Skip to content

Method

what the readings are

The framework of The Architecture of Money is diagnostic, not predictive. I publish readings of it on those terms: each is dated before outcomes are known and published with its data, and the reading of the four network dimensions is also scored. Appendix C defines each dimension by what high and low levels indicate (Table C.1). The three-step scale, its cut points and the direction rule are this site's, chosen to follow those poles; they are not thresholds from the book. The reading of cultural legibility is scored by judgment, against Appendix A4's own markers. The readings are my own, and so are the scores (disclosure). A reading is not independent validation, not evidence that the framework is correct, and not a track record.

Three kinds of reading are published, each on its own dated page: the four network dimensions (Appendix C), scored high, medium or low against the leading currency; the Conclusion's four signs, shown as figures against each system's own history and regulators' conventions, without scores; and cultural legibility (Appendix A4), scored high, medium or low by judgment. All of them are listed on the readings page. From the book's publication, a dated reading is never revised, only superseded, with the change logged.

the Conclusion's lens

From the Conclusion:

"It offers a lens for reading vulnerability. Four signs often come before a rupture: private credit outrunning output for several years running, the most reliable precursor in the historical record; persistent divergence between external references—energy, time, ecological capacity—and core prices; rising rollover dependence and duration mismatch—more debt falling due sooner than the income to meet it—under placid aggregates; and widening arbitrage between constraint layers that survives beyond the business cycle. They say where a system is exposed, not when it will break or whether it must: erosion sets the vulnerability, not the date—a sound franc was attacked in 1992–1993—and Japan's rollover dependence has stayed lit for more than fifteen years without a break (Appendices C4 and D). What the lens commits to is the reading: made before the outcome is known, dated, and published, so that the record can show where it saw clearly and where it did not. It is meant to be checked, not admired."

Appendix C4 maps each sign to its series:

"The Mapping: Of the Conclusion's four signs, credit outrunning output and rollover dependence are read from the series above; external-reference divergence and arbitrage between constraint layers belong to the anchors of Table 7.1 and the official–market spreads of Chapter 12, and the series behind them are published at thearchitectureofmoney.com. None of these reads cultural legibility, which remains an ordinal judgment (Appendix A4), or cheating cost directly (Appendix E, Objection 10)."

the four network dimensions: method

The reading scores the four network dimensions the book uses to read network dominance (Chapter 3, Appendix C), dated so that each later reading can be compared with this one. The reading dated November 8, 2026 is on its own page.

Scoring against the leader.

Appendix C defines each dimension by what high and low levels indicate (Table C.1). The three-step scale, its cut points and the direction rule are this site's, chosen to follow those poles; they are not thresholds from the book. Each dimension is scored from group D series. Scores are relative to the leading currency on each dimension; a currency with structural strength can score Medium when one leader dominates.

  • High: at least half the leader's level.
  • Medium: a tenth to a half.
  • Low: under a tenth.

Conversion friction is scored as friction, so low is favorable:

  • Low: within twice the leader's cost.
  • Medium: two to five times.
  • High: above five times, or where capital controls bind.

Confidence.

Confidence follows Appendix C2:

  • liquidity depth: medium-high;
  • global reserve share: high for direction, medium for magnitude;
  • trade settlement share: medium;
  • conversion friction: high for relative comparison.

Figures, thresholds and gaps.

Each score is shown with the figures behind it. A score within 10 percent of a threshold is marked "within 10 percent of a threshold." A cell with no comparable series is not scored. Where proxies disagree, the score says so. These thresholds stay the same from one reading to the next; any change is dated and logged in the corrections log.

Direction.

Direction compares each figure with the same series about five years earlier: rising or falling if it moved by more than a tenth of its earlier value, otherwise steady. Direction is shown, not scored. A direction is also marked "within 10 percent of the threshold" when the change is within one percentage point of the ten percent rule, that is, between 9 and 11 percent. On conversion friction the arrow shows the direction of the cost, so a falling arrow means a cheaper conversion.

Bitcoin.

The four network dimensions were built for currencies; bitcoin is read on them as Appendix C3 reads it, and only where a comparable series exists.

the Conclusion's four signs: method

The reading shows each sign as figures, without scores, for each system; the reading dated November 10, 2026 is on its own page. Its systems are in alphabetical order.

Percentiles.

Each figure is shown with its percentile in the system's own record since 2000: the share of the system's quarterly observations of the same series, from the start of 2000 to the latest one, that are at or below the latest figure. The number of observations (n) is in the data file. A percentile is not computed with fewer than 20 observations, or for a yes/no series. A percentile measures the figure against the system's own past, not its exposure. A condition can be chronic, and so look unremarkable against itself.

Reference lines.

The reference lines are regulators' conventions, shown only where they apply. Appendix C4: "The levels are regulators' conventions, not this framework's thresholds; use them to judge whether a signal is loud, not whether a system will break." On the reading, a figure above one carries a thin rule beneath it.

Not read.

"Not read" means that no free series covers the cell with an observation from 2025 on. The table gives the reason for each such cell. In this reading and in the reading of the four network dimensions, no cell is filled by judgment, and the reading of the four signs is not scored. The reading of cultural legibility is different: every cell in it is scored by judgment (its method).

