Fed Dollar Swap Lines: Who Actually Borrows the Dollars?

A headline says the Fed supplied dollars overseas. It is tempting to imagine the Fed buying euros as a directional trade or lending directly to a foreign commercial bank. The transaction has two central banks, two exchange legs and a separate lending relationship overseas. Following those three pieces changes the interpretation.

Does this number describe outstanding funding or a new operation? Start by checking Wednesday stock, counterparty and operation maturity in the source documents. The calculations below distinguish the published record from an illustrative customer transaction and show how to check the result in dollars.

Key insight. A central bank dollar liquidity swap gives a foreign central bank temporary access to dollars in exchange for its currency, with reversal at the same agreed exchange rate. The foreign central bank can provide dollars to institutions in its jurisdiction and bears the credit risk on its onward lending.

Draw the two exchanges and the separate loan

Suppose the Fed supplies $120 million at 1.20 dollars per euro in an illustrative swap. The foreign central bank supplies €100 million on the initial leg. At maturity the parties return the principal currencies at the agreed rate and the foreign central bank pays the applicable interest. A move in spot EUR/USD while the swap is outstanding does not change those principal amounts.

Onward lending is a separate transaction under the foreign central bank’s rules, including eligibility, collateral and pricing. The Fed’s counterparty is the foreign central bank. This separation matters when interpreting who is exposed to an individual commercial bank default. A swap line’s existence also differs from an actual drawing on a date.

Read a weekly balance without calling it trading volume

The Fed’s balance sheet reports central bank liquidity swaps outstanding. A stock of $10 billion outstanding across two weekly snapshots does not mean only $10 billion was offered or that no new loans were made. Maturing transactions and fresh drawings can offset. To study usage, distinguish a point-in-time balance, a weekly average, an allotment and cumulative gross activity.

Dollar demand abroad can rise during funding stress, but an increase in drawings can also reflect pricing, auction frequency and demand for liquidity insurance. Compare money-market conditions and announced operations separately. A strong dollar and rising balances can share a cause, while an effective liquidity facility can alter funding conditions without producing a predictable EUR/USD candle.

Why interest and FX translation need separate lines

For an illustrative seven-day $120 million drawing at 4.8% simple annual interest on an assumed 360-day basis, the interest is $112,000. The two exchanges of principal use the agreed rate. The interest convention and maturity come from the operation terms; the quoted illustration does not establish historical pricing. Revaluing all dollars outstanding into euros at today’s rate may help a reader understand scale. It does not tell the reader the actual foreign-currency principal returned at maturity or the Fed’s earnings.

Suppose one operation matures Wednesday while another of almost the same size starts the same day. Outstanding balances can stay roughly flat despite considerable gross activity. A later report may show a sharp drop as maturity dates cluster. Analyze the specific operations or foreign central bank allotments before treating an outstanding series as a gauge of new stress. Then investigate whether funding spreads, demand for dollars and FX pricing moved at the time new information became public.

Three different measures of demand for central bank dollars

The announced availability of a facility, the amount allotted at an operation and the balance outstanding on a reference date provide three different measures. A facility can be available with very little take-up. A large single auction can replace maturing loans and leave the balance nearly unchanged. A fall in outstanding swaps can reflect a scheduled maturity after earlier stress had already passed. Line up operation dates and maturities before describing an observed change as a new borrowing decision.

Suppose $10 billion is outstanding on Wednesday, $7 billion matures on Thursday and $8 billion is newly drawn before the next snapshot. The new stock is $11 billion. New borrowing was $8 billion and the change in the stock was only $1 billion. Summing the two Wednesday observations would give $21 billion, a third number with no direct interpretation as weekly new lending. The example uses dollars throughout. Converting both snapshots into euros at different rates would add a valuation effect to the stock comparison.

For the counterparty chain, draw the foreign commercial bank’s loan from its central bank separately from the central bank’s swap with the Fed. Interest rates, collateral conditions and access criteria on the local lending operation need not be the same as on the central bank swap. The commercial bank does not become the Fed’s direct swap counterparty merely because the received dollars originated through that facility. The Federal Reserve’s FAQ identifies the foreign central bank’s responsibility for repayment and its separate credit exposure on local lending.

