Simulator · methodology
How the simulator works, and where it falls short.
Trade disruption
The trade model
Shocks and time
A scenario closes or restricts places (straits, canals, ports, airports, airspace, rail lines, crossings, pipelines), blockades countries, sets tariffs, bans or sanctions, cuts production or makes routes dearer. Every shock starts at once and lasts some months; the year is split into periods, each solved as a steady state at annual rates, and the year is their duration-weighted mean, with the rest of the year at the 2024 baseline.
Routing
Each flow (exporter, importer, HS4 product) is split by mode and given its cheapest modelled route. A flow whose route is shut chooses among the stored routes of its mode that are still open, plus air for manufactures, paying the added freight, time, congestion delay and, where a node fills, a capacity rent. A share with no open route cannot move.
Markets
Each product market clears again (an Armington model in the style of the USITC's GSIM): buyers turn to other suppliers as delivered prices rise, by the product's published trade elasticity, and buy less as their import price rises. This decides, flow by flow, whether trade keeps going on its new route, switches supplier or stops.
Routing and markets together
Steps 2 and 3 repeat until the loads on ports and canals, their capacity rents, sea freight rates and every market settle, up to 80 rounds. Results say when they didn't, and which places were still over capacity.
Supply chains
The year's price and supply changes pass through the OECD input-output tables (80 economies and the rest of the world, 22 sectors) to each industry's costs and output, to consumer prices and to GDP. First order: fixed input shares, no income, exchange-rate or factor-market feedback.
Transport costs are per tonne, so the same $30 a tonne reroute is 5% on crude oil and 0.1% on machinery. A flow’s break-even price rise is its added cost per tonne over its value per tonne: the price increase at which it just pays for the new route. A capacity rent is the price that rations a full port, canal or pipeline. Buyers pay it, in fees, queues or freight premiums, but it is a transfer rather than a resource cost, so results report it apart from the real cost of rerouting; the model credits it to the country that owns the place, though who collects it in practice depends on how slots are allocated. Rents on ports that carried little of the cargo in 2024 are the least certain figures in the model. Real income is GDP plus the terms-of-trade gain or loss from the prices of what a country exports and imports.
Every result page lists the assumptions it ran under, the share of exposed trade it modelled, the data versions and any warnings, and every table on it downloads as CSV.
Data sources
- Trade flows
- CEPII BACI, release 202601, HS 2022: every product traded between every pair of 238 countries and territories in 2024, in US dollars and tonnes, aggregated to 1,229 HS4 headings.
- Transport network and routes
- Over 2,300 ports, airports, straits and canals, airspaces, rail lines, border crossings and pipelines, with 2024 throughput or design capacity, freight and handling costs and transit times. Each country pair gets its cheapest route by each mode and, for every node on it, the cheapest route avoiding that node. Mode splits are observed in customs data where they are published and estimated from value-to-weight ratios elsewhere.
- Elasticities
- Trade (Armington) elasticities per HS heading from CEPII's tariff-based estimates (Fontagné, Guimbard and Orefice 2022); demand and supply elasticities by sector from the literature, documented in the parameters build.
- Input-output tables
- OECD Inter-Country Input-Output tables, 2025 edition, for 2022, aggregated to 81 areas and 22 sectors. Countries outside its 80 economies form one rest-of-world area.
Validation against past disruptions
Five disruptions replayed on 2024 data, each check set against a published figure. Some gap is expected: the baseline is 2024 rather than the episode’s year, there are no stockpiles, and some published figures measure something slightly different (volume against value, one kind of ship against all cargo). The last column says how far apart the two are. The checks run on every change to the engine, with broad bands that catch it drifting; they are not a measure of its accuracy, which the summary below the table gives.
