 bps of operating-margin expansion over three years, and about $2.60 of 2028 EPS. Walker cited consensus of $2 in 2027 — roughly 15x on a $30 stock, with about a $7.5B market cap and $10B EV. Returning to the pre-Kirkland multiple would produce about a 25% yield. The long-dated kicker is the Monster template: international sales went from about 5–10% to 40% in 10 years, and Tolani called Asia “a very low-hanging fruit.” Okada said he was waiting for Celsius and Alani to reach Japan, where they were not yet available.
• Walker’s honest hedge: the absolute case isn’t “smashing me in the face alpha,” but the relative case is clean — Coke forecasts 8% growth at a higher multiple, Monster trades at roughly 34–35x with similar growth, and “the competition concerns play double for Monster cuz Monster’s got more of the market.” Tolani’s close: even discounting growth, margin expansion gives “double-digit earnings growth… of 12–15% at a lower than market multiple — still a unique opportunity.”