#CXMTDrops7.7%AtOpen CXMT Drops 7.7 Percent At Open

CXMT opened down 7.7 percent this morning.

The stock started the day at 24.78 after closing yesterday at 26.85. Volume in the first 30 minutes was heavy at 410 million shares.

This is the first major down day since the IPO and the 90.1 billion volume debut we saw last week. I want to break down what happened, why it matters, what the data says, and what to watch next. No panic. Just context.

First, the facts from the open

Opening price 24.78, down 2.07 from yesterday’s close.

Low so far 24.10. High so far 25.40.

As of 10:30 Eastern the stock is trading around 25.05, down 6.7 percent on the day.

Volume is 3.2x the 20 day average for this time of day.

Options activity spiked. Put volume is 4x normal.

The broader semiconductor index is down 1.2 percent. The Nasdaq is down 0.8 percent. So CXMT is underperforming the sector by a wide margin.

What caused the drop

There is no single headline. It is a combination of 4 factors that hit at once.

One, profit taking after the IPO pop.

CXMT debuted last week and closed up 9.1 percent on day one with 90.1 billion in volume. That brought in a lot of traders. When a stock runs that fast, some of those holders sell on the first red day. We saw that in the pre market. Institutional flow data shows net selling from short term funds.

Two, memory price concerns.

Overnight, two memory distributors published contract pricing for August. DDR5 contract prices are flat month over month. That is a surprise because the market was expecting a 3 to 4 percent increase.

HBM pricing is still up, but the pace of increases has slowed. Analysts took that as a sign that the memory cycle may be cooling faster than expected. CXMT is a pure play memory company, so it got hit hardest.

Three, supply chain commentary.

One of CXMT’s large customers issued guidance last night. They said they are building 2 weeks of extra inventory to be safe on geopolitics. That sounds positive, but the market read it as demand being pulled forward. If customers are stocking up now, they may order less in Q4.

The company did not mention CXMT by name, but CXMT is listed as a supplier in their 10K. The stock reacted.

Four, macro and rates.

We got a hotter than expected PPI number this morning. That pushed rate cut expectations out. Growth and cyclical names sold off. Semiconductors are high beta, and CXMT is new and volatile, so it moved more.

None of these are company specific. There was no news from CXMT overnight. No guidance change. No fab issue.

What the business fundamentals look like right now

It is important to separate price action from the business.

Revenue last quarter was 4.9 billion. Guidance for this quarter is 5.5 to 5.8 billion.

Gross margin was 38 percent. Management reiterated the target of 42 percent by year end as the new fab ramps.

They are still cash flow positive.

HBM3E samples are with 3 customers. Qualification is on track for Q4.

The new fab is 28 percent complete and on schedule.

So the long term story has not changed in the last 12 hours. What changed is expectations and positioning.

On the memory cycle

This is the key debate right now.

Bulls say AI demand is structural. Every new AI training cluster needs HBM. Every AI PC and phone needs LPDDR5X. That keeps memory in a multi year upcycle.

Bears say we are seeing the classic memory pattern. Prices rise, everyone adds capacity, then there is oversupply 12 to 18 months later. The flat DDR5 pricing this morning feeds that fear.

The truth is probably in the middle. HBM is still supply constrained. That is 25 percent of CXMT’s revenue and 40 percent of gross profit. DRAM for servers and PCs is more cyclical. If PC demand softens, that hurts.

CXMT’s advantage is that they are ramping new capacity just as HBM demand is growing. Their disadvantage is they have less pricing power than the largest incumbents.

What analysts are saying today

Three notes came out in the last hour.

One firm cut their near term price target from 32 to 28 citing the flat DRAM pricing. They kept a buy rating.

Another firm said this is a buying opportunity and reiterated 35. They pointed to HBM ramp and valuation at 15x forward earnings now.

A third firm went neutral and said they want to see Q3 orders before getting more constructive.

The consensus is still that the cycle is intact, but the slope may be less steep.

What this means for investors

If you bought at the IPO at 24.60, you are still up slightly.

If you bought at yesterday’s close at 26.85, you are down.

If you are considering buying today, here is the framework.

The bull case. You believe HBM demand stays strong for 18 months. CXMT hits yield targets. The stock re rates to 20x earnings. That implies significant upside from here even after the drop.

The bear case. Memory prices fall 10 percent in Q4. Margins compress. The stock trades to 12x earnings. That implies more downside.

The base case. Memory is flat to up 5 percent in H2. CXMT executes on the fab. The stock trades sideways to up 15 percent into year end.

Volatility is part of owning a new cyclical name. 7.7 percent down days will happen. 7.7 percent up days will happen too. Position size matters.

What this means for customers and the industry

For customers, a lower stock price does not change supply. CXMT’s contracts are multi year. Deliveries are on schedule.

For competitors, this puts pressure on all memory names. The other two large DRAM suppliers are also down 3 to 4 percent today. The market is repricing the sector.

For suppliers to CXMT, nothing changes in the near term. Capex plans are funded. The 6.2 billion from the IPO is on the balance sheet.

Technical view

Technically this is a test of the IPO price at 24.60. If it holds, buyers will likely step in. If it breaks and closes below 24.50, the next support is 22.80, which is where the IPO book was anchored.

RSI is now at 38, down from 72 last week. That is not oversold but it is cooling.

Volume is high, which is healthy. It means there are two sides to the trade.

What to watch next

Earnings in October. That is the next real data point.

HBM qualification updates. Any announcement of a design win would matter.

September memory contract pricing. If that comes in up 2 to 3 percent, the narrative flips back positive.

Fab construction milestones. On time is priced in. Early is upside.

On risk management

If you own CXMT, ask yourself why you own it.

If it is for a 6 month trade on memory prices, this volatility is expected.

If it is for a 3 year bet on AI infrastructure, one down day does not change the thesis.

Do not add on margin just to catch a falling knife. Wait for the price to stabilize and for a catalyst.

A note on the IPO and liquidity

The 90.1 billion of volume on debut was a good thing. It means the stock is liquid. Today’s 410 million shares in 30 minutes proves that again.

Liquidity cuts both ways. It lets you get out fast, but it also means price can move fast. That is normal for a new listing in a cyclical sector.

Final thoughts

CXMT dropping 7.7 percent at the open is not fun if you own it. But it is not unusual.

New IPOs are volatile. Cyclical stocks are volatile. Memory stocks are especially volatile.

The business did not break overnight. The cycle may be moderating. The market is repricing that.

If you are long term, watch execution. If HBM ramps and margins expand, this price will look low in 12 months.

If you are short term, watch memory pricing and positioning. This could bounce or it could drift.

I will keep tracking the data and update as we get more. If you have questions about the memory market, CXMT’s business, or how to think about volatility in new IPOs, drop them here.

For now, the key level is 24.60. Hold that and the story is intact. Break it and we test lower.

Let us see how the day closes.