A US trader evaluating dollar funding stress can compare facility use with actual private dollar funding spreads or cross-currency basis under stated conventions. Access to a backstop can matter even when take-up is limited. Conversely, a surge in take-up may coincide with policy changes to price or auction frequency. A chart of aggregate balances establishes the quantity outstanding. It cannot reveal the unobserved funding conditions that would have prevailed without the facility, nor which institutions would otherwise have sold euros to raise cash.

Historical workbook: Euro-equivalent size of total dollar swaps in billions

The workbook follows dollar liquidity outstanding under central bank swaps from a US dollar perspective. Its question is specific: Does this number describe outstanding funding or a new operation? Keep three columns beside the dates. Column one is fed liquidity swaps outstanding in billions of usd. Column two is usd per euro at the observation date. The calculated column shows euro-equivalent size of total dollar swaps in billions. Using the same definitions in every row makes it possible to explain an outlier by returning to the inputs instead of changing the story after the price has moved.

Take the last available Wednesday balance in the month on a date with a matching EUR/USD observation. Convert SWPT from millions to billions of dollars and divide by the euro rate. This is a size illustration across all counterparties. Actual swaps can involve several foreign currencies at independently agreed reversal rates.

There are 120 monthly observations from 2015-01 through 2024-12. The historical FX observations come from DEXUSEU. For the independent dollar comparison below, one unit of EUR is always expressed in US dollars. Where the published quote gives foreign currency per dollar, the comparison takes its reciprocal. This keeps a rise in the comparison series synonymous with appreciation of the named foreign currency against USD. The calculation of the main worksheet stays in its original units.

Selected records from the workbook. The table displays rounded values. Statistics and reconstructions use the unrounded inputs.
PeriodFed liquidity swaps outstanding in billions of USDUSD per euro at the observation dateEuro-equivalent size of total dollar swaps in billions
2015-010.00101.13420.0009
2015-040.00001.11740.0000
2019-120.07901.11130.0711
2020-05448.94601.0969409.2862
2022-040.21401.05490.2029
2024-120.08401.04690.0802

Read one row from the source to the answer

Start with 2019-12, near the chronological middle of the sample. The two inputs were 0.0790 for fed liquidity swaps outstanding in billions of usd and 1.1113 for usd per euro at the observation date. Applying the stated rule gives 0.0711 for the translated outstanding amount. A reader reconstructing the published number should carry more precision than the table displays. The market comparison used an FX mark on 2019-12-18, quoted at 1.1113 in the ordinary EUR/USD convention. It is not necessarily the observation time of the other inputs.

Outstanding dollar swaps were 0.079 billion dollars on the selected Wednesday. At the matched spot reference of 1.11130 USD per euro, their illustrative scale is 0.071 billion euros. If all these funds were mistakenly assigned to the euro area, a dollar number would be turned into an unsupported claim about the geographic borrower. The Fed series covers the total across relevant operations. An actual allocation requires the official operation-level or counterparty records.

Suppose $2.4 billion is drawn, then $1.7 billion of that new amount matures before the next observation. Outstanding balances increase by $0.7 billion, bringing the example to 0.779 billion under those assumptions. Gross lending activity is larger than the simple change in stock. Measure actual allotments when that is the object of the study, and keep the foreign currency agreed on each swap separate from the table’s illustrative euro translation.

Which input explains the largest change?

Across adjacent months the largest absolute change in the translated outstanding amount occurred between 2020-03 and 2020-04. The calculated value went from 190.1717 to 404.4531, a signed difference of 214.2815. A headline describing the two end values is accurate but incomplete. To identify what generated the change, first update only the first column. Then update the second while holding the new first column fixed. This procedure accounts for the whole change without assuming which input moved first in real time.

The first input went from 206.0510 to 438.9530. With the second held at its starting value of 1.0835, that gives a hypothetical intermediate result of 405.1251. The first step therefore contributes 214.9534 to the measured change. Replacing the remaining old input with 1.0853 contributes another -0.6719. Add the two signed contributions to recover 214.2815 before display rounding. A negative contribution offsets a positive one even if both raw inputs rose.

