Methodology

Data and updates

Series come from official sources: INDEC's data through the Argentine government's time-series API (datos.gob.ar); the BCRA's statistics API and published tables; the Ministry of Economy's Finance Secretariat (debt and placements) and Treasury Secretariat (fiscal accounts); and the US Treasury (bill yields). The exceptions are the IPC Congreso used in the long-run CPI (see below), three hand-kept files for net reserves (see Net reserves), each row with its source, and market prices no official body publishes, which come from market-data services and are marked as unofficial: the informal and bond-market dollars and dollar futures (see Exchange Rate). An automated job fetches every series twice a day, checks it and publishes the site.

Each download is checked before it is accepted: no missing months or quarters, no dates moving backwards, no published observations disappearing, no non-positive index values, and rates between 0 and 100. A series that fails keeps its last good data and its chart shows a notice. Large revisions and unusually large moves are accepted but logged for review.

Sources revise their data, sometimes years back. Every value the site has ever received is stored with the date it was received, so any chart can be rebuilt as it looked on a past date. The full revision histories are in the project repository.

Seasonally adjusted series are INDEC's. Where INDEC publishes only the original series (for example, EMAE by sector), the site shows changes that compare like periods (year over year, accumulated) and does not show month-on-month changes, which would mostly reflect seasonality.

Changes and averages

m/m, q/q

Change from the previous month or quarter: X_t / X_t−1 − 1.

y/y

Change from the same month or quarter a year earlier.

YTD (year to date)

Change since December of the previous year. In December it equals the December-to-December change.

Accumulated y/y

INDEC's variación acumulada: the sum (equivalently, the average) of the months published so far in the year, over the same months a year earlier. For GDP this is annual growth once all four quarters are in.

Rates

Changes in a rate (capacity utilization, investment rate) are differences in percentage points, not percent changes.

Rebasing

Any index can be rebased so that a chosen month (for example, the month before a presidency starts) equals 100. Rebasing changes levels only, never percent changes.

Carryover

The statistical carry-over (arrastre estadístico): next year's growth if activity stayed for the whole year at the level of this year's last month or quarter (seasonally adjusted). For the year in progress, the table shows this year's growth if the remaining periods stayed at the latest level.

Summary dashboard

The home page summarizes about 25 indicators over the last seven months. Month on month uses the seasonally adjusted series where INDEC or the BCRA publish one (quarter on quarter for quarterly data, shown in the quarter's last month); rows marked † have no seasonally adjusted series and are shown as published, and rows with strong seasonality and no adjusted series (trade, capacity utilization, unemployment) appear only in the 12-month change view, which uses the original series. Daily series are averaged by month (reserves: the month's last value); the current month is marked as partial. Changes in rates are in percentage points. Color measures how unusual each change is: its distance from the median of the same change over the previous ten years, divided by a robust measure of dispersion (1.4826 times the median absolute deviation), capped at three. Green is better than usual and red worse, given the direction that counts as an improvement for each indicator (lower is better for inflation, country risk, the exchange-rate gap and unemployment); grey marks a large move in an indicator with no better or worse direction (the exchange rate, interest rates, base money). Changes within half a deviation of the median are left uncolored. The comparison is with each series' own recent history, so a figure can look good against a decade of high inflation and still be high by international standards.

Presidencies

Charts can be colored and summarized by presidential term. A term's first period is the first full month (or quarter) after the inauguration; the month of a handover belongs to the outgoing president, since the incoming one has had no time to change anything. Milei, inaugurated on 10 December 2023, starts in January 2024 (index base December 2023 = 100), and December 2023 counts for Alberto Fernández. Between Fernando de la Rúa's resignation on 20 December 2001 and Eduardo Duhalde's inauguration on 2 January 2002, three interim presidents served for a few days each; December 2001 stays with De la Rúa and Duhalde's term starts in January 2002.

Term statistics use each full term, whatever time range a chart shows:

• Average of m/m or q/q changes: geometric, (∏(1 + r_t))^1/n − 1. This is the constant rate that compounds to the term's cumulative change; an arithmetic mean of compounding rates overstates it.
• y/y views: inflation or growth annualized over the term, (X_end / X_base)^12/m − 1, where m is the number of months from the base month to the latest month. An average of overlapping y/y rates is not the term's inflation.
• Levels: start (base month), latest or final value, change, average, minimum and maximum.

Party colors: Peronism (PJ, Frente para la Victoria, Frente de Todos) light blue; UCR and the Alianza red; PRO and Cambiemos gold; La Libertad Avanza purple.

Episodes

Some charts can mark dated shocks ("Episodes"), with a short neutral description of each. With episodes switched on, the presidency table also shows the average monthly rate without the marked months, so readers can see how much a few months drive a term's average. Episodes are off by default; which months to exclude is a judgment the reader can inspect and switch off. The current list:

1. Post-primary devaluation (Aug–Sep 2023): the official exchange rate was raised about 22% the day after the August 2023 primaries.
2. Plan Platita and the December devaluation (Nov–Dec 2023): pre-election fiscal and monetary expansion; the official exchange rate then more than doubled on 13 December 2023.
3. Correction of relative prices (Jan–Apr 2024): pass-through of the December devaluation and the adjustment of regulated prices.

Weights and incidence

EMAE sector weights are each sector's share of 2004 GDP at constant prices, the base year of the index, from INDEC's quarterly value added by industry. Net taxes on products are the remainder. With these weights the sector changes add up to the change in total EMAE.

CPI weights are recovered from the published indices. A fixed-base (Laspeyres-type) index with every component on the same base satisfies I_t = Σ_i w_i I_i,t in every month, so regressing the headline index on its components without a constant returns the weights. The recovered weights reproduce the published headline index within 0.05% (categories) and 0.22% (divisions) since December 2016. They are the weights at December 2016 prices.

Incidence is a component's contribution to the change in the total, in percentage points:

incidence_k = w_k (X_k,t − X_k,b) / Σ_j w_j X_j,b × 100

where b is the comparison period: the previous month (m/m), the same month a year earlier (y/y) or the previous December (YTD). For the accumulated change the same formula is applied to year-to-date sums. Incidences add up to the total's change.

GDP

Contributions to growth use the original (not seasonally adjusted) series at constant 2004 prices, which are additive. The contribution of a component C to y/y GDP growth is (C_t − C_t−4) / GDP_t−4 × 100. A residual group (inventories and statistical discrepancy on the demand side; other services and net taxes on the supply side; other assets in investment) closes the gap to total growth exactly.

Real GDP since 1993 chain-links INDEC's accounts at 1993 prices (quarterly growth through the first quarter of 2004) with the accounts at 2004 prices (from the second quarter of 2004). Growth rates are kept from each base; the level is that of the 2004 base. INDEC's 2004 rebasing raised the measured level of GDP and changed growth in the overlapping years; the splice does not revise the 1993–2003 figures.

Quarter tracker: the seasonally adjusted EMAE averaged over each quarter, next to seasonally adjusted quarterly GDP. INDEC reconciles the two, so for published quarters their q/q changes coincide (within 0.03 percentage points since 2004). For the quarter in progress the tracker uses the months published so far: their average against the previous quarter's (q/q), what the quarter would show if the remaining months stayed at the latest month's level, and the y/y change of the same months in the original series.

Investment rate: gross fixed capital formation over GDP, both at current prices, summed over the last four quarters to remove seasonality.

