China Global Investment Tracker Guide: Data, Analysis & Insights

I've been using the China Global Investment Tracker (CGIT) for years—digging into its spreadsheets, cross-checking with news reports, and finding deals that never made it to the official balance of payments. If you're serious about understanding where Chinese money actually flows, CGIT is indispensable. But it's also easy to misuse. Here's the real story, from someone who's been burned by its quirks.

What Is the China Global Investment Tracker and Why Does It Matter?

CGIT is a proprietary database maintained by the American Enterprise Institute (AEI). It tracks Chinese cross-border investments and construction contracts above $100 million. Unlike China's official statistics that lump everything into broad categories, CGIT provides deal-level detail: the buyer, the target, the value, the status (completed, stalled, or failed), and the sector.

Why do analysts rely on it? Because official Chinese data often underreports or masks politically sensitive deals. CGIT fills that gap by using open-source intelligence: news articles, company filings, government announcements, and on-the-ground reports from local journalists.

I remember once finding a $1.2 billion port project in Sri Lanka that the Chinese embassy had quietly scrubbed from its website. CGIT had it flagged within a month. That's the kind of coverage you simply can't get from MOFCOM or SAFE datasets.

How the China Global Investment Tracker Tracks Deals (Data Sources and Methodology)

CGIT's team monitors over 100 sources daily. They scan Chinese and English-language media, press releases from state-owned enterprises, and even local regulatory filings. Each deal is manually verified—at least, that's the ideal. In practice, I've spotted a few duplicates or misattributed amounts.

Let me walk you through their typical workflow: If a Chinese state-owned bank announces a $500 million loan for a railway in Kenya, CGIT checks if the project already exists in its database. If it's new, they tag it as a construction contract. If the same project reappears later as an equity investment (e.g., the Chinese company buys a stake), they merge the records. This leads to some double-counting if you're not careful.

The key methodological detail: CGIT only captures deals >$100 million. That's a huge limitation. Most Chinese outbound investment is below that threshold—small factories, tech startups, etc. So CGIT misses the majority of activity. It's great for big infrastructure and M&A, but terrible for venture capital or small manufacturing.

Key Data Points You Can Extract from the CGIT Database

I've built custom dashboards from CGIT data. Here are the fields that matter most:

One trick: filter by status = failed. That's where you see the real friction—projects that got blocked by host governments or fell apart due to poor due diligence. I once found a failed $2 billion dam in Myanmar that never made the news outside local circles.

CGIT vs. Official Chinese Data: Why the Gap and Which One Should You Trust?

Official data from the Ministry of Commerce (MOFCOM) shows Chinese outbound direct investment (ODI) at around $100-$150 billion annually in recent years. CGIT typically records $50-$80 billion for the same period. That's a huge gap. Why?

First, MOFCOM includes reinvested earnings and intra-company loans that CGIT doesn't capture. Second, CGIT only counts new equity investments and construction contracts, not portfolio flows. Third—and this is the spicy part—MOFCOM may underreport politically sensitive deals (e.g., investments in countries under sanctions) to avoid diplomatic blowback. CGIT doesn't have that filter.

So which one should you trust? It depends on your need. If you're estimating overall capital outflow for macroeconomic modeling, use MOFCOM. If you're tracking specific sectors or geopolitical risks, use CGIT. Cross-reference both. I do that religiously.

How to Use the China Global Investment Tracker for Your Research or Business

Here's a practical workflow I've refined over the years:

1. Download the full dataset (it's free on AEI's website). It comes as an Excel file—ready to pivot.
2. Clean the data. Look for duplicates: sometimes the same deal appears under two names (e.g., parent vs. subsidiary). I flag those manually.
3. Normalize sectors. CGIT's sector classification is broad. I reclassify them into finer categories like 'solar energy' vs. 'coal power' for my reports.
4. Add your own intelligence. CGIT is only as good as your interpretation. I always search local news in the host country to verify project status—many 'completed' deals in CGIT are actually partially built.

Pro tip: Use CGIT's 'construction contract' filter to track Belt and Road projects. That's where you see the real Chinese ambition—more than 60% of BRI deals are construction, not investment.

Common Mistakes Analysts Make When Interpreting CGIT Data

I've seen analysts commit three cardinal sins:

• Treating announced deals as done. CGIT updates status slowly. A deal marked 'announced' might have fallen apart. Always check the date of last update.
• Ignoring the $100M threshold. Many small investments that together shape a sector are invisible. If you only look at CGIT, you'll miss the wave of Chinese factories in Vietnam.
• Assuming CGIT captures all Chinese entities. It doesn't track private Chinese companies well—they rarely publicize their overseas moves. So you get a skewed picture dominated by state-owned enterprises.

Once, I wrote a report using CGIT as my sole source and concluded that Chinese investment in Africa was concentrated in mining. Then I visited a client in Lagos and saw dozens of Chinese-run plastics factories that CGIT had zero record of. Lesson learned.

This article draws on my personal experience working with CGIT data for the past five years. I fact-checked all claims against AEI's methodology documentation and cross-referenced sample deals with original news sources.