Two providers can show different financing amounts for the same pair, and a morning estimate can differ from the later booking. Before attributing the difference to markup, match the product, direction, notional, rate unit, date and financed days. This guide explains what those comparisons can establish and which component data are still needed.

Where does a swap rate come from in the first place?

For retail forex, swap is a contractual financing debit or credit. Wholesale currency-funding conditions can influence the provider’s reference rate, but the policy-rate gap alone does not determine the applied amount or sign. Check dated Long and Short rates: both can be negative. The carry trade on USD/JPY describes exchange-rate exposure as well as financing.

Covered interest parity is a textbook no-arbitrage relation between comparable funding rates, spot and forward prices under its assumptions. Forward points are the difference between the forward and spot rate for a specified maturity. BIS documents persistent deviations associated with hedging demand and balance-sheet constraints; actual wholesale quotes need not equal a simple central-bank differential. Nor do they prescribe a retail tariff.

How does the market turn a rate gap into a daily point?

Institutional tom-next rolls settlement between consecutive value dates. Many spot pairs use T+2 settlement, while exceptions and currency holidays affect dates. A retail rolling CFD's financing follows the provider's contract; it is not a physically deliverable spot position that must automatically be closed and reopened in the client's account every night.

MT5 documents several ways of accounting for a rollover, including separately accumulated swap and configured reopening methods; it does not make a retail financing entry the same contract as a wholesale FX swap. The BIS survey from April 2022 provides historical instrument-turnover context, rather than today’s retail rate. The institutional FX swap explains the near and far currency exchanges.

The final amount does not reveal provider markup

If the method separates a market component and a fee, assessing the fee requires both. Fictional A has a −5 USD market component and −1 USD fee; B has −4 and −2 USD. All values already use one common unit: one lot and one day.

Scroll sideways to see the full diagram

The final amount does not reveal provider markupA: market −5 plus fee −1 totals −6 USD per lot per day. B: −4 plus −2 also totals −6. At 0.50 lot over two days, both charge 6 USD, but the fee component is 1 USD in A and 2 USD in B.Same total, different fee shareA: market −5 + fee −1 = −6Fee: 1 of 6 USD per lot per dayB: market −4 + fee −2 = −6Fee: 2 of 6 USD per lot per day0.50 lot × 2 days: both total −6 USDThe final amount alone hides the fee.
Different component combinations can produce the same client charge.

Both total charges are −6 × 0.50 × 2 = −6 USD. The fee component over that period is −1 USD in A and −2 USD in B. Knowing only the final −6 USD does not uniquely reconstruct these components. IG's document describes the relationship tom-next + administration fee; this diagram uses different, invented monetary values. A difference between final swap rates alone does not prove a markup difference. Check the disclosed components and their units.

Model assumptions: Same product, direction, date, lot and USD currency; final component values in an additive model, before applying the multiplier. A/B do not describe real providers. Mechanism source: IG Australia — How is funding on forex positions calculated?. This is an illustrative educational example, not current market data.

Where does the broker step in, and why do rates differ?

There is no single binding retail-financing table for every provider. IG Australia describes a tom-next market rate with an administration charge; other contracts can use different references or methods. Source quotes, adjustments and calculation terms can all differ. A final rate alone does not reveal which component changed, just as the spread between the bid and the ask must be examined with its pricing terms. Do not infer a larger markup merely from a smaller credit.

The BIS analysis of interest parity and the cross-currency basis explains why wholesale forward pricing can depart from textbook parity. At the retail level, the provider’s methodology and component disclosure matter. Some final tables show only Long and Short totals; where components are disclosed, compare their same-date units. A difference between finished values does not identify the difference in fees without matched reference data.

What does the difference look like in numbers?

The accompanying card isolates one parameter in an entirely hypothetical model: the market component is assumed equal, while the fees differ. Both net rates are stated for the same chosen long exposure and the same financing day unit. The 30-day arithmetic assumes constant inputs; it does not describe elapsed nights or observed provider rates. Those supplied assumptions let the model attribute its difference to fees.

Fictional A: assumed market credit USD 0.65 minus fee USD 0.10Net +USD 0.55 per financing day for the chosen position; 30 day-equivalents at constant inputs give +USD 16.50, before other transaction costs.
Fictional B: assumed market credit USD 0.65 minus fee USD 0.60Net +USD 0.05 per financing day for the same chosen position; 30 day-equivalents at constant inputs give +USD 1.50, before other transaction costs.
What it meansThe model assigns the same market component and different fees, so the net credits have an 11:1 ratio. With only the finished totals, you could not infer those components; their equality and fees are assumptions here.

A separate unit comparison assumes long 0.25 lot EUR/USD, 100,000 EUR per lot, point 0.00001 USD per EUR, USD accounts and three financing days. The invented rates are final and net of fees, before the multiplier. A quotes −8 points per lot per day: one lot-point is USD 1, so −8 × 1 × 0.25 × 3 = −USD 6. B quotes −USD 7 per lot per day: −7 × 0.25 × 3 = −USD 5.25. B costs USD 0.75 less in this model. The totals do not tell us how much of either rate is a provider fee. The accompanying card is a separate model with explicitly assumed components, not observed provider data.

What does this teach an individual investor?

Read the provider’s rate for the correct date and product, then reconcile the booked amount. A brief position can cross the cutoff, while a longer one can accumulate several entries with changing inputs. The glossary entry on the swap gives the basic definition. Duration labels alone do not prove that financing is irrelevant or that one component always dominates.

A policy-rate difference can illustrate a simplified funding model but cannot reproduce the retail quote without its methodology. OANDA Corporation publishes indicative rates and later final rates, and says funding rates may change daily. Save the reference date and status, then compare the final historical rate and booking rather than assuming that the pre-entry screen number is guaranteed. The forex swap explains rate-to-cash conversion.

What to do tomorrow

  1. Open your broker swap table and write down the long and short point for the pairs you trade. Most platforms show it in the instrument specification or on a terms page. Note the values for two or three of your main pairs, separately for the long and the short side, so you have them at hand before you click buy or sell.
  2. Calculate financing from the instrument's actual schedule. Record long/short rate, billing unit, rollover time, multi-day multipliers and holiday adjustments. Apply the specified day weights rather than assuming a triple Wednesday charge for every instrument.
  3. Match the comparison before attributing its cause. Use the same product, direction, quantity, date, rate status, financed days and currency. Convert differing points or percentage units correctly. The final cost difference is not “exactly the markup difference”; fee attribution requires separate matching component data.
  4. Check whether your broker uses a triple swap day other than Wednesday. Look in the specification or ask support, because a few brokers shift the triple charge to a different day for selected instruments. It takes five minutes and saves an unpleasant surprise when you plan an entry right before the weekend.