World Economics: Bitcoin and Macro Data
Bitcoin next to policy rates and bond yields in the US and Germany, the dollar, the money supply, central bank balance sheets and US federal debt. The page shows data and measured relationships; it contains no forecasts.
Where things stand
Each tile shows the latest value of a series and its percentile rank, that is, the share of all values since 2000, or since the series began, that were lower. Depending on the series, these are trading days, weeks or months. A rank of 90% means the series was lower than today on 90% of all trading days. The line shows the last two years, the bar the rank between the lowest and the highest value. Clicking a tile takes you to the chart of the series.
Higher than on 94.9% of all trading days since 2000
Higher than on 69.1% of all trading days since 2000
Higher than on 94.1% of all trading days since 2000
Higher than on 66.6% of all trading days since Aug 2000
Higher than on 99.6% of all trading days since Jan 2003
Higher than on 32.5% of all trading days since 2000
Higher than on 27.8% of all trading days since 2000
Higher than on 85.1% of all trading days since Jan 2006
Lowest since the series began (Jul 2002) · 54.8% below the high of Aug 4, 2020
Higher than on 44.4% of all months since Jan 2001
Higher than on 78.9% of all weeks since Dec 2002
Bitcoin compared
Pick a series and overlay it on the Bitcoin price, divide Bitcoin by it, or see how closely the weekly changes of both moved together. Hover over or tap the chart to see the values on a given day.
The ratio is not available here. A price divided by an interest rate is not a meaningful quantity.
Bitcoin on the left, log scale, in US dollars (toggle at the top). US Treasury yield, 10 years on the right, linear. Source: FRED DGS10.
Interest rates
Policy rate and two-year government bond yield
Source: FRED (DFF, DGS2, ECBDFR), Deutsche Bundesbank (term structure, BBSIS). Grey areas: US recessions (NBER). Chart: 21 Data Lab.
10- and 30-year yields: US and Germany
Source: FRED (DGS10, DGS30), Deutsche Bundesbank (BBSIS, Svensson term structure). Hatched: from Feb 18, 2002 to Feb 9, 2006 the US Treasury published no 30-year yield; FRED filled in the values later. Chart: 21 Data Lab.
Real yield and yield curve
Source: FRED (DFII10, T10Y2Y), Deutsche Bundesbank (BBSIS, 10 minus 2 years). Hatched: US yield curve below zero. Calculation and chart: 21 Data Lab.
Bonds and the dollar
TLT: long-dated US Treasuries
Source: Yahoo Finance (TLT, closing price without distributions). Bitcoin: 21 Data Lab (prices). Chart: 21 Data Lab.
Broad Dollar Index
Source: FRED (DTWEXBGS, Nominal Broad U.S. Dollar Index, January 2006 = 100). Chart: 21 Data Lab.
Money and government
Money supply and central banks
Source: FRED (M2SL seasonally adjusted, WALCL in USD, ECBASSETSW in EUR). Grey areas: US recessions (NBER). Calculation and chart: 21 Data Lab.
US debt and budget balance
Budget balance by fiscal year (surplus green, deficit red)
Source: U.S. Treasury Fiscal Data (Debt to the Penny), FRED (FYFSD, fiscal year ending June 30 until 1976, September 30 since). Chart: 21 Data Lab.
US money market funds
Source: FRED (MMMFFAQ027S, Financial Accounts Z.1, holdings at quarter end). Chart: 21 Data Lab.
Bitcoin ratios
The market cap is the price times the estimated circulating supply and follows the currency toggle. The ratios are computed in US dollars and therefore do not depend on it. The gold tile measures a single bitcoin, the two percentage tiles the market cap.
Ratios since 2010
Source: 21 Data Lab (prices), Yahoo Finance (GC=F), FRED (M2SL), U.S. Treasury (Debt to the Penny). Supply estimated from block height. Calculation and chart: 21 Data Lab.
