LEARN MORE
What these numbers actually mean — and why they affect you
Every figure behind the score, in plain language, with the exact arithmetic beside it — so you can follow how five published numbers become one rank, or check it yourself.
THE METHOD AT A GLANCE
How five published figures become one rank
Nothing here is hidden and nothing is estimated by us. Each step is shown below, and every constant it uses is published further down this page.
- Today’s published figures, grouped into threeAustralia, as a worked example
- Inflation3.5%38% of itvs 2.5% target → 0.95 away
- Unemployment4.5%62% of it
Conditions80 percentile+0.82 sd- GDP growth2.1%
Direction51 percentile+0.16 sd- Government debt50.7%50% of it
- Policy rate4.35%50% of it
Capacity84 percentile+1.06 sd- Average the three, then rank85.1better than 85% of the last 26 years
- 01Start with what is publishedFive official figures, each from a named national source — plus each country’s own inflation target.
- 02Compare with historyNot "is 4.35% high?" but "where does it sit among 148 readings from these six economies since 2000?"
- 03Group into threePrices and jobs. Which way it is moving. What is left in reserve.
- 04Put the three levelEach is divided by its own spread, so one cannot shout over the others.
- 05Rank itThe average becomes a place in the historical record. That is the score.
WHY THESE THREE GROUPS
The indicators that move together, and the ones that do not
Putting five figures into three dimensions asserts that each group belongs together and that the groups measure different things. Measured across the whole reference record, here is whether that is true.
Pairs of indicators, strongest first
- Govt debtPolicy ratemove together · same dimension
- GDP growthPolicy ratemove together · across dimensions
- GDP growthGovt debtloosely linked · across dimensions
- UnemploymentGovt debtloosely linked · across dimensions
- InflationGovt debtloosely linked · across dimensions
- UnemploymentPolicy rateloosely linked · across dimensions
- InflationUnemploymentloosely linked · same dimension
The three dimensions, strongest link first
- DirectionCapacitymove together · 12% shared
- ConditionsDirectionbarely related · 2% shared
- ConditionsCapacitybarely related · under 1% shared
SCORED MARKETS
The five indicators behind every MMAI score
One card each, under the dimension it feeds. Oil and the method section are on this page too, listed separately because neither is one of the five.
Inflation & CPI — the price of everything you buy
The Consumer Price Index (CPI) tracks the price of a representative basket of everyday goods and services. Inflation is how fast that basket's total cost is rising year over year.
How far inflation sits from the country’s own target is compared with its average across the 148 market-years in the record: 1.1 percentage points away. Closer than that scores above zero, further away scores below, and every 1.2 points away from the average moves this indicator’s score by one.
For example, inflation half a point from target scores +0.53; four points from target scores −2.40.
Show the exact formula
X is the published figure. 0 is average distance from target across the record; positive is closer to target than average.
- United States 2%
- Euro Area 2%
- Canada 2%
- United Kingdom 2%
- Japan 2%
- Australia 2.5%
- Sweden 2%
- Norway 2%
- South Korea 2%
The target is each central bank's own published objective, not one number applied to everyone. Every scored market aims at 2% except Australia: the Reserve Bank targets a 2–3% range and sets policy so inflation returns to the midpoint, so MMAI scores Australia against 2.5%. The United States is measured on the PCE price index rather than CPI, because that is the index the Federal Reserve defines its 2% target on.
When CPI rises faster than your income, your money buys less. This is why inflation directly shapes wage negotiations and why central banks raise interest rates specifically to slow it down.
GDP Growth — is the economy actually expanding?
Gross Domestic Product (GDP) is the total value of everything a country produces. GDP growth measures whether that total is expanding or shrinking.
GDP growth is compared with its average across the 148 market-years in the record: 1.8% a year. Faster than that scores above zero, slower scores below, and every 2.3 points away from the average moves this indicator’s score by one.
For example, growth of 3% scores +0.53; a 1% contraction scores −1.19.
Show the exact formula
X is the published figure. 0 is average growth across the record; positive is faster than average.
When growth is strong, businesses hire more and wages tend to rise. When it shrinks for two consecutive quarters, that's a recession.
Unemployment Rate — how easy is it to find or keep work?
The share of people actively looking for work who can't find it, as a percentage of the labour force.
Unemployment is compared with its average across the 148 market-years in the record: 6.1%. Lower than that scores above zero, higher scores below, and every 2.1 points away from the average moves this indicator’s score by one.
For example, unemployment of 3% scores +1.49; 8% scores −0.93.
Show the exact formula
X is the published figure. 0 is average unemployment across the record; positive is lower than average.
Low unemployment means workers have leverage — more job offers, easier to negotiate a raise. High unemployment flips that.
