Transparency

How every number is computed

AGI is built on one principle: every score is explainable, every formula is public, and every input is labelled live or seed. Nothing here is a black box — hover the ⓘ beside any number on the site to see the same definition you'll find below.

01

Data provenance

● Live

Big Mac prices & currency valuations

The Economist Big Mac index, cycle 2026-01-01, plus 26 years of history (2000–2026). Real prices in local currency and USD, and implied over/undervaluation vs the dollar.

● Live

Inflation, unemployment, health, lethal violence & GDP growth

Real national figures from World Bank Open Data (latest year): consumer-price inflation, the modelled ILO unemployment rate, life expectancy at birth (which sets the health-outcome score), the UNODC intentional-homicide rate (which sets the lethal-violence score — homicide only, not total crime), and real GDP growth (3-year average, which drives the opportunity score). These feed the risk, purchasing-power, opportunity and life-index scores directly. The same series power the /trends explorer and the per-country profiles.

Seed

All other indicators

Income, safety, mobility, education and the rest are illustrative sample values for the MVP, transitioning to licensed live feeds (World Bank, IMF, WHO, UNODC and others).

Normalized

0–100, comparable

Scores are min-max normalized across the current 22-market seed set. Inverted metrics (cost of living, inflation, unemployment, housing, climate risk) are flipped so a better raw value always scores higher.

Independence. Scores are never influenced by partners, sponsors, or the markets being ranked. The engine is deterministic and the methodology is versioned in the open-source repository — github.com/bigbigboss67/agi. When a formula changes, the change is in the commit history.

02

Composite scores

The 0–100 headline scores, each a weighted blend of the indicators below.

Overall Score

A single 0–100 score summarizing how attractive a market looks across quality of life, spending power, growth opportunity, risk and business conditions.

How it's calculated
overall = 0.30·GLI + 0.25·purchasingPower + 0.20·opportunity + 0.15·(100−risk) + 0.10·businessScore — all five inputs are 0–100 scores built from the indicators below (health, inflation and unemployment are live World Bank 2024; the rest seed), min-max normalized across the 22 markets.

Global Life Index

A broad read on quality of life in the market — affordability, safety, health, environment, infrastructure and more, in one number.

How it's calculated
gli = mean of 20 pillar scores (affordability, purchasing power, currency-value score, income, employment, safety, health, education, happiness, housing affordability, infrastructure, governance, climate, digital, business environment, consumer readiness, tourism, mobility, air quality, population-growth), each 0–100 and min-max normalized across the 22 seed markets.

Purchasing Power

How far a resident's income actually stretches locally, combining how much they earn with how expensive the market is.

How it's calculated
purchasingPower = 0.6·incomeScore + 0.4·affordabilityScore, where incomeScore comes from net monthly income (USD, seed data) and affordabilityScore is cost-of-living inverted — both min-max normalized across the 22 seed markets.

Big Mac Affordability

How many Big Macs an average local salary can buy versus other markets — a plain-language everyday-affordability gauge.

How it's calculated
bigMacAffordability = min-max normalize(incomeNetUSD ÷ bigMacUSD) across the 22 markets; income is seed data, the Big Mac USD price is live (The Economist, Jan 2026).

Opportunity

How promising the market looks for growth, blending real GDP growth, business climate, spending power and tourism.

How it's calculated
opportunity = 0.3·gdpGrowthScore + 0.3·businessEnv + 0.25·purchasingPower + 0.15·tourism. The growth signal is LIVE real GDP growth (World Bank, 3-year average), min-max normalized; businessEnv and tourism are seed.

Risk

How much could go wrong operationally in this market — inflation, joblessness, climate exposure, a weakening currency, weak governance and crime.

How it's calculated
risk = 0.22·inflationScore + 0.18·unemploymentScore + 0.15·climateRisk + 0.18·currencyRisk + 0.12·(100−governance) + 0.15·crime, where currencyRisk = clamp((0.5 − currencyTrend/2)×100); inflation, unemployment (World Bank 2024) and crime (real homicide) are live, the rest seed; min-max normalized where applicable.

Economic Strength

The near-term strength of the local economy — income, jobs, inflation and how business-friendly conditions are.

How it's calculated
economic = 0.3·incomeScore + 0.25·employmentScore + 0.25·inflationScore + 0.2·businessEnv, each 0–100 (inflation and unemployment live from World Bank 2024, income and businessEnv seed; min-max normalized across the 22 markets).

Consumer Readiness

How ready and able local consumers are to spend — purchasing power, income and digital adoption.