Nominal GDP: prices, output and money. Monthly nominal GDP, seasonally adjusted, in trillions of pesos at annual rates: INDEC's real GDP, seasonally adjusted, times the GDP deflator (nominal over real GDP, original series), placed at each quarter's middle month and interpolated log-linearly between quarters. For the months after the last published quarter it is a proxy, the seasonally adjusted EMAE times the CPI, chained to the last quarter; it is replaced when INDEC publishes the quarter. Prices and output: the year-over-year change of nominal GDP, split into real growth and the deflator, (1 + n) = (1 + r)(1 + p). The dashed line is the nominal growth the market expected twelve months earlier: (1 + expected CPI inflation over the next twelve months) × (1 + expected real growth) − 1, from the REM, with expected real growth weighting the survey's forecasts for the current and the next calendar year by the months of each it covers. Money and velocity: by the equation of exchange, nominal GDP growth is money growth times velocity growth, (1 + n) = (1 + m)(1 + v). Money is private M2 (currency held by the public plus private current and savings accounts in pesos), monthly average of daily data; velocity is nominal GDP at an annual rate over private M2. Both splits are descriptive: they say whether spending grew with output or with prices, and with the money people hold or with how fast they spend it (in 2023 velocity rose as money demand collapsed; in 2025 it fell as people rebuilt peso balances), not which caused which.

EMAE trend-cycle in real time

INDEC publishes three versions of the EMAE: the original series, the seasonally adjusted series and the trend-cycle, a smoothed version of the seasonally adjusted series. The trend-cycle is easier to read, but its last months are revised as new data arrive, because the smoothing filter needs data on both sides of each month. How Fast the Trend-Cycle Calls a Turn measures that delay with the trend-cycle as each release published it.

Data. The m/m change of the trend-cycle in the latest month of each release, from INDEC's technical reports (Cuadro 2) from August 2016 onwards; the September 2018 report is missing. Later releases come from the copies of the series the monitor stores at each update. Changes are rounded to one decimal, as published.

Turning point. Dated on today's trend-cycle: the last month of a phase of rising months (a peak) or of falling months (a trough). Phases shorter than four months are merged into their neighbours, and a zero change continues the phase in course. Today's dating can move as the latest months are revised.

First signal. The first release after the turning point whose latest month changes with the sign of the new phase. Confirmed: the first release that starts a run of three releases in a row, all with that sign. Delays are counted two ways: months of data (from the turning point to the last month in the release) and calendar months (from the turning point to the release date, which adds INDEC's publication lag of about two months).

Reversed signal. A release whose latest month changes sign while today's trend-cycle shows the opposite sign for that same month: a turn the real-time series called and later took back.

That release dated the turn at. In the confirming release, the month of the highest (peak) or lowest (trough) trend-cycle level within six months of today's date for the turn; left blank when that release shows no turn near the month.

CPI since 1943

From January 2007, INDEC was intervened by the national government, under Secretary of Domestic Trade Guillermo Moreno, and its CPI understated inflation. In February 2013 the IMF issued a declaration of censure against Argentina over the quality of its CPI and GDP data (IMF Press Release 13/33). After the change of government, Decree 55/2016 declared an administrative emergency of the national statistical system in January 2016 (InfoLEG), and INDEC suspended the CPI while it was rebuilt. A new CPI for Greater Buenos Aires started with May 2016 and the national CPI with December 2016; the IMF removed the censure in November 2016 (IMF Press Release 16/497).

The chart shows two series, chain-linked by monthly changes:

• Official (as published): INDEC's historical CPI to December 2006; CPI-GBA (2007–2013); the national urban IPC-NU (January 2014 – October 2015, the last figure published before the change of government); CPI-GBA base April 2016 (May–December 2016); and the national CPI from January 2017. There is no official CPI for November 2015 – April 2016.
• Corrected: the same, with the IPC Congreso, a composite of private inflation estimates, for January 2007 – April 2016, including the November 2015 – April 2016 gap. While INDEC was intervened, members of Congress published the average of private estimates and kept the private sources confidential, to shield them from the fines and lawsuits the Secretariat of Domestic Trade brought against private estimators.

Both series are equal through December 2006, so in levels the distance between them is the inflation the official figures left out: from January 2007 to October 2015, official CPI inflation was 159.6% and corrected inflation 631.8%, so the corrected price level ends 2.8 times the official one. Across the November 2015 – April 2016 gap, the official level carries on with the corrected series' changes so that later levels stay comparable, but the gap months are left blank and no change spanning them is shown.

CPI with the 2017/18 basket

The national CPI weights each of its twelve divisions by spending shares from the 2004/05 household expenditure survey (INDEC, CPI methodology). INDEC prepared a new index with weights from the 2017/18 survey, due to start with January 2026 data; its introduction was postponed. The estimate applies the new division weights to the published division indices from December 2025:

CF_t = H_t × Σ_i w^new_i r_i,t / Σ_i w^old_i r_i,t,   r_i,t = I_i,t / I_{i,Dec 2025}

H is the official headline index; w^old are the current basket's weights (recovered as above) at December 2025 prices; w^new are the 2017/18 division weights, taken as applying in December 2025. The denominator rebuilds the official index from the same division data, so the gap between the two series is the effect of reweighting alone (the rebuilt index matches the official one within 0.02%). The estimate captures only the shift in weights across divisions; the new index would also change the items and weights within divisions and by region, which cannot be reproduced from published data.

Market expectations (REM)

The BCRA's Relevamiento de Expectativas de Mercado asks about 40 forecasters (banks, consultancies, universities) each month for their forecasts of inflation (headline and core), the official exchange rate, the TAMAR rate, GDP, unemployment, trade and the primary fiscal balance. The survey closes in the last days of the month and is published early the next; we date it by the month it closes and read every figure from the BCRA's historical workbook, which reports the median, the mean, percentiles and the number of answers.

Expected paths. Several charts extend the data with the latest survey's expected path: monthly inflation (next seven months), TAMAR (monthly average, next six months), GDP (quarter on quarter, seasonally adjusted, next three quarters), unemployment (next three quarters) and the official exchange rate (monthly average, on the futures curve). The dashed line is the median; the band, the middle half of the answers (25th to 75th percentile). The path starts at the first period not yet published, so it can include the month the survey was taken.

Disagreement. The spread of 12-month inflation expectations across forecasters (10th, 25th, 75th and 90th percentiles around the median) is a simple measure of how uncertain or credible the disinflation path looks.

Forecast revisions. For each calendar year, the median forecast in every survey that asked about it (up to about three years ahead), against the outcome: December-to-December CPI inflation (INDEC), the change in the year's average real GDP (INDEC quarterly GDP), the December average of the official wholesale rate (BCRA, Communication A 3500) and the primary balance of the national non-financial public sector, cash basis (Secretaría de Hacienda; the survey's figure, in billions of pesos). The survey's GDP and primary-balance definitions follow INDEC's and the Secretaría's, but first estimates are revised, so outcomes may move after the fact.

Money & Banking

Data come from the BCRA's statistics API, daily. Exchange rates, interest rates and reserves are shown daily. For daily data, a month-on-month or year-on-year change compares with the last business day on or before the same date a month or a year earlier.

Monthly averages, not end of month. Money stocks (base money, aggregates, deposits, loans) are shown as monthly averages of daily data by default, as in the BCRA's own monetary reports, for three reasons. Month-end figures swing with the payments calendar (wages paid around the turn of the month, the June and December bonus, holidays). Banks manage their reserve requirements on monthly averages, so the average is the figure that reflects their behavior. And the CPI and activity data measure a whole month, so real money balances should compare the month's average money with the month's average prices; with month-end stocks, the timing mismatch is half a month of inflation, which in 2023 and early 2024 was several percentage points. End-of-month stocks are right where flows have to add up to a change in a stock, as in the sources of base money, and base money can also be shown at month end.

Aggregates. M1 = currency held by the public + current accounts; M2 = M1 + savings accounts; M3 = currency + all peso deposits of the public and private non-financial sectors; private M2 uses private deposits only. All in pesos.

Sources of base money. The BCRA attributes each day's change in base money to its operations (factores de explicación). Each month's sum of each group is shown in percentage points of base money at the end of the previous month, so the bars add up to the month's change in end-of-month base money.

