It refutes a common misconception—**"the storage price hike is a malicious price gouging by manufacturers, sucking blood from downstream."** His argument can be divided into three levels:

### 1. Price is determined by supply and demand, not by what manufacturers arbitrarily decide

The author uses the metaphor of "adding two zeros to a pancake" to illustrate a basic economic principle: **price is not unilaterally decided by the seller, but the result of a game between buyers and sellers in the market.** If storage chip manufacturers could set prices at will, why did they suffer losses and see price drops in the past? Because demand was insufficient. The current price increase is essentially due to explosive growth on the demand side (AI data centers, large model training), while the supply side (DRAM/HBM production lines) cannot keep up in the short term.

### 2. High prices are signals for resource allocation, not "exploitation"

This is the core economic viewpoint in the post:

- **Price signal**: When storage chips are in short supply, high prices tell us two things—(1) downstream users need to use resources more sparingly, prioritizing the highest-value scenarios; (2) upstream manufacturers have profitable expansion opportunities and should increase investment.

- **More efficient than queuing or administrative allocation**: If price competition is not used, the alternatives are "first come, first served" or "allocation by an authority." Both methods lead to resource misallocation—companies that urgently need chips may not get them, while those that don't may hoard them. Price competition directs resources to the highest bidders (i.e., those who value them most), achieving the highest economic efficiency.

### 3. High prices themselves are a problem-solving mechanism

The author points out that "if prices didn't rise, where would the motivation to expand production come from?"—this is precisely the process of market self-regulation:

1. Demand explosion → Short supply → Price increase

2. High profits attract manufacturers to expand production, new capacity comes online

3. 2-3 years later, new capacity is released → Supply increases → Price falls back

This is the classic **cobweb model / hog cycle** logic: price fluctuations themselves are a means to balance supply and demand, not a conspiracy.

---

## In the context of the current market

From today's market information, this discussion is very timely:

- Analysts point out that the market is discussing **whether the current memory cycle has peaked**, with the game between AI demand and supply constraints expected to last until 2028-2029 still ongoing.

- On one hand, investors worry about "chip inflation" and cyclical peak risks; on the other hand, the structural demand for storage from AI data centers remains bullish.

- SK Hynix's plan to list in the US (valuation about $29 billion) also confirms that capital is chasing the storage sector.

---

## One-sentence summary

The author's viewpoint, summarized in economic terms, is: **The storage price hike is not "manufacturers being evil," but a natural market equilibrium under supply constraints during a demand explosion; high prices themselves are both the optimal solution for resource allocation and a signal driving future supply expansion.** Unless AI demand slows down, storage chips will maintain high prices for at least the next 2 years before new capacity is released, which is a market law, not manipulation.$Micron Tech(MU.US)$Sandisk(SNDK.US)$Roundhill Memory ETF(DRAM.US)$SK Hynix(SKHY.US)$XL2CSOPHYNIX(07709.HK)