How it's calculated
consumerReadiness = 0.4·purchasingPower + 0.3·incomeScore + 0.3·digital, all 0–100 (digital and income score are seed/derived indicators).

Business Environment

How favorable local conditions are for operating a business — regulation, ease of doing business and infrastructure.

How it's calculated
businessScore = 0.6·businessEnv + 0.4·infrastructure, both illustrative 0–100 seed indicators.

Data Confidence

How reliable the underlying data for this market is thought to be — lower means treat the numbers with more caution.

How it's calculated
dataConfidence is a seed value (0–100) set per market in the dataset; it is not derived from other scores and is illustrative for the MVP.
03

Big Mac · live

Sourced live from The Economist's Big Mac dataset — real prices and currency valuations.

Big Mac Price (local)

The price of a McDonald's Big Mac in the market's own currency.

How it's calculated
Live figure from The Economist Big Mac Index, Jan 2026 cycle (see lib/bigmac.ts).

Big Mac Price (USD)

The Big Mac's price converted into US dollars, so it can be compared across countries.

How it's calculated
Live local price converted to USD at the same-cycle exchange rate, from The Economist Big Mac Index, Jan 2026.

Macs per Salary

How many Big Macs an average local monthly salary could buy — a quick gut-check on real purchasing power.

How it's calculated
macsPerSalary = incomeNetUSD ÷ bigMacUSD (income is seed data; the Big Mac price is live, The Economist Jan 2026).

Work Minutes per Mac

How many minutes of work at the local average wage it takes to earn one Big Mac.

How it's calculated
workMinutes = bigMacUSD ÷ (incomeNetUSD ÷ 10,560), assuming a 22-day × 8-hour work month (10,560 minutes); income is seed data, the Big Mac price is live.

Currency Valuation

Whether the local currency looks over- or under-valued, based on Big Mac prices versus the US benchmark.

How it's calculated
currencyValuation = (bigMacUSD − US bigMacUSD) ÷ US bigMacUSD × 100 (%); positive = overvalued vs. the US dollar, negative = undervalued. Big Mac prices are live (The Economist, Jan 2026).
04

World Bank · live

Real national figures from World Bank / UNODC (latest year) — these feed the scores directly.

Inflation

The annual rate at which prices are rising in this market.

How it's calculated
inflation — LIVE from World Bank Open Data (consumer prices, % annual, 2024); used inverted in most scoring formulas so higher inflation lowers the score, and directly (higher = worse) at 22% weight in risk.

Unemployment

The share of the local workforce that is out of work.

How it's calculated
unemployment — LIVE from World Bank Open Data (modelled ILO estimate, % of labour force, 2024); used inverted in scoring so higher unemployment lowers the score, and directly (higher = worse) at 18% weight in risk.

Health

Population health outcomes, 0–100 — how healthy and long-lived people are in this market.

How it's calculated
health — LIVE, derived from real life expectancy at birth (World Bank SP.DYN.LE00.IN, 2024) mapped to 0–100 on an absolute scale (50 years → 0, 85 years → 100). Used directly in GLI and as the Healthcare industry-fit driver.

Lethal violence

Lethal-violence level, 0–100 (higher = worse). Based on the real homicide rate — this measures lethal violence specifically, NOT total or property crime, so low-homicide markets score low even where other crime exists.

How it's calculated
LIVE — from the real intentional-homicide rate (UNODC via World Bank VC.IHR.PSRC.P5, per 100k, latest year) mapped to a 0–100 index on an absolute scale (~28.6 per 100k → 100). Used directly at 15% weight in the risk formula and as (100−value) in the relocate score.

GDP growth

How fast the economy is expanding — the real (inflation-adjusted) annual growth rate.

How it's calculated
gdpGrowth — LIVE, real GDP growth from World Bank (NY.GDP.MKTP.KD.ZG), a 3-year trailing average (2022–2024) for stability. Drives the growth term of the opportunity score (30% weight), min-max normalized across markets.
05

Raw indicators

The underlying inputs. Currently illustrative seed values pending licensed live feeds.

Net Income

The average net monthly salary in this market, converted to US dollars.

How it's calculated
incomeNetUSD — an illustrative seed figure per market (USD/month), not a live feed.

Cost of Living

How expensive everyday life is in this market, on a 0–100 scale (higher = pricier).

How it's calculated
costOfLiving — an illustrative seed index (0–100) per market; inverted when used in scoring so a cheaper market scores higher.

Population Growth

How fast the local population is growing (or shrinking) each year.

