Method overview: one scale for every flow

The map gives every country two numbers for 2022 to August 2026, each against what it would have done without the war: support for Ukraine and Russia's earnings from trade (what Russia keeps: taxes and profit above a normal return, not the full price). Their sum, with the return Russia lost on its frozen reserves, is the net. Worldwide, Russia's net gain from buyers who cut fell by EUR 353.6bn and rose by EUR 216.4bn from buyers who bought more (2022–August 2026).

The core idea in four steps

1. Change, not totals. Each flow is set against the level expected without the war (counterfactual). A country that kept buying the same oil shifted nothing.
2. Volumes, valued at each year's prices. Paying more for the same oil is not a change in behaviour, so trade is compared in volumes (prices). The extra Russia earned from higher prices is a part of its own, the war-time windfall.
3. Summed from 2022 to the chosen year: whole calendar years, so 2022 includes January and February, before the invasion (little changed then), and 2026 counts January–August. 2021 is the pre-war reference.
4. As % of one year's GDP, the 2022–2024 average. It is a total for about 4.7 years, not a yearly rate.

One rule for what counts. A country's figure reflects its own choices and the foreseeable answers to them. Things forced on it that it did not choose — Ukraine's strikes on Russian refineries, Russia's sudden export bans — count neither for it nor against it. Embargoes, sanctions, switching suppliers, Russia's 2022 gas cut-off and its counter-sanctions, and the war-time prices paid by buyers who kept buying all count (what counts).

Choices and their consequences

• Buying from Russia is a choice. What Russia earns on it, at the price of the day, is a consequence of that choice. So the war-time price spike on what a country still bought counts to it, even though it did not set the price.
• No credit for others' sanctions. Where other countries' sanctions pushed Russia's prices down (the oil discount, the price cap), the buyer that kept buying paid less. That saving is not credited against it: it came from others' choices.
• There is always an alternative: another supplier, or, where none exists, sending money to Ukraine to balance it. So buyers who had little choice are counted too, including those who get Russian oil inside refined fuel bought from a third country.
• Forced losses stay neutral. When a buyer wanted to buy and could not (Ukraine's strikes on refineries, Russia's sudden export bans, a pipeline hit in the war), the fall in its purchases counts neither for nor against it.
• Russia's gas cut-offs count to the buyer. Russia cut gas to countries that had sanctioned it or refused its terms (paying in roubles); the cut is a foreseeable answer to the buyer's own choices, so the fall counts, labelled "cut off by Russia".
• The end of gas transit through Ukraine (1 January 2025) was Ukraine's decision not to renew its transit contract with Gazprom, not a cut by Russia. It was announced more than a year ahead and the buyers could switch, so it is counted like the cut-offs: the gas that stopped counts, labelled, only for gas still flowing in 2024.

Every part of the net shift

Each part in a country's panel opens its own method. In short:

Labels that change no total: gas that Russia cut off itself (Poland, Bulgaria, Germany and others in 2022; Austria in 2024) or that stopped when Ukraine ended gas transit in 2025 (counted only for gas still flowing the year before), and the share of a cut required by sanctions (sanctions).

Markers outside the totals

Shown on the map and in the panel, added to nothing:

• Critical goods (◆): sales to Russia of goods its weapons need; the money is already in the trade number (marker).
• Wages from Russia (△): migrants' remittances. They are individuals' wages, not a country's choice, and would swamp the ranking: Tajikistan would be about −40% of GDP; without them it is −0.40% (remittances).
• Taken by Russia (■): company assets Russia seized and forced-exit costs. Russia took them; the country did not choose to move them (taken).

One rule for every seller

Every sale counts at the seller's net gain × the share that is genuinely extra.