Dates.

Each column gives its latest observation date; a cell from an earlier date shows its own. The source of every cell is named under the table and in the data file.

sign 1: credit outrunning output

Series. The BIS credit-to-GDP gap, as Appendix C4 names it: "The BIS credit-to-GDP gap—private non-financial credit as a share of GDP, less its long-run trend." It covers credit from all lenders and is given in percentage points, at the latest quarter.

Reference lines. Appendix C4: "Basel III's countercyclical-buffer guide starts adding bank capital at a gap of 2 percentage points and reaches its maximum at 10."

Not read. Where the BIS publishes no credit-to-GDP gap for a system.

sign 2: external-reference divergence

Series and transformation. Energy (Brent crude) and gold, anchors of Table 7.1, each priced in the system's own currency and set against its core consumer prices. The figure is the change over twelve quarters in the log of the anchor's local-currency price, less the change over the same twelve quarters in the log of core prices, times 100, at the latest quarter with both. Monthly prices are averaged to quarters, and no peer is subtracted. The other anchors in the series register are not read here.

Log change × 100. The change in the natural logarithm of a series, times 100. For small changes it is close to the percent change; unlike a percent change, it is the same size for a rise and for the fall that undoes it, and changes over successive periods add up.

Consumer prices. Where no current core series exists, headline consumer prices are used, and the cell carries a footnote letter that says so. A cell from an earlier quarter is at the latest quarter for which all three of its series are published. In the reading dated November 10, 2026, consumer prices come from the OECD, the IMF, the World Bank Global Inflation Database or the BIS long series of consumer prices, as each cell's row in the data file names.

Reference lines. No regulator names a convention for this series, so none is shown.

sign 3: rollover dependence

Two series, one for the sovereign and one for the economy as a whole.

The sovereign. General government gross financing need, in percent of GDP, from the IMF's Fiscal Monitor (in the reading dated November 10, 2026, the Fiscal Monitor, April 2026, Tables A23 and A24). The figures for the latest year are IMF staff projections; each edition gives one year, so no percentile is computed.

Fallback. Where the tables give no figure, a fallback follows Appendix C4's own series, "debt maturing within twelve months plus the fiscal deficit, as a share of GDP": short-term government debt plus long-term government debt due within one year (World Bank, Quarterly Public Sector Debt), plus the general government deficit (IMF, World Economic Outlook, October 2026). It carries a footnote letter and is not called a gross financing need. It is a quarterly series, so it has a percentile where it has 20 observations or more. An exact zero in either debt component is a non-reporting placeholder in that database and is treated as missing. Where neither the tables nor the fallback give a figure, the cell is not read.

The economy. Short-term external debt of all sectors over official reserve assets excluding gold, in percent, at the latest quarter with both: debt from the World Bank's Quarterly External Debt Statistics (SDDS), reserves from the IMF's International Reserves and Foreign Currency Liquidity. Where a system does not report its debt by remaining maturity, the figure is at original maturity: debt first contracted for a year or less, without long-term debt that falls due within the year. No free series gives the foreign-currency share of all external debt for these systems, so it is not read.

Reference lines and their scope. Appendix C4: "The IMF flags sovereign financing needs above 15 percent of GDP in emerging markets and 20 percent in advanced economies; short-term external debt exceeding reserves is the Greenspan–Guidotti line." The IMF gives both financing-need lines in International Monetary Fund, Staff Guidance Note for Public Debt Sustainability Analysis in Market-Access Countries (May 9, 2013), Table 7, Debt Burden Benchmarks, p. 32. Advanced or emerging follows the IMF World Economic Outlook groups of October 2026. The 100 percent line applies to pegged exchange rates and foreign-currency borrowers only, so it is not shown for reserve-currency issuers that are not pegged or for free floaters, whose figures and percentiles are shown without it.

The classification. Pegged, floating and reserve-currency issuer follow the Ilzetzki–Reinhart–Rogoff fine classification for September 2025: free floaters are its code 13; the euro area, a reserve-currency issuer, has no code of its own.

sign 4: arbitrage between constraint layers

Chapter 12's free spreads, in three parts.

Interbank minus policy rate. An overnight interbank or money-market rate less the policy rate, in percentage points, averaged over the latest complete quarter. For the euro area, Japan, the United Kingdom and the United States, the central bank's own daily series are used, averaged to complete quarters, and the percentile uses the same series since 2000 or its first observation:

  • United States: Effective federal funds rate (FRED EFFR) minus the target-range midpoint (FRED DFEDTARU, DFEDTARL; before December 16, 2008, DFEDTAR); FRED (Federal Reserve Bank of St. Louis).
  • Euro area: €STR minus the deposit facility rate; ECB Data Portal (EST.B.EU000A2X2A25.WT; FM.D.U2.EUR.4F.KR.DFR.LEV).
  • Japan: Uncollateralized overnight call rate (BoJ FM01 STRDCLUCON) minus the BoJ policy target as compiled by the BIS (WS_CBPOL, D.JP); Bank of Japan; BIS.
  • United Kingdom: SONIA minus Bank Rate; Bank of England Database (IUDSOIA; IUDBEDR).