How it's calculated
populationGrowth — an illustrative seed figure (annual %), min-max normalized across the 22 seed markets when used in GLI (weight 1 of 20). It is also the fallback growth signal for opportunity if a market has no live GDP-growth figure.

Safety

How safe the market feels for residents and visitors, 0–100 (higher = safer).

How it's calculated
safety — an illustrative seed score (0–100) per market, used directly (already 0–100) as one of GLI's 20 pillar inputs and as an industry-fit driver.

Education

The quality of the local education system, 0–100.

How it's calculated
education — an illustrative seed score (0–100) per market, used directly in GLI and as the Education industry-fit driver.

Happiness

A general well-being / life-satisfaction score for residents, 0–100.

How it's calculated
happiness — an illustrative seed score (0–100) per market, used directly as one of GLI's 20 pillar inputs.

Infrastructure

The quality of roads, utilities and physical infrastructure, 0–100.

How it's calculated
infrastructure — an illustrative seed score (0–100); feeds GLI directly and businessScore at 40% weight (0.4·infrastructure).

Governance

How stable and well-run local institutions and government are, 0–100.

How it's calculated
governance — an illustrative seed score (0–100); used directly in GLI, and as (100−governance) at 12% weight in the risk formula.

Digital Readiness

How advanced local digital infrastructure and adoption are — internet, payments, connectivity — on a 0–100 scale.

How it's calculated
digital — an illustrative seed score (0–100); used directly in GLI and at 30% weight in consumerReadiness (0.3·digital).

Business Environment

How friendly local regulation and market conditions are for running a business, 0–100.

How it's calculated
businessEnv — an illustrative seed score (0–100); used directly in GLI, at 30% weight in opportunity, 20% in economic, and 60% in businessScore.

Tourism

How much tourist traffic the market sees — a proxy for visitor-driven demand, 0–100.

How it's calculated
tourism — an illustrative seed score (0–100); used directly in GLI, at 15% weight in opportunity, and as an industry-fit driver for Tourism, Retail and Food & Beverage.

Climate Risk

Exposure to climate-related hazards such as heat, flooding and storms, 0–100 (higher = worse).

How it's calculated
climateRisk — an illustrative seed score (0–100, higher = worse); inverted when it feeds GLI, and used directly at 15% weight in the risk formula.

Mobility

The quality of public transit and ease of getting around the city, 0–100.

How it's calculated
mobility — an illustrative seed score (0–100) per market, used directly as one of GLI's 20 pillar inputs.

Air Quality

How clean the local air is, 0–100 (higher = cleaner).

How it's calculated
airQuality — an illustrative seed score (0–100), used directly as one of GLI's 20 pillar inputs.

Migration

How attractive the market is to inbound migrants, 0–100 (higher = more inbound migration).

How it's calculated
migration — an illustrative seed score (0–100) per market; shown on map/detail views but not currently used inside the overall, GLI or risk formulas.

Interest Rate

The central bank's policy interest rate in this market.

How it's calculated
interestRate — an illustrative seed figure (annual %) per market; shown on map/detail views but not currently used inside the overall, GLI or risk formulas.

Public Debt

Government debt as a share of the size of the economy.

How it's calculated
publicDebt — an illustrative seed figure (% of GDP) per market; shown on map/detail views but not currently used inside the overall, GLI or risk formulas.

Housing Price

The purchase price of housing, per square meter, in US dollars.

How it's calculated
housingPerSqmUSD — an illustrative seed figure (USD/sqm); inverted and min-max normalized across the 22 seed markets to form the housing-affordability pillar of GLI.

Monthly Rent

The average monthly rent for a typical home, in US dollars.

How it's calculated
rentMonthlyUSD — an illustrative seed figure (USD/month) per market; shown on detail views but not currently used inside the overall, GLI or risk formulas.
06

Derived concepts & engines

How the scores combine into decisions, suggestions and comparisons.

Industry Fit

How well a specific industry (e.g. Retail, Healthcare) suits this market, based on the pillars that matter most to it.

How it's calculated
industryFit = mean of that industry's driver scores (e.g. Retail = consumerReadiness, purchasingPower, opportunity, tourism), each already a 0–100 seed-derived score; the driver set per industry is a fixed, illustrative weighting from the strategic plan.

Forecast

A 3-scenario outlook (Conservative / Base / Growth) for how the overall score could move, with rough probabilities for each.

How it's calculated
volatility band = clamp((inflation×1.2 + climateRisk×0.15 + max(0, −currencyTrend)×20) / 100, 0.05, 0.35); conservative = overall×(1−band), growth = overall×(1+band); scenario probabilities tilt from momentum = populationGrowth + currencyTrend, then are renormalized to sum to 1. All inputs are seed indicators.