Real terms. Peso amounts can be shown deflated by the CPI, in pesos of the latest month with CPI data. For 2007–2016 the deflator is the corrected CPI (see CPI since 1943), since deflating by the official figures would overstate real growth in those years.

BCRA balance sheet. From the BCRA's historical table "Balance del Banco Central de la República Argentina", monthly since 1940, in its analytical presentation (sources of base money and absorption), grouped and valued for monetary analysis. Each month is converted from the currency of its time (peso moneda nacional, peso ley, peso argentino, austral) to current pesos with the BCRA's own factors. Changes in the valuation of the Treasury's non-transferable bills (below) and the May 2026 change in the accounting of repos create breaks in some lines.

Weekly balance sheet table. The BCRA's weekly summary balance of assets and liabilities (balances on the 7th, 15th, 23rd and last day of each month, since 2002), line by line, in the BCRA's own wording and grouping for each year; pesos of each date, dollars at each balance's exchange rate, or shares of total assets. When two dates are compared, lines are matched by name, so a line renamed or regrouped between the two dates shows no change. Where the BCRA's file has a blank or an error cell, the table shows a dash.

Valuation of the Treasury's non-transferable bills (Letras Intransferibles). The BCRA has received these dollar bills from the Treasury since January 2006 in exchange for reserves. Their book value depends on the valuation criterion in force, which has changed three times:

• Technical value (face value plus accrued interest), from 2006 to the weekly balance of 15 July 2018.
• Estimated present value (Board Resolution 209/18, 5 July 2018, applied from the 2017 financial statements): cash flows discounted at the yields of comparable Argentine-law bonds, from 23 July 2018 to 23 December 2019. The switch cut their book value from about USD 49 bn to 37 bn; after the August 2019 primaries it fell to about USD 10 bn. During 2019 the BCRA also reported net equity with the bills at technical value.
• Technical value again from 31 December 2019 (Law 27,541, art. 62; Board Resolution 2/2020), on the grounds that the BCRA would hold them to maturity; the reversal added about $2.1 trillion pesos.
• Amortized cost, capped at recoverable value (Board Resolution 128/24, recorded in the 2023 financial statements and in the weekly balance from 7 May 2024): their book value fell from about USD 69 bn to 18 bn. In April 2025 the Treasury repurchased bills with a face value of about USD 16.9 bn using IMF funds (Decree 179/25).

Sources: BCRA financial statements for 2018, 2019, 2023 and 2025; BCRA, "El BCRA adecua el valor contable de las Letras Intransferibles"; dates of each change as they appear in the weekly balances.

Inflation expectations. The median 12-month-ahead expectation in the BCRA's Survey of Market Expectations (REM), dated by the month of the survey, against CPI inflation over the following 12 months and over the previous 12 months.

Net reserves

Net reserves are estimates, not official statistics. Gross reserves include dollars the BCRA owes or does not control: banks' reserve requirements on dollar deposits, the swap with the People's Bank of China, repos, loans from international agencies. "Net reserves" subtract some of these, but there is no official or universal definition. Analysts, the BCRA and the IMF subtract different liabilities, so the same balance sheet gives different numbers. The site shows three conventions and, for the latest balance, a table with each line of each one.

• Standard (market): the estimate most private analysts publish: gross reserves minus the BCRA's foreign-currency liabilities that sit inside them, and minus the Treasury's dollar deposits, which belong to the Treasury.
• Liquid: the standard estimate minus gold, which cannot be used for payments without being sold. SDR holdings cannot be separated in the weekly balance, so they are not deducted.
• IMF-style: an approximation of net international reserves as defined for the 2025 Extended Fund Facility (Technical Memorandum of Understanding, May 2026 staff report, Attachment II, ¶11): gross reserves minus swaps, reserve requirements, deposit insurance (SEDESA), non-resident deposits, foreign-currency liabilities with an original maturity of one year or less, and the Fund's cumulative net purchases since the program started. The Treasury's dollar deposits are not deducted. The weekly balance does not show deposit insurance, BOPREAL maturing within a year, forwards or the US Treasury swap, so this line runs above the IMF's published figures: −0.5 against −2.4 bn at end-2024, and −9.2 against −11.8 bn at end-2025. The IMF also values non-dollar items at fixed program exchange rates; the site uses current rates.

Data. The BCRA's weekly summary balance sheet (balances on the 7th, 15th, 23rd and last day of each month, since 2003), converted to dollars at each balance's exchange rate: gross reserves, banks' current accounts in foreign currency, obligations with international agencies, gold. The Treasury's dollar deposits come from the monthly balance sheet (the latest month-end in force at each weekly date). The China swap is valued with the BCRA's own yuan quote. Month-end balances are lower than mid-month ones because banks move their dollar reserve requirements around the turn of the month.

Hand-kept files. Three items are not in any machine-readable source; they are kept in the repository with a source for every row (data/manual):

• China swap booked in reserves (CNY bn): CNY 70 bn from October 2014, credited in tranches; CNY 130 bn since the CNY 60 bn expansion was credited on 17 December 2018, renewed since (most recently in August 2026, for five years). The BCRA announced the tranches only through January 2015; the build-up from February to September 2015 is read from the BCRA's weekly balance, where the swap was recorded with repo liabilities (net of peso repos). That path matches the announced tranches and settles at CNY 70 bn. The activated tranche (CNY 35 bn since 2023) is part of the CNY 130 bn, not in addition to it.
• BCRA repos with foreign banks (USD m outstanding): January–November 2016, and from January 2025. The 2017–2019 repos were contracted by the Treasury, not the BCRA, and enter through Treasury deposits. Since 2025 all of the BCRA's repos have had original maturities above one year, so the IMF-style line does not deduct them. BIS loans are in the weekly balance's obligations with international agencies.
• IMF purchases under the 2025 program (USD m), net of repurchases.

The weekly balance's repo-liability line cannot replace the repo file: before July 2024 it mixes in peso repos and, through 2022, the China swap, and since then it carries other items of USD 1–3 bn. The pipeline flags any move in that line of more than USD 500 m that the repo file does not explain, so the file is updated when the BCRA contracts or repays a repo.

Banks

Sources. Two BCRA publications. The statistical annex to the monthly Informe sobre Bancos gives the financial system and bank groups (public banks; private banks, domestic and foreign; non-bank financial institutions): a standardized balance sheet, the BCRA's indicators of liquidity, credit, loan quality, profitability and capital, and loan quality by type of loan, monthly since December 2009 (Decembers since 2002). The monthly archive Información de Entidades Financieras has each institution's balance sheet, account by account in the BCRA's chart of accounts, monthly since November 1994; it omits the public banks before June 1996, so concentration and the largest institutions start then. Both are unconsolidated (each institution on its own) and are revised; the latest month is provisional.

Exposure to the public sector. Treasury and other public sector: the BCRA's indicator "credit to the public sector" (government securities, including the Treasury's LEFI, loans to the public sector and other claims), as a share of assets. Central bank: BCRA notes in banks' own portfolio plus net repos with the BCRA. Repos are measured as the difference between the BCRA's liquidity ratios with and without repos, times deposits; from July 2024 to June 2025 that difference was the Treasury's LEFI, which replaced BCRA repos, so it is not counted twice. The BCRA's indicator excludes its own paper because the central bank is not part of the non-financial public sector; for banks, both are claims on the state.

Liquidity and reserve requirements. The BCRA publishes three nested ratios over total deposits: liquidity (cash and current accounts at the BCRA); liquidity with repos with the BCRA and LEFI; and broad liquidity, which adds BCRA notes and Treasury bonds counted toward reserve requirements. Treasury bonds used for reserve requirements are broad liquidity minus liquidity with repos and LEFI, minus BCRA notes in own portfolio. The difference is zero before 2018, as expected if no bonds counted toward requirements then. The ratios cover pesos and dollars together.

Dollar deposits. Dollar items are the all-currency balance-sheet line minus the pesos-only line. Banks' dollars at the BCRA (reserve requirements) are also part of the BCRA's gross reserves.