For the others, the IMF's money-market rate (Monetary and Financial Statistics, MFS_IR) less the BIS policy rate, quarterly averages.

Dual or multiple rates. The IMF's Annual Report on Exchange Arrangements and Exchange Restrictions (AREAER 2023 edition, Summary Features) classifies each exchange-rate structure as unitary, dual or multiple. Multiple currency practices are recorded separately in the same report and are not part of this cell. It is a yes/no series, so it has no percentile.

Parallel-market premium. Read only where a free official series exists.

Reference lines. No regulator names a convention for these series, so none is shown.

cultural legibility: method

The reading scores seven systems, the US dollar, the euro, the Chinese yuan, bitcoin, Tether, USD Coin and the e-CNY, on the four criteria of Appendix A4: symbolic familiarity, narrative fit, usability fluency and trust memory. It follows A4's Assessment Approach. Step 1: "Score each criterion on a qualitative high–medium–low scale." Step 2: "Identify where gaps are most acute." Step 3: "Cross-check whether legibility gaps align with constraint erosion." The reading dated November 12, 2026 is on its own page.

Unlike the other two readings, it is scored by judgment. Appendix A4: "None of these is a series; each is a judgment, and a system can read well on all of them and still break, as Lebanon's peg did (Chapter 4)." Each score is shown with its reason, the evidence it rests on and the date of that evidence. The scores are mine (disclosure).

The rules are this site's, not the book's, and the reading page gives them in full. In short:

  • Each cell is judged against A4's own High and Low markers, not against the other systems. There is no ranking, no count and no total.
  • Trust memory is judged against the promise each system asks users to rely on: stable purchasing power and legal status for state currencies, redemption at par for stablecoins, par with physical renminbi for the e-CNY, and, for bitcoin, the store-of-value claim its proponents make.
  • An intermediary counts against usability fluency only when it is a step users would not otherwise take, such as an exchange.
  • A digital form of a familiar unit reads Medium on symbolic familiarity: its unit has a cultural precedent, and it lacks A4's High markers, which are features of notes and coins.
  • A sign that was not read is named, not filled in.

The cross-check of step 3 sets each system's legibility gaps beside what the signs reading shows for its constraints. It is made only for the systems the signs reading covers, and it does not say that one explains or will produce the other.

the series register

These are the series behind my readings: every series that Table 7.1, Chapter 12, the Conclusion and Appendices C2 and C4 name or imply, with its publisher and access. Where the text names an example, it is labeled "book's example"; where it does not, the series is my choice, labeled "site's choice." The full register, with the passage each series answers to and its latest observation, is here as data: series-register.csv.

The Conclusion's four signs

The Conclusion names four signs that "often come before a rupture." External-reference divergence is read from the series in group A, arbitrage between constraint layers from group B, and rollover dependence from the last three series of group C. Credit outrunning output is read from the BIS credit-to-GDP gap, which group C also lists.

Their dated reading is the Conclusion's four signs, November 10, 2026.

  • Source: Bank for International Settlements (BIS), Credit-to-GDP gap (WS_CREDIT_GAP).

    book's example · free · quarterly

    Book: Conclusion: "private credit outrunning output for several years running, the most reliable precursor in the historical record"; Appendix C4: "Credit Outrunning Output (Chapter 12): Series: The BIS credit-to-GDP gap—private non-financial credit as a share of GDP, less its long-run trend."

    Same series as group C, Credit Outrunning Output. The other three signs are read from groups A, B and C.

A. Table 7.1's anchors, each read against core prices

The second of the Conclusion's four signs is "persistent divergence between external references—energy, time, ecological capacity—and core prices." Appendix C4's Mapping places it in "the anchors of Table 7.1." The last three series are the core prices the anchors are read against.

  • Source: Ministry of Commerce and Industry, Government of India (TradeStat), India merchandise imports of gold (monthly trade statistics).

    book's example · free · monthly

    Book: Table 7.1, Chapter 7: Metric "Gold price stability"; Examples "Pre-1971 USD gold peg, India gold imports"

    Listed first because Table 7.1 names it. It measures demand for the anchor, not the stability of its price. The table's other example, the pre-1971 USD gold peg, is a regime that ended in 1971, not a series.

  • Source: World Bank, Gold, monthly average price, U.S. dollars per troy ounce (Commodity Price Data, "Pink Sheet").

    site's choice · free · monthly

    Book: Table 7.1, Chapter 7: Metric "Gold price stability"

    Table 7.1 gives no gold price series; I use this one because it gives the price whose stability the metric reads, free and monthly. The World Bank's definition changed in June 2025: "Gold, spot average of daily rates, from June 2025; previously (UK), 99.5% fine, London afternoon fixing, average of daily rates."

  • Source: U.S. Energy Information Administration (EIA), Europe Brent spot price FOB, U.S. dollars per barrel (series RBRTE).

    book's example · free · daily

    Book: Table 7.1, Chapter 7: Metric "Oil or energy index"; Examples "1970s oil shocks, Brent crude benchmark"

    Table 7.1 names the Brent benchmark, not a publisher; EIA publishes the Brent spot series free. "1970s oil shocks" is a historical episode, not a series.

