Backtesting FX: Spread, Swap, Sunday Gaps

A backtest without costs is a novel.

Advanced 5 min read Course 34 of 60

Course 34 of 60 in the forex hub. The object is FX tests that include spread, swap, Sunday gaps.

A Backtest Without Spread and Swap Is a Novel

A Backtest Without Spread and Swap Is a Novel. The honest one-sentence object of this lesson is FX tests that include spread, swap, Sunday gaps. If you cannot say that without opening a platform, you will size the wrong clock. The failure mode that actually hurts accounts is testing mid prices with zero cost. Write the object, then size. Educational only.

Analog, not identity: costs are the first filter. Crypto pairs and listed index futures are a different lesson. Different machine, different hours, different ruin path. Contrast the object with futures hub rather than treating every product as the same machine.

1. Spread

Spread is the first working definition. FX tests that include spread, swap, Sunday gaps. People skip this because a chart is easier than a specification. A chart is not a spec. If you cannot explain Backtesting FX: Spread, Swap, Sunday Gaps to a skeptical friend without a screenshot, you do not understand it yet.

Keep a crib note: object, invalidation, dollar cap. Testing mid prices with zero cost is how cribs get skipped. Do not skip. For the arithmetic habit, use stop calculator until dollars are boring.

A backtest without spread, swap, and Sunday gaps is a novel net pips = gross − spread − slippage − swap − gaps Asia backtest must use Asia spreads London backtest must use London spreads If net ≤ 0 there is no system Fixed 0.2 pip assumption is a children's book

2. Swap

Swap. Convert every pretty statement into dollars. The arithmetic below is illustration — live ticks, pips, and margins change. Re-read the live schedule.

If round-turn cost is 1.4 pips and edge is 1.1, $219 stays paper. The test failed.

If that arithmetic already exceeds your cap, the lesson is over: pass or step down in size. Passing is a position. Confirm the dollar translation with margin calculator so the notebook and the statement agree.

Multi-day systems live and die on swap markup Include Nightly swapTriple WednesdayDealer markup vs interbank Ignore and you Invent carry that is not yoursHold losers 'for swap'Fit a fantasy

3. Sunday gaps

Sunday gaps. Context is not a trigger. Costs are the first filter. Use context to veto, not to force a click.

When in doubt, name FX tests that include spread, swap, Sunday gaps again. If the sentence changed, you changed trades without admitting it. Related structure: free calculators.

Friday close to Sunday open is a return. Model it or do not swing Friday close Sunday open If your stop would have been skipped, the backtest must skip it too.

4. Overfit

Overfit. Process beats mood. Write the rule that fires without a debate at the worst moment. If the rule is 'I'll see how I feel,' you do not have a rule.

Stops, flatten policies, and session boundaries belong in the same notebook as the thesis for Backtesting FX: Spread, Swap, Sunday Gaps. If the stop is a price, convert it with listed venues after you already know the tick or pip.

Twelve filters on EURUSD 2017–2019 is a costume, not an edge Do not automate when The equity curve only exists after spread is removed.You cannot say the rule in one sentence.Walk-forward fails and you 'tweak' until it passes.

5. When not to automate

When not to automate. Fair warning: specs, leverage caps, and dealer rules change over time. The live document wins. Educational only. Not NFA, tax, or a solicitation.

Re-read primary docs before you add size on the object of Backtesting FX: Spread, Swap, Sunday Gaps. See also previous lesson when the confusion is the venue layer, not the chart.

6. Mistakes, limits, takeaways

Mistakes: testing mid prices with zero cost; copying size from a stream; ignoring costs; mixing this machine with crypto pairs or listed index futures. Another: treating Backtesting FX: Spread, Swap, Sunday Gaps as advanced because the vocabulary is long rather than because the dollar cap is written. If the sister asset class is the real mix-up, next lesson before you add size.

Maps go stale. Costs are the first filter. If this lesson and the live spec or statement disagree, the live document wins.

Include spread, swap and Sunday

A forex backtest that ignores spread, swap and weekend gaps tells you how a strategy would perform in a market that doesn't exist. Spread costs alone can turn a profitable short-term system into a losing one.

Use realistic spreads for the hours you'll trade, including wider spreads around news and at the daily rollover. Add swap for any trade held overnight, and model Sunday open gaps. Test on one period and confirm on another you haven't seen. If results fall apart out of sample, the strategy was fitted to noise.

Worth writing down

  • Object: FX tests that include spread, swap, Sunday gaps.
  • Failure: testing mid prices with zero cost.
  • Dollars first, leverage last.
  • Skip the window you cannot survive.

Backtesting FX: Spread, Swap, Sunday Gaps can remain a useful lesson and a poor live habit at the wrong size. Educational only. Not a recommendation to buy, sell, or hold any contract or pair.