Delight in watching development of robotics

At the recent World Humanoid Robot Games in Beijing, I happened to find myself a few seats away from Wang Xingxing, founder of Unitree Robotics. When the opening ceremony ended, most of the corporate executives in the row got up and left. Wang stayed. He kept watching. And then he laughed.

Not the polite, camera-ready smile executives wear at public events. He actually laughed out loud — several times — as robots sprinted, stumbled, crashed, scrambled back to their feet and tried again.

That caught my attention because I had seen Wang at another milestone only days earlier: Unitree's stock market debut.

For most entrepreneurs, ringing the listing bell is the kind of moment that is celebrated with champagne, hugs and photographs destined for office walls. Wang looked almost unnervingly calm.

He barely smiled.

The contrast was striking. The man who showed little emotion while ringing the bell was visibly delighted watching a robot fall over.

At the listing ceremony, Wang was surrounded by capital: valuations, price-to-earnings ratios, market capitalization and stock charts. At the robot games, he was watching machines run, jump, fail, recover and improve.

Since Unitree listed on Shanghai's STAR Market, investors have experienced both extremes.

On its first trading day, excitement surrounding humanoid robots pushed Unitree's market value above 440 billion yuan ($65.56 billion) at one point. But the rally quickly reversed. Within about two weeks of listing, more than 230 billion yuan had been erased from its peak valuation.

That kind of swing makes for irresistible headlines. It also makes it easy to confuse a stock chart with an industry.

The pullback does not necessarily mean Unitree's fundamentals suddenly deteriorated. New shares often have limited freely tradeable supply, and enthusiasm for humanoid robots had already pushed expectations extraordinarily high before the company even reached the market.

Unitree's IPO was priced at a price-to-earnings ratio of 92.92 times, embedding a substantial growth premium. When investors eventually stopped asking how high the stock could go and started asking how much of the future had already been priced in, some correction was almost inevitable.

The more important question is what lies underneath the share price.

Humanoid robots remain an early-stage industry. The path to mass commercialization is still unclear. Costs remain high, useful applications are limited and a robot that looks brilliant in a demonstration may still struggle to perform the same mundane task reliably for eight hours on a factory floor.

That is the uncomfortable reality behind the excitement. But it is also why Unitree matters.

The company is one of the relatively small number of humanoid-robot makers globally that has reached meaningful shipment scale while also generating a profit. Its revenue approached 1.7 billion yuan in 2025, and its humanoid robot shipments ranked among the world's highest.

In an industry where many companies are still burning cash to prove that their machines can walk, Unitree has already demonstrated strengths in vertically integrated research and development, rapid product iteration and supply-chain management.

There is still a caveat.

Much of Unitree's revenue today comes from universities, research institutes and other professional customers. The truly enormous markets — industrial production and household use — remain largely in pilot stages. Robots that companies or families will buy repeatedly, at scale and for clear economic reasons, are still being developed.

That is why Unitree's listing may ultimately matter more as a benchmark than as a spectacle.

The next phase of the humanoid-robot race will increasingly be judged not by how impressive a demonstration looks, but by orders, deployment, reliability, repeat purchases and economies of scale.

On the track, robots kept making mistakes. Some ran beautifully. Others looked awkward. A few fell hard enough to trigger laughter from the audience.

Wang laughed too.

And suddenly his expression at the IPO ceremony made a little more sense to me.

A stock can rise 400 percent and then lose hundreds of billions of yuan in market value. Those numbers can change in days.

A robot learning how to run, fall and stand back up again is slower. But for a technology obsessive, that may be the part that actually matters.