  • Source: UBS, UBS Prices and Earnings (working time needed to buy staples, including bread, by city).

    book's example · discontinued · about every three years (1971–2018)

    Book: Table 7.1, Chapter 7: Metric "Labor-time cost of staples"; Examples "U.K. 'loaf of bread' metric, UBS Prices and Earnings"

    Last published 2018.

  • Source: Office for National Statistics (ONS), RPI average price: bread, white loaf, sliced, 800g (series CZOH).

    book's example · discontinued · monthly (until January 2025)

    Book: Table 7.1, Chapter 7: Metric "Labor-time cost of staples"; Example "U.K. 'loaf of bread' metric"

    Table 7.1 names a metric, not a series; this is the official U.K. bread price nearest to it. It gives a price, not working time: an earnings series would be needed as well. ONS changed how it aggregates food prices from March 2025 and archived the earlier average prices (ONS ad hoc release 2724).

  • Source: U.S. Bureau of Labor Statistics (BLS), via FRED, Average price, bread, white, pan, per pound (APU0000702111) divided by average hourly earnings, total private (CES0500000003): working time per pound of bread.

    site's choice · free · monthly

    Book: Table 7.1, Chapter 7: Metric "Labor-time cost of staples"

    Substitute for UBS Prices and Earnings: the "loaf of bread" metric in its literal form, the working time needed to buy bread, from official, current, monthly data. Covers the U.S. dollar only. BLS advises using average prices for levels, not for price change.

  • Source: International Monetary Fund (IMF), Reserves as a percent of the IMF's Assessing Reserve Adequacy (ARA) metric.

    book's example · free · annual

    Book: Table 7.1, Chapter 7: Metric "FX reserve adequacy ratio"; Examples "IMF reserve adequacy metrics, HK currency board"

    Same series as group C, Reserve Adequacy. Covers emerging markets only (78 in the 2023 workbook), so not the dollar or the euro.

  • Source: Hong Kong Monetary Authority (HKMA), Currency Board Account: backing assets against the monetary base (backing ratio).

    book's example · free · monthly

    Book: Table 7.1, Chapter 7: Metric "FX reserve adequacy ratio"; Example "HK currency board"

    Same series as group C, Cover Ratio Drift. Applies to the Hong Kong dollar's currency board only.

  • Source: United Nations Framework Convention on Climate Change (UNFCCC), NDC Registry (submitted national targets).

    book's example · free · irregular (submissions in five-year rounds)

    Book: Table 7.1, Chapter 7: Metric "CO₂ budget adherence"; Example "Paris Agreement nationally determined contributions (NDCs)"

    NDCs give the budget; reading adherence needs observed emissions (next row).

  • Source: European Commission, Joint Research Centre (JRC), Greenhouse gas and CO₂ emissions by country (EDGAR).

    site's choice · free · annual

    Book: Table 7.1, Chapter 7: Metric "CO₂ budget adherence"

    Table 7.1 implies this series without naming it: adherence to an NDC is read from emissions against the target, and EDGAR gives annual emissions for every country on one method.

  • Source: Swift, Global Currency Tracker: currency shares of international payments by value (renamed from RMB Tracker in February 2026).

    book's example · free · monthly

    Book: Table 7.1, Chapter 7: Metric "Cross-border settlement share"; Example "USD in SWIFT payments"

    Same series as group D, TSS (SWIFT payment-message statistics). Document centre: https://www.swift.com/products/renminbi-tracker/document-centre.

  • Source: International Monetary Fund (IMF), Renminbi share of trade invoicing (IMF invoicing-currency dataset, WP/25/178: 132 countries, 1990–2023).

    book's example · free · annual (released in irregular vintages)

    Book: Table 7.1, Chapter 7: Metric "Cross-border settlement share"; Example "RMB share in global trade"

    Table 7.1 names the measure, not a publisher. Appendix C2 points to "IMF and ECB invoicing-currency data" for the same dimension, and this IMF dataset added renminbi coverage in its 2025 update. Same source as group D, TSS (IMF).

  • Source: U.S. Bureau of Labor Statistics (BLS), via FRED, CPI for all urban consumers, all items less food and energy, seasonally adjusted (CPILFESL).

    site's choice · free · monthly

    Book: Conclusion: "persistent divergence between external references—energy, time, ecological capacity—and core prices"; each Table 7.1 anchor is read against core prices

    No core-price series is named; I use the standard U.S. core measure.

  • Source: Eurostat, HICP, overall index excluding energy, food, alcohol and tobacco (dataset prc_hicp_minr, coicop18=TOT_X_NRG_FOOD), euro area.

    site's choice · free · monthly

    Book: Conclusion: "persistent divergence between external references—energy, time, ecological capacity—and core prices"; each Table 7.1 anchor is read against core prices

    The standard euro-area core measure. From the January 2026 data, Eurostat uses ECOICOP version 2; the old dataset (prc_hicp_manr) ends at 2025-12.

  • Source: National Bureau of Statistics of China (NBS), CPI excluding food and energy, year-on-year (row in the monthly CPI release).

    site's choice · free · monthly

    Book: Conclusion: "persistent divergence between external references—energy, time, ecological capacity—and core prices"; each Table 7.1 anchor is read against core prices

    The standard Chinese core measure, published as a row of the monthly CPI release rather than as a downloadable series.