| Episode | Check | Published | Engine | Gap | Source |
|---|---|---|---|---|---|
| Suez 2021 (6 days) | Trade through Suez a day ($bn) | $9.6bn | $6.0bn | 1.6× smaller | Lloyd's List, 25 March 2021 |
| Extra days round the Cape (median rerouted flow) | up to 14 | 12.5 | in range | Maersk, March 2021 | |
| World trade over the year (%) | −0.2 to −0.4 pp of growth per week, mostly delay (−0.17 to −0.34 for six days) | −0.005% | 34× smaller | Allianz Research, 26 March 2021 | |
| Black Sea 2022 (Odesa ports, 5 months) | Ukraine's wheat exports while shut (%) | about −65 to −75% | −94% | 1.3× larger | Ukraine Ministry of Agrarian Policy via Ukrinform; CGIAR/IFPRI |
| Ukraine's maize exports (%) | (same) | −58% | 1.1× smaller | ||
| World wheat import price (%) | +24.5% Feb→Mar, +33.7% Feb→May | +24.3% | about the same | World Bank Pink Sheet (US HRW) | |
| World maize import price (%) | +14.7% Feb→Mar | +27.6% | 1.9× larger | World Bank Pink Sheet | |
| Red Sea 2024 (container and car carriers avoid it for a year) | Trade through Suez (%) | −50% by volume | −67% by value | 1.3× larger | IMF Blog, 7 March 2024 |
| Tonnage through Suez (%) | container tonnage −82% | −10% (all cargo) | 8.2× smaller | UNCTAD, February 2024 | |
| Trade round the Cape (%) | +74% by volume | +93% by value | 1.3× larger | IMF Blog, 7 March 2024 | |
| Tonnage round the Cape (%) | +60% | +8% | 7.5× smaller | UNCTAD, February 2024 | |
| Extra days on diverted voyages | 10 or more | 10.3 | in range | IMF Blog, 7 March 2024 | |
| Container freight rates (%) | +115% in the first week of 2024, more than double through the year | +23% | 4.3× smaller | Drewry WCI Shanghai–Rotterdam, 4 January 2024 | |
| Euro-area consumer prices after a year (pp) | +0.4 pp per 100% freight rise; +0.7 pp per doubling | +0.04 to +0.06 pp | 6.7× smaller | OECD Interim Outlook, February 2024; IMF WP/22/61 | |
| Panama 2023–24 (65% for 6 months) | Canal traffic while restricted (%) | −21% transits in FY2024, −36% daily slots at the trough | −33% | in range | Panama Canal Authority FY2024; advisories A-48-2023, A-54-2023 |
| US grain to Asia, added cost ($ a tonne) | about +$4/t | +$7.3/t | 1.8× larger | USDA AMS Grain Transportation Report | |
| Hormuz (3 months) | Crude import price while shut (%) | Brent up about 50% by March 2026 with stocks released; $110–150 in earlier scenarios | +99% (year average +25%) | 2.0× larger | EIA STEO, March 2026; Goldman Sachs scenarios |
| Saudi crude exports | bounded by the 5 mb/d East-West pipeline | −28% | not scored | EIA, 16 June 2025 |
Of the 17 checks with a figure on both sides, every one moves the way the published figure did. The engine lands inside the published range on 3, within a factor of two on 9 more, off by two to five times on 1, and off by more than five times on 4; those misses are listed below, each with its reason. One check is a bound rather than a figure, and not scored.
Why the engine is far off where it is:
- Suez 2021: World trade over the year (34× smaller). In 2021 most ships queued at Suez for the canal to reopen, and Allianz's figure is mostly that delay; the engine reroutes rather than queues, so it has almost none of it.
- Red Sea 2024: Tonnage through Suez (8.2× smaller). Only container ships and car carriers divert in the scenario; tankers and bulk carriers, which carry most of the tonnage, keep using Suez, where in 2024 many of them diverted too. UNCTAD's figure is for container ships alone.
- Red Sea 2024: Tonnage round the Cape (7.5× smaller). The same: the heavy tanker and bulk cargo stays on Suez. And for Asia to northern Europe the engine's stored alternative to the Red Sea runs through the Panama Canal, not round the Cape, so some of the diverted cargo never reaches the Cape.
- Red Sea 2024: Container freight rates (4.3× smaller). Rates follow the extra tonne-kilometres the fleet sails, and too little cargo diverts for that to double them (some goes via Panama or by air). The model also adds no war-risk premium and gives carriers no market power, and both added to the 2024 rise.
- Red Sea 2024: Euro-area consumer prices after a year (6.7× smaller). Mostly the freight-rate miss above: scaled to the engine's own +23% freight rise, the published estimates give about +0.1 to +0.2 pp, much nearer the engine's figure.
