A multi-day plan needs an answer to “what happens while I cannot manage the account?”, as well as an entry condition. This is one explicit process to test, not a demonstrated profitable strategy. The 2–7 day label is a planning horizon, not a requirement to hold a losing position for two days. An evening review or a stop order does not remove overnight execution or operational risk.

How this process differs from swing trading in general

The guide to swing trading looks like for someone with a day job introduces the style; this article supplies a reproducible test version. Compare it with position trading by its actual rules, not a universal number of sessions separating labels. Choose the instrument, account, price feed and review time first. If required cancellation or exits cannot be handled during work, this version may be operationally infeasible.

Choosing timeframes top down: D1 for the trend, H4 for the entry

Our long-only example uses EUR/USD Bid bars from one feed, with one predetermined H4 close per day. Record server time zone and D1 construction. For the last completed D1 require Close > SMA20 and SMA20 > its previous value; SMA20 averages the last twenty closed D1 prices. This is a selected filter, not a definition of every real trend or an optimal setting. The trading strategies and trend-following discussions offer other rules to evaluate.

Do not use the eventual Close of an unfinished D1 when the H4 decision arrives earlier. Higher-timeframe values can change until confirmed. Keep the signal time, information then available and delay before an executable order separately.

The entry pattern: a pullback to the trend and a trigger candle

Our H4 signal requires the previous completed bar to have Close < Open and the signal bar to have Close > Open and Close above the previous High. Equality fails the colour condition. No signal means no order. We add no news filter in this version and do not hold through weekends; changing either policy creates another version to test, rather than a hindsight exception.

Submit a Buy Limit at the signal Close only if it lies below current Ask and the platform permits the order. Expiry is thirty minutes from submission. Freeze SL at the lower of the two bar Lows minus two conventional pips, and TP at the entry limit plus twice the limit-to-SL distance. These parameters are choices for testing, not mandatory values. A limit may remain unfilled; a better purchase price changes the distances to the frozen exits.

Position sizing with a wider stop

In a separate fictional USD-account execution model, buy 10,000 EUR at Ask 1.1002, with SL 1.0942 and TP 1.1122. These supplied prices are for arithmetic, not a reconstructed candle signal. With a 0.0001 pip, stop distance is 60 pips, target distance 120 and pip value USD 1. No universal 50–80-pip stop or obligatory width follows from holding several days.

Assume USD 2 round-turn commission and USD 2 financing reserve. Loss at an exact 1.0942 exit would be USD 64, not a guaranteed maximum. With strictly proportional fees, a USD 64 budget gives 64/(0.0060 + 0.0002 + 0.0002) = 10,000 EUR. Check minimum fees, volume step, margin and combined exposure separately. sizing positions for different stop widths explains the chosen distance and size; an available leverage cap does not determine this price sensitivity or ensure the planned fill. A lower size may preserve a specified budget when distance widens, under the same cost assumptions.

Where to set the target, and what holding overnight costs

The two-times-distance target above is a fixed experiment rule, not a promise of profitable odds. The trailing stop are covered separately mechanism would create another management version if added here; an ordinary trailing stop is not guaranteed. Freeze the decision rule before evaluating outcomes and calculate returns from actual fills, including fees.

OANDA Corporation’s US financing explanation uses its own 17:00 ET cutoff, directional rates and administration adjustment; funding can cover multiple settlement days and holidays. These are named-provider terms, not the tariff assumed below or a global Wednesday rule. The what a forex swap is guide separates financing from a central-bank rate gap. Sum actual postings rather than multiplying today’s displayed rate by past nights.

For our separate arithmetic, charges 0.40, 0.40 and 1.20 total USD 2. Three postings are not three identical day units. Ordinary stop execution can worsen around gaps; reopening dates and product hours belong to the provider and time zone. Record those constraints before choosing a weekend policy.

Five days is a limit, not a promised holding time

From a first fill Wednesday10:00,a5×24h limit falls Monday10:00. A second condition,Friday30 minutes before close,gives Friday19:30. The earlier deadline applies:57h30min after entry,although120h has not passed.

The earlier of two swing deadlinesHours since Wednesday10:00 fill. Friday19:30 is57.5h later;Friday20:00 is58h;Monday10:00 is120h. Our rule ends observation at the first deadline,57.5h.Clock starts at the first fillWednesday10:00 → Friday19:30Earlier:57h30minLater:120h,Monday10:00Deadlines use one shared clock
The green segment is the permitted horizon here. The dashed120h boundary is later,not additional time after Friday.

Record each limit’s deadline separately and select the minimum. If the provider closes earlier that week,use its actual calendar; do not apply fictional20:00 to every account. A review deadline does not guarantee execution at precisely that time,so retain the actual closing time in the journal too.

Assumptions: A fictional week and one fixed clock without a daylight-saving change. First fill Wednesday10:00;market close Friday20:00. Observation ends at the earlier of30 minutes before close or5×24h. No prior stop or target exit.

Deadlines, partial fills and separate closing outcomes

Our holding deadline is the earlier of 5 × 24 hours from first fill and thirty minutes before the instrument’s Friday close. SL/TP may close earlier. In the preserved diagram’s fictional fixed clock, Wednesday 10:00 first fill and Friday 20:00 market close produce a Friday 19:30 deadline after 57.5 hours, before Monday’s 120-hour limit. Use the actual provider calendar; a deadline is not a guarantee of execution exactly then.

Do not submit if the thirty-minute expiry would reach or exceed that Friday deadline. Cancel all unfilled orders or remainders by it; skip entry if cancellation cannot be handled. If the five-day deadline falls in a scheduled holiday closure, set an earlier exit before that closure or skip the trade. Permit only one position or pending order for this version on the instrument. Protect filled volume after any partial fill; the remainder retains the original expiry.

Independent fictional exits after the same USD 2 financing and USD 2 commission
ExitExecuted sell pricePrice P/L, USDNet P/L, USD
Target1.1122+120+116
Stop at its level1.0942−60−64
Worse stop fill1.0922−80−84
Time exit1.1032+30+26

Every row uses 10,000 × (executed sale − 1.1002) − 2 − 2 and the same three funding charges before another accrual. Spread is already included in Ask/Bid fills and is not deducted again. Earlier closure needs its actual financing history. The worse-stop loss of USD 84 exceeds the planned USD 64; changed future fees can increase it further.

Separately, suppose only 4,000 of a 10,000 EUR limit fills at 1.1002 before cancellation of the remaining 6,000 is confirmed. The filled position remains open. If the plan calls for closing it and the closing sale executes at Bid 1.0997, price P/L is 4,000 × (1.0997 − 1.1002) = −USD 2 before charges, not zero because the remainder was cancelled. Record requested, filled, remaining and cancelled units and confirmed SL/TP. A cancellation request alone does not exclude an additional fill before acceptance. Use the account’s correct close procedure; a new opposite position is not necessarily a close.

What to do tomorrow

  1. Fix the data and review schedule. Record instrument, feed, server zone, D1 boundary and the one H4 close used by this test.
  2. Apply the completed D1 SMA20 filter and the two-bar H4 condition exactly; do not use a future higher-timeframe Close or select an attractive past reaction.
  3. Check permissible order prices, volume step, margin, costs and worse fills. The price-distance budget is a planned scenario, not a maximum loss or a universal 1% safety rule.
  4. Prepare Friday and holiday deadlines plus cancellation and closing procedures. After fills, verify the actual volume and active protection; an alert does not close a position.
  5. Keep signal, order, fill and plan deviations separate, including unfilled limits and skipped cases. Evaluate fixed rules on unused data with costs and execution assumptions; demo consistency alone cannot certify live readiness.