B. Chapter 12's official–market spreads

The fourth of the Conclusion's four signs, arbitrage between constraint layers, is read from "the official–market spreads of Chapter 12": the items among its "Hallmarks of the Stress Stage" and "Common Stress Signals" that set a market price or rate against an official one.

  • Source: World Bank; Board of Governors of the Federal Reserve System, Commodity Price Data ("Pink Sheet") in U.S. dollars, converted to other currencies at Federal Reserve H.10 rates.

    site's choice · free · monthly (Pink Sheet); daily (H.10)

    Book: Chapter 12, Hallmarks of the Stress Stage: "Commodity prices, when quoted in the system's currency, start rising faster than prices in peer currencies."

    Chapter 12 names no index; I use the Pink Sheet already in group A (gold, Brent) and the Fed's free daily rates, which cover the euro and the yuan. The Pink Sheet URL is in group A.

  • Source: S&P Global Ratings; Moody's; Fitch Ratings, Sovereign long-term foreign-currency credit ratings and rating actions.

    book's example · free with registration · irregular (rating actions)

    Book: Chapter 12, Hallmarks of the Stress Stage: "…despite stable sovereign ratings."

    Chapter 12 names sovereign ratings, not an agency.

  • Source: Federal Reserve Bank of New York; Board of Governors of the Federal Reserve System (via FRED), Secured Overnight Financing Rate (SOFR) against the FOMC federal funds target range (DFEDTARU, DFEDTARL).

    site's choice · free · daily

    Book: Chapter 12, Hallmarks of the Stress Stage: "Interbank or inter-sovereign lending rates diverge from policy targets, reflecting counterparty suspicion."

    Dollar. SOFR is a market repo rate; the target range is the policy target.

  • Source: European Central Bank (ECB), Euro short-term rate (€STR) against the ECB deposit facility rate.

    site's choice · free · daily

    Book: Chapter 12, Hallmarks of the Stress Stage: "Interbank or inter-sovereign lending rates diverge from policy targets, reflecting counterparty suspicion."

    Euro. €STR is the market rate; the deposit facility rate is the policy floor.

  • Source: China Foreign Exchange Trade System (ChinaMoney); People's Bank of China (PBOC), Overnight Shibor against the People's Bank of China open-market reverse repo rate.

    site's choice · free · daily

    Book: Chapter 12, Hallmarks of the Stress Stage: "Interbank or inter-sovereign lending rates diverge from policy targets, reflecting counterparty suspicion."

    Yuan.

  • Source: Office of Financial Research (OFR), U.S. Department of the Treasury, Tri-party repo transaction volume and rates by collateral type (Short-term Funding Monitor).

    site's choice · free · daily

    Book: Chapter 12, Hallmarks of the Stress Stage: "Clearing systems or repo markets—where institutions borrow short-term cash against pledged securities—begin refusing classes of collateral once considered safe."

    Dollar repo market only. Shows when a class of collateral loses volume or must pay more, which is the refusal Chapter 12 describes; it does not show haircuts.

  • Source: Federal Reserve Bank of New York, SOFR percentile distribution (1st, 25th, 75th, 99th) and volume.

    site's choice · free · daily

    Book: Chapter 12, Historical Diagnostic: Common Stress Signals: "Short-Term Funding Volatility: Overnight and repo rates spike suddenly despite central bank injections."

    Dollar. The 99th percentile against the target range shows spikes within a day's trading.

C. Appendix C4's five series

Appendix C4 names one series for each physical and institutional signal, with the level practitioners conventionally use for it: "The levels are regulators' conventions, not this framework's thresholds; use them to judge whether a signal is loud, not whether a system will break." Of the five, Credit Outrunning Output and Rollover Dependence are among the Conclusion's four signs.

  • Source: Bank for International Settlements (BIS), Credit-to-GDP gap (WS_CREDIT_GAP).

    book's example · free · quarterly

    Book: Appendix C4: "Credit Outrunning Output (Chapter 12): Series: The BIS credit-to-GDP gap—private non-financial credit as a share of GDP, less its long-run trend."

    Appendix C4's convention: the Basel III buffer guide starts at a gap of 2 percentage points and reaches its maximum at 10.

  • Source: Bank for International Settlements (BIS), Real effective exchange rate, broad basket (WS_EER).

    book's example · free · monthly

    Book: Appendix C4: "Real Appreciation (Chapter 12): Series: The real effective exchange rate (BIS or IMF) against its own ten-year average."

    The ten-year average is computed from the series. Covers the dollar, the euro and the yuan. Appendix C4's convention: "Sustained double-digit deviations above the average are where overvaluation disputes begin; Argentina after 1999 sat well beyond them, and Britain's 1925 parity, wrong by a tenth, sat at their edge."

  • Source: International Monetary Fund (IMF), Real effective exchange rate (IMF effective exchange rates dataset).

    book's example · free · monthly

    Book: Appendix C4: "Real Appreciation (Chapter 12): Series: The real effective exchange rate (BIS or IMF) against its own ten-year average."

    The alternative Appendix C4 names alongside the BIS.

