The primary market for AI is still talking about scarcity, while the secondary market has already taken a hit on profit expectations for optical modules.
That’s the glaring misalignment today.
In the OKX community, the hottest topic over the last few minutes isn’t just how hot a new stock story is, but rather the two different prices emerging on the same AI chain: pre-listed assets are being pumped up by imagination, while already traded hardware companies are being revalued based on cash flow and order rhythm. In the after-hours Nasdaq data, AAOI closed at $162.88, down 17.17% for the day; COHR closed at $355.94, down 11.44%; MRVL closed at $266.88, down 7.61%. These aren’t just small-cap noise; they play roles in the AI computing power supply chain, such as optical communication, interconnects, or chip supply.
What the numbers really express isn’t just the drop.
If AI demand is truly as perpetually scarce as the primary market narrative claims, the companies closest to the computing power bottleneck shouldn’t all be getting slaughtered in valuation. What the market is currently doubting is the time lag between order fulfillment speed, inventory cycles, and capital expenditure rhythms. Stories can be priced in a year in advance, but revenues cannot.
This gap will spill into the crypto market.
$BTC isn’t an optical module stock, but it operates under the same risk budget. Binance's 24-hour market shows BTCUSDT around $61,528, down 1.97% for the day; ETHUSDT around $1,630, down 2.15%. The drop is much milder than for optical modules, indicating that the crypto space isn’t the main stage for declines, but it hasn’t completely decoupled either. When funds see AI hardware starting to wobble, the first reaction is usually not to buy higher volatility assets, but to first reduce leverage.
Here’s an ironic point: the hotter the IPO expectations for AI, the more one needs to watch the cold shoulder of the secondary market. This is because primary market pricing relies on a sense of scarcity, while secondary market pricing relies on marginal orders. When both sides appear at the same time, one side grabs quotas while the other side cuts positions; the real signal isn’t about who’s right or wrong, but that funds are starting to differentiate levels within the AI narrative.
The most vulnerable to being mispriced are the small assets with AI in their name but no income verification. The most likely to be repriced are the hardware chains that already have orders but whose growth slope is doubted by the market. AI concept coins in crypto are caught in the middle, lacking the constraints of earnings reports from US stocks or the scarcity tickets of top pre-listed assets, relying mainly on liquidity and sentiment.
So today, this line shouldn’t just be seen as a drop in chip stocks. It’s more like a risk budget check-up: when AI hardware starts being asked when profits will materialize, the on-chain AI narrative will also face the same question—users, revenue, costs, retention—what can actually support valuation?
That initial misalignment still exists. The primary market talks about the future, while the secondary market does the math. As more people start doing the math, all assets inflated by imagination will be forced to hand in their homework.
That’s the glaring misalignment today.
In the OKX community, the hottest topic over the last few minutes isn’t just how hot a new stock story is, but rather the two different prices emerging on the same AI chain: pre-listed assets are being pumped up by imagination, while already traded hardware companies are being revalued based on cash flow and order rhythm. In the after-hours Nasdaq data, AAOI closed at $162.88, down 17.17% for the day; COHR closed at $355.94, down 11.44%; MRVL closed at $266.88, down 7.61%. These aren’t just small-cap noise; they play roles in the AI computing power supply chain, such as optical communication, interconnects, or chip supply.
What the numbers really express isn’t just the drop.
If AI demand is truly as perpetually scarce as the primary market narrative claims, the companies closest to the computing power bottleneck shouldn’t all be getting slaughtered in valuation. What the market is currently doubting is the time lag between order fulfillment speed, inventory cycles, and capital expenditure rhythms. Stories can be priced in a year in advance, but revenues cannot.
This gap will spill into the crypto market.
$BTC isn’t an optical module stock, but it operates under the same risk budget. Binance's 24-hour market shows BTCUSDT around $61,528, down 1.97% for the day; ETHUSDT around $1,630, down 2.15%. The drop is much milder than for optical modules, indicating that the crypto space isn’t the main stage for declines, but it hasn’t completely decoupled either. When funds see AI hardware starting to wobble, the first reaction is usually not to buy higher volatility assets, but to first reduce leverage.
Here’s an ironic point: the hotter the IPO expectations for AI, the more one needs to watch the cold shoulder of the secondary market. This is because primary market pricing relies on a sense of scarcity, while secondary market pricing relies on marginal orders. When both sides appear at the same time, one side grabs quotas while the other side cuts positions; the real signal isn’t about who’s right or wrong, but that funds are starting to differentiate levels within the AI narrative.
The most vulnerable to being mispriced are the small assets with AI in their name but no income verification. The most likely to be repriced are the hardware chains that already have orders but whose growth slope is doubted by the market. AI concept coins in crypto are caught in the middle, lacking the constraints of earnings reports from US stocks or the scarcity tickets of top pre-listed assets, relying mainly on liquidity and sentiment.
So today, this line shouldn’t just be seen as a drop in chip stocks. It’s more like a risk budget check-up: when AI hardware starts being asked when profits will materialize, the on-chain AI narrative will also face the same question—users, revenue, costs, retention—what can actually support valuation?
That initial misalignment still exists. The primary market talks about the future, while the secondary market does the math. As more people start doing the math, all assets inflated by imagination will be forced to hand in their homework.