What the dollar actually did, over every window, and what nobody published about the other side

“A strong dollar hurts emerging markets” is a two-sided claim and only one side was obtainable. The Federal Reserve’s broad trade-weighted dollar index is published daily and free back to January 2006, so this page measures it exhaustively: 5,164 daily values, 4,906 overlapping twelve-month windows with their full decile distribution, and twenty-one calendar years one by one. For the emerging market side, no free official series of prices, flows or debt service was obtained, and no figure for it appears here.

The file, and the four numbers that frame it

The nominal broad US dollar index runs daily from 2 January 2006 and the file read on 13 August 2026 has 5,164 values to 7 August 2026. Its lowest reading is 85.4692 on 26 July 2011 and its highest 130.0413 on 13 January 2025. The latest value is 119.0649.

From trough to peak the index rose 52.2% over thirteen and a half years, and from that peak it has fallen 8.4% in the nineteen months since. Both moves are large and both are slow, which is the first thing the daily file establishes: this is not a series that does much in a day.

Every twelve-month window, not the memorable ones

The right way to answer “how much does the dollar move?” is to take every day in the file, compare it with the same day a year earlier, and look at the whole distribution. That gives 4,906 overlapping twelve-month windows. Overlapping windows are not independent observations and no statistical test is run on them here; they are used as a description of the range that has actually occurred.

The extremes of that distribution are +21.66%, in the window ending 9 March 2009, and −12.30%, ending 3 March 2010 — almost exactly one year apart, which is what an overlapping-window measure does when a single sharp move enters and then leaves the comparison. The median window is +1.02%, so a slight upward drift over these twenty years, and 38.4% of all windows exceed 5% in absolute value.

That last figure is the practical one. A move of more than five percent in a year is not a dollar crisis: it happened in nearly two windows in five. The latest window, ending 7 August 2026, is −1.19%— between the 40th and 50th percentiles of the distribution, which is to say entirely ordinary.

Twenty-one calendar years, in order

Calendar years are an arbitrary window and are given here because they are the unit almost every commentary uses. Each figure is the first observation of the year against the last.

Twelve of the twenty complete years are positive and eight negative. The largest rise is +10.15% in 2015 and the largest fall −7.63% in 2007, closely followed by −7.57% in 2017 and −7.50% in 2025. There is no run longer than four years in the same direction, and the sign changes eleven times in twenty years.

The last three complete years are worth reading together, because they are the ones under discussion: −3.01% in 2023, +8.39% in 2024 and −7.50%in 2025, with 2026 flat to 7 August at −0.45%. A single narrative that covers all four is describing something other than this series.

Day to day, almost nothing happens

The same file measured at daily frequency gives 5,163 changes and a very different impression from the headlines. The file itself contains 5,375 dated rows, of which 5,164 carry a value; the other 211 are published as blanks and are market holidays, so treating a blank as a zero would invent two hundred days on which the dollar did not move at all.

In twenty years there are fourdays on which this index moved more than two percent, and not one on which it moved three. Eighty-eight and a half percent of days moved less than half a percent. The 52.2% rise from the 2011 low to the 2025 high was assembled almost entirely out of days nobody noticed — which is why a distribution of twelve-month windows says more about this series than any single session ever will.

The half of the subject that was not obtainable

The claim this page is named after has an emerging market side, and measuring it needs at least one of four things: an index of emerging market equity or bond prices with a published methodology; a series of capital flows; a measure of dollar-denominated debt outstanding by borrower country; or a series of local currency exchange rates against the dollar with an official publisher. None was obtained free for this page.

The nearest thing that isfree and official is the European Central Bank’s daily euro reference rate table, which this site uses elsewhere. It carried twenty-nine currencies on 13 August 2026 and its own notice records that the Bulgarian lev was removed when Bulgaria adopted the euro on 1 January 2026. That table is a reference published for information purposes only, with transaction use strongly discouraged, and it is euro-based, so any dollar pair from it must be derived by division and inherits the rounding of two published figures. It is a usable source and it is not an emerging market dataset.

Consequently this page publishes no figure of any kind about emerging market returns, flows, defaults, reserves or debt service. Not a range, not an example, not an anecdote. The claim in the title of this subject remains untested here, and saying so is more useful than a number nobody can check.

