The Legendary Debate

The Investment Decision Council: PDD Holdings

Phase 1: Business Quality & Long-Term Moat

Warren Buffett: folds his hands "The customer can always walk across the street. That's the brutal truth of retail—Sears learned it, Kmart learned it, and every marketplace operator eventually confronts it. So my question for this room is simple: what makes 900 million Chinese consumers NOT walk across the street to Alibaba, JD, or Douyin? What keeps them on Pinduoduo—and will whatever that is still work in ten years?"

Charlie Munger: "I'll answer that directly. What keeps them on Pinduoduo is the same thing that keeps people at Costco: they know they're getting the lowest price, full stop. PDD built its entire architecture around connecting Mr. Wang's USB cable factory in Yiwu directly to a consumer in Chengdu, eliminating four layers of distributors who each took a margin. That's not brand loyalty in the Nike sense—it's structural price leadership. The analogy I keep coming back to is Walmart in 1985. Sam Walton didn't have the prettiest stores or the best service. He had the lowest prices, and that was enough because there is always—always—a larger pool of price-sensitive consumers than premium ones. Pinduoduo understood something about China that Alibaba's Tmall missed: most Chinese consumers aren't aspirational shoppers looking for brand validation. They're pragmatists looking for the best deal on laundry detergent and phone chargers. That psychological insight built into the platform's DNA is harder to replicate than any technology."

Dev Kantesaria: "Charlie, I respect the Walmart parallel, but let me invert it. Walmart's moat wasn't just low prices—it was an integrated supply chain with distribution centers, real estate, and vendor relationships that took thirty years to build. A competitor couldn't replicate it without spending $50 billion and two decades of execution. PDD's moat is an algorithm and a user interface. The merchants selling on PDD sell simultaneously on Taobao, on JD, on Douyin's livestream commerce. I've studied toll booth businesses for eighteen years, and the test is brutally simple: can the essential activity occur without paying this company's toll? A merchant in Guangzhou can absolutely list products on five platforms today and shift advertising spend to whichever one delivers the best ROI this quarter. That's not a toll booth—that's a competitive marketplace where your position must be defended every single day through subsidies, fee reductions, and merchant incentives. The RMB 100 billion support program isn't a sign of strength—it's a sign that management knows the switching costs are lower than bulls want to admit."

Pulak Prasad: "Dev raises the competitive question, but I want to go deeper on something more fundamental—the corporate structure itself. We're sitting here debating whether PDD has a durable moat, and nobody has mentioned that we don't actually own the business. We own shares in PDD Holdings Inc., a Cayman Islands shell company that has contractual arrangements—not equity ownership—with the Chinese operating entities where the consumers, merchants, cash, and algorithms actually reside. When I apply my Darwinian framework, I ask: can this organism survive the worst conceivable shock? For PDD, the worst shock isn't a competitor—it's the Chinese government deciding that VIE structures no longer serve the national interest. We saw exactly this playbook in 2021 when Beijing destroyed the entire private tutoring industry overnight. New Oriental went from a $30 billion company to essentially zero in its core business within months. The government didn't need a reason that satisfied Western legal standards. They just did it."

David Tepper: "Pulak, you're conflating two completely different situations. The tutoring crackdown was about social policy—the government viewed private education as exacerbating inequality. PDD is the exact opposite: it serves the broadest base of Chinese consumers with affordable goods, which is precisely what Beijing wants. If anything, PDD is aligned with the 'common prosperity' agenda, not opposed to it. I've traded Chinese assets through three separate crackdown cycles—2015, 2018, and 2021—and the pattern is always the same: Western investors panic, sell everything with a Chinese ticker at fire-sale prices, and then the stocks recover when the regulatory dust settles. The VIE risk is real but it's priced in at 8.6x earnings. Meta trades at 25x, Amazon's marketplace at 30x-plus. You're getting a 60-75% discount for the China risk. The market isn't stupid—it knows about VIE structures—but the question is whether that discount is appropriate or excessive."

Mohnish Pabrai: "David's right about the discount, but I want to steer us back to Warren's original question about the moat because there's a subtlety everyone is missing. PDD didn't just build a low-price marketplace—they pioneered a fundamentally different discovery model. Traditional e-commerce is search-based: you know what you want, you type it in, you compare prices. Pinduoduo is recommendation-based: the algorithm shows you things you didn't know you wanted at prices that trigger impulse purchases. That's a meaningfully different consumer behavior pattern, and it's why PDD's engagement metrics—time spent per session, purchase frequency—consistently exceed Taobao's despite Alibaba having twenty years of head start. The group-buying mechanic, the gamification, the social sharing—these create habitual behavior loops that are genuinely sticky. Dev, you asked whether merchants can switch platforms, and they can. But the more important question is whether consumers switch their daily browsing habit. That's harder to change than you think."