Loan quality, capital and profitability. As defined by the BCRA. Non-performing loans: debtors in situations 3 to 6 of the debtor classification, over loans to the private sector; households are loans to individuals except for commercial purposes. Capital: regulatory capital over risk-weighted assets (credit, market and operational risk, since February 2013). Return on equity: year-to-date results over average equity, annualized, in constant currency (balances restated for inflation since 2020, IAS 29).

Concentration and the largest institutions. Shares of total assets (accounts 1 and 2 of the chart of accounts), deposits (31) and loans (13) of each institution in the system's total; the Herfindahl-Hirschman index is the sum of squared shares in percent. The composition of the ten largest groups accounts by the chart of accounts: cash and deposits in banks (11); securities (12), split by the account's description into BCRA notes, private securities and government securities (the rest, including LEFI); loans (13) to the non-financial public sector, the non-financial private sector and others (financial sector, non-residents), including interest and net of provisions; repos with the BCRA (accounts 141143, 141144, 141222, 145106, 145129, 145222), added to BCRA notes; other claims from financial intermediation (the rest of 14); and the rest. Funding: deposits in pesos (311, 312, 318) and in foreign currency (315, 316, 319), other liabilities, and net equity including the year's results. Shown as shares of each institution's assets. The groupings are ours and approximate the BCRA's, so they differ slightly from the group data in the annex.

Who holds the claims on the state. Each bank's share of the system's government securities plus loans to the non-financial public sector (national, provincial and municipal, which the chart of accounts does not separate), and of BCRA notes plus repos with the BCRA, from each institution's balance sheet at book value. The eight banks named are the largest holders of Treasury paper today; merged or closed banks are counted in "All others" throughout.

Exposure relative to capital. Claims on the Treasury, the rest of the public sector and the BCRA (as in the exposure chart, in pesos) over net equity, by bank group.

Where profits come from. The BCRA's monthly results as an annualized percentage of netted assets, in constant currency (IAS 29), averaged over the last twelve months: net interest (income minus expense), CER adjustments, results on securities plus repo premiums, exchange-rate differences, net fees, loan-loss charges, costs (administrative expenses and taxes other than income tax), the monetary result (the loss on net peso positions from inflation), and other items including income tax. They add up to the total result.

Credit by type and UVA. Financing to the private sector by type (overdrafts, promissory notes, mortgage- and pledge-backed loans, personal loans, credit cards, export finance, other) as shares of the total, pesos and dollars together; and loans in UVA (indexed to the CPI) as a share of financing to the private sector and to households, from the BCRA's loan-quality data.

Credit depth and spreads. Loans to the private sector over nominal GDP, both averaged over the last four quarters (INDEC's quarterly GDP is at annual rates). Spreads: personal-loan and overdraft rates minus the 30-day time-deposit rate, daily, in percentage points.

Fiscal & Debt

Sources. Fiscal accounts: the national non-financial public sector on a cash basis (Secretaría de Hacienda, Información Mensual de Ingresos y Gastos), monthly since 2015, and economic subsidies, both via datos.gob.ar. Debt: the Finance Secretariat's monthly debt bulletin (central administration gross debt, stocks by instrument and currency, flows, and payments, monthly since January 2019, in US dollars at each month's exchange rates), its quarterly debt report (the maturity profile) and its yearly files of debt placements (colocaciones, since 2021). The pipeline finds the latest file of each on the Secretariat's pages, since file names change with each release. The latest months of the bulletin and the placements file are preliminary.

Ratios to GDP. Every "% of GDP" on this page is a flow summed over the last twelve months, divided by nominal GDP of the same twelve months, so numerator and denominator cover the same period and the same prices. Nominal GDP comes in three steps:

1. Quarters published by INDEC. INDEC publishes quarterly GDP at current prices at annual rates, so the average of four consecutive quarters is the GDP of the twelve months ending in the last month of the fourth quarter. Example: the average of Q3 2025 to Q2 2026 is GDP from July 2025 to June 2026 (about $985 trillion).
2. Months between quarter ends. The twelve-month GDP for, say, the year to April is interpolated between the values at March and June, linearly in logs (a constant monthly growth rate between the two).
3. Months after the latest quarter published. GDP is released about two and a half months after a quarter ends, so the latest fiscal months usually have no GDP yet. For those months the latest twelve-month GDP is carried forward with prices: multiplied by the change in the twelve-month average of the CPI since that quarter's last month. Example: with GDP published through June 2026, the twelve-month GDP for August 2026 is the June figure times the ratio of the CPI averaged over September 2025 to August 2026 to the CPI averaged over July 2025 to June 2026 (about $1,034 trillion). Each card says which months use this estimate.

The estimate assumes that real GDP stays flat for those few months. With real growth of 4% a year it understates GDP by about 0.3% per month of extension, so a ratio of 5% of GDP would be overstated by less than 0.05 percentage points. When INDEC publishes the quarter, the estimate is replaced and the latest ratios are revised. In the yearly view, each year is the December value of the twelve-month series (that year's total over that year's GDP); the year in progress shows the twelve months to the latest month, marked with an asterisk, and its GDP is partly estimated in the same way.

Spending by item. Each item deflated by the national CPI. The contribution of an item to the real change in primary spending is its real change from the same month a year earlier over total primary spending a year earlier, so contributions add up to the total. Provinces: current plus capital transfers; capital spending excludes capital transfers to provinces; "Other" is the residual (goods and services, transfers to universities and public companies, and the rest).

Revenue by tax. The same method for revenue: each tax deflated by the national CPI, and its real change from a year earlier over total revenue a year earlier. Taxes are the national government's share (after automatic transfers to the provinces), from the monthly fiscal report (IMIG). "Other taxes" groups import duties, fuel, excise, personal assets and the rest, including the PAIS tax on foreign-currency purchases (December 2019 - December 2024); "non-tax and capital revenue" is total revenue minus all taxes: property income (BCRA profit transfers, the pension fund's income), fees and capital revenue. The yearly view compares the real sum of each calendar year with the year before.

Capitalized interest. The budget records interest when it is paid. Capitalizing letters and bonds (LECAP, BONCAP) pay principal and interest together at maturity, and the whole payment is recorded as amortization of debt, a financing item below the line; the interest never appears in the financial balance. We estimate the interest that accrues on them each month in three steps:

1. Stock. The stock of LECAP and BONCAP at the end of the previous month, from the Finance Secretariat's monthly debt bulletin. The bulletin values it in US dollars at that month-end's exchange rate; multiplying by the same wholesale rate (Communication A 3500) returns it to pesos, with capitalized interest to date included.
2. Rate. The average monthly yield (TEM) at which the Treasury placed fixed-rate LECAP and BONCAP over the last six months, weighted by cash raised. Each placement's yield is implied by its issue price: the payment at maturity per 1,000 of face value, 1,000 (1 + coupon TEM)^{days from issue to maturity / 30}, over the price, expressed per month. The yield is used rather than the coupon because a reopening sold above or below its accrued value changes what the Treasury will pay relative to what it received.
3. Interest. Stock times rate. Example: at the end of July 2026 the stock was US$86.8 billion, or $129 trillion at $1,488 per dollar; at an average yield of 2.14% a month, interest accrued in August 2026 is about $2.8 trillion (US$1.9 billion).

Monthly amounts are summed over twelve months and divided by GDP as above. Not used: the bulletin's own "capitalization of bonds, letters and other loans" line. It also records the capitalization already accrued on securities sold in reopenings, which the buyer paid for in cash, so it is not a cost to the Treasury; in months of large reopenings and conversions it is several times the estimate (US$14 billion in July 2025, with the end of the LEFI). Biases: the estimate covers gross debt, including LECAP and BONCAP held by the BCRA, the pension fund (FGS) and Banco Nación, whose interest the budget would net out, so it overstates; it takes the TAMAR-linked BONCAP and the dual (fixed or TAMAR) BONCAP from the January 2025 swap, which the bulletin counts in its BONCAP line, at the fixed-rate average; and it leaves out the LEFI (July 2024 - July 2025) and letters sold at a discount, which understates it. For 2024 it gives $8.2 trillion (1.4% of GDP); the Congressional Budget Office (OPC) reported about $14.1 trillion for that year, closer to the bulletin's capitalization line.

