Amazon Faces Its Q3 Test With Shorts Rebuilding and Retail Attention Spiking

AMZN reports Q3 results on Thursday after the close, with the stock at $253.10, down 3.6% over the past week and still roughly 7% below the all-time high set after Q2's blowout numbers. The key tension is not positioning, which remains light, but whether the quarter can clear the demanding bar Q2 set.

Short interest has risen meaningfully this week but remains too small to be a structural story. Roughly 94.7 million shares are borrowed against the float, up nearly 9% over the past seven days after a step-change higher on October 23. Even so, that puts SI at less than 1% of the free float. The borrow market confirms the same picture: availability is essentially unconstrained, with more than 10 billion shares available to lend, and cost to borrow at 0.26% is among the cheapest levels possible for a name of this size, though it has risen 27% over the week from a very low base. The options market is equally relaxed. The put/call ratio at 0.73 is almost exactly in line with its 20-day average and shows none of the defensive skew that tends to appear when investors genuinely fear a miss. There is no structural short-side pressure, and no unusual demand for downside protection. What there is, separately and worth noting, is a sharp spike in public attention: Wikipedia page views are running more than 2.4 standard deviations above Amazon's own 90-day baseline as of October 22, the highest retail attention reading of the period.

The analyst community's direction is firmly bullish, but the texture of recent moves reflects some caution on near-term valuation. Morgan Stanley's Brian Nowak raised his target from $335 to $350 on October 19, maintaining Overweight, a move that runs counter to Wells Fargo and Cantor Fitzgerald, both of which trimmed targets on October 12. The Street mean has drifted to $332.41, implying roughly 31% upside from current levels. Factor scores support the broad bullish lean: Amazon ranks in the 94th percentile on EPS surprise, and analyst recommendation differentiation scores at the maximum 100, meaning the coverage universe is as uniformly positive as it gets. The weak spot is value. The EV/EBIT factor ranks at just the 13th percentile, and the trailing PE at 25.3x and EV/EBITDA at 10.5x are not stretched by Amazon's own history but are also not cheap heading into a print where growth expectations are high. The bear case, as articulated by the company's own strategic disclosures, centres on capital intensity: AWS infrastructure, fulfillment, logistics expansion, and the Globalstar transaction are all consuming capital at a rate that may keep earnings leverage constrained even as revenue grows.

The Globalstar angle adds a secondary layer worth tracking. ORTEX's Index Rebalance Radar forecasts an 84% probability that Amazon's share count increase from the agreed GSAT acquisition flows through into the Russell 1000 and Russell 3000, with a 65% probability for the MSCI World and FTSE All-World. The estimated passive buying flows are $206 million for MSCI World and $117 million for FTSE All-World, though all dates are estimated and no index provider has confirmed them. These are mechanical flows contingent on deal completion, not near-term catalysts.

Insider activity over the past 90 days has been one-directional. Net sales over the period total roughly $18.3 million across 70,600 shares, with the most recent disclosed trade being Douglas Herrington's routine 1,000-share sale under a 10b5-1 plan on October 1. Jeff Bezos's transactions in August were gift transfers, not market sales. Every open-market sale in the window was made under a pre-arranged plan, which substantially reduces the informational content of the activity. The institutional register shows broadly stable ownership: BlackRock added 12.8 million shares in the last reported period, JP Morgan added 9.8 million, and Fidelity added 6.9 million, while Bezos himself trimmed slightly. Net institutional direction is additive, but none of the moves represent a departure from index-driven drift.

The last three earnings reactions on record are striking: the most recent two Q2 readings each show a next-day move of roughly 20%, with a five-day move of similar magnitude. Q2 2026 produced a 20.6% one-day gain. That history illustrates how binary Amazon's earnings events have become at this size, and Thursday's print is the first opportunity since then for the data to either confirm or complicate the story Q2 told. The setup going in is calm positioning, universal analyst support at prices well above where the stock trades, and a stock that has spent three months failing to recapture its post-Q2 peak despite the Street's targets sitting a third above current levels.

What matters on Thursday is not whether AWS grew, but whether the growth rate sustained the acceleration Q2 logged, and whether operating margin held or expanded given the capital investments bearing down on the income statement.