The price of the exact second
On the announcement's second the book empties and the spread opens. The price an order fills at in that instant is not the price the history holds — and that difference is a large part of each operation's result.
Tuurt Trading · In measurement
Heimdall
A news robot for MetaTrader 5. Seconds before a scheduled high-impact announcement — US data — it places two pending orders, one above and one below the price. The announcement moves the market, one of them fills and the other is cancelled. It never guesses direction: the announcement picks it. Its edge is not demonstrated, and this page explains why a simulation cannot demonstrate it.
In measurement
2 live operations
High-risk instrument
MetaTrader 5
- Platform
- MetaTrader 5MQL5, Windows terminal
- Market
- GoldXAUUSD, on the announcement's second
- Model
- Two pending ordersThe announcement picks the side
- Live operations
- 2Both winners · +7.79 USD
Live record
Two operations. Both winners. +7.79 USD.
That is Heimdall's entire live record: 2 closed operations, both winners, +7.79 USD on one of our real accounts. Two operations.
Two wins in a row do not separate a good robot from a lucky one, and the amount says the same thing: it is under eight dollars. It is here because it is the number we have, not because it means something.
How it works
It holds no opinion on the data. It stands on both sides and waits.
A high-impact macro announcement moves gold within seconds, and nobody knows which way before it lands. Heimdall does not try: it places both orders and lets the announcement answer.
Seconds before the announcement, two pending orders
One above the price and one below, placed before the data lands. There is no indicator, no bias, no forecast of the figure.
The announcement picks the side
The market moves, one of the two orders fills and the other is cancelled. The direction is not the robot's decision: it is the announcement's result.
It adds to the winning side while the move continues
If the impulse keeps going the way it entered, it stacks positions on that same side. It adds to what is going for it, never to what is going against it.
It closes by chasing the peak as it fades
A news move burns out; the exit trails that peak while it fades, rather than waiting for a fixed target that may never come back.
The point of this page
Its edge is not demonstrated, and a simulation cannot demonstrate it.
This is not a polite caveat: it is a property of the strategy. The two things that define its risk are precisely the two a backtest does not contain.
The market that lurches both ways
The other way to lose is the price lurching in both directions and hitting both orders: it fills upward, turns around, fills downward, and both lose. A clean move is the good case; the whipsaw is what decides whether the strategy is worth anything.
A backtest of this strategy is a test in which the ways of losing do not exist
Stored historical data preserves the path of the price and loses the impact of the order. It reproduces where the market went; it does not reproduce what happens to your order when it enters on exactly that second.
You can see it in the simulation's own output: across 186 simulated announcements there was not one event without a breakout and not one whipsaw. That is how you know the simulation is not the market — in a real market both happen, and they are exactly what loses money.
How an honest number is obtained
An observer mode: it decides everything and sends no order.
That is why Heimdall ships with an observer mode: it takes every decision and records all of it, on the real announcement and on the real second, without sending a single order.
That is the honest route to a number, and it is the route we are taking. Once enough announcements are recorded live, whatever gets published here will come from those and not from a simulation in which the whipsaw does not exist.
Risk disclosure
Heimdall is in measurement. Nothing on this page is an invitation to trade, and it is not investment advice.
- Its edge is not demonstrated. The live record is 2 closed operations, both winners, +7.79 USD. Two operations support no claim about its profitability.
- A simulation cannot demonstrate it. Stored history preserves the path of the price and loses the impact of the order, so a backtest of this strategy contains neither the slippage of the announcement's second nor the whipsaw that hits both orders.
- Across 186 simulated announcements there was not one event without a breakout and not one whipsaw. That is not a good result: it is the sign that the simulation does not represent the market.
- It trades at the worst moment of liquidity in the day. On the announcement's second the spread opens and the execution price can land a long way from the one requested.
- It stacks positions on the side going its way, so exposure grows during the move and an abrupt reversal finds it open.
- It trades a high-risk instrument. Leveraged gold via CFDs. It can cost the entire capital in an account.
- Nothing published here is investment advice or a promise of returns. No past result, measured or projected, guarantees a future one.
- Trading leveraged CFDs carries a high risk of losing money rapidly, including the entire deposit.
Shall we talk with these numbers in front of us?
Write to us if you want to follow it. What exists today is two operations and an observer mode recording; when there is more, we publish it here.
trading@tuurt.com