Policy Interest Rate — the cost of money, and the room to move it
The benchmark rate set by a central bank. It ripples out to set the tone for every other interest rate in the economy.
The policy rate is compared with its average across the 148 market-years in the record: 2.0%. Higher than that, meaning more room to cut, scores above zero, lower scores below, and every 2.0 points away from the average moves this indicator’s score by one.
For example, a rate of 4% scores +1.00; 0.5% scores −0.79.
Show the exact formula
X is the published figure. 0 is average policy rate across the record; positive is a higher rate, meaning more room to cut.
The policy rate is no longer measured against an estimated neutral rate. It is scored against its own history across the six economies, inside Capacity, where the question is how much room a central bank has to cut if trouble arrives. That removed the least defensible number in the old method: a neutral-rate estimate for Australia dating from a 2017 study, because the Reserve Bank publishes its current figures only in charts. Nothing in the score now depends on anyone’s estimate of neutral.
This number sets the floor for mortgage rates, car loans and credit card rates, so a rise makes borrowing more expensive for you. Inside the score it reads the other way round: a central bank at 4% can cut if trouble arrives, and one at 0.5% has almost nothing left to give. Both are true at once — one is what a high rate costs you today, the other is what it buys the economy later.
Government Debt — how much room a country has left
What a government owes, measured against the size of its economy. A country with low debt can borrow to cushion a recession; a country with high debt has less room to act and spends more of its budget on interest instead of services.
Government debt is compared with its average across the 148 market-years in the record: 83% of GDP. Lower than that scores above zero, higher scores below, and every 46 points of GDP away from the average moves this indicator’s score by one.
For example, debt of 45% of GDP scores +0.82; 176% scores −2.00.
Show the exact formula
X is the published figure. 0 is average debt across the record; positive is lower than average.
There is no threshold and no cap: every point of debt counts the same, above the average or below it. Earlier versions of MMAI gave full marks to anything under the European Union's 60% reference value and penalised debt above it only gently; that method was retired on 5 September 2026 and is not used. Japan shows what the straight-line rule does. Its debt, far above the record's average, puts its Capacity near the bottom of the range — even though that debt is owed mostly to its own citizens, in its own currency. MMAI does not soften the arithmetic for that. It publishes all three dimensions beside the headline, and the homepage warns when the headline is hiding a weak dimension.
High government debt tends to show up in your life slowly — as higher taxes, tighter public spending, or less help from the state when the next downturn arrives.
Brent Crude Oil — shown, but no longer scored
The global benchmark oil price. Even if you never buy gasoline directly, oil is embedded in shipping, manufacturing, and heating costs.
Oil used to be MMAI's fifth pillar and stopped being one on 1 September 2026. The reason is simple: there is only one oil price in the world, so every country received exactly the same score for it. It moved every score up and down together and could never make one country look better or worse than another — while taking 12.5% of the weight. Government debt replaced it because every country has its own. Oil is still shown in the weekly feed, where a single global price is exactly what it claims to be.
Oil spikes show up weeks later as higher prices at the pump, higher heating bills, and higher shipping costs baked into everything you buy.
How MMAI Combines All of This Into One Score
Every indicator is first expressed as how unusual it is against one historical record — how far from average, in standard deviations, across 148 market-years from these same six economies, 2000–2025. The signs are aligned so that positive always means better. Those five numbers form three dimensions: Conditions (inflation and unemployment), Direction (GDP growth), and Capacity (government debt and room to cut rates). Each dimension is standardised against the same history, the three are averaged equally, and the result is published as a percentile — a rank, not a mark out of 100. A score of 85 means better than 85% of that record.
Every figure describes what has already been published by an official source. MMAI contains no prediction, no model and no view about what happens next. It is also why the yield curve is not scored despite the data being collected: its standing in economics is almost entirely as a predictor of future recessions, and an index that disclaims forecasting cannot lean on one.
Averaging three numbers means a collapse in one can be offset without limit by the other two. Japan is the live case: its Conditions score is the highest of every market-year in the record — inflation on target, unemployment near 2% — while its Capacity is close to the lowest, with debt near 180% of GDP and almost no room to cut rates. Those genuinely offset, and a headline near the middle describes neither. Where that is happening the site says so beside the score, and all three dimensions are always shown.
Inside Capacity a higher policy rate scores better, which is the opposite of how a rate is usually read. That is deliberate, and it only makes sense alongside the question Capacity asks: how much room is there to respond to trouble. A central bank at 4% can cut; one at 0.5% cannot. Read without that framing the sign looks simply wrong, which is why the question appears beside the number everywhere it is shown.
Most published indices give you a number and ask you to trust the brand. These are the checks MMAI ran on itself, and it publishes the answers whether or not they flatter it.
