Airwallex waits out a 2026 calendar already carrying $132bn of equity supply

Start with the supply side, because that is where the number sits. Four offerings on Fortune's 2026 tally add up to about $132 billion of new paper, and SpaceX's $85.7 billion is roughly 65% of that on its own ^{[4][6][7][1][4]}. Airwallex's December and June rounds together come to $650 million ^[2], which is under half a percent of the same figure ^[3], and the two sums are not chasing the same money in any tidy way: $650 million is placed with a handful of late-stage funds in a negotiated round, while a mid-cap listing needs an order book from exactly the accounts being asked to absorb Seoul, Shanghai, Hong Kong and whatever SpaceX's syndicate demanded ^[4][6][7].

Liu's own framing is that capital exists and wants evidence, that investors are moving later-stage and want a track record before deploying ^[11]. That is a repricing of risk rather than a shortage of it, and the private market has an instrument for it, which is another letter: Airwallex's June round was a Series H, and Databricks recently closed what would have been a Series M ^[9][17]. Each additional letter is another year in which a company's marks are set by negotiation with a dozen counterparties instead of by a book of hundreds.

What the wait actually buys is more interesting than what it avoids. The $650 million is going into agentic commerce and automated bookkeeping ^[10], and into Mexico, South Korea, Brazil and the U.S. behind a headquarters move from Australia to Singapore and San Francisco ^[16], rather than into underwriting fees and two quarters of management time spent on a roadshow. It also leaves the 2015 cap table unliquid, including Liu's original $1 million seed cheque, twice what the founders asked for ^[18].

This is probably wrong in one specific way. If the constraint were really calendar congestion, then Klook, which filed in New York last November, slid from end-2025 to early 2026 and has been silent since ^[12], should price the moment the tape clears, and Airwallex should follow. If instead Klook stays quiet through an emptier winter, the binding constraint was never the calendar but the standard Liu described, and every mid-cap waiting for a window is waiting for the wrong thing. Watch which one happens before assuming the door is shut.

At a claimed $1 billion annualized run rate serving more than 675,000 businesses ^[2], $650 million is close to eight months of revenue sitting on the balance sheet ^[5]. That is a lot of patience for sale.