ChipMOS reports Q2 revenue of $231.8 million, up 28.7% YoY

ChipMOS reported Q2 revenue of NT$7,383.1m ($231.8m), up 6.5% QoQ and 28.7% YoY, with June revenue hitting NT$2,538.4m ($79.7m), +6.5% vs May and +37.2% vs June 2025. The company said growth is tied to demand for AI-related applications and highlighted that both quarterly and monthly revenue are the highest since 2014. With the stock trading at a 6.85 P/E and noted as undervalued on InvestingPro, the update is modestly supportive for sentiment, though it’s based on unaudited figures.

Analysis

The market may be underpricing how much of this move is about utilization leverage rather than pure demand. For an OSAT like IMOS, a sustained AI-related uptick can matter far more on margin than on headline revenue: once fixed test/assembly capacity is absorbed, incremental revenue can drop through quickly, especially if long-term supply commitments reduce the risk of idle lines. That said, the first rerating usually happens on confidence in forward bookings, not on one quarter of top-line strength.

The key second-order effect is competitive rather than company-specific: if AI demand is genuinely tightening backend capacity, peers with similar bottlenecks should see better pricing discipline, while equipment vendors into test/handler capacity could get a delayed benefit as customers fund expansion. The flip side is that larger OSATs and captive IDMs can eventually redirect capacity, which caps IMOS’s moat if this is a cyclical spike rather than a structural mix shift. In other words, the market may be overestimating duration and underestimating how fast supply can catch up.

Contrarian view: the consensus is likely extrapolating revenue momentum into earnings power too quickly. At this valuation, the stock can work if gross margin and free cash flow inflect over the next 1-3 quarters, but it can also remain a value trap if capacity additions front-run demand or if the growth is mostly low-margin substrate/test fill rather than true AI pricing power. The thesis is falsified if sequential growth slows materially in the next quarter, or if margin expansion does not show up despite continued revenue growth.

Key Decisions for Investors

• Small tactical long IMOS on a pullback after the print, sized as a special-situation value/quality blend; target is a rerating from sub-7x earnings toward 8.5-9.5x if the next update confirms margin leverage. Stop if the next quarterly sequential revenue growth falls below low-single digits or gross margin fails to improve.
• Relative-value hedge: long IMOS / short SOXX for 1-3 months to isolate backend-OSAT alpha from broader AI beta. This works only if the market continues to reward under-owned laggards with visible cash yield; cover the short if semicap multiples re-expand faster than expected.
• Watchlist, not action yet: ASMPT, TER, COHU as delayed beneficiaries if IMOS and peers continue expanding capacity. Add only on evidence that customers are funding additional test/handler spend, because the capex read-through is the real confirmation signal for a multi-quarter AI backend cycle.