Now reverse the order. Updating the second input first produces 189.8563. Its first-step contribution under this order is -0.3154, and the remaining contribution is 214.5969. Addition and subtraction formulas give the same attribution either way. Products and quotients can distribute a joint interaction differently. If the two allocations differ, report the order you used or average the two contributions for each input. Both reconciliations describe an identity. Neither establishes an economic cause. Translate the stock for scale without inferring which central bank borrowed the dollars. The weekly change can reflect new drawings, maturities and other movements in the outstanding balance.

Average of both replacement orders. The two allocations sum to the full change.
Measured changeAllocated to first inputAllocated to second input
214.2815214.7751-0.4937

Use a typical period to test a dramatic headline

Ordering the calculated record by level gives a minimum of 2015-04 (0.0000) and a maximum of 2020-05 (409.2862). The distance between them is 409.2862 in the units of the worksheet. The mean is 12.8859 and the median is 0.1810. A mean gives every period equal weight and includes the influence of extreme observations. A median locates the middle of the ordering. Neither measures the experience of a specific US account with transactions concentrated on one or two dates.

A second question is how much the translated outstanding amount generally changed between adjacent observations. Its median absolute change was 0.0408. The transition from 2024-11 to 2024-12 lies closest to that amount, going from 0.1210 to 0.0802. In that interval the first input moved by -0.0440 and the second by -0.0106 in their respective units. Compare those with the largest-change interval. The same final-sized change can come from modest moves in both inputs or large movements that largely offset.

For the high observation, ask what the same calculation would have returned if only the first input had taken its median sample value. Using 0.1940 in place of 448.9460 while keeping the second input at 1.0969 gives 0.1769. This is a deliberately constructed comparison. Those values need never have occurred together. If an input is contractually fixed or mechanically related to the other one, interpret the substitution as an arithmetic sensitivity exercise rather than a market scenario that could have been traded.

Does the dollar chart agree with the worksheet?

The largest absolute adjacent change in the dollar value of one EUR occurred from 2022-03 to 2022-04. The market comparison changed from 1.116300 to 1.054900 USD per unit, or -5.50%. By contrast, the worksheet moved by -0.0041 in its own units. A percentage currency return and a change in the translated outstanding amount are different measurements. Comparing their signs is meaningful only after specifying the direction and recognizing when one input is itself a transformation of the exchange rate.

Of the 119 adjacent pairs, 119 had a measurable change in both series. In 61 they moved in the same signed direction, and in 58 they moved in opposite directions. The remainder had a flat result in at least one column at the stated numerical tolerance. These counts describe a sample, with no assumption of independence between observations. Where the translation formula already uses the exchange rate, part or all of the association can be mechanical. An agreement count alone is unsuitable as a trading win rate.

Compare a period where the signs explicitly disagreed. From 2017-12 to 2018-01, the dollar value of one EUR moved 4.42% and the worksheet changed by -9.5475. The first input ended at 0.6730 and the second at 1.2428. Record the observation even if it challenges your preferred narrative. Then return to federal reserve h.4.1 and swap faq to test whether the disagreement reflects definitions, different sampling dates or a genuinely different economic exposure. The outcome by itself does not choose between these explanations.

What was knowable before the last observation?

Keep the last month outside the record when setting a reference. The earlier 119 observations had a median worksheet value of 0.1865. The last result was 0.0802 in 2024-12. A simple baseline using the immediately preceding value, 0.1210, misses the last result by -0.0408 on an actual-minus-baseline convention. The earlier median misses by -0.1062. These baselines answer different questions about stability and representativeness.

Extend the last-value test across every adjacent pair and take the mean absolute difference. The result is 7.6982, compared with a median absolute difference of 0.0408. The mean gives unusually large misses their full weight. The median describes a typical miss. When you evaluate a real budget, multiply a per-unit figure only by its proper exposure, and account for any amount the formula already includes. Multiplying a dollar total by the entire invoice a second time would produce a fictitious level of risk.

These tests run on one documented set of historical observations. A revised series, an annual value compiled after year-end or a market quote taken after a release can include information unavailable at the hypothetical decision. To turn the worksheet into a real-time study, save the exact version that was available, the timestamp at which it became available and the quote at which a decision could be implemented. Keep operation announcements, allotments and the dated balance sheet with that file. A comparison that relies on a later release remains a historical reconciliation even if it happens to fit the chart perfectly.