The Red Sea misses share a cause: the scenario diverts only container ships and car carriers, and the engine’s stored alternative for Asia to northern Europe runs through Panama rather than round the Cape (see the known limits). So treat the engine’s Red Sea tonnage, freight-rate, and consumer-price figures as well below what happened in 2024. Ukraine’s wheat exports fall further than in 2022 because the Danube ports and rail crossings have their 2024 capacities, not the expanded ones of mid-2022.
Known limits
No stockpiles
No inventories or strategic reserves are drawn and nothing carries over between periods. Price rises while a shock lasts are the steady state with no stocks, so for crude oil and grain, where stocks are large, they overstate the immediate spike; the year average is the better guide.
One stored alternative per place
The route build stores, for each node on a best route, the cheapest route avoiding it. When several places close at once, only stored routes that avoid all of them remain, so flows that could reroute may stop. For Asia–northern Europe the stored route round the Red Sea runs through the Panama Canal, not the Cape, so a Red Sea scenario loads Panama and pushes some high-value goods to air.
Freight rates respond less than in 2024
Sea freight rates move with each fleet's tonne-kilometres, with a supply elasticity set so that 10% more tonne-km roughly doubles rates. Because the Red Sea scenario diverts only container and car carriers, some via Panama and air, its container rates rise by about a fifth, not the doubling of 2024. That check is not independent of the setting.
Coverage
A scenario downloads at most 30 MB. The products carrying most of the exposed trade are loaded first. The presets model between about half (the Red Sea, Malacca and Taiwan straits, which carry many products) and 85% of the trade they expose, and every result shows its share. What is left out is in no figure, so totals understate the shock.
Small ports and convergence
When a restriction pushes cargo onto ports that carried little of it in 2024, the rationing between them and the restricted route can cycle without settling, as in the Panama drought preset. Results then say so and name the places still over capacity. Ports with under 10,000 tonnes a year of capacity and no modelled 2024 traffic take no diverted cargo.
Countries outside the input-output tables
Venezuela, Iran, Iraq, Kuwait, Qatar, and about 150 other economies are one rest-of-world area in the OECD tables, so they have no GDP or consumer-price figures. Their trade impact (exports and imports lost, switched, stopped, or rerouted, and its share of their whole trade) comes straight from the flow results and is marked “trade impact only”.
Queues, risk and partial equilibrium
A short closure reroutes rather than waits (in 2021 most ships at Suez waited), so delays after reopening are missing. War risk is a cost or closure the scenario sets. Products don't substitute for one another except as buyers cut back above twice the baseline price, and the input-output step is first order, which understates losses from large shocks to essential inputs.
Out of scope
The trade mode models goods trade between countries. It does not cover:
- Services trade
- Finance, payments and exchange rates
- Stockpiles and strategic reserves
- Month-by-month paths
- Probabilities of any scenario
- Domestic transport and sales
Scenarios are not forecasts: see the research standards. Open the simulator.
Macro scenarios
The macro model
A transparent semi-structural model of one economy at a time, in the family of the IMF’s quarterly projection models: a handful of equations whose parameters are all listed below. It runs for the 28 economies with a full country profile, in your browser, over five years. It is backward-looking and linear, and it shows how the levers move an economy away from the IMF’s baseline, not where the economy will go.
Demand and the output gap
The output gap (how far GDP is from potential) is persistent, and is pushed by the levers' direct demand effects (government spending, taxes, business investment, and the income lost or gained from energy prices, tariffs and the terms of trade), pulled by the real policy rate above its neutral level and by the real exchange rate, and opened by any change in potential that demand doesn't match at once.
Inflation
Headline inflation is core inflation plus the change in cost-push price levels: energy (its weight in the basket, the share of a world price change reaching retail prices, plus indirect effects), tariffs, the exchange rate and a trade-mode price shock. Core inflation is persistent, rises with the output gap and takes a second-round share of cost shocks. Expectations adapt; anchoring is how quickly inflation returns to target.
The policy rate
A smoothed Taylor rule on core inflation and the gap, with a neutral rate that moves with trend growth and that the central bank corrects while a gap persists. Euro-area members share the ECB's rate, which reacts to a member's own conditions by its share of the euro area; Saudi Arabia's follows the dollar peg. The policy-rate lever holds the rate where you set it, then the rule takes over, above an effective lower bound.
