The Trading Engine — Business Overview

Plain-English guide to how the simulator trades · for the strategy / business team

Subject: the part of the simulator that makes the trading decisions. Audience: no coding knowledge needed. Purpose: a shared understanding so we can review together — what's right, what's missing, what should change.

Contents
  1. In one paragraph
  2. The strategy in everyday terms
  3. The life of a trade
  4. Safety & risk controls
  5. What the engine gives us
  6. Assumptions & limitations
  7. Discussion points
  8. Feedback log
  9. Glossary

1. In one paragraph

The engine is a simulator. You give it a price history for an instrument (e.g. Gold) and a set of strategy settings (a "profile"), and it replays the market one candle at a time, making the same decisions a live trading robot would make. When a strong move breaks out, it places a two-sided "straddle" of pending orders — a ladder of buy orders above the price and a ladder of sell orders below it — then lets the market choose the direction, builds into the winning side, holds the losing side open as a built-in hedge, and banks the whole basket once it is a set percentage ahead. It never touches a real broker — it produces a report (equity curve, trade list, profit/loss, win rate) so we can judge whether the strategy would have worked before risking real money.

2. The strategy in everyday terms

The core idea is "cover both directions at the breakout, then let the market pick the winner."

A useful mental image: it's like backing both sides of a race, then doubling down on whichever horse breaks ahead while your other tickets cap the downside — and cashing the whole book once you're a set amount up.

3. The life of a trade (step by step)

  1. Watch & wait — no breakout, no trade. Filters can also block trading (see §4).
  2. Confirmed breakout — both ladders (7 buy orders above + 7 sell orders below) are deployed instantly around the breakout price.
  3. Direction decided — the first rung to fill sets the "main" direction; the opposite ladder becomes the hedge.
  4. Building & hedging — as price runs, more winning-side rungs activate (averaging up); rungs on the other side fill too and act as the hedge.
  5. Break-even protection — once the main ladder is fully filled and ahead, all its safety levels move to break-even (a small locked profit). Nothing is ever closed at a loss.
  6. Keeping it full (7 + 7) — if a protected rung is closed in profit, the engine re-places that same order at its level when price returns, so both ladders always stay complete.
  7. If it stalls — if a fully-built basket sits without hitting its target for a set time, the engine adds a few extra rungs on both sides (once), to catch a renewed move either way.
  8. Target hit — when total floating profit (after costs) reaches the target %, everything closes and profit is realised.
  9. Cooldown & repeat — after a 15-minute pause, the engine is ready for the next breakout.

4. Safety & risk controls

5. What the engine gives us (the report)

6. Important assumptions & limitations

7. Discussion points for the meeting

Areas where the current behaviour may need a business decision:

  1. Profit target = 5%. Is 5% of the account the right basket target across all instruments and account sizes?
  2. Ladder size & spacing. Are 7 orders per side, 40 pips apart, right across instruments (Gold, Crude, Nasdaq) with their very different price scales?
  3. Fixed vs. adjustable settings. Are we comfortable with the hard-coded items (always-on hedge side, 7+7 refill, stall extension), or should any be dashboard-controllable?
  4. Trend definition. The engine confirms trends on the 4-hour chart from price averages (EMA 40 / EMA 110) + slope + strength. Does this match how the desk actually defines a tradeable trend?
Resolved (team review, 2026-06-17): at every breakout the engine places both ladders — 7 buy orders above and 7 sell orders below — and keeps them full (re-placing any rung closed in profit). The first side to fill is the main direction, the other is the hedge, always the same size. There is no stop-loss, no take-profit, no daily-loss stop, no drawdown limit, and no hedge expiry; a position is never closed in the red, and the basket is banked only at +5% net (after spread, commission and swap).

8. Feedback log

Fill this in during/after the review (print this page or copy the table). The notes come back to the team that maintains the engine for corrections to the engine and/or this document.
#Raised byArea (§)Comment / concernDesired changePriority
1
2
3
4

Overall verdict: ☐ Approved   ☐ Approved with changes   ☐ Needs rework

Appendix — Glossary

Candle / bar
One time-slice of price — open, high, low, close. Here the trend is read on 4-hour candles and entries on 1-hour candles.
Trend
A sustained directional move; here, confirmed on the 4-hour chart by several signals agreeing.
Breakout
Price decisively pushing through the trend line, confirmed on a 1-hour candle close.
Straddle
Placing both ladders at once — buy orders above and sell orders below the price — so either direction is covered.
Ladder / pyramiding / averaging-up
Building a position in steps that trigger only while the trade keeps moving the right way.
Main vs. hedge side
Whichever ladder fills first is the "main" direction; the opposite ladder is the "hedge," held as protection (same size).
Basket
The whole group of related trades from one setup (both ladders), managed and closed together.
Refill (7 + 7)
Re-placing a rung that was closed in profit, so both ladders always keep their full count of orders.
Break-even
Moving an order's safety level to its entry (plus a small buffer) so it can no longer turn into a loss.
Pip
The standard small unit of price movement for the instrument.
Spread
The cost difference between buy and sell prices.
Commission / swap
Per-trade dealing cost and the overnight financing charge for holding a position; both are deducted before the +5% target is judged.
Drawdown
The decline from a previous high point in account value.
Profit factor
Total winnings divided by total losses (above 1.0 is profitable).