Cross-check. The bulletin's flows by instrument allow an independent estimate. For LECAP and BONCAP, capitalization is the change in the stock (in pesos) minus issuance plus amortization; since mid-2025 this residual matches the bulletin's capitalization line almost exactly. Subtracting what buyers paid above face value at auction (the accrued interest they bought in reopenings) leaves the interest that accrues to holders. For 2024 it gives $7.9 trillion, against $8.2 trillion in our estimate; in the twelve months to August 2026, $48.0 trillion against $41.6 trillion (the bulletin's line: $56.8 trillion). The cross-check runs somewhat high in months of off-auction swaps and conversions (January and July 2025, December 2025, April 2026), whose accrued interest cannot be separated from the new principal. Our estimate is the smoother and the more conservative of the two; the gap between them is a reason to read it as a floor.

Real component and an outside check. Part of the yield on capitalizing letters compensates holders for inflation. The interest card's dotted line shows the real component: each month's capitalized interest less the month's CPI inflation on the previous month's stock, that is, the stock times (1 + TEM) / (1 + inflation) − 1, summed over twelve months over GDP. For 2025 it is $9.5 trillion, 1.1% of GDP, against 3.7% for the nominal estimate; it was near zero or negative in early 2025, when monthly inflation exceeded the placement yield, and rose in the second half, when yields jumped before the midterm elections. The IMF's May 2026 staff report (footnote 2 to ¶7) states that including the real component of capitalized interest above the line would turn the 2025 overall surplus of 0.2% of GDP into a deficit of about 0.8%, a real component of about 1% of GDP, close to ours. The bulletin-flow cross-check above gives a real component of about 3.5% of GDP for 2025, far above the IMF's; its swap months inflate it. IMF staff report.

The BCRA's interest cost (quasi-fiscal deficit). The interest the central bank pays on its own liabilities: repos (pases pasivos), LELIQ and banks' remunerated current accounts, from the base-money factor "interest, premiums and remuneration of current accounts", monthly sum, twelve months over GDP. It matches published figures: $692.7 billion in 2019 (3.2% of GDP) and $9.74 trillion from 10 December 2019 to 14 June 2023. It is the quasi-fiscal deficit in the narrow sense: the BCRA's income (interest on its Treasury securities) and exchange-rate valuation gains are not netted, and they belong to its annual financial statements, not to a monthly series. It starts in 2018: LEBAC, wound down that year, were discount letters whose interest was paid at maturity with the principal and does not appear in the factor, so 2018 is partial. Consolidated interest adds it to the Treasury's interest, paid and capitalized. In 2023 it was about 10% of GDP, 8.4 points of it at the BCRA; from July 2024 the BCRA's repos became Treasury debt (LEFI, then LECAP and BONCAP), and the cost moved with them. Interest on the LEFI (July 2024 – July 2025) is in neither line: the LEFI factor nets placements and redemptions, so their interest cannot be separated.

CER adjustment. The stock of CER-linked peso debt at the end of the previous month (in pesos) times the month's change in the CER (end of month). Like capitalized interest, it is paid with the principal at maturity and never recorded as interest. It is shown on its own, not added to interest or to the adjusted balance: it is mostly compensation for inflation, and with inflation above 200% it exceeded 10% of GDP. Counting it as a cost would only be consistent in an inflation-corrected (operational) balance that also credits the government with the erosion of its fixed-rate peso debt; part of the yield on capitalizing letters is inflation compensation too, so the capitalized-interest estimate is a nominal measure, like the budget's own interest.

Operational balance. The financial balance corrected for inflation (Tanzi's operational balance): interest is counted in real terms. Starting from the accrual balance (cash balance minus capitalized interest), we add the inflation component of interest on peso debt that is not indexed: each month, CPI inflation times the stock of non-indexed peso debt at the end of the previous month (monthly debt bulletin, in pesos at the month-end exchange rate), excluding the BCRA's temporary advances. The CER adjustment of indexed debt, which is pure inflation compensation, is not counted, and the real coupon on CER bonds (small, often zero) is ignored. Interest on foreign-currency debt is counted in full. In high inflation the correction is large (in 2023-24 several points of GDP): the government's peso debt shrank in real terms, a transfer from its holders to the Treasury, not fiscal effort, so the operational balance measures the stance of policy, not the cash the Treasury must find.

What the estimates cover. Gross debt of the central administration, including securities held by the BCRA, the pension fund (FGS), Banco Nación and other public agencies. The budget's net interest excludes interest between public agencies, so the adjustment overstates what accrues to holders outside the public sector; the omissions above understate it. Neither the stock data nor the placements say who holds each security.

Country risk. "Riesgo país" is the spread of Argentina's sovereign dollar bonds over US Treasuries in JP Morgan's Emerging Markets Bond Index (EMBI), in basis points. JP Morgan does not publish it freely; we take the daily series compiled by ArgentinaDatos from Ámbito Financiero (since 1999), drop weekend rows (filled with Friday's value in some years), and mark it as market data, not an official statistic. Because the index is built from the bonds outstanding, it breaks when they change: after the 2005 swap and the 2020 restructuring, the new bonds entered the index and the spread fell by half or more in a day. Read levels across those dates with care. A new one-day move of more than 40% is flagged in the daily run so it can be checked and marked. Days found wrong in the source are corrected in a hand-kept file, each with a note: so far, 25–31 May 2023, when the feed repeated one value for five business days, are dropped. Large one-day moves that the source got right are kept, such as 30 July 2014 (a rally on hopes of a deal the day before the default).

Debt as a share of GDP. The stock of gross debt at the end of each month (monthly bulletin, US dollars) times the wholesale exchange rate of that month's last day, over GDP of the last twelve months (see Ratios to GDP). This is the Finance Secretariat's "end-of-period exchange rate" convention; it moves at once with a devaluation, while GDP catches up over the following year, so the ratio jumps and then falls back (as in early 2024). The Secretariat also reports the ratio at the year's average exchange rate, which smooths this. By holder: the quarterly debt presentation splits gross debt among public-sector agencies (securities, loans and advances held by national public bodies: the BCRA, the pension fund, Banco Nación and others), multilateral and bilateral lenders, and the private sector (the remainder). We take each quarter-end's figures from the presentation's table, value them at the quarter-end exchange rate and divide by GDP as above. Debt held by public agencies is a claim of one part of the state on another; debt held privately and by multilaterals is what the state owes outside itself.

Treasury deposits. Official deposits at the BCRA from its monthly balance sheet: the Treasury's foreign-currency deposits (valued at the balance sheet's exchange rate) and, as the rest, deposits in pesos, which include those of other national public bodies. End of month; can be shown in pesos of the latest month.

Maturities. The quarterly report's profile of capital and interest by month (the next eighteen months) and by year, for debt in normal payment status at the report's date, in US dollars at that date's exchange rates. Groups: peso debt not indexed (excluding the BCRA's temporary advances) and CER-linked, from the report's currency totals; the IMF; other multilaterals and bilateral lenders; foreign-currency bonds and other debt (the foreign-currency total minus the above and the BCRA's non-transferable letters); and the BCRA (temporary advances and non-transferable letters). Debt issued after the report's date is not included. Months already past show the total actually paid from the monthly bulletin, for all debt.

Payments in foreign currency. Capital and interest paid in foreign currency (monthly bulletin); the IMF and other multilaterals (IDB, World Bank, CAF) from the payments by creditor; bonds and other debt are the rest, including payments to public agencies.