Robert Vinall: "Mohnish, that's an important distinction, but I want to raise something that hasn't been addressed yet—Temu. Because when we talk about PDD's future, we're really talking about two businesses. Domestic Pinduoduo is a mature, dominant marketplace with proven economics. Temu is a three-year-old international experiment burning capital to acquire customers across fifty-plus markets while facing tariff headwinds, regulatory scrutiny under the EU's Digital Services Act, and competition from Amazon, Shein, and local incumbents. I want to apply the Guy Spier 'away from desk' test here: if I bought PDD today and couldn't check the price for twelve months, would I sleep well? For Pinduoduo alone, probably yes—it's a proven business with extraordinary returns on capital. But Temu introduces a wildcard that could consume billions in cash with uncertain returns. Management's own language on the earnings call was remarkably candid—Chen Lei said there are 'significant uncertainties exposing the company to risks that are unpredictable and difficult to quantify.' When a CEO tells you uncertainty is unpredictable and difficult to quantify, that's not boilerplate—that's a warning."

Warren Buffett: "Robert makes an excellent point, and it connects to something that's been nagging me. I've always believed that the best businesses don't require heroes—they produce great results even with mediocre management. Coca-Cola can have a bad CEO for five years and barely notice. Could PDD survive a bad CEO? The honest answer is I don't know, because the business is only ten years old, has had essentially one leadership team, and operates in an environment where government relationships matter enormously. That opacity troubles me. American companies I invest in, I can read the proxy, attend the shareholder meeting, talk to board members. With PDD, I'm trusting a Cayman Islands entity to faithfully represent the interests of foreign shareholders when those interests might conflict with Chinese regulatory priorities. That's not a fatal flaw, but it's a genuine disadvantage that deserves a real discount."

Phase 2: Financial History & Long-Term Growth

Charlie Munger: "Let's move to the numbers, because the financial trajectory here tells a story that either confirms or contradicts everything we just debated. Revenue went from $268 million in 2017 to $54 billion in 2024—a compounding rate of roughly 115% annually over seven years. That's one of the fastest revenue ramps in the history of commerce, faster than Amazon, faster than Alibaba at the equivalent stage. But what's truly remarkable isn't the revenue growth—it's the profit trajectory. This business went from losing $1.5 billion in 2018 to earning $15.4 billion in 2024. ROIC expanded from 6.3% to 32.8% in just three years. When a business grows revenue tenfold and simultaneously improves capital efficiency fivefold, that's the financial fingerprint of a genuine platform flywheel. The marketplace creates operating leverage because the marginal cost of another transaction approaches zero—the algorithm and payment infrastructure that serves ten million orders a day can serve fifty million without adding a single server."

Dev Kantesaria: "Charlie, I'll grant you those are extraordinary numbers. But let me point to something troubling in the recent data. The TTM ROIC has already dropped from 32.8% to 24.2%. Operating margins peaked at 27.5% in 2024 and have compressed to 22.4% in the trailing twelve months. R&D spending surged 41% in Q3. And Co-CEO Zhao Jiazhen explicitly said on the call that investments similar to the RMB 100 billion support program 'will continue in the long run' and that these investments 'will affect the sustained performance of revenue and net profit.' That's management telling you in plain language that the ROIC peak is behind us. The margin trajectory from here is not up—it's flat to down as competitive reinvestment absorbs the operating leverage. I've seen this pattern in platform businesses before: extraordinary returns in the scaling phase, followed by margin compression as competitors force a reinvestment cycle. The financial history is spectacular. The financial future may be very different."

David Tepper: "Dev, you're reading those numbers through the most pessimistic possible lens. Let me reframe. Even at 22.4% operating margins and 24% ROIC—the supposedly 'compressed' figures—PDD is generating returns that would place it in the top decile of all publicly traded companies globally. The depreciation line tells the real story: $362 million against $54 billion in revenue. That's a capital intensity ratio below 0.7%. This business converts virtually every dollar of revenue into cash rather than replacing worn-out assets. And look at the cash flow history—FCF per share went from $1.73 in 2017 to $11.97 in 2024, a 31.8% CAGR. Even if FCF growth decelerates to 10-15% from here—a fraction of historical rates—you're still getting paid an absurd starting yield at today's price. The 2024 reported FCF was distorted by heavy investment activity, which is why you see $122 billion in operating cash flow but only $3.6 billion in reported FCF. Normalize for the investment deployment, and true owner earnings are running around $15-17 billion."