Correlation with Bitcoin
Correlation of weekly changes since July 2010 (Pearson): Bitcoin and levels as log returns, rates as changes in percentage points. US series are paired with Bitcoin in dollars, German series and the ECB balance sheet with Bitcoin in euros. Values near 0 mean no linear relationship. The right-hand column counts how often the 52-week correlation was negative. A 52-week window only counts if both series moved in at least 20 weeks. No correlation is shown for Fed Funds, ECB rate, M2, M2 YoY: policy rates only change at meetings, M2 only once a month.
| Series | Since 2010 | Last 52 weeks | 52-week windows negative |
|---|---|---|---|
| US Treasury yield, 2 years | 0.00 | +0.09 | 35% |
| US Treasury yield, 10 years | +0.01 | +0.15 | 42% |
| US Treasury yield, 30 years | +0.02 | +0.13 | 40% |
| US real yield, 10 years (TIPS) | -0.07 | -0.04 | 55% |
| US yield curve, 10Y minus 2Y | +0.01 | +0.06 | 52% |
| German Bund yield, 2 years | -0.05 | +0.10 | 58% |
| German Bund yield, 10 years | -0.03 | +0.19 | 51% |
| German Bund yield, 30 years | 0.00 | +0.13 | 50% |
| German yield curve, 10Y minus 2Y | +0.03 | +0.10 | 45% |
| Broad Dollar Index (Fed) | -0.11 | -0.26 | 67% |
| TLT (US Treasuries 20+ years) | -0.01 | -0.11 | 58% |
| Fed balance sheet | +0.06 | +0.02 | 58% |
| Eurosystem balance sheet (ECB) | +0.01 | 0.00 | 45% |
| US federal debt | -0.02 | -0.08 | 44% |
| Gold (USD per troy ounce) | +0.09 | +0.11 | 32% |
Methodology & data
Frequently asked questions
01Why does the page show the Broad Dollar Index and not the DXY?
The DXY is an index of the ICE exchange and is licensed. It measures the dollar against six currencies, with the euro making up more than half. The Federal Reserve's Broad Dollar Index measures it against the currencies of the main US trading partners, weighted by trade, and is freely available via FRED (series DTWEXBGS, January 2006 = 100). The levels of the two indices cannot be compared directly: 120.3 points on the Broad Dollar Index is not a DXY level.
02What does the percentile rank in the tiles mean?
The percentile rank is the share of all values of a series since 2000, or since it began, that were lower than the latest one. Depending on the series, these are trading days, weeks or months. The 10-year US yield stood at 5.24% on Oct 1, 2026; it was lower on 94.9% of the 6,691 trading days since 2000. The rank compares the latest value with all earlier values of the same series. It does not say whether the value stays high or falls.
03Does a correlation mean one series moves the other?
No. A correlation measures how strongly the weekly changes of two series move together linearly, as a number from -1 to +1; large moves weigh more than small ones. Both series may react to the same trigger, such as a central bank decision, or the co-movement may be chance. Over the whole period since 2010, 14 of the 15 values in the table lie between -0.1 and +0.1, close to zero. Over 52 weeks they swing more widely, for TLT between -0.37 and +0.39. When a relationship changes sign that often, no rule can be drawn from a single period.
04What is TLT?
TLT is the iShares 20+ Year Treasury Bond ETF, an exchange-traded fund (ETF) holding US Treasuries with more than 20 years to maturity. Its price rises when long-term yields fall and falls when they rise. We show the closing price without the monthly distributions, so price performance only. On Oct 2, 2026 it stood at $77.48, 54.8% below its high of Aug 4, 2020.
05Why are some series not up to date?
Each series runs up to its source's latest release. US yields appear on the next business day, the Bundesbank's German yields on the same trading day, the Broad Dollar Index once a week for the previous week (latest Sep 25, 2026), M2 on the fourth Tuesday of the following month (latest Aug 2026), money market funds about ten weeks after quarter end (latest as of Jun 30, 2026) and the budget balance once a year (latest fiscal year 2025). We do not estimate missing values. Every tile states the date of its value, and in the weekly grid of the charts a value carries forward from its reference date for at most 10 days (daily series), 14 days (weekly series) or 95 days (monthly series).
06How do you compute the Bitcoin market cap?
We multiply the price by the circulating supply. The supply follows from the block height because the issuance schedule is fixed. We estimate the block height of a day from measured heights: four times a month from July 2010 to 2012, on the 1st, 8th, 16th and 24th, twice a month in 2013 and 2014, on the 1st and 16th, and on January 1 and July 1 afterwards, interpolating linearly in between. Checked against actual block heights on 566 dates, the estimated supply was off by at most 0.2% in 2010 and 2011 and by less than 0.1% from 2012. On Oct 3, 2026 that gave about 20.09 million BTC and a market cap of $1.70tn.