Data center backlash led me to create a stock analysis workflow

I spent the weekend tinkering and dug through Bloomberg's report on the political backlash against AI data centers from start to finish. Recently, I wanted to give my AI infrastructure positions a health check, so I organized this analysis flow. Beginners can follow it and complete all steps in about an hour.

First, the background. Data centers, simply put, are buildings filled with servers; AI training and inference rely on them. The problem is these buildings consume massive amounts of electricity and occupy land, leading to strong public backlash in the US. Gallup polls show that a majority of Americans oppose building data centers near their homes, and this is one of the few issues opposed by both parties. Republican Senator Josh Hawley said: "We are watching a handful of companies consolidate capital, information, and political power." On the Democratic side, it's more direct; someone proposed a national moratorium on data center construction in the NYT. This is essentially a political issue that will directly impact the valuation logic of related companies.

When I wrote about Cambricon news last week, I said reading financial news requires calculating first. Same here; let's get the tools ready.

Preparation: Three things are enough

1. A browser, Chrome or Edge works fine.

2. Software to view market quotes, such as Tonghuashun, Xueqiu, or your broker's app. Use it to watch stock prices and recent volatility of companies you're interested in.

3. Patience. You'll need to find English materials later; I used the browser's built-in translation, which was sufficient.

Getting Started: Five Steps to Complete the Assessment

Step 1: Check where the company builds data centers. Open any search box and search for "Company Name + data center locations." For example, if you want to check Microsoft or Amazon, you'll get a bunch of English pages. Find the official website or major media outlet link and click in. You're looking for a list of cities or states; note it down. Beginners often get overwhelmed at this step. Remember one principle: only look for explicit addresses; vague statements like "plans to build ten data centers" don't count.

Step 2: Search for "Location Name + data center opposition" one by one. This step is core. For example, searching for "Ohio + data center opposition" will show news about local resident protests or county council votes rejecting projects. In my tests, every state yielded at least one similar result. Both parties have rarely stood together on this issue; Politico reported that the industry knows its PR problem is huge, with SpaceX, OpenAI, and Meta learning to deal with this situation.

Judgment criterion: If the news mentions "resident protests," "council rejection," or "construction pause," mark this area as high risk. If it only says "planned expansion," keep looking; don't rush to flag it.

Step 3: Check the Capital Expenditure (CapEx) item in the company's financial reports. Open your market software, find the financial statements or data, and click in. Or simply search for "Company Name + 10-K" to find SEC filings (annual reports from the US Securities and Exchange Commission, essentially the company's self-summary). Look at two numbers: total CapEx for the past year, and whether the report mentions terms like electricity costs or Power Purchase Agreements (PPAs). PPAs are long-term electricity purchase contracts signed between companies and power plants; since data centers consume a lot of power, these contract amounts are substantial. If electricity costs are broken out separately, it indicates a significant proportion, making the company sensitive to rising electricity prices. Beginners often get confused here; skip fields you don't understand and just focus on these two numbers.

Step 4: Combine the results from the previous three steps. Make a simple 2D assessment:

• Is public backlash strong in the locations where the company builds data centers?
• Is the proportion of electricity-related costs high in the company's financial reports?

If both are "High," the company is a target in this backlash. If only one is High, risk is medium. If both are Low, it's temporarily safe. This step doesn't require professional tools; paper and pen are enough.

Step 5: Observe recent stock price reactions. Go back to your market software, search for the company code, and look at the trend over the last month. No need to analyze K-lines; just see if there were single-day drops exceeding 2% due to news like "data center pause" or "rising electricity prices." If yes, it means the market is already pricing in this risk, and your safety margin needs to be tightened further.

Pitfalls: Where beginners most easily stumble

Pitfall 1: Treating polls as stock price predictors. The Gallup poll says "Americans don't like data centers in their backyard," not "everyone wants to sell AI stocks." Transmitting public opinion to stock prices involves legislation and regulation in between; it could take half a year, a year, or fizzle out. Initially, I cleared half my position based on poll data, but the stock didn't move for half a month, causing me to miss a rebound wave.

Pitfall 2: Only looking at national news and ignoring local information. CNBC saying AI prospects are bright is useless; you need to look at the local forum of that county in Iowa. Data centers affect approval in the specific county where they are built, triggering resident protests there. National media rarely reports this, but stock prices loosen starting from these specific nodes. I later narrowed my search scope from states to counties to identify risk points.

Pitfall 3: Treating politician statements as enacted policy. Many headlines say "Lawmaker calls for pause," but clicking in reveals it's just a proposal, not passed. There's a huge gap between statements and legislation. Filter using Step 4 criteria before concluding; don't act just by looking at headlines.

Next Steps After Learning This

Try applying this flow to utility stocks instead of tech stocks. Power companies are more directly affected by data center construction: they benefit from rising demand but face backlash from residents over electricity prices. The same news event triggers two sets of reaction logic. Running through this gives you a feel for "how to price political risk."

---

📌 This article is compiled from Bloomberg Tech. Original: https://www.bloomberg.com/news/newsletters/2026-08-19/ai-data-center-political-opposition-to-bring-stock-market-winners-and-losers

All rights reserved by the original authors. This is a compilation and independent analysis based on public reports.