JD Pre-Earnings: A Wide Cohort In A Choppy China-ADR Setup

JD's Cohort Reflects The China-ADR Volatility Pattern

JD.com's pre-print cohort is one of the wider ones in our China-ADR coverage. The cohort_analyze pull on 2026-05-06 retrieved 207 historical analogs with a tightness score of 0.22. The 5-day forward median is -0.80% with IQR [-5.2%, +3.3%], hit-rate 44%. The 10-day median extends to -1.65%, IQR [-7.5%, +5.1%].

Modestly negative bias, wide IQR — the cohort is telling you that JD's pre-print pattern doesn't separate winners from losers cleanly. Some historical analogs ripped, some collapsed, and the central tendency is mildly down.

The XLU Feature (And What It Means For ADRs)

The top 5-day feature was sector_etf=XLU with positive direction (importance 1.83). When Utilities were outperforming relative-to-cohort, JD's forward returns were stronger. That's the risk-off rotation paradox in China ADRs: utilities outperformance signals US-investor risk-off, and those periods historically produced ADR weakness in the immediate term but ADR strength on the multi-week mean reversion. The signal is real but the timing isn't 5-day-clean.

• n=207, 5d median -0.80%, IQR [-5.2%, +3.3%], hit-rate 44%
• 10d median -1.65%, IQR [-7.5%, +5.1%]
• Cohort tightness 0.22 — wide, choppy
• Top 5d feature: sector_etf=XLU (positive direction, importance 1.83)

Print-Day Signals

JD's post-call cohort match weights three signals: GMV growth excluding annualized, JD Logistics segment margin trajectory, and any color on the cross-border (international) marketplace contribution. GMV ≥6% YoY with logistics margin expansion historically maps to the +5% upper band; GMV sub-3% with flat logistics margin maps to the cohort's central -1% scenario.

Like AMAT and NVDA, this is a print where the pre-print cohort isn't decisive. The trade is the print, not the drift.