The historical record changes data source partway through: figures to 2018 come from the World Bank’s annual series, and from 2019 onward from each country’s own statistics office. Rebuilding the entire history on a single source moves any published score by up to 2.3 percentile points, though it does not change the order. Countries closer together than that are reported as level rather than ranked, so the site never claims a lead it cannot support. GDP growth is where the two sources disagree most, by about a quarter of a point on average — later revisions, not a mistake by either.
The 6 judgment calls, stated in fullEvery index of this kind rests on choices like these. Open to read each one.
The data and the arithmetic are checkable. Every figure comes from a named official source, the 148-row reference table is published in full, and any reader can reproduce a score by hand — one already has, working only from the published files and without being told the answer. What follows is the layer above that: the choices about how those figures are combined. No experiment settles what inflation is worth against unemployment, so every index declares an answer instead. What is unusual here is stating them, and measuring what they actually do.
No index of this kind can measure what inflation is worth against unemployment, because there is no experiment that would settle it — so every one of them declares a weighting instead. MMAI does too, and says which is which. The inflation-to-unemployment split follows published research on the wellbeing cost of each, though that research measures the cost of a higher inflation RATE while MMAI applies it to distance from a target; those are not the same quantity, so the split is described as informed by that evidence rather than measured for this use. The two weights inside Capacity are declared outright. Nothing was fitted to MMAI’s own data, which is what would make a score circular.
Weighting the three dimensions equally is the most common approach for an index like this and the default in the OECD’s handbook. That is a reason to adopt it, not evidence the three matter equally. Testing shows it is also the choice the ranking is most sensitive to: vary the other weights by half and the order never changes, but vary these and it does. And as measured above, equal weighting does not produce equal influence, because the three dimensions move together to different degrees. All three are published beside the score so a reader who disagrees can weigh them differently.
MMAI penalises inflation one point above target exactly as much as one point below. The evidence does not support that, and does not agree on the alternative either: estimates of what central banks actually do find overshoots penalised about three times as heavily, while work on what happens when interest rates are near zero implies the opposite. Adopting either would mean picking a side of a live argument. Symmetry is kept because the European Central Bank states it in its own published strategy — negative and positive deviations equally undesirable — a citation to a stated objective rather than a measurement. Choosing differently would move scores by around six percentile points.
Unemployment is scored against one historical distribution for all nine markets, so a country with structurally higher unemployment scores lower for reasons that are not about current conditions. Japan and the euro area do not plausibly share a single definition of full employment. Using each country’s own estimate would be more accurate, but those estimates are themselves uncertain and revised often, which would trade one problem for another.
The reference distribution holds 148 market-years rather than a clean 26 for each of the six markets in it. Japan contributes 18, because for eight of those years its central bank targeted the quantity of money rather than an interest rate, so no policy rate exists to record. Those years are absent at source, not dropped for convenience — but it does mean Japan shapes the historical average slightly less than the others, and the years missing are its most unusual ones.
The 148-row reference distribution was built from six markets and frozen. Sweden and Norway and South Korea joined afterwards, so their scores are a rank against other countries' history rather than against their own. That is deliberate: re-cutting the distribution every time a market is added would move every published score for reasons having nothing to do with the economy. It does mean the comparison is to a shared yardstick, not a home one — and the yardstick is six large economies over 2000–2025.
Equal thirds is what the method DECLARES. Measured across the reference history (as of 13 September 2026), the effective influence is Conditions 25.5% · Direction 39.6% · Capacity 34.9% — a spread of 14.1 points where the declared spread is zero, with Direction doing the most work. The dimensions move together to different degrees, so the composite tracks some more closely than others. Standardising them equalised how much each VARIES, which is what it was for; it does not equalise how much each MATTERS, and nothing can while they stay correlated.
A measure every country scores alike cannot tell them apart, whatever weight it carries — the reason the global oil price stopped being scored. Asked of the three dimensions, across the nine scored markets, highest first: Conditions 27.2 · Direction 25.7 · Capacity 21.8. Conditions does the most to distinguish one country from another; Capacity does the least. Re-measured rather than assumed, and it will shift as cycles diverge.
You can take the raw numbers for any market, run them through the formulas shown here, and get the exact score shown on the Overview page — no hidden adjustments. The full 148-row reference table is published too, so the averages themselves can be checked rather than taken on trust. A percentile describes conditions for an average household and business in that economy against the last 2000–2025 — manageable inflation, available jobs, reasonable borrowing costs — not investment returns, and not a government’s policy performance. MMAI is not investment advice.
BEFORE YOU RELY ON IT
Common questions
What the score is for, what it can tell you, and how it holds up against a year everybody remembers.
Isn't this just public data collected in one place?
The data is public and there is no point pretending otherwise. But putting it in one place is the one thing you must not do, because these numbers do not mean the same thing.