How much simple interest would a small hypothetical drawing accrue?

Use the calendar days in the selected reference month as the assumed term of a hypothetical $75 million drawing, with a 3.6% simple annual rate and a 360-day basis. Multiply principal by rate and days divided by 360. The interest is independent of the euro translation of aggregate swaps.

The output column below is usd interest on $75 million at 3.6% for an assumed tenor. This is a second calculation on the same dated records. State the assumption before reading the result so it is clear which parts of the comparison come from observations and which come from the proposed decision. The smallest calculated result occurs in 2015-02, at 210,000.0000. The largest occurs in 2024-12, at 232,500.0000. A negative result, where the formula allows one, means the order of the comparison matters. It does not mean that the source supplied a negative market price.

USD interest on $75 million at 3.6% for an assumed tenor. Repeated selections are shown once.
PeriodFirst inputSecond inputOriginal worksheet resultSecond calculation
2015-020.00301.13630.0026210,000.0000
2024-120.08401.04690.0802232,500.0000

For a central case use 2024-12, where the result is 232,500.0000, against a full-sample median of 232,500.0000. The associated FX observation was 1.04690 in the EUR/USD convention on 2024-12-18. Compare it with the largest case rather than simply picking the most recent observation. If an assumed quantity, policy or percentage in the calculation would have been different at the time, describe the result as a controlled comparison across historical conditions. Avoid presenting it as a transaction the source actually recorded.

The first 60 records, 2015-01 through 2019-12, average 228,250.0000 on this second calculation. The later 60 records, 2020-01 through 2024-12, average 228,375.0000. The signed difference is 125.0000. This gives equal weight to each retained observation and asks how the same rule behaved over two consecutive intervals. It does not adjust for changes in business activity or prove that the start of the later interval marks a regime change. This tests the sensitivity of interest to days of funding. Actual swap operations use the announced maturity and pricing, and the aggregate outstanding figure cannot establish either input for one counterparty.

Can a change be reconciled to an operation?

A useful funding note specifies the foreign central bank, currency, auction amount, maturity and whether the quoted figure is offered, allotted or outstanding. If the cited series is only the Fed aggregate, label the statement accordingly. Perfect arithmetic in a euro translation cannot convert an aggregate into a counterparty breakdown.

Compare the release timestamp with the date for which the outstanding amount was measured. A backtest that uses a later published balance as though known at the start of that Wednesday can overstate apparent predictive power.

How to identify the price setup on a EUR/USD chart

US dollar swap drawings are reported in an official balance-sheet time series. A release-day EUR/USD chart answers a different question about what the market traded. The schematic tests a prior swing and does not pretend a central bank put orders at a particular five-minute price.

Use a five-minute demo chart with a named time zone, such as America/New_York. The first decision is where an existing swing could attract orders. In the drawing the prior low sits at 1.10944. Mark it while the chart still ends before the next challenge. The subsequent wick reaches 1.10865, below the earlier reference, and the market returns through it. ICT describes that sequence as a liquidity sweep. Candles show traded or quoted prices on a feed, so the sequence does not establish any individual institution’s identity or the number of orders at the swing.

The next test is displacement through the earlier internal swing at 1.11144. In this schematic the outer wicks of bars five and seven leave a three-candle fair value gap between 1.11081 and 1.11130. Write down both boundaries before testing the retest. A wick across the initial level without displacement is a different observation. A displacement without a later return can leave a correctly identified setup unfilled. Preserve each outcome so a replay does not silently select only the attractive-looking retracements.

Chart 1. Identify the sweep and three-candle imbalance
EUR/USD | hypothetical spot pricesPrior low 1.10944Sweep 1.10865Prior swing 1.11144Gap 1.11081 / 1.11130Green: up Red: down Blue: reference Purple: sweep Gold: gap
Schematic EUR/USD bars. Compare the wick of bar five with the wick of bar seven. The middle bar provides displacement. The shaded interval records the gap between the outside wicks.