  • Source: Hong Kong Monetary Authority (HKMA), Currency Board Account: backing assets against the monetary base (backing ratio).

    book's example · free · monthly

    Book: Appendix C4: "Cover Ratio Drift (Chapter 12): Series: The anchor at market value divided by the claims it backs—gold over notes and sight liabilities under a gold standard; usable reserves over pegged liabilities under a board or peg."

    Appendix C4 names no publisher; I use the HK currency board, Table 7.1's own example. Applies to boards and pegs only: the dollar, the euro, the yuan and bitcoin have no anchor in this sense. Chapter 12's other examples ("dollar liabilities against U.S. gold, broad money against reserves") are not read as series. Appendix C4's convention: "Direction matters more than level; a ratio falling for three consecutive years while issuance grows is the signal."

  • Source: International Monetary Fund (IMF), Reserves as a percent of the IMF's Assessing Reserve Adequacy (ARA) metric.

    book's example · free · annual

    Book: Appendix C4: "Reserve Adequacy (Chapter 7): Series: Reserves as a share of the IMF's Assessing Reserve Adequacy metric."

    Same series as group A, FX reserve adequacy ratio. Appendix C4's convention: 100 to 150 percent is adequate; below 100, pegs become contestable.

  • Source: International Monetary Fund (IMF), General government gross financing need, percent of GDP (Fiscal Monitor statistical tables).

    book's example · free · semiannual (April, October)

    Book: Appendix C4: "Rollover Dependence (Conclusion): Series: For the sovereign, debt maturing within twelve months plus the fiscal deficit, as a share of GDP, against its own ten-year mean…"

    Appendix C4's convention cites the IMF ("sovereign financing needs above 15 percent of GDP in emerging markets and 20 percent in advanced economies"). Covers selected economies only. The ten-year mean is computed from the series.

  • Source: World Bank (with the IMF), Gross external debt position, all sectors, short-term, all instruments (DT.DOD.DSTC.CD.US), Quarterly External Debt Statistics (SDDS).

    site's choice · free · quarterly

    Book: Appendix C4: "Rollover Dependence (Conclusion): Series: …for the economy as a whole, short-term external debt against reserves."

    Appendix C4 names no publisher; I use QEDS, the standard cross-country source for short-term external debt. The World Bank last updated it on November 3, 2026.

  • Source: International Monetary Fund (IMF), Official reserve assets (International Reserves and Foreign Currency Liquidity template).

    site's choice · free · monthly

    Book: Appendix C4: "Rollover Dependence (Conclusion): Series: …short-term external debt against reserves." Convention: "short-term external debt exceeding reserves is the Greenspan–Guidotti line."

    Appendix C4 names no publisher; I use the IMF's standard reserves template, reported by member countries.

D. Appendix C2's proxies for the four network dimensions

Appendix C2 gives a primary and a secondary proxy for each of the four network dimensions, with a confidence level for each. The reading of the four network dimensions uses these series.

  • Source: Bank for International Settlements (BIS), Triennial Central Bank Survey: foreign exchange turnover by currency.

    site's choice · free · every three years

    Book: Appendix C2, For Liquidity Depth (LD): "Primary proxy: Daily volume/price volatility ratios from public exchanges"

    Not a daily ratio, so not a substitute for the primary proxy. I use it because it is the standard measure of turnover by currency, and the Bibliography lists it (Data Sources and Reports).

  • Source: Coin Metrics (Community data), Bitcoin spot trading volume, all reported exchanges, U.S. dollars per day (volume_reported_spot_usd_1d), and price (PriceUSD).

    site's choice · free · daily

    Book: Appendix C2, For Liquidity Depth (LD): "Primary proxy: Daily volume/price volatility ratios from public exchanges"

    The free tier gives reported volume, which is unfiltered; the filtered ("trusted") volume is paid. Spot only, while the BIS Triennial counts spot, forwards and swaps.

  • Source: International Monetary Fund (IMF), Currency Composition of Official Foreign Exchange Reserves (COFER).

    book's example · free · quarterly

    Book: Appendix C2, For Global Reserve Share (GRS): "Primary source: IMF COFER quarterly data (free, lag of three to six months)"

    As Appendix C2 notes: "Data often revised retrospectively." Bitcoin is not a reported reserve currency.

  • Source: Individual central banks, Reserve currency composition disclosed in central banks' annual reports.

    book's example · free · annual

    Book: Appendix C2, For Global Reserve Share (GRS): "Secondary: Central bank annual reports"

    No consolidated series; as Appendix C2 notes, detailed compositions are "often classified." No single URL.

  • Source: International Monetary Fund (IMF), Classification of crypto assets without a corresponding liability (Release of New Standards for Macroeconomic Statistics, BPM7).

    site's choice · free · not applicable (a statistical standard)

    Book: Appendix C2, For Global Reserve Share (GRS): "Primary source: IMF COFER quarterly data (free, lag of three to six months)"

    Not a series. The IMF states such assets "do not meet the definition of financial assets and therefore cannot be considered part of a country's reserve assets"; this is why COFER cannot show bitcoin. The IMF says the classification "will be reviewed."