Rollover. Issuance over amortization of peso-denominated marketable bonds and letters, from the bulletin's flows by instrument (valued in dollars on each operation's date). Letters placed with public agencies, the LEFI and dollar or dollar-linked instruments are excluded; swaps and conversions count on both sides.

Placements and auctions. Each market placement of bonds and letters in the yearly files (the "Bonos" and "Letras" sheets: auctions, their second rounds and direct placements), classified by instrument: fixed-rate pesos; CER-linked; floating (TAMAR, BADLAR); dual; dollar-linked, paid in pesos; and hard dollar. Letters placed with public agencies, swaps and conversions are left out. The files report the cash value of dollar-linked placements in pesos through 2024 and in dollars from 2025; dollar values are converted at the month's average wholesale rate. Implied yields for capitalizing fixed-rate instruments and discount letters: the payment at maturity per 1,000 of face value, 1,000 (1 + TEM)^d/30 with d the days from issue to maturity (zero-coupon letters pay 1,000), over the issue price, expressed as an effective monthly rate, (payoff / price)^30/n − 1, and an effective annual rate, with n the days from settlement to maturity. Real yields for zero-coupon CER-linked instruments (BONCER and LECER "tasa cero"): the price per 1,000 of original face value equals 1,000 times the CER ratio between settlement and issue, discounted at the real yield, so the real yield is (1,000 × CER_settlement / CER_issue / price)^365/n − 1, with the CER taken ten business days before each date (approximated as fourteen calendar days), as the bonds' terms specify. Coupon-paying CER bonds are left out. Average life: time to maturity for letters and bullet bonds; as reported for amortizing bonds.

Exchange Rate

Market exchange rates. The official rates are the BCRA's: the wholesale reference rate (Communication A 3500) and the average retail selling rate. The others are not published by any official body. MEP and CCL are the pesos per dollar implied by buying a bond with pesos and selling it for dollars, delivered in Argentina (MEP) or abroad (CCL, "contado con liquidación"); the informal ("blue") rate is quoted by street dealers; the card rate is the official retail rate plus the taxes on card spending abroad; the crypto rate is the price of dollar stablecoins in pesos. We take their daily history from ArgentinaDatos and today's quote from DolarAPI, two open-source projects that compile quotes published by the financial press. Each run compares the two sources on the same day and rejects a series if they differ by more than 5%, or if today's quote jumps more than 25% from the last close; the site then keeps the last validated data and flags the series. Selling prices throughout. Treat individual days with care: these are quotes, not transaction averages.

The exchange-rate band. Since 14 April 2025 the official rate floats within a band whose floor and ceiling the BCRA publishes daily (series "Régimen de bandas cambiarias", lower and upper limits, in its statistics API), set in advance through the end of each month (shown here through the latest day). The band started at 1,000 and 1,400 pesos per dollar; through December 2025 the floor fell and the ceiling rose 1% a month, prorated by day, and since January 2026 the monthly widening has followed a different rule set by the BCRA. We take the limits as published rather than recomputing them from the rule, so a change of rule needs no change here. Room to the ceiling: the ceiling over the official wholesale rate (Communication A 3500) of the same day, minus one.

Gaps. Each market rate over the official wholesale rate of the same day, minus one.

Carry trade. The dollar return of converting dollars to pesos, holding a 30-day time deposit renewed each month, and converting back. Each month's return is (1 + i × 30/365) × S_t−1 / S_t − 1, with i the 30-day deposit rate (annual nominal) at the end of the previous month and S the exchange rate at the end of each month; twelve months are compounded. It is shown with the official rate, which most savers could not use under exchange controls, and with the MEP rate. Before taxes and fees; a month is taken as 30 days.

Expected devaluation. Four measures, all for the official wholesale rate twelve months ahead. (1) Survey: the median expected rate for the month twelve months ahead in the BCRA's Relevamiento de Expectativas de Mercado, over the official rate's average in the survey month. (2) Auctions: when the Treasury places, on the same day, a fixed-rate capitalizing letter and a zero-coupon dollar-linked letter maturing within 45 days of each other, the breakeven devaluation is (1 + TEA_pesos) / (1 + y_USD) − 1, where TEA_pesos is the peso letter's annual yield implied by its price and y_USD = (1,000 / price)^365/n − 1 is the dollar yield of the linked letter; the month shows the average over its auctions, and months without such pairs are empty. (3) Interest parity: (1 + TEA_LECAP) / (1 + y_{US 52-week bill}) − 1, with the average yield of the month's fixed-rate letter placements and the US Treasury 52-week bill (coupon-equivalent yield, month-end). Parity includes the risk premium on peso assets, so it is an upper bound on expected depreciation. (4) Futures: the twelve-month dollar-futures price (below) at the end of the month over the official rate that day, F₃₆₅ / S − 1. The actual change in the official rate over the following twelve months is shown for comparison once it is known.

Dollar futures. Daily settlement prices of the DLR contracts traded on A3 Mercados (the exchange formed by the merger of MATba-Rofex and MAE): monthly contracts on the US dollar, settled in pesos against the BCRA's wholesale rate (Communication A 3500) of the expiry month's last business day, US$1,000 each. The data come from the exchange's public market-data API, which is not documented and is not an official statistic; each run validates the series like any other source, and our own vintages keep the history if the API changes. History starts in 2020. Constant-horizon prices: each day, the settlement prices are interpolated log-linearly in days to expiry (each contract taken to expire on the last calendar day of its month) at 182 and 365 days; beyond the longest contract the last two contracts are extrapolated, but only when the longest reaches at least 82% of the horizon, and the day is left empty otherwise. Implied devaluation: F₁₈₂ / S annualized by compounding, and F₃₆₅ / S − 1, with S the official wholesale rate of the same day. A futures price is a forecast plus a risk premium (the price of insurance against devaluation), and it can be moved by the BCRA's own sales of futures, which it does not publish as a series; read it as an upper estimate of the expected rate. Open interest is the number of contracts outstanding across all maturities, in millions of US dollars. On the curve chart, the band's ceiling is extended to each expiry at its average monthly change over the last three months, an illustration of where it would be if it kept that pace, not the BCRA's announced rule.

Real exchange rate. The BCRA's multilateral real exchange rate index (ITCRM) and its bilateral indices, daily since 1997, 17 December 2015 = 100, from the BCRA's workbook. The index is the nominal exchange rate against each partner's currency times the partner's prices over Argentina's prices, weighted by each partner's share of Argentina's trade; a rise is a real depreciation of the peso. The BCRA deflates with INDEC's CPI (and, during the 2007–2015 INDEC intervention, provincial indices) and partner CPIs; for the latest months, not yet covered by the CPI, it uses the inflation expected in its REM survey, so recent values are revised. At the CCL rate: the index times the CCL over the official rate, the price relevant to most private transactions while exchange controls were in force (November 2011 to December 2015, and September 2019 to April 2025, with restrictions on firms lasting longer); before November 2011 it equals the official index, and for November 2011 to December 2012 there is no CCL series.

The calculator. If the real exchange rate index stood at R_ref at a reference date or on average over a reference period, and stands at R_today today, the exchange rate that would restore the reference with today's prices is E* = E_today × R_ref / R_today, with E the official rate (or the CCL, with the index at the CCL rate). Averages, medians and percentiles use weekdays only. This is relative purchasing-power parity, and it inherits its assumptions: that the reference was an equilibrium and that the fundamentals that determine the equilibrium real exchange rate (terms of trade, productivity relative to partners, net foreign assets, the fiscal stance, capital flows) have not changed since. Argentina's terms of trade and energy balance have changed substantially since the 2000s, which argues for a stronger equilibrium peso than the historical averages imply. Model-based estimates, such as the IMF's External Balance Assessment, attempt to control for these fundamentals; their results are reported in the IMF's Article IV consultations.

