What stands out to me is not just that JD’s 2025 profit collapsed, but how violently the earnings architecture broke despite a still-huge and still-growing top line. Revenue went from $171.0B to $193.2B in 2025, a very respectable 13% gain for a business already this large, and gross profit rose from $27.1B to $31.0B. So the customer franchise did not crack. The problem is that nearly all of that incremental gross profit was consumed below the gross line: operating income fell from $5.84B to just $545M, and net income dropped from $6.10B to $2.90B. That is not ordinary retail volatility; it implies an aggressive expense step-up or subsidy cycle that management chose, not something forced by core demand weakness alone. When a company adds $22B of revenue and $3.9B of gross profit yet destroys over $5B of operating profit, the key question is no longer “is the business growing?” but “can management stop spending?” On that score, JD currently looks less like a mature compounder than a company temporarily opting out of margin discipline.

The other thing I see is that valuation is cheap, but not automatically mispriced. At $41.8B market cap against $193.2B revenue, 0.24x sales is low by any standard; EV/revenue at 0.20x is lower still given the $20.3B cash and $15.8B debt. Book value of $33.2B means the stock is only around 1.4x equity, and a 3.2% dividend yield is unusual for an internet retailer. But the balance sheet strength is being asked to underwrite a strategic experiment at exactly the wrong time in the cycle. ROIC of 1.7% and operating margin of 0.28% say the current capital allocation is poor, full stop. I do not read the numbers as a broken franchise; I read them as a healthy core business whose economics are being obscured by management’s willingness to burn margin. That distinction matters because it means solvency is not the issue, but it also means investors should not casually capitalize 2024 earnings as if they are the true run rate. The stock is cheap on assets and sales, but only conditionally cheap on earnings.

I agree with Opus’s central pushback against the “extreme dislocation” narrative. Opus argues that “the market may simply be correctly pricing a capital-allocation regime change,” and I think that is exactly right. A stock on 16.3x trailing earnings and 6.9x EV/EBITDA is not screamingly mispriced when operating income has just fallen 91% year over year and operating cash flow has dwindled to $2.8B. I also agree with Opus that gross margin stability matters: gross margin actually improved to 16.0% from 15.9%, which strongly suggests the core commercial engine is intact and the damage sits in operating expenses. And I agree that the insider tape is a non-signal; option exercises and in-kind/tax-related entries tell us nothing. Where I diverge is on the degree of comfort embedded in the “fairly valued near $31 with asymmetric optionality” framing. That sounds slightly too forgiving to me. If 2025 operating income is only $545M on a $193B revenue base, even a modest further deterioration can erase EBIT entirely. Once that happens, the market stops valuing “normalized” earnings generously and starts demanding proof. Optionality exists, yes, but it is management-dependent optionality, and that deserves a discount.

I disagree more directly with Opus’s suggested fair value range of $32-42. That upper end assumes investors will look through the current P&L damage and underwrite a rebound before it is visible in reported results. I doubt that. The raw data do not show a business troughing; they show a business still actively sacrificing economics. Net income of $2.90B against a $41.8B market cap is not a distressed multiple if earnings are falling by 52%. In fact, if one haircuts earnings further toward $2B or below, the stock quickly ceases to look obviously cheap on P/E grounds. Put differently: 0.24x sales is attractive only if management eventually restores something like the 2024 operating margin of 3.4%; at the 2025 margin of 0.28%, low sales multiples are deserved. I do agree with Opus that if management were to pull back on low-return initiatives, earnings could normalize sharply and the stock would rerate. But that is a contingent turnaround case, not present-value fact. So I would not call JD fairly valued with upside optionality; I would call it modestly undervalued only below a clearer margin of safety, because the current year’s earnings power is too fragile.

A careful skeptic of both my view and Opus’s would say we are both over-interpreting one ugly annual print without segment detail. They would argue 2025 could include temporary investments, accounting noise, or timing effects that annual aggregates exaggerate, and that a company with $20.3B cash, net cash of roughly $4.5B, and a history of profitability should not be judged on trough optics. They would also note the opposite risk: both of us may be too eager to “normalize” profits at all, because Chinese e-commerce competition may have permanently lowered acceptable margins and JD’s integrated model may simply earn less than investors once thought. That skeptic is useful. But with the data provided, I land on a simpler conclusion: the franchise is real, the spending is self-inflicted, and until management proves it can convert $31.0B of gross profit into something better than $545M of operating income, the stock deserves caution rather than a recovery multiple.