Example B: oil and gas

Which countries would be hit hardest if oil and gas from the Gulf or Russia stopped flowing?

Read the method, then the table.

The exposure ranking

Higher means more imports at risk relative to GDP, not a forecast of lost output.

Read the ranked table below.

Charts need JavaScript and the chart download. The table and CSV remain available.

Top 25: at-risk imports as a share of GDP
RankCountryRussia %Gulf %Total %Data
1Brunei3.83%8.42%12.25% Importer
2Kyrgyzstan6.07%0.01%6.07% Importer
3Mongolia5.64%0.00%5.65% Importer
4Mozambique0.00%5.46%5.46% Importer
5Mauritius0.00%5.46%5.46% Importer
6Singapore0.50%4.70%5.20% Importer
7Greece1.77%2.78%4.55% Importer
8Armenia3.78%0.65%4.43% Importer
9Madagascar0.00%4.37%4.37% Importer
10Slovakia4.21%0.02%4.23% Importer
11Jordan0.03%4.19%4.22% Importer
12South Korea0.26%3.95%4.21% Importer
13Thailand0.08%4.04%4.12% Importer
14Namibia0.00%3.97%3.97% Importer
15Pakistan0.00%3.87%3.87% Importer
16Lithuania1.86%1.92%3.77% Importer
17India0.89%2.85%3.74% Importer
18Belarus3.64%0.01%3.65% mirror: 2021, 2022, 2023
19Bulgaria2.87%0.30%3.17% Importer
20Taiwan0.22%2.85%3.08% Importer
21Malaysia0.24%2.54%2.77% Importer
22Hungary2.72%0.02%2.74% Importer
23Lebanon1.41%1.28%2.68% Importer
24Serbia1.58%1.10%2.68% Importer
25Senegal1.65%1.00%2.66% Importer

ASEAN and G7

Members present in this dataset, with their overall rank
RankCountryGroupExposure
1BruneiASEAN12.25%
6SingaporeASEAN5.20%
13ThailandASEAN4.12%
21MalaysiaASEAN2.77%
37JapanG71.84%
51VietnamASEAN1.10% (mirror)
54ItalyG71.05%
55PhilippinesASEAN0.95%
64FranceG70.66%
69IndonesiaASEAN0.47%
79United KingdomG70.33%
80GermanyG70.30%
97United StatesG70.12%
101CanadaG70.10%
110MyanmarASEAN0.05%
123CambodiaASEAN0.01%
138LaosASEAN0.00% (mirror)

Data: UN Comtrade (importer-reported, mirror where noted), IMF WEO April 2026 GDP. Fetched: 2026-10-10.

Try it yourself

A CSV is a plain-text table. In exposure.csv, russia_usd and gulf_usd are imports in US dollars. gdp_usd is GDP in US dollars. exposure_pct_gdp is the at-risk share of GDP. mirror_years lists years taken from the seller’s report.

01

Download the data

✅ Free pending test

Save exposure.csv in your Downloads folder.

Prompt
I saved a file named exposure.csv. Tell me how to confirm it is in my Downloads folder and that the name is exactly exposure.csv. Do not change the file.
What you should see

You see a file named exposure.csv in your Downloads folder.

02

Ask for a ranking and chart

⚠️ Free, limit: free file-upload and message limits pending test

In a new chat with your AI, upload exposure.csv and send this prompt.

Prompt
Use only the attached exposure.csv. Do not invent values.
exposure_pct_gdp is imports at risk as a percent of GDP.
Show the 10 countries with the highest exposure_pct_gdp, highest first.
Make a horizontal bar chart split into Russia and Gulf.
The Russia part is russia_usd divided by gdp_usd, times 100.
The Gulf part is gulf_usd divided by gdp_usd, times 100.
Label both parts as a percent of GDP.
If mirror_years is not blank, mark that country as mirror.
If you cannot draw a chart in this chat, give the same 10 countries in a table with those two percentages.
What you should see

You see ranks 1 to 10 from the table on this page, each split into a Russia percent of GDP and a Gulf percent of GDP.

03

Make the AI explain its method

⚠️ Free, limit: free message cap pending test

In the same chat, send this prompt.

Prompt
Explain this method in plain words.
Average the 2020 to 2024 imports of HS 2709 (crude oil), HS 2710 (refined fuel) and HS 2711 (natural gas, including LNG).
The suppliers are Russia, Saudi Arabia, the United Arab Emirates, Iraq, Kuwait, Qatar, Iran, Oman and Bahrain.
Divide that average by average GDP.
GDP is the size of an economy.
Name two weaknesses.
Explain why the result is not a prediction of lost GDP.
What you should see

You see a reply that names the imports being added, GDP as what they are divided by, two weaknesses, and a statement that this is not a prediction of lost GDP.

04

Separate the two supply shocks

⚠️ Free, limit: free message cap pending test

In the same chat, send this prompt.

Prompt
Use only the attached exposure.csv. Do not invent values.
Rank countries by russia_usd divided by gdp_usd.
Then rank them by gulf_usd divided by gdp_usd.
Compare each rank with the rank by exposure_pct_gdp.
Name the countries whose rank changes most if only Russia stops supplying.
Name the countries whose rank changes most if only the Gulf stops supplying.
For those countries, show the old rank and the new rank.
What you should see

You see two lists of country names, one for a Russia-only stop and one for a Gulf-only stop, each with an old rank and a new rank.

05

Write a short ASEAN briefing

⚠️ Free, limit: free message cap pending test

In the same chat, send this prompt.

Prompt
Use only the attached exposure.csv. Do not invent numbers or countries.
Use only the rows where the group column says ASEAN.
Write a briefing of about 300 words on the five with the highest exposure_pct_gdp.
For each of the five, give the country name as written in the file, the exposure as a share of GDP, and the Russia share versus the Gulf share.
Explain the method in two sentences.
Name two limitations.
Keep the numbers separate from any policy suggestion.
If mirror_years is not blank, name the country and the years.
Do not forecast GDP or prices.
What you should see

You see a briefing that names Brunei, Singapore, Thailand, Malaysia, and Vietnam, with a share of GDP and a Russia versus Gulf split for each, plus two limitations.

06

Spot-check two numbers

✅ Free pending test

In the same chat, send this prompt. Replace each bracket with a country name from the table on this page.

Prompt
Use only the attached exposure.csv. Do not change a number to force a match.
For [first country] and [second country], show russia_usd, gulf_usd and gdp_usd.
For each country, calculate (russia_usd + gulf_usd) / gdp_usd * 100.
Show the file's exposure_pct_gdp beside that result.
Say whether the two percentages match after rounding.
Then say whether those percentages match the briefing you wrote above.
What you should see

You see russia_usd, gulf_usd, gdp_usd, and a calculated percentage for each country that matches exposure_pct_gdp apart from rounding.

Open the explorer when you want to move a Russia cut and a Gulf cut yourself.

Advanced: how the data was pulled

An API is a web address that returns data. This UN Comtrade preview needs no password.

The address below asks for 2023 imports of crude oil from Russia. partnerCode 643 is Russia. cmdCode 2709 is crude oil. flowCode M means imports.

https://comtradeapi.un.org/public/v1/preview/C/A/HS?partnerCode=643&period=2023&cmdCode=2709&flowCode=M

The same request was repeated for each supplier, for each year from 2020 to 2024, and for product codes 2709, 2710 and 2711.

Where the buyer filed no report, the seller’s export report was used for that year and marked mirror.

One call returns about 500 rows. A request that hits that cap has to be split. The free preview allows a limited number of calls each hour, so the file was downloaded once and saved.