Two equilibrium dates. If the real exchange rate was at equilibrium on two dates, A and B, its change between them is a change in the equilibrium rate, and the calculator asks how much relative productivity would have to change to account for it. Under the Balassa–Samuelson mechanism (the law of one price for tradables, wages equal across sectors, and the same weight α of non-tradables in prices at home and abroad), Δln RER = −α · Δln[(A_T/A_N) / (A_T*/A_N*)], where A_T and A_N are labor productivity in tradables and non-tradables, and starred values are trading partners'. The required change in the productivity gap is exp(−Δln RER / α) − 1 in total and exp(−Δln RER / (α · years)) − 1 a year. α is an input (0.3 to 0.7, default 0.5, roughly the weight of services and other non-tradables in the CPI); the answer scales with 1/α. Attributing the whole change to productivity makes it an upper bound: other fundamentals, above all the terms of trade (shown alongside, INDEC's quarterly index for the quarters containing A and B), also move the equilibrium rate.

BCRA purchases. The BCRA's daily series of the change in international reserves from its purchases and sales of foreign currency, summed by month. Treasury purchases in the market (block trades) and the BCRA's positions in dollar futures are not published as series and are not included.

Balance cambiario. The BCRA's monthly record of foreign-exchange transactions between authorized entities and their clients, by concept, in millions of US dollars: a cash view of the balance of payments (dollars that change hands in the market in a given month, which can differ from when goods are shipped or services provided). We group it into goods; services; income (interest, profits, dividends and transfers); purchases of banknotes and foreign currency with no specific purpose ("dollars for holding", the formation of external assets of individuals and firms); the IMF and other official lenders; and the rest of the capital and financial accounts. The groups add up to the change in international reserves from these transactions; each run checks that identity and stops if the workbook's layout changes.

International Trade

Source. INDEC's monthly Intercambio Comercial Argentino (customs records of goods; exports FOB, imports CIF), its exports and imports by country, and its quarterly indices of prices and quantities of exports and imports (2004 = 100), all through datos.gob.ar. The annual series since 1910 is INDEC's historical series of exports and imports in current US dollars. Services are not included.

Seasonality. Monthly trade follows the harvest (soybeans, corn and wheat ship mostly in the second and third quarters) and the winter demand for imported energy, so the charts use twelve-month sums or calendar-year totals for trends. Growth contributions compare the last twelve months with the twelve before: the contribution of a component is the change in its twelve-month sum over the previous twelve-month total, so they add up to the change in the total.

Prices against volumes. INDEC's value index equals its price index times its quantity index (divided by 100). With p and q the year-over-year changes in prices and quantities, value grows (1 + p)(1 + q) − 1 = p + q + pq. The interaction term pq is split equally between the two, so prices contribute p(2 + q)/2 and quantities q(2 + p)/2; the split does not depend on which comes first. The terms of trade are INDEC's: the export price index over the import price index, times 100.

Trading partners. Rest of Mercosur is INDEC's Mercosur total minus Brazil; rest of the world is total trade minus Mercosur, China, the United States and the European Union. On datos.gob.ar the series labeled "total EU-28" is far too small to be the bloc's total, and the series labeled "rest of the EU-28" matches INDEC's published EU totals (2024: exports US$8,239 million, imports US$8,986 million), so the site uses the latter.

Goods shipped against dollars settled. Customs exports and imports against the BCRA's records of export receipts and import payments for goods in the foreign-exchange market (balance cambiario), each as twelve-month sums. The two measure different things: customs records when goods cross the border; the foreign-exchange market records when dollars change hands, including prefinancing and advance payments before shipment and deferred payments after it. Their gap is a rough measure of trade credit, which swings with exchange controls and with expectations of devaluation.

Balance of Payments

Source and conventions. INDEC's quarterly balance of payments and international investment position, compiled under the sixth edition of the IMF's Balance of Payments and International Investment Position Manual (BPM6), in millions of US dollars, through datos.gob.ar. The balance of payments starts in 2006 and the investment position in 2016 on this basis. Transactions are recorded on an accrual basis (when goods change owner or services are provided), which is why they differ from the BCRA's balance cambiario, a cash record of the foreign-exchange market.

Reading the financial account. Under BPM6 the financial account is net acquisition of foreign financial assets minus net incurrence of liabilities, so a positive balance is net lending to the rest of the world. An increase in reserves, residents' purchases of dollars or deposits abroad, or the repayment of foreign debt all add to it; foreign investment in Argentina and new foreign borrowing subtract from it. By construction, current account + capital account + net errors and omissions = financial account. On the financing chart the bars are the financial account's components plus errors and omissions with its sign reversed, so they add up to the line (the current and capital accounts).

Statements. The statement tables add up quarters: the selected quarter, the four quarters ending in it, the year to date or a calendar year, each against the same period a year earlier. Investment income from reserve assets, and any income INDEC does not assign to direct, portfolio or other investment, is shown as the residual "Reserve assets and other". The investment position table compares end-of-quarter stocks; changes include valuation effects (prices and exchange rates), not just flows.

Current account, % of GDP. The current account of the last four quarters over GDP in dollars of the same four quarters, each quarter's nominal GDP (INDEC) divided by four (INDEC publishes quarters at annual rates) and converted at that quarter's average official wholesale exchange rate (BCRA, Communication A 3500). Under exchange controls the official rate overvalues the peso, which raises dollar GDP and shrinks the ratio.

Accrual against cash. The goods-and-services balance of the balance of payments against the goods and services lines of the balance cambiario (monthly, summed by quarter), each over the last four quarters.

Labor & Poverty

Labor market. INDEC's Encuesta Permanente de Hogares (EPH), the continuous household survey of 31 urban areas, quarterly since 2003. Activity and employment rates are over total population; unemployment and underemployment over the labor force. The series are not seasonally adjusted. In 2016 INDEC warned that its 2007-2015 survey results should be used with caution (they rested on outdated population projections and suffered from weak survey quality); the charts shade that period rather than drop it. Registered private employment is the Ministry of Labor's count of wage earners in the social security system (SIPA), seasonally adjusted.

Wages. INDEC's wage index (October 2016 = 100) for registered private, public and unregistered workers, and the RIPTE, the average taxable wage of workers who have been with the same employer for 13 months. Real wages divide by the CPI, corrected for 2007-2016 as in the long-run CPI, and are expressed in pesos of the latest month with CPI data (or rebased to 100 at any month). Wages in dollars divide the RIPTE by the month's average official rate (one peso per dollar during convertibility) and, for the CCL version, by the month's average CCL during exchange controls.

Minimum wage and pension since 1965. datos.gob.ar publishes the minimum wage (salario mínimo, vital y móvil) and the minimum pension in the currency of each period. We convert them to current pesos at the official parities: peso ley 18.188 from January 1970 (100 pesos moneda nacional), peso argentino from June 1983 (10,000 pesos ley), austral from June 1985 (1,000 pesos argentinos; the monthly series switch in July) and peso from January 1992 (10,000 australes). In real terms they are deflated by the CPI since 1943, corrected for the intervention.

Poverty and indigence. INDEC's incidence of poverty and indigence, persons and households, in the urban areas of the EPH, by semester, kept by hand from INDEC's reports (Cuadro 1 of each release; each row cites its report). We show everything INDEC published: H1 2003 to H1 2013, and H2 2016 on. During the intervention (2007-2013) the poverty and indigence lines were the cost of baskets priced with the intervened CPI, so they rose far less than actual prices and the measured rates fell to implausible levels; INDEC stopped publishing after H1 2013 and did not estimate H2 2007. In 2016 it resumed with updated baskets and equivalence scales, so the series before and after are not strictly comparable. Two alternatives are shown for comparison. CEDLAS's comparable series (Tornarolli, 2018) applies INDEC's 2016 method to the same household survey from H2 2003, with the 2007-2015 baskets repriced using private price indices; it coincides with INDEC from H2 2016 and ends in H1 2017. The UCA's Observatorio de la Deuda Social measures poverty with its own annual survey (EDSA, fieldwork in the third quarter; shown at mid-year), spliced across three survey designs and, before 2010, reconstructed backwards from INDEC's survey; its level differs from INDEC's even in years when both are reliable, so it reads best as a trend. Other reconstructions (Zack, Schteingart and Favata, 2020) tell a similar story.