Every country headlines a different growth figure. The United States and Canada take a quarterly number and multiply it up as if the year continued at that pace. The United Kingdom's headline compares against the previous three months, not the same period last year. Australia's is not adjusted for inflation at all. Collect those four, display them side by side, and you have built a comparison that looks entirely reasonable and is wrong.
MMAI rejects all four headlines and rebuilds each one on the same basis. Central banks also aim at different targets — five at 2%, Australia at 2.5% — so inflation is scored against each bank's own objective rather than one number applied to everybody. Collection is the raw material, not the product. The judgment that makes 9 countries genuinely comparable is the work.
Does it predict the market?
No. It has never been tested against market returns, and no claim is made that it forecasts them. Anyone who tells you otherwise is overselling it.
MMAI measures the economic environment, not asset prices. Those are related but they are not the same thing, and they routinely move in opposite directions — markets often rise while conditions are still deteriorating. Whether the score leads or lags market returns is a legitimate research question. It has not been done, so it cannot be claimed.
What is it actually worth to someone who is not an economist?
It tells you what kind of environment you are making decisions in, before you make them. Most people already feel the economy — rent is up, the job market feels tight or loose, a mortgage costs more than it did. What is hard to tell is whether that is your situation or the situation, and whether it is getting better or worse.
It can answer: is borrowing unusually expensive right now, or does it just feel that way? Is my country's job market genuinely stronger than its neighbours'? Are conditions improving over the last few years? It cannot answer: should I buy a house, what will rates do next, where should I put my savings, is my job safe. It is context, not a signal.
Has it been tested?
Run across the 148 market-years behind every score, the index falls hard through the 2008 crisis for every country that had a recession — and not for the one that did not. Ranks, so the change is in percentile points:
| Market | 2007 | 2009 | Change |
|---|---|---|---|
| United Kingdom | 92 | 8 | -84 |
| United States | 81 | 3 | -78 |
| Canada | 75 | 5 | -70 |
| Euro Area | 68 | 3 | -65 |
| Japan | 36 | 0 | -36 |
| Australia | 99 | 80 | -19 |
Australia is the important row. An index that simply reacted to global bad news would have dropped Australia with everyone else. It stayed high, because Australia's own conditions held — it did not have a recession. That is evidence the index reads each country rather than following the headlines.
And the limit of that test, before you ask. 2007, 2008 and 2009 sit inside the 148-row reference window the score is ranked against. So this is a consistency check — the index reproduces the crisis and separates the country that avoided it — not an out-of-sample forecast test. A genuine out-of-sample test has not been run.
GULF WATCH (NON-SCORED)
Six more countries, tracked but deliberately unscored
Cost of Living — inflation across the Gulf
The same measure used everywhere else on MMAI: year-over-year change in the price of a representative basket of everyday goods and services, tracked separately for UAE, Bahrain, Qatar, Saudi Arabia, Kuwait, and Oman.
A rising CPI in a Gulf market squeezes salaries, expat remittances, and local business costs the same way it does anywhere else — even though most of these currencies are pegged to the US dollar and largely import their monetary policy from the US Federal Reserve rather than setting it independently.
Investment Confidence — where property markets are heating up or cooling down
Property transaction volumes and values, tracked differently in each country depending on what its land authority or statistics office actually publishes — from Dubai's raw transaction registry to Qatar's real estate trading index. Every Gulf Watch card labels its exact metric rather than showing a generic "price" number.
Gulf economies lean heavily on real estate and foreign investment as part of their push to diversify away from oil. Rising transaction volumes tend to signal investor confidence in a country's direction; a slowdown often shows up here well before it reaches slower-moving indicators like GDP.
Currency Stability — FX reserves and the current account
FX reserves, measured in months of import cover, show how long a country could sustain its current imports using only its foreign currency reserves — a standard IMF stability benchmark. The current account balance (as % of GDP) shows whether a country is a net earner or net spender of foreign currency overall.
Most Gulf currencies are pegged to the US dollar, so unlike a floating currency, strain on the peg doesn't show up as a falling exchange rate — it shows up here first, in shrinking reserves or a widening current account deficit, well before the peg itself would ever be tested.
Why Gulf Watch Has No Composite Score
Gulf Watch deliberately does not combine these figures into one score the way MMAI does for its nine scored markets. The reason is comparability, not missing data: Gulf unemployment largely reflects how migrant labour is counted, so it does not carry the same meaning as in the scored markets, and a score built on it would rank unlike things against each other. Instead of producing that number anyway, Gulf Watch shows exactly what each government publishes, with its own date, and marks the rest "N/A."
Every figure on Gulf Watch traces back to its source on the Sources page, same as every scored market. The missing score here isn't a gap in the site — it's the same honesty rule that keeps every other number on MMAI trustworthy, applied consistently even when it means showing less rather than more.