  • Source: International Monetary Fund (IMF), Invoicing-currency shares of exports and imports (IMF dataset, WP/25/178: 132 countries, 1990–2023).

    book's example · free · annual (released in irregular vintages)

    Book: Appendix C2, For Trade Settlement Share (TSS): "Primary proxy: IMF and ECB invoicing-currency data; SWIFT payment-message statistics"

    Same source as group A, RMB share in global trade.

  • Source: European Central Bank (ECB), The International Role of the Euro (annual report; euro share of invoicing and other indicators).

    book's example · free · annual (June)

    Book: Appendix C2, For Trade Settlement Share (TSS): "Primary proxy: IMF and ECB invoicing-currency data; SWIFT payment-message statistics"

  • Source: Swift, Global Currency Tracker: currency shares of international payments by value (renamed from RMB Tracker in February 2026).

    book's example · free · monthly

    Book: Appendix C2, For Trade Settlement Share (TSS): "Primary proxy: IMF and ECB invoicing-currency data; SWIFT payment-message statistics"

    Same series as group A, USD in SWIFT payments. Two charts: "Global payments currency" (all payments) and "International payments currency - Excluding payments within Eurozone." Bitcoin does not appear. Document centre: https://www.swift.com/products/renminbi-tracker/document-centre.

  • Source: World Bank, Remittance Prices Worldwide (cost of sending remittances, by corridor).

    book's example · free · quarterly (historically)

    Book: Appendix C2, For Conversion Friction (CF): "Primary proxy: World Bank remittance cost data"

    The derived annual indicators in World Development Indicators (SI.RMT.COST.IB.ZS, SI.RMT.COST.OB.ZS) run to 2023 (World Bank API). Measures corridors into and out of countries, not bitcoin.

  • Source: World Bank (World Development Indicators), Average transaction cost of sending USD 200 from a country (SI.RMT.COST.OB.ZS) and to a country (SI.RMT.COST.IB.ZS), percent; derived from Remittance Prices Worldwide.

    book's example · free · annual

    Book: Appendix C2, For Conversion Friction (CF): "Primary proxy: World Bank remittance cost data"

    The World Bank remittance cost data named in Appendix C2, in the form the World Bank publishes through its open API. Outbound costs exist for the United States and euro-area countries; China has only an inbound series. There is no euro-area aggregate.

  • Source: State Administration of Foreign Exchange (SAFE), China, China: annual quota on individual foreign exchange purchases and sales, USD 50,000 per person.

    site's choice · free · not applicable (a rule)

    Site's scoring note: "High: above five times, or where capital controls bind." (conversion friction)

    Not a series and not in Appendix C2. It is my source for "capital controls bind" in the scoring note. SAFE's FAQ: "it is specified in the Detailed Rules that the respective annual quota shall be USD 50,000."

Not read: no free series

These series have no free source. They are not read, and no proxy stands in for them.

  • Source: commercial: LSEG (Refinitiv), Bloomberg, Forward exchange rates against the official or spot rate.

    Book: Chapter 12, Hallmarks of the Stress Stage: "Exchange rates in forward markets—where contracts for future currency delivery are priced—or black markets deviate persistently from official pegs, revealing what participants actually believe the currency is worth."

    There is no free series of forward rates. The official side is free (central-bank official rates; Federal Reserve H.10 for spot rates). No proxy is used.

  • Source: none consolidated, Parallel-market (black-market) exchange rates against the official rate.

    Book: Chapter 12, Hallmarks of the Stress Stage: "…or black markets deviate persistently from official pegs…"

    Where parallel rates are published, it is country by country, mostly by private outlets rather than official statistics. No proxy is used.

  • Source: commercial: S&P Global Market Intelligence (Markit), Bloomberg, LSEG, ICE Data Services, Sovereign credit default swap spreads (typically five-year).

    Book: Chapter 12, Hallmarks of the Stress Stage: "Credit default swap spreads (the price of insuring against default) widen despite stable sovereign ratings."

    No free series. No proxy is used.

  • Source: none public, Market repo haircuts by collateral class.

    Book: Chapter 12, Hallmarks of the Stress Stage: "Haircuts rise, shortening the leverage cycle."

    There is no free market-wide haircut series; OFR publishes volumes and rates by collateral, not haircuts. Central-bank haircut schedules (ECB; Federal Reserve discount window) are free, but they are official terms, not market haircuts. No proxy is used.

  • Source: commercial: LSEG (Refinitiv), Bloomberg, EBS (CME Group), Kaiko (crypto-assets), Bid-ask spreads (currency and crypto-asset markets).

    Book: Appendix C2, For Liquidity Depth (LD): "Secondary: Bid-ask spreads during high-volume periods"

    There is no free historical series. Exchanges show live order books, but not their history. No proxy is used.

  • Source: commodity exchanges (various), Announcements by commodity exchanges of contracts settled in a given currency.

    Book: Appendix C2, For Trade Settlement Share (TSS): "Secondary: Commodity exchange settlement announcements"

    Appendix C2 names announcements, which are free but are not a time series. No proxy is used.

  • Source: none, Bitcoin's share of trade settlement.

    Book: Appendix C2, For Trade Settlement Share (TSS): "Primary proxy: IMF and ECB invoicing-currency data; SWIFT payment-message statistics"

    No free series measures bitcoin's trade settlement share. Swift's statistics and the IMF and ECB invoicing data do not cover it, and no official source measures it. No proxy is used.

