Second Wind Index EA
- Experts
-
Pitt Petruschke
Algo trader and EA developer from Germany. Founder of AlgoVerdict (algoverdict.com) – independent tests of Expert Advisors, brokers and signal platforms based on live accounts. Author of Dawnbreak, a session range breakout EA for MT5. My own EA park runs on real accounts; my signals are listed in - Versão: 1.0
- Ativações: 10
Second Wind trades the Turnaround Tuesday strategy, one of the oldest and best documented calendar effects in stock indices. Weekend news often pushes indices lower on Monday. When that happens while the index is already in a pullback, Tuesday has tended to recover. Second Wind buys on Monday while price is below its moving average and sells at a fixed time on Tuesday.
At most one trade per week and symbol. Long only. No grid, no martingale, no averaging.
How it trades
- Monday: from the start time (for example 01:05 server time) Second Wind checks the index once per minute for the whole day.
- Filter: it buys only while the bid is below the moving average (default: 20-day SMA, current bar). A Monday that starts above the MA can still trigger later in the day, as soon as price falls below it.
- Hold: one long position, held overnight into Tuesday. Stop loss, take profit, break-even and trailing stop are optional.
- Tuesday: the position is closed at a fixed time, for example 23:50. Then Second Wind waits for the next Monday.
Open day and close day are inputs, so you can also test Monday to Wednesday or any other weekday pair.
Position size your way
- Fixed lots or lots per x money (for example 0.01 lot per 1,000 of balance).
- % of the balance or a money amount: with a stop loss this is the amount lost at the stop. Without a stop loss it sets the position value: 50 % means a position worth half of the balance.
- The lot size respects the broker's minimum and maximum lot, the symbol volume limit and the free margin, so an order is reduced rather than rejected. On start the Experts log shows the lot size, the position value and what a 10 % index drop would cost.
Features
- All inputs of the classic Turnaround Tuesday EA: wait for a new daily high, four volume modes, TP / SL / break-even / trailing stop in % or points, start and close time, open and close day, moving average filter (timeframe, periods, method, applied price)
- Extra entry filters: red day (only below the previous day's close), latest entry time, maximum spread
- Restart-safe: after a terminal restart it knows whether this week's trade was already made and never buys twice
- Never more than one position per chart; works on hedging and netting accounts
- Checks the symbol's trading sessions on start and warns if the start or close time lies outside them
- Entry and exit drawn on the chart
- Live dashboard: week status, turnaround window with countdown, distance to the moving average, next lot size and position value, open position, today's result and track record
- Three dashboard themes, compact mode, any chart corner
Ready-made set files
- DE40 (DAX): Monday 09:05 – Tuesday 22:55, SMA 40
- US30 (Dow Jones): Monday 01:05 – Tuesday 23:50, SMA 25
- USTEC (Nasdaq-100): Monday 01:05 – Tuesday 23:50, SMA 9
- US500 (S&P 500): Monday 01:05 – Tuesday 23:50, SMA 20
All sets: position value 0.5 × balance, no stop loss, server time UTC+2 in winter / UTC+3 in summer. Other server times: shift the start and close time by the difference. The set files are available in the comments section.
MetaTrader 5 backtests with real ticks
Strategy Tester, "Every tick based on real ticks", IC Markets data, January 2025 – October 2026, 20,000 USD, fixed 0.1 lot, the set values above:
- US30, SMA 25: 37 trades, 73 % winners, profit factor 2.30, +732.55 USD, maximum equity drawdown 1.58 %
- DE40, SMA 40: 34 trades, 74 % winners, profit factor 3.03, +632.67 USD, maximum equity drawdown 0.97 %
- USTEC, SMA 9: 50 trades, 58 % winners, profit factor 2.16, +575.86 USD, maximum equity drawdown 1.12 %
A short and strong period for index dips — the 21-year check below is the more sober picture.
Long-term check (21 years, four indices)
Python simulation on daily data (index futures for the US indices, DAX cash index), January 2005 – October 2026, the set values above, position value 0.5 × balance, compounding. Spread and one night of swap (0.03 % of the position value) deducted on every trade.
- DE40, SMA 40: 446 trades, 57 % winners, profit factor 1.39, about 2.5 % a year, maximum drawdown 9.0 %
- US30, SMA 25: 468 trades, 58 % winners, profit factor 1.50, about 2.7 % a year, maximum drawdown 9.6 %
- USTEC, SMA 9: 588 trades, 56 % winners, profit factor 1.42, about 3.2 % a year, maximum drawdown 13.2 %
- US500, SMA 20: 469 trades, 59 % winners, profit factor 1.48, about 2.9 % a year, maximum drawdown 11.1 %
What the test also shows: the moving average filter beats "buy every Monday" on all four indices (profit factor 1.16–1.33 without it). Any MA length from 5 to 200 days stays profitable, so the result does not hang on one lucky setting. Both halves of the test period are profitable on every index; with doubled costs the profit factor is still 1.34–1.43. The returns are modest because the position is only half the balance and the EA is in the market about one day a week. More size means more return and a proportionally deeper drawdown.
Recommended use
- Stock index CFDs with low spreads and swaps. The position is held for one night, so the long swap is part of the cost.
- A hedging account is recommended. On a netting account Second Wind shares one position per symbol with every other EA; it then skips the week instead of buying into someone else's position.
- The indices move together. If you run several index charts on one account, their risk adds up almost one to one; reduce the position value per chart.
- Test every symbol in the Strategy Tester with your broker's data, ideally "Every tick based on real ticks", and on a demo account before trading live.
Trading CFDs involves a high risk of losing money. Past results, including backtests and simulations, do not guarantee future results. The default inputs and set files are examples, not recommendations.