What incomes buy. The minimum wage, the minimum pension (excluding the fixed bonus) and the RIPTE over INDEC's poverty line for its reference household of four in Greater Buenos Aires.

Inequality and the labor share. The Gini coefficient of per capita household income from the EPH, and INDEC's generation-of-income account: employee compensation, mixed income and gross operating surplus as shares of GDP at factor cost.

Institutions

Confidence surveys. Universidad Torcuato Di Tella publishes two indices from one monthly national telephone survey of about 1,000 adults. The index of confidence in government (ICG, Escuela de Gobierno, since November 2001) averages five questions about the national government (overall assessment, whether it governs for the general interest, efficiency, honesty, capacity to solve problems), each on a 0-5 scale. The consumer confidence index (ICC, Centro de Investigación en Finanzas, since 2001) averages questions on households' personal situation, the country's macroeconomic situation and conditions for buying durables and real estate, on a 0-100 scale, and is also reported by region. Both are opinion measures: they move with elections and with changes of government. We take them from datos.gob.ar; UTDT's own releases come out a few days earlier.

Tenure of officials. The list of presidents of the BCRA (since its creation in 1935) and ministers of the economy (since 1946, under the ministry's successive names) is kept by hand in the repository, with the dates each took and left office. Days in office are counted from the start date to the end date, or to today for the official in office. Averages use completed tenures only. Bars are colored by the presidency in which each official took office (since December 1983; officials appointed in the interim presidencies of late December 2001 are grey). The pipeline checks twice a day that the official in office in the list is still the one named on the BCRA's board page and on the Ministry of Economy's site, and flags a warning when they differ; the list itself is only changed by hand. Short gaps between consecutive officials (for example, in 1962 and 1966) are periods for which the list has no minister.

Economic Freedom of the World. The Fraser Institute's index rates about 165 countries from 0 to 10 in five areas (size of government; legal system and property rights; sound money; freedom to trade internationally; regulation), built from some 45 components taken from the IMF, the World Bank, the World Economic Forum's surveys and other sources; the summary rating is their average. Ratings are every five years from 1970 to 2000 and annual since; each edition, published in the autumn with data to two years earlier, revises past years. We use the summary and area ratings and the rank as published in the edition's "EFW Ratings" sheet (area 2 with the legal-gender adjustment, as in the summary). The Fraser Institute also publishes a chain-linked panel for comparisons over time; for Argentina the two differ by about 0.01. The workbook's address changes with each edition and is updated by hand.

V-Dem. The Varieties of Democracy project (University of Gothenburg) codes several hundred indicators per country and year since 1789 from the ratings of country experts, combined with a Bayesian measurement model that adjusts for each coder's reliability and scale. We show six high-level indices, on a 0-1 scale: liberal democracy, electoral democracy, judicial and legislative constraints on the executive, rule of law, and political corruption (higher is more corrupt). The values are V-Dem's point estimates as republished by Our World in Data; the model's confidence intervals, which are wide in some years, are not shown. Legislative constraints are not defined in years without a legislature. The Latin American median (here and for EFW) is taken over Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, Cuba, the Dominican Republic, Ecuador, El Salvador, Guatemala, Haiti, Honduras, Mexico, Nicaragua, Panama, Paraguay, Peru, Uruguay and Venezuela, among those with data each year.

References

• INDEC. Índice de precios al consumidor. Cobertura nacional. Metodología (2019). PDF; English version (2017): PDF.
• INDEC. Metodología del estimador mensual de actividad económica (2016). PDF.
• INDEC. Cuentas nacionales. Metodología de estimación. Base 2004 (Metodología INDEC N° 21). PDF.
• INDEC. ¿Cómo usar un índice de precios? Preguntas frecuentes (2022). PDF.
• Decree 55/2016, administrative emergency of the national statistical system and INDEC. InfoLEG.
• IMF. Press Release 13/33, declaration of censure (February 2013). imf.org.
• IMF. Press Release 16/497, removal of the declaration of censure (November 2016). imf.org.
• IMF. Argentina: 2026 Article IV Consultation and Second Review under the Extended Arrangement, Country Report, May 2026; Attachment II, Technical Memorandum of Understanding. imf.org.
• BCRA. Informe sobre Bancos and statistical annex. bcra.gob.ar.
• BCRA. Información de Entidades Financieras (monthly data on each institution). bcra.gob.ar.
• BCRA. Weekly summary balances of assets and liabilities. bcra.gob.ar.
• Secretaría de Finanzas. Monthly and quarterly public debt data; debt placements (colocaciones de deuda). argentina.gob.ar; placements.
• Secretaría de Hacienda. Información Mensual de Ingresos y Gastos (national non-financial public sector, cash basis).
• BCRA. Informe de Evolución del Mercado de Cambios y Balance Cambiario, statistical annex; Relevamiento de Expectativas de Mercado (REM). bcra.gob.ar.
• ArgentinaDatos (argentinadatos.com) and DolarAPI (dolarapi.com): market exchange-rate quotes (unofficial).
• INDEC. Intercambio comercial argentino (monthly report) and Índices de precios y cantidades del comercio exterior. indec.gob.ar.
• IMF. Balance of Payments and International Investment Position Manual, sixth edition (BPM6), 2009. imf.org.
• BCRA. Índice Tipo de Cambio Real Multilateral (ITCRM), daily series and methodology. bcra.gob.ar.
• IMF. Argentina: 2026 Article IV Consultation and Second Review under the Extended Arrangement, staff report, May 2026: external sector assessment. imf.org.
• INDEC. Incidencia de la pobreza y la indigencia (semiannual reports) and Pobreza e indigencia, primer semestre de 2013. indec.gob.ar.
• Tornarolli, L. (2018). Series comparables de indigencia y pobreza: una propuesta metodológica. CEDLAS Documento de Trabajo 226. cedlas.econo.unlp.edu.ar.
• Observatorio de la Deuda Social Argentina (UCA). Pobreza: serie histórica 2004-2023; Deudas sociales en la Argentina del siglo XXI (2004-2024). uca.edu.ar.
• Zack, G., Schteingart, D. and Favata, F. (2020). Pobreza e indigencia en Argentina: construcción de una serie completa y metodológicamente homogénea. Sociedad y Economía 40. univalle.edu.co.
• A3 Mercados. Dollar futures (DLR), daily closing prices, market-data service (unofficial). a3mercados.com.ar.
• U.S. Department of the Treasury. Daily Treasury Bill Rates. treasury.gov.
• Universidad Torcuato Di Tella. Índice de Confianza en el Gobierno (Escuela de Gobierno) and Índice de Confianza del Consumidor (Centro de Investigación en Finanzas). utdt.edu.
• Fraser Institute. Economic Freedom of the World: 2026 Annual Report and dataset. efotw.org.
• Coppedge, M. et al. (2026). V-Dem Country-Year Dataset v16. Varieties of Democracy (V-Dem) Project. v-dem.net; republished by Our World in Data: ourworldindata.org.
• Pemstein, D. et al. The V-Dem measurement model: latent variable analysis for cross-national and cross-temporal expert-coded data. V-Dem Working Paper 21. v-dem.net.
• BCRA. Directorio (board of directors). bcra.gob.ar.
• BCRA. Relevamiento de Expectativas de Mercado (REM), historical results. workbook.
• ArgentinaDatos. Country risk (riesgo país), daily, from Ámbito Financiero. argentinadatos.com.
• Datos Argentina. Time-series API. datos.gob.ar.

Corrections and questions are welcome on GitHub.