What the index is and is not

• It is an index, so its level is meaningless alone. January 2006 equals 100 by construction; 119.0649 means 19% above that day and nothing else.
• It is nominal, not real. No price adjustment is applied, so a change in the index and a change in purchasing power are different things.
• Its weights were not read. The composition is set out in a methodology document that was not opened, so nothing here says which currencies matter most to it.
• It is one series among several.A broad index, a major-currencies index and a single bilateral rate will disagree, and any of the three can be called “the dollar”.
• It has no causal content. The publisher asserts nothing about what the index does to anything, and neither does this page.

What these sources do not say

• No emerging market figure of any kind appears here. No index, flow, reserve, default or debt service series was obtained free, so none is reported.
• The index composition was not read. Nothing here says which currencies it contains or in what weights.
• The index is nominal. No price adjustment is applied, so it is not a measure of purchasing power.
• Overlapping windows are not independent. They are used as a description of what has occurred and no statistical test is run on them.
• The ECB reference rates are not transaction rates. Their publisher says they are for information only and strongly discourages transaction use.
• No causal claim. Nothing read for this page asserts that the dollar causes anything, and neither does this page.

How to verify this yourself

1. Download fredgraph.csv?id=DTWEXBGS and count the rows. There should be 5,164 with a value, beginning 2 January 2006.
2. Find the minimum and maximum. They should be 85.4692 on 26 July 2011 and 130.0413 on 13 January 2025.
3. For every date, find the nearest available observation twelve months earlier and compute the change. You should get 4,906 windows.
4. Sort those windows and read the deciles. The median should be about +1.02%.
5. Count how many windows exceed 5% in absolute value. You should get 38.4% of them.
6. Take the first and last observation of each calendar year and compute the change. 2015 should give about +10.15%.

Where to go next on this site

• The trade balance — the other side of the external accounts, counted over 414 months.
• Interest rates and asset prices — the US rates read on the same days, in full.
• The Federal Reserve — the policy setting behind those rates, with every dated change.
• All macroeconomics guides — the area index, one click away.

Sources opened for this page

Each source below was opened and read on 13 August 2026. Verified on 13 August 2026.

• Federal Reserve Bank of St. Louis — FRED series DTWEXBGS, nominal broad US dollar indexUsed for: all 5,164 daily values from 2 January 2006 to 7 August 2026, the minimum of 85.4692 on 26 July 2011 and maximum of 130.0413 on 13 January 2025, the 4,906 overlapping twelve-month windows with their full decile distribution and extremes of plus 21.66% and minus 12.30%, the share of 38.4% exceeding five percent, the latest window of minus 1.19%, and every calendar year change from 2006 to 2026. Read 13 August 2026.
• European Central Bank — Euro foreign exchange reference ratesUsed for: the twenty-nine currencies listed on 13 August 2026, the removal of the Bulgarian lev when Bulgaria adopted the euro on 1 January 2026, and the statements that the rates are published for information purposes only and that transaction use is strongly discouraged. Read 13 August 2026.
• Federal Reserve Bank of St. Louis — FRED series DFF and DGS10, read in the same sessionUsed for: the effective federal funds rate of 3.63% and the ten-year Treasury yield of 4.70% for 11 August 2026, recorded so that the dollar figures on this page carry the US rate context of the same days without any relation between them being asserted. Read 13 August 2026.
• Board of Governors of the Federal Reserve System — H.15 Selected Interest Rates, release of 12 August 2026Used for: confirmation of the US rates for the same five business days at the original publisher, and the statement that the release is posted daily Monday to Friday at 4:15pm and not on holidays. Read 13 August 2026.

Change log

• 13 August 2026— page rewritten from scratch against the full FRED broad dollar index file, the ECB reference rate page and two FRED rate series, all opened that day. Removed from the previous version: statements about emerging market returns, capital flows and dollar-denominated debt with no source behind any of them; a description of the dollar index composition that no methodology document supports; the claim that a rising dollar causes emerging market stress, asserted without data on either the cause or the effect; a single dollar level quoted as 'strong' with no base year; and calendar year moves given without the full distribution of windows that shows how ordinary most of them are.

Vextor Capital is not authorised under MiFID II as an investment firm, is not registered with the SEC, FINRA, the CFTC or the NFA, is not an investment adviser, holds no position in any instrument named on this page, and receives no payment from any bank, broker, exchange, index provider or data vendor. Every outbound link on this page points to a statistical agency, a central bank or a public data series.