Jobs, the budget and trade
Unemployment follows the gap by each country's Okun coefficient. The budget balance changes with the tax and spending levers, automatic stabilisers, tariff revenue (and energy revenue for exporters) and the interest bill as debt rolls over at the new rate; debt follows from the balance and nominal GDP. Import volumes follow output, the exchange rate and tariffs; exports the exchange rate.
Potential output and sectors
Potential output combines productivity, the labour force and capital (which closes slowly on its long-run level and grows with extra investment), less the long-run costs of tariffs, dearer energy and higher taxes. Sectors move with potential plus their usual share of the cycle: industry swings more than services, agriculture least.
The baseline and the link to trade
Each result is a quarterly change from the IMF World Economic Outlook, added to its annual paths from CK's country data; the policy rate's baseline is the latest rate held flat. A trade-mode result can be sent in: its first-year GDP effect enters as lost supply, its consumer-price effect as a cost shock and its terms-of-trade effect as income, for the scenario's first year.
Parameters and their sources
Rows marked published rest on a published estimate; those marked judgement are CK’s choices within published ranges, tested only through the checks below. Advanced and emerging economies differ, and Okun coefficients and budget sensitivities are each country’s own where a study gives one. Openness, energy trade, energy’s weight in prices and debt maturity are each economy’s own. The values are read from the model’s code.
| Block | Parameter | Value | Source |
|---|---|---|---|
| Demand | Gap persistence and the real policy rate | Lags 0.95 and -0.15 (advanced), 0.9 and -0.15 (emerging); 0.10 (advanced), 0.08 (emerging) per point of real rate | Within the ranges of the IMF's quarterly projection models: Berg, Karam and Laxton (2006), IMF WP 06/81; Carabenciov et al. (2008), IMF WP 08/278 (U.S. lag 0.85, real-rate gap 0.13) |
| Fiscal | Spending and tax multipliers | Direct effects: spending 0.65 (advanced), 0.30 (emerging), taxes 0.25 (advanced), 0.25 (emerging), business investment 0.60 (advanced), 0.50 (emerging); plus 0.1 of the direct effect spent again the next quarter; all scaled by import leakage, from ×1.1 for closed economies to ×0.5 for the most open (imports of 30% of GDP: ×1) | Batini, Eyraud, Forni and Weber (2014), IMF Technical Notes 14/04: first-year spending 0.75 and revenue 0.25 in advanced economies, spending 0.1–0.3 and revenue 0.2–0.4 in emerging ones; Ilzetzki, Mendoza and Végh (2013) on open economies |
| Fiscal | Automatic stabilisers (budget semi-elasticity) | 0.21–0.66 by OECD member; 0.5 for other advanced economies (Singapore, Taiwan) and 0.3 for other emerging ones (an assumption) | Price, Dang and Botev (2015), OECD Economics Department WP 1275, Table 10 |
| Prices | Phillips curve: persistence and slope | Persistence 0.85 (advanced), 0.85 (emerging); slope 0.07 (advanced), 0.12 (emerging) per 1% gapjudgement | Set within the ranges of IMF WEO April 2013, ch. 3 (a flat, well-anchored curve in advanced economies) and WEO October 2018, ch. 3 (emerging markets: output-gap coefficient about 0.16), then checked against the central banks' responses below |
| Prices | Exchange-rate pass-through | 0.13 (advanced), 0.39 (emerging) in the long run; about three-quarters of it within two years | Carrière-Swallow, Gruss, Magud and Valencia (2016), IMF WP 16/240 (0.13 and 0.39 after two years) |
| Prices | Energy prices | 20% of a world energy price change reaches retail energy prices (less where prices are administered); indirect effects add 60% within about a year | ECB Monthly Bulletin July 2004, Box 3 (10% on oil moves consumer energy prices about 1 point) and April 2011, Box 1 (10% on oil: euro-area prices +0.2% in a year, +0.45% after three); energy's weight in each basket from national statistics (Eurostat HICP weights, BLS relative importance) |
| Prices | Tariffs | 40% of the tariff reaches consumer prices, in proportion to imports' share of GDP | Calibrated to the price effects of McKibbin, Hogan and Noland (2024) and the Yale Budget Lab (2025), checked below |
