What is a semiconductor?
A semiconductor is a material with electrical conductivity between that of a “conductor” and an “insulator.” It can be imagined as a smart faucet rather than a regular water pipe.
Conductor → Like an open copper pipe, water (current) can flow freely at any time. For example, metals like copper and silver.
Insulator → Like a solid wall, water (current) cannot pass through at all. For example, rubber and ceramics.
Semiconductor → Like a smart faucet, we can precisely control the flow of water (current) on and off, or adjust its volume by “turning the valve” (applying voltage, light exposure, changing temperature, etc.).
In simple terms, in the financial world, if data is the “oil” of the new era, then semiconductors are the “crude oil” and “engine” that drive the entire digital world. At the same time, the semiconductor industry has strong cyclicality and high capital barriers.
The most typical financial characteristic of the semiconductor industry is its “strong cyclicality.” It resembles a wildly fluctuating wave line, cycling between “chip shortages and price increases, prosperity and expansion” and “overcapacity and price crashes.”
Upward Cycle: When emerging technologies (such as 5G, AI, electric vehicles) explode, market demand surges, and chips become scarce. At this time, chip prices soar, semiconductor companies’ profits skyrocket, and their stock prices soar, prompting them to invest heavily in building new fabs to expand capacity.
Downward Cycle: When the market becomes saturated or the economy cools, demand slows. The previously expanded capacity begins to release, leading to oversupply and a sharp drop in chip prices. Company profits plummet, stock prices correct, and the industry enters a “winter,” beginning a new round of reshuffling and consolidation.
Moreover, the semiconductor industry has extremely high “capital barriers.” Building an advanced fab can cost hundreds of billions of dollars.
Suppose a sudden gold rush for “smart robots” (analogous to the AI revolution) erupts in the market. Everyone wants to create the smartest robot to get rich. There are mainly two types of participants:
“Gold Diggers” (Tech Companies): For example, various robotics companies and internet giants. They are desperately developing algorithms and designing products, all wanting to strike “AI gold.”
“Shovel Sellers” (Semiconductor Companies): Semiconductor companies are the ones selling “shovels”—that is, AI chips (like GPUs)—to all the gold diggers.
[Story Development]
Initial Boom (Beginning of Upward Cycle): The robot concept is booming, and countless “gold diggers” are pouring in. Anyone wanting to strike gold must first buy a good “shovel” (powerful AI chips). Thus, the demand for “shovels” skyrockets. “Shovel sellers” (like NVIDIA and TSMC) receive an overwhelming number of orders for “shovels,” and prices rise sharply. Company financial reports shine, profits double, and their stocks become the brightest stars in the market, with investors scrambling to buy.
Peak of the Boom (Cycle Peak): Seeing the incredible profits from “shovel selling,” others want to get a piece of the pie. They begin to invest heavily in building new “shovel factories” (fabs) to produce more “shovels.”
Bubble Burst (Downward Cycle Arrives): After a while, the market realizes that not all “gold diggers” can succeed. Many robotics companies face unsold products, bankruptcy, and closure. The surviving “gold diggers” also start cutting expenses and no longer purchase as many new “shovels.” At this point, the newly built “shovel factories” have just started production, leading to a severe oversupply of “shovels” in the market. To clear inventory, “shovel sellers” have to engage in fierce price wars, resulting in significant profit shrinkage. Their stock prices also plummet from their peaks, causing heavy losses for investors who bought in at high points.