  • Source: commercial: LSEG (Refinitiv), Bloomberg, Spot against forward exchange rates.

    Book: Appendix C2, For Conversion Friction (CF): "Secondary: Spot/forward spread samples"

    Spot rates are free (Federal Reserve H.10); forward rates are not. Same gap as group B, forward markets. No proxy is used.

data files, license and citation

  • signs-2026-11-10.csv: the Conclusion's four signs, reading dated November 10, 2026; one row per system and series, with the figure, its unit, series, source and address, observation date, percentile and number of observations, flags, reference line, and retrieval date. Column guide.
  • reading-2026-11-08-v2.csv (with direction) and reading-2026-11-08.csv (as first published): the four network dimensions, reading dated November 8, 2026; one row per system and dimension, with the score, the figures behind it, series, observation date, source address and confidence; the first adds the earlier figure, its date and the direction. Column guide.
  • legibility-2026-11-12.csv (scores and reasons) and legibility-evidence-2026-11-12.csv (evidence list): cultural legibility, reading dated November 12, 2026; one row per system and criterion, with the score, the reason, the evidence it cites, the sign note and the data date, and one row per evidence item, with what it shows, its figure and date, its source and address, and its retrieval date. Quoted source material belongs to its owners.
  • series-register.csv: the series register above; one row per series, with the book passage it answers to, the series used, publisher, address, access, frequency, latest observation date, status and notes.

License. The data files are my compilations, published under CC BY 4.0. The underlying series remain subject to their publishers' terms. The text of the site is not covered by that license.

How to cite.

  • Meneses Falconi, Felipe. "The Conclusion's four signs: reading dated November 10, 2026." The Architecture of Money. https://thearchitectureofmoney.com/readings/signs/2026-11-10/. Data CC BY 4.0.
  • Meneses Falconi, Felipe. "The four network dimensions: reading dated November 8, 2026." The Architecture of Money. https://thearchitectureofmoney.com/readings/dimensions/2026-11-08/. Data CC BY 4.0.
  • Meneses Falconi, Felipe. "Cultural legibility: reading dated November 12, 2026." The Architecture of Money. https://thearchitectureofmoney.com/readings/legibility/2026-11-12/. Data CC BY 4.0.

what the framework cannot see

  • The political substrate. Chapter 5: "In a decade when sanctions, freezes, and mandates are doing much of the rebuilding, these may not stay the edge cases they have been. The lens reads what they do to money; it does not read them."
  • Bank and shadow money. In Appendix D4 I list "Systems the framework does not model (shadow banking, informal exchange)." Of the dealer system, I concede in Appendix E, Objection 2, that "this is a genuine layer the framework does not engage"; of bank credit creation, in Objection 3: "This objection identifies a significant gap in the book."
  • Cultural legibility, as an ordinal judgment. Appendix C4: "None of these reads cultural legibility, which remains an ordinal judgment (Appendix A4)." Appendix A4: "None of these is a series; each is a judgment, and a system can read well on all of them and still break, as Lebanon's peg did (Chapter 4)."
  • Cheating cost, read only by proxy. Appendix E, Objection 10: "Cheating cost is defined as a cost but never measured as one; the book infers that it fell from the symptoms that follow—parallel premiums, reserve drain, rule exceptions normalized—and those symptoms are usually known only once a system has failed." And: "Where none is available, the reading is a judgment, and this book has tried to say so each time it makes one."

corrections log

Corrections to a reading's content come first; changes to the site that alter no reading follow. Each is dated and listed newest first.

corrections to readings

  • . The scoring note attributed the three-step scale to Appendix A4's Assessment Approach, which in the book applies to the legibility criteria. Appendix C defines each dimension by its high and low levels; the scale, its cut points and the direction rule are this site's. No score, figure or date changed.
  • . Direction added to the reading of November 8, 2026 (reading-2026-11-08-v2.csv); in the same file, the bitcoin liquidity caveat was corrected (the 2019 base was inflated, so the measured rise understates the change) and four source links were replaced with links that return the cited figures. The original, reading-2026-11-08.csv, is kept as first published.

By reading: cultural legibility, November 12, 2026: none; the Conclusion's four signs, November 10, 2026: none; the four network dimensions, November 8, 2026: two.

changes to the site

  • . Before the book's publication, the site and its readings were brought into line with the book's final text (rev. 19), including Appendix C4's statement of the IMF's sovereign financing-needs threshold (15 percent of GDP for emerging markets, 20 percent for advanced economies), now shown in the signs reading.
  • . The key findings of the reading dated November 8, 2026 now lead with the direction of each currency's figures over about five years, as Appendix C3 advises; the lines are generated from the reading's own data file, and no figure, score or date changed. The readings page lists the two kinds of reading newest first, as its table does, and its table no longer repeats what each reading measures.
  • . The readings moved to their own pages under /readings/; presentation only, no figure or date changed.
  • . The quotation from the Conclusion and Appendix C4's Mapping now follow the locked manuscript (rev. 17), which revised that paragraph before publication; the series register lists credit outrunning output under the Conclusion's four signs.