| Policy rate | Taylor rule | 1.5 on core inflation, 0.5 on the gap, smoothing 0.7 | Taylor (1993); smoothing as in Carabenciov et al. (2008), 0.71 |
| Labour | Okun's law | Country estimates, 0.015 (China, Singapore) to 0.57 (Australia); group means 0.39 and 0.2 where a country has none | Ball, Furceri, Leigh and Loungani (2019), Open Economies Review 30(5), Table 1 |
| Supply | Capital share and capital-output ratio | 0.4 and 4.5 (advanced), 0.47 and 4 (emerging) | Penn World Table 10.01 (labour shares and capital stocks, 2019, group medians) |
| Supply | Tariffs' long-run cost | Potential output falls 0.25 × imports' share of GDP per point of tariff, before capital adjusts | Calibrated to Furceri, Hannan, Ostry and Rose (2018), NBER WP 25402 (1 point of tariff: output −0.12% after five years), checked below |
| Judgement | Other links | demand matches 50% of a change in potential at once; 50% (60% emerging) of an energy or terms-of-trade income change is spent, 35% of that for an energy exporter's windfall; budgets take 70% (SAU), 40% (RUS), 20% (MYS), 20% (MEX) of energy export income; net exports lose 0.1 × openness of GDP per 1% real appreciation; sectors' sensitivity to the gap: agriculture 0.3, industry 1.5, manufacturing 1.8, services 0.85; tariffs cut demand by 0.2 × imports' share per point and 70% of their revenue is kept; second-round price effects 0.15 (advanced), 0.30 (emerging); the neutral-rate correction 0.05 a quarter per 1% gap; real exchange rate 2% per point of real-rate differential (half in the euro area, none under a peg); unemployment adjusts 50% a quarter; capital closes 2% of its gap a quarter; energy's long-run supply cost 0.1 × its weight in prices per 1%, taxes' 0.1% of potential per 1% of GDP; imports 1.5 × output and 0.8 × prices, exports 0.6 × the real exchange rate; trade shares capped at 60% of GDP (re-exports); lower bounds 0% (-0.5% euro area, -0.1% Japan, -0.75% Switzerland)judgement | CK's judgement within the ranges of the IMF's projection models and the literature above; the validation below checks the combined result |
Validation against published responses
The model’s responses to the experiments central banks, the IMF and studies have published, run on this page as it is built. Each must fall in a band from half the smallest published figure to twice the largest (for the one ratio, the second-year multiplier over the first, from 1.0 to 1.6): right in direction and within a factor of two, which is about how far published models differ from each other. The suite also checks, for every economy and lever, the textbook signs and shapes (a rate rise lowers output before inflation and is neutral in the long run; supply shocks move output and inflation in opposite directions; temporary shocks fade). It runs on every change.
| Experiment | Check | Published | Model | Band | Source |
|---|---|---|---|---|---|
| UK: policy rate +1 point for a year | Largest fall in GDP | −0.20% to −0.35%, after about 5 quarters | −0.55% | −0.70 to −0.10 | Bank of England (1999), “The transmission mechanism of monetary policy”, Quarterly Bulletin, May, p. 170 |
| Largest fall in annual inflation | −0.2 to −0.4 points, after about 9 quarters | −0.18 pt | −0.80 to −0.10 | Bank of England (1999), “The transmission mechanism of monetary policy”, Quarterly Bulletin, May, p. 170 | |
| Australia: cash rate +1 point for four quarters, then the usual rule | GDP after six quarters | about −0.8% | −0.49% | −1.60 to −0.40 | Ballantyne et al. (2019), “MARTIN Has Its Place”, RBA Research Discussion Paper 2019-07, section 5.1 |
| Largest rise in unemployment | about +0.3 points | +0.25 pt | +0.15 to +0.60 | Ballantyne et al. (2019), “MARTIN Has Its Place”, RBA Research Discussion Paper 2019-07, section 5.1 | |
| Annual inflation after two years | a little less than −0.2 points | −0.15 pt | −0.40 to −0.10 | Ballantyne et al. (2019), “MARTIN Has Its Place”, RBA Research Discussion Paper 2019-07, section 5.1 | |