How to trade the retest on demo

For the hypothetical long position, the order waits at 1.11104 after the gap has formed. The protective stop is 1.10819, beyond the swept extreme. The target is 1.11702, which requires the quoted pair to move higher. Cancel the order if price reaches invalidation before entry, the agreed observation window expires, or the executable spread makes the original risk budget infeasible. A chart mid touching a line does not prove the customer’s bid or offer filled. In a live record those two events need separate timestamps.

The second drawing shows one favorable branch solely to establish the accounting. A losing branch exits at the stop under an exact-fill assumption. A third branch never enters and incurs no modeled round-trip charge. If stop and target both lie within one observed bar, the wick alone does not reveal their order. Seek finer time-stamped evidence or retain an ambiguous result rather than labeling the trade a winner. For an event study, select the exit, expiry and treatment of ambiguous sequences before viewing the release-day chart.

If the Wednesday observation becomes public later in the week, use the actual publication time in an event study. Dating a backtest entry to the balance-sheet reference day can introduce look-ahead information.

Chart 2. Price the entry, protective stop and target
EUR/USD | hypothetical spot pricesEntry 1.11104Stop 1.10819Target 1.11702Green: up Red: down Blue: reference Purple: target
Illustrative continuation after a retest. Lines show the planned long order. The favorable path is one possible sequence and carries no probability estimate.

Worked trade in a dollar-denominated demo account

Illustrative position. The prices are separate from the historical observations and the outcome has no assigned probability.
FieldHypothetical execution
Pair and sideEUR/USD long
Size22,000 EUR
Entry1.11104
Protective stop1.10819
Target1.11702
Gross risk and potential reward$62.70 and $131.56
Assumed round-trip charges$3.33

First subtract the entry and stop in absolute price units. The distance is 0.00285 USD per EUR, compared with 0.00598 to the target. Multiply each by 22,000 base units. That yields 62.7000 USD of gross risk and 131.5600 USD of potential gross gain. The figures are already in the USD account currency. Keep the full numerical precision through this step.

Gross planned risk is $62.70 and potential gain is $131.56. Divide gain by risk: $131.56 / $62.70 = 2.10. On a risk-to-reward convention that is 1:2.10. Now assume $3.33 in total round-trip charges in either completed branch, with executable entry and exit prices in the example. Net risk is $62.70 plus $3.33, or $66.03. Net potential reward is $131.56 minus $3.33, or $128.23. The net R:R is 1:1.94 from $128.23 / $66.03. These ratios use full-precision amounts before display rounding.

If the stop slipped one further pip, the modeled exit would move to 1.10809 and the completed loss including the assumed charge would become $68.23. Compare this with the original $66.03. Slippage, funding and market access can all affect an actual instrument. The purpose of this branch is to make the size and account-currency conversion explicit. A promising reward-to-risk ratio says nothing about the frequency with which this particular entry, protective exit and target are actually executable.

To annotate the rehearsal, mark the existing swing first, then identify displacement and the interval between the outer candle wicks. For definitions and session context, see the school’s lessons on fair value gaps, order blocks, the London session, market structure shifts, and the Judas swing.

Sources and calculation notes

Federal Reserve liquidity swaps. Federal Reserve swap line FAQ.

Historical series and original records. DEXUSEU series, units and frequency · SWPT series, units and frequency. The calculations use observations available when checked on November 15, 2026. Subsequent revisions can change retrospective comparisons.

Frequently asked questions

Does the Fed face the foreign bank that takes an onward loan?

Its swap counterparty is the participating foreign central bank, which bears the onward borrower credit risk.

Why reverse at the same FX rate?

The agreed reversal fixes the principal amounts for the swap legs and addresses the Fed’s exchange-rate exposure on that structure.

Are outstanding drawings the size of the facility?

No. Drawings are the amount used under the reporting measure, distinct from available arrangements.

Can I add weekly outstanding balances to get annual lending?

That would repeatedly count balances carried across weeks. Gross operations require transaction data.

Where can I check the accounting and counterparties?

Use the Federal Reserve liquidity swap FAQ, H.4.1 references and the relevant central bank’s operation records.

To take the release from the data table to a dated trading record, use SOFR for Forex Traders, Quantitative Tightening, the news-release filter, and hindsight-free replay.