Robert Vinall: "The reinvestment math is what I keep coming back to. PDD has essentially two reinvestment engines running simultaneously. Engine one: domestic Pinduoduo, where incremental investment goes into merchant support, agricultural supply chain, and logistics partnerships—this reinvestment earns extremely high returns because it deepens the ecosystem that generates 900 million repeat buyers. Engine two: Temu, where capital is being deployed into customer acquisition across Western markets at returns that are currently negative but improving. The question for any compounder is whether the reinvestment runway extends far enough at attractive enough returns to justify holding for a decade. For Pinduoduo domestically, I think the answer is clearly yes—Chinese consumption is still under-penetrated in rural areas, and PDD's agricultural commerce model has structural advantages. For Temu, the returns are genuinely uncertain, and the tariff environment—particularly the potential elimination of the $800 de minimis threshold in the U.S.—could impair the unit economics permanently. When I look at the EPS trajectory from negative $2.02 in 2018 to positive $11.13 in 2024, I see a business that has proven it can monetize at scale. But the trajectory from here depends heavily on how much of that earnings power gets reinvested at uncertain returns versus returned to shareholders."

Mohnish Pabrai: "Robert, let me add a dimension to your reinvestment analysis that makes the current valuation even more absurd. Look at the balance sheet: stockholders' equity went from essentially nothing in 2017 to $313 billion RMB—roughly $43 billion USD—in 2024. Book value per share went from negative $0.18 to positive $18.55 in seven years. The business is not only growing earnings at extraordinary rates but accumulating tangible equity at an even faster pace. And yet the stock trades at roughly 0.4x the book value of the equity in USD terms. Even if you apply a heavy VIE haircut—say the cash and equity are worth 50 cents on the dollar because of the structural questions Pulak raised—you're still buying the stock near liquidation value while the business earns 24% on capital. That's the kind of disconnect I've only seen a handful of times in my career."

Phase 3: Valuation & Final Verdicts

Warren Buffett: "Let's bring this to the question that matters: at $96.19, what are we actually paying? The trailing EPS is roughly $11.13 per ADS when you properly convert the RMB financials and divide by the 1,382 million ADS outstanding. That's an 8.6x P/E. The cash and short-term investments on the balance sheet represent roughly $20-25 per ADS, depending on which quarter you reference—roic.ai shows $45 to $54 billion in the working capital detail across recent quarters. Strip out the cash and you're paying about $71-76 for the operating business, which puts the ex-cash P/E around 6.5x. Now, I've been doing this for seventy years, and I can count on one hand the number of times I've seen a business earning 24% ROIC with 57% gross margins trading at 6.5x ex-cash earnings. The problem is that every one of those situations had a genuine structural risk that explained the cheapness—and here, that risk is the VIE structure and the China regulatory overhang. I'd want to own this, but at $88 or below—a price that gives me 30% margin of safety against a fair value I'd estimate around $125 using a conservative 12x multiple on normalized $10 EPS."

Charlie Munger: "I'm in a similar place as Warren but with slightly more conviction. Applying inversion: for this investment to destroy capital, you need one of three things—VIE invalidation, permanent margin destruction, or Temu becoming a cash incinerator. The combined probability of any one of those proving fatal within five years is perhaps 15-20%, and at 8.6x earnings, that risk is already embedded in the price. My fair value is around $130—13x normalized earnings, which is still a massive discount to Western platform peers. I'd start buying at $90 and build a position over time. The quality of the underlying business is genuinely exceptional; the uncertainty is entirely about whether foreign shareholders will ever capture that value."

Dev Kantesaria: "I'm going to be the disciplined voice here. PDD fails my inevitability framework, and no amount of cheapness changes that. A merchant in Shenzhen can list on five platforms simultaneously and shift advertising spend quarterly based on ROI. There's no mandatory checkpoint, no licensing bottleneck, no structural lock-in. The RMB 100 billion in merchant subsidies proves my point—if you had true toll booth economics, you wouldn't need to pay merchants to stay. I avoid this stock entirely. Not because the business is bad—it's actually quite impressive operationally—but because competitive marketplaces require constant reinvestment to maintain position, which makes the five-to-ten-year earnings stream inherently unpredictable. That's outside my framework."