| Euro area (Germany here): ECB rate +1 point for two years | GDP in the second year | −0.38% (national models, euro area), −0.71% (Area-Wide Model); Germany's own model −0.33% | −0.54% | −1.42 to −0.16 | van Els, Locarno, Morgan and Villetelle (2001), ECB Working Paper 94, Tables A.1–A.2 (national models and the Area-Wide Model) |
| Price level in the fourth year | −0.40% (national models), −0.49% (Area-Wide Model); Germany −0.56% | −0.51% | −1.12 to −0.20 | van Els, Locarno, Morgan and Villetelle (2001), ECB Working Paper 94, Tables A.1–A.2 (national models and the Area-Wide Model) | |
| U.S.: government spending +1% of GDP for two years | GDP in the first year (the first-year multiplier) | 0.75 on average across 41 studies of advanced economies | +0.78% | +0.37 to +1.50 | Batini, Eyraud, Forni and Weber (2014), IMF Technical Notes 14/04, pp. 3–9 |
| Second-year multiplier over the first | 10% to 30% larger | 1.05× | +1.00 to +1.60 | Batini, Eyraud, Forni and Weber (2014), IMF Technical Notes 14/04, pp. 3–9 | |
| U.S.: taxes +1% of GDP | GDP in the first year | −0.25 (revenue multiplier, advanced economies) | −0.31% | −0.50 to −0.12 | Batini, Eyraud, Forni and Weber (2014), IMF Technical Notes 14/04, pp. 3–9 |
| India: government spending +1% of GDP for two years | GDP in the first year | 0.1 to 0.3 (emerging economies) | +0.34% | +0.05 to +0.60 | Batini, Eyraud, Forni and Weber (2014), IMF Technical Notes 14/04, pp. 3–9 |
| UK: average tariff +3.6 points (one standard deviation in the study) | GDP after five years | about −0.4% (151 countries; about −1% for advanced economies) | −0.49% | −2.00 to −0.20 | Furceri, Hannan, Ostry and Rose (2018), NBER WP 25402, p. 12 |
| U.S.: average tariff +10 points, no retaliation | Inflation in the first year | +0.4 points (IMF GIMF) to +0.6 points (G-Cubed) | +0.40 pt | +0.20 to +1.20 | IMF WEO October 2025, Box 1.2; McKibbin, Hogan and Noland (2024), PIIE WP 24-20 |
| Largest rise in the price level | about +0.8% (G-Cubed by 2028; Yale Budget Lab, scaled) | +0.84% | +0.40 to +1.60 | McKibbin, Hogan and Noland (2024); Yale Budget Lab, November 2025 | |
| Germany: energy prices +10%, lasting | GDP after three years | about −0.25% for the euro area, from about 0 to −0.4% by country (a 10% oil price rise) | −0.19% | −0.50 to −0.12 | ECB Monthly Bulletin, April 2011, Box 1 (from ECB Occasional Paper 113) |
| Price level in the first year | about +0.2% | +0.30% | +0.10 to +0.40 | ECB Monthly Bulletin, April 2011, Box 1 | |
| Price level after three years | about +0.45% | +0.49% | +0.22 to +0.90 | ECB Monthly Bulletin, April 2011, Box 1 |
Monetary policy lowers output by more than the Bank of England’s 1999 model and by less than the Reserve Bank of Australia’s MARTIN, and inflation bottoms out sooner than the Bank of England found. The energy checks use a 10% rise in all energy prices against the ECB’s estimate for oil alone.
Known limits
- A small model of one economy: the rest of the world doesn't react (no retaliation, no foreign slowdown), and there are no financial crises, defaults or confidence effects.
- Expectations adapt to what has happened; they don't jump on announcements, so pre-announced or credible policies may work faster than shown.
- Parameters are typical published estimates for advanced or emerging economies, adjusted per country where an estimate exists: the true effects are uncertain, often by half or more.
- Effects are linear: twice the shock gives twice the effect, except where the policy rate hits its lower bound and for tariff revenue (which shrinks as imports fall).
- The baseline is the IMF's, taken as given; the scenario shows only how the levers move the economy from it.
- The budget is simplified: energy subsidies cost nothing and extra debt carries no extra interest, so debt paths are approximate.
- A few economies fit the model badly: Argentina and Türkiye (inflation far above target), Russia (sanctions), Saudi Arabia (a peg and oil revenue), China and Vietnam (managed exchange rates), Singapore (an exchange-rate instrument) and Ireland (GDP inflated by multinationals). Their results say so.
Out of scope
- DSGE or FRB/US-scale models
- Market prices, returns or predictions
- Other countries' reactions and retaliation
- Financial crises, defaults and confidence
- Forecasts or probabilities of any scenario