• Net gain: the taxes the seller's state collects plus profit above a normal return on capital, as a share of the price. Wages, materials, transport and a normal return are real costs and do not count. For other countries' sellers it comes from OECD industry accounts (2021–2023).
• Genuinely extra: the share of a sale that would not exist without it (weapons about 0.6, sales lost to Russia 0.2–0.45, because many firms found other buyers).
• In full: aid, refugee hosting, unpaid weapons, taxes and profits paid to Russia, and the return on frozen reserves.

Manufacturing earns little above a normal return, so US weapons sold for Ukraine count at about 4% of their value.

The net-gain rule

Aid is a gift: Ukraine gets EUR 1 and gives nothing back. Trade is an exchange: the seller gives up goods that cost money to make. So trade counts at the seller's net gain. The idea came from a reader on Reddit.

Russia's net gain per euro, central values:

Each country's payments take the gain of what it actually bought: its own mix of fuel and other goods, from its customs records. Russia's normal return is 11.4% a year on its capital (5% real plus Russia's risk premium). Ukraine's net gain is lower, about 0.03–0.05 per euro in 2022–2025: war-time logistics take much of the price and investors need a high return.

Two numbers

• Support for Ukraine: aid, EU share, refugees, purchases from Ukraine above or below the no-war level, donations, new multilateral lending, minus money flowing back.
• Russia's earnings from trade: everything else. Positive means the country cut what Russia earns; the panel says it in words ("cut Russia's earnings from trade by 1.6%"). It has two lines that add up to it: Russia's own gain (what Russia itself kept more or less: fuel and goods it sold, war-time prices, taxes it collected, Russian firms' profits) and the country's own gain from Russia (its sales, services, shipping and its firms' profits there). Both count; the split is on the trade page.

A negative support for Ukraine means the country earned more from the war effort than it gave, not money toward Russia; on the map such a country is grey.

The cost view

What the war cost the country itself (aid at the lender's cost, the extra paid to replace Russian energy, income lost on sales to Russia, minus discounts gained on Russian oil). The site does not show it; the full data download keeps it (Download the data).

Gross flows

The map shows changes, not gross totals, because gross totals include normal trade. Gross flows (paid to Russia, bought from Ukraine, aid) are in each panel's small charts, in the downloads and as an option in Filters.

How the map shows it

One colour rule: blue toward Ukraine, orange toward Russia, grey no clear change (or support below zero), faded where the direction is not very likely. Money streams show the two numbers in euros, never gross trade. What's counted lets you switch parts off; every total, range and grade follows, and the page says "Custom mix, not the official figures".

Rules that apply everywhere

• Counted where used: gas where it is burned, fuel resold by hubs at the final buyer, goods re-exported by hubs at the country that made them.
• Counted once: automated checks test for double counting.
• Rankings leave out economies under EUR 5bn GDP. Each headline number says how likely its direction is, from simulating its parts: very likely at 95% or more with every method choice agreeing and the likely range on that side; otherwise ranked faded, labelled likely (at least 65%), too close to call, near zero or essentially zero (the whole range within ±0.1% of GDP). Among 164 economies above EUR 5bn, 95 are not very likely on the net.
• Parties to the war, Belarus and North Korea, are listed under the ranking, not ranked. Russia invaded Ukraine from Belarus (UN resolution ES-11/1); North Korean troops fought for Russia in Kursk from late 2024.

No data

A trade part counts only where the country has its own observation for the war years: a figure it or its partners report for 2022–2025, or a country-specific record from the research (tanker tracking, national statistics, a refinery's exports). A trend carried forward, or Russia's import gap split by shares, is not an observation.

Why: without an observation the number would be our guess about that country, not its record. A part with no observation counts as zero and is named under the trade section as "No data". A country with nothing observed at all is hatched on the map, with no tags and no ranking. This affects 74 economies and territories, mostly small; the list is in the research files.

Decisions behind the method

The main choices, why they were made and what they did to the numbers when they came in (the panels show today's figures). The full changelog has the evidence for each.

How it works · Limitations · Changelog · Open in the map

Cite as: War Balance, figures to August 2026, https://warbalance.org/method/overview/