Decision Score

The Sector Decision Report's headline score for entering a given industry in a given market, blending fit, opportunity and risk.

How it's calculated
decisionScore = 0.55·industryFit + 0.25·opportunity + 0.20·(100−risk); all three inputs are 0–100 scores derived from seed indicators.

Affordability Check

Whether a specific product's price is realistic for local buyers, expressed as months of local income needed to afford it.

How it's calculated
months = productPriceUSD ÷ incomeNetUSD (monthly net salary, seed data); ≤1 month = mass-market, 1–6 = attainable, 6–24 = stretch/financing-dependent, 24–120 = high-end, >120 = elite/foreign-buyer territory. Price is whatever you enter, not dataset-sourced.

Weighted Score (WeightLab)

WeightLab's custom score — lets you re-weight the pillars yourself to match what matters to you, instead of the fixed overall formula.

How it's calculated
weightedScore = (Σ weight·pillarValue) ÷ Σ weight, over 7 seed-derived pillars (Global Life, Purchasing Power, Big Mac Affordability, Safety, Opportunity, Data Confidence, and risk entered as 100−risk so higher is always better); each weight is a 0–100 slider you set, default 50 (balanced).

Entry Window

How open the opportunity currently looks for entering a given sector in a given market, and which way it is moving.

How it's calculated
window level = the Decision Score (0.55·industryFit + 0.25·opportunity + 0.20·(100−risk)); state: open ≥72 with risk<45, ajar 61–71, narrow 50–60, closed <50. Direction = forecast probability tilt (P(growth) − P(conservative)): opening >+0.10, closing <−0.10, else steady. Supply-side saturation data is a roadmap input, not yet included.

Relocate Score

How far a salary actually goes in a city, combined with day-to-day quality of life — for comparing places to live and work.

How it's calculated
relocateScore = 0.40·spendingPower + 0.35·lifeQuality + 0.25·(100−crime), where spendingPower = min-max-normalized (salary ÷ live Big Mac USD price) with rent share as a visible sub-metric, and lifeQuality = mean(safety, health, mobility, airQuality). Salary is whatever you enter (or the local seed average); Big Mac price, health and lethal-violence (homicide) are live, the other quality inputs are seed data.

RE Entry Score

How favorable it looks to enter this property market now (buy / invest / develop), 0–100.

How it's calculated
reEntry = 0.35·valuation + 0.25·demand + 0.20·financing + 0.12·title + 0.08·(100−climateRisk). Bands: ≥68 Open, ≥55 Ajar, ≥42 Narrow, else Closed. Only computed where OECD valuation exists.

Valuation

Whether homes are cheap or expensive relative to the market's own history — the overvaluation read.

How it's calculated
LIVE — mean of OECD standardised house price-to-income and price-to-rent (100 = the market's long-run average). valuation = clamp(200 − that mean), so at/below the norm scores high, overvalued scores low.

Demand

The underlying pull on housing — economic momentum and people moving in.

How it's calculated
demand = 0.5·GDP-growth score (live, World Bank) + 0.25·migration + 0.25·population-growth score, min-max normalized across markets.

Financing

How cheap it is to borrow to buy or build — the cost of capital.

How it's calculated
financing = policy interest rate, inverted and min-max normalized (lower rate → higher score). Rate is a current seed indicator.

Map Coverage

The shading shown for countries that aren't one of the 22 fully-scored markets, so the whole world map has some color.

How it's calculated
coverage = a modeled prosperity index by World Bank income-group tier (High=78, Upper-middle=58, Lower-middle=40, Low=24) plus a small deterministic nudge (country numeric ISO code mod 9, minus 4) for visual texture. The income-group tier is a real public classification; the numeric score itself is an illustrative proxy, not a scored market.

Normalization

How raw numbers (like income or inflation) get turned into comparable 0–100 scores.

How it's calculated
normalize(value) = (value − min) ÷ (max − min) × 100, computed across the current 22-market seed dataset and clamped to 0–100; 'inverted' metrics (cost of living, unemployment, inflation, housing, climate risk) flip the direction so a better raw value always scores higher.

The Big Mac Index is a supporting indicator. It does not measure housing, does not prove causation, does not cover every country, and is never used alone to reach a conclusion. It is combined with income, inflation, rent, purchasing power and the other indicators above.

Explore it in action: the map & scoreboard, sector reports, entry windows, or the relocate calculator.