David Tepper: "Dev, I respect the framework, but you're applying a steady-state compounding lens to what is fundamentally a catalyst-driven asymmetric opportunity. At $96.19, the market is pricing in perpetual FCF decline of 1.6% annually for a business that compounded FCF per share at 31.8% over seven years. The perception gap is over 33 percentage points—one of the widest I've seen in large-cap equities. My catalysts are concrete and near-term: first, the $10 billion buyback authorization, which at current prices would retire 3-4% of shares annually. If the Q4 share count drops below 5,500 million ordinary shares from 5,679 million in Q2, that's proof of cash accessibility through the VIE—and the stock re-rates immediately. Second, Temu approaching contribution-margin breakeven in at least one major market. I'm buying now, full position, $96.19."

Mohnish Pabrai: "I'm with Tepper on the timing, and here's why the math works despite the $134 billion market cap. Normally I'm skeptical of mega-caps because you need heroic assumptions for a triple. But PDD is trading at 8.6x earnings—strip out $30 billion in cash and the operating business is valued at $105 billion. If margins stabilize and the P/E normalizes to 15x on $11 earnings, that's a $165 stock—72% upside. Add Temu optionality reaching profitability and you're looking at $180-200. Meanwhile my downside is $85-90 in a bear case where margins compress to 18% and the multiple stays depressed—that's 10% down. I'm risking a dime to make seventy cents. The asymmetry is extreme. Buy now."

Robert Vinall: "I'm more measured than David and Mohnish. The reinvestment runway is extraordinary, but the visibility on Temu's returns is genuinely poor, and management's warning about unpredictable profitability fluctuations deserves respect—these are not executives sandbagging for a beat, they're preparing you for margin volatility. I'd estimate fair value around $135 using a sum-of-parts framework: $120-130 billion for domestic Pinduoduo at 10x operating profit, $15-20 billion for Temu optionality, plus net cash. I'd start accumulating at $92—a 32% margin of safety that compensates for both China risk and reinvestment uncertainty."

Pulak Prasad: "My position hasn't changed. The underlying economics are exceptional—I freely admit that a U.S.-listed company with these metrics would warrant serious attention. But the VIE structure introduces an existential fragility that cannot be priced with any confidence. We don't have a base rate for VIE invalidation because it's never happened at this scale, which means assigning '5% probability' to it is false precision. When I can't model the downside distribution, I don't invest. I'll watch the $10 billion buyback execution as a real-world test of cash accessibility—if PDD can demonstrably move cash from Chinese entities to Cayman-level share repurchases over the next year, that's genuine evidence against the VIE bear case. Until then, I avoid."

Phase 4: Synthesis & Conclusion

Warren Buffett: surveys the room "Let me try to pull together where this discussion has taken us, because I think we've surfaced something genuinely important about what separates a great business from a great investment. On the qualitative side, there's broad agreement in this room that PDD has built something remarkable. The direct-from-manufacturer model that eliminates intermediary layers, the recommendation engine that creates habitual engagement, the 900 million buyer base that gives merchants no choice but to participate—these are real competitive advantages. Charlie's Walmart analogy resonates: the largest pool of consumers will always be the price-sensitive ones, and PDD owns that segment in the world's largest consumer market.

Where we diverge is on the critical question of whether foreign shareholders can actually capture the value this business creates. Dev and Pulak make a structurally sound argument: the VIE arrangement means we own contractual claims on a Cayman shell, not equity in the Chinese business that generates $15 billion in annual profit. That's a genuine structural handicap, not a sentiment-driven discount. The earnings call reinforced this concern—when Co-CEO Zhao says profitability 'will continue to fluctuate' and investments 'will continue in the long run,' he's telling us that near-term earnings are not the right metric for valuation.

But here's what the bears must contend with: the financial trajectory is almost unprecedented. ROIC from 6% to 33% in three years. FCF per share compounding at 32% annually. Depreciation of $362 million on $54 billion in revenue—a capital intensity ratio that would make any platform investor salivate. Even at the 'compressed' TTM margins of 22.4%, PDD generates returns in the top decile globally. The $10 billion buyback program is the first concrete test of cash accessibility, and David's right that the Q4 share count will be the leading indicator.

Five of us would own this business at some price—three at modestly lower levels, two today. Dev and Pulak would avoid it entirely, and their reasoning isn't dismissible. When reasonable people with decades of experience look at the same facts and reach opposite conclusions, the honest answer is that the uncertainty is genuinely irreducible. My position is Buy Lower at $88—a price where the margin of safety compensates for the structural unknowns. At today's $96, the odds favor the buyer, but not by enough for a man who sleeps best when he's paid for his patience."