January 13, 2025
About the Company
Nanya is a small Taiwanese DRAM manufacturer. They produce DRAM on trailing-edge process nodes (licensed from Micron) but are developing their own more advanced, technology. Nanya has about a 1.6% share of the DRAM market, a distant fourth, behind Samsung, Hynix, and Micron. The top three companies control a combined 96.4% of the DRAM market.
I don’t own any shares of Nanya and would not buy any, because they lack the scale and the technology to be competitive in DRAM. I keep an eye on their quarterly releases as a data point on the DRAM market.
Market Commentary and Q4 2024 Performance
Eight quarters have passed since Nanya stopped including the time trend of their gross margin with their earnings. The weak non-AI DRAM market hit Nanya full on this quarter. Their ASPs decreased low-teens % and their bit shipments were down high-single digits %. Combining these two gives a drop in revenue of 19.2% sequentially, on a flat exchange rate. This is pain. For the last twelve quarters, Nanya’s ASPs have been down low-teens %, up mid-single digits %, up low-teens %, up high-single digits %, up low-single digits %, down high-single digits %, down mid-single digits %, down mid-single digits %, down high-single digits %, down mid-20s%, down low-20s%, down mid-single digits %, down mid-single digits %. The four consecutive quarters of rising ASPs for the company has come to an end. If this drop doesn’t prove to be temporary – and I don’t think it will be temporary – this will go down as the shortest DRAM upturn in at least thirty years. For the non-AI and non-leading-edge segments, at least. The bottom of the downturn for Nanya was somewhere between July and September of 2023 and it seems the peak was between July and September of 2024. This short and shallow upcycle is an existential threat to Nanya. They did not make nearly enough cash to recover what was lost in the prior downturn, and now it appears the company is entering another valley. If the Chinese suppliers keep downward pricing pressure on the DRAM market, this may be the valley of death for Nanya.
Bit Shipments and Capital Expenditures
Two quarters ago, Nanya raised their 2024 capital expenditures from NT$20B to NT$26B. Last quarter, they lowered it back down to NT$20B. They pulled back further in the fourth quarter, such that total capital expenditures in 2024 came in at $16.1B. This is a dramatic change in two quarters, the result of the rapidly deteriorating legacy DRAM market conditions. The company needs to conserve cash. That fact makes their claim that 2025 bit shipments will increase by more than 20% over 2024. This is following a flat year for total bit growth in 2024 compared to 2023. Subject to Board approval, the company is planning for NT$20B in CapEx in 2025. I’ll take the under on this.
DRAM Market Outlook for 2025
Demand for “general” PCs, mobiles, and consumer products is still weak. The market “may” bottom in the first half of 2025 with a recovery in the second quarter. The company is pinning their hopes for this on “regional economic improvements by stimulation policies.” This means they are hoping China will pour money into their economy to fight their deflation problem, leading to higher consumer demand for electronics. Memory executives don’t know anything more than the rest of us do more than one quarter out, least of all the executives at Nanya. They are a clueless bunch. HBM and DDR5 DRAM supply is going to grow in 2025 while “standard” DRAM products (all the rest) will see continued inventory reductions.
- · Server Market: Positive growth continues for both AI and conventional servers.
- · Mobile Market: Inventory levels are normalizing, hopefully leading to better supply-demand balance in the first half of calendar 2025.
- · PC Market: They recycled the commentary from the other memory makers that PC demand will recover because of enterprise replacements/upgrades and “AI PCs,” whatever those are.
- · Consumer Market: “Potential recovery” starting in Q2’25. See my comment above about memory executives not knowing anything beyond one quarter out.
Analyst Call
Prepared Remarks
- · There were $NT600M of idle costs in the current quarter financials.
- · The company has negative cash flow from operations. Then on top of that, they are spending the high required capital expenditures. Their cash position still increased in the quarter because they patched this cash flow hole by taking on more debt.
- · They have sampled DDR5 parts and expect to ramp this product in 2025.
- · Total bit output of their second generation 10nm node will be more than 30% for the whole of 2025. As of today, 20% of their total bit output is this second generation (1b) DRAM.
Analyst Q&A
- · They are hoping for some improvement of non-AI DRAM pricing in the second half of 2025. Memory executives don’t know much beyond one quarter out, so this is a wish.
- · The March quarter is still forecasted to be experiencing inventory digestion and associated demand weakness. They are hoping that Q2 will start to see some recover in their memory prices.
- · One analysts reminded us that at the beginning of 2024 Nanya forecasted 20% bit growth in 2024 and it ended up being flat.
- · There are “some signs” that Q1-25 may show some improvement in bit shipments compared to Q4-24.
Summary
I think Nanya is in real trouble. Their ASPs only rose for four quarters and are now headed back down. The company executives are hoping that inventory reductions and demand improvement will stabilize the non-AI DRAM market in the second quarter of 2025 and beyond. I don’t believe this will happen, because I think it is indigenous Chinese supply that is causing the lower demand in the PC, mobile, and consumer segments. CXMT continues to ramp and I think this will keep downward pressure on legacy DRAM prices, which is all that Nanya makes. For Micron investors, the takeaway here is that the non-leading edge DRAM market (non-DDR5, non-HBM) has now passed the peak of pricing and ASPs are declining. This is early (one year) and typically the upturn lasts longer than that. But CXMT coming online may be disrupting that paradigm. If this is the case, Nanya is going to get squeezed out of existence. With the US expansions of fab capacity underway, it is less likely than it has been in the past that Micron will be interested in acquiring Nanya’s fabs when they go bankrupt, but there is a price that is low enough that Micron will buy them up. A lot of blood will need to be spilled first.
-Smooth Hughes (no MU position)