An indicator is not a strategy
An indicator turns price into a number. A strategy is the rule that decides what to do with that number, including the entry, the exit, and the stop loss.
A technical indicator turns a price series into another number. A strategy is the decision that uses that number: when to enter, when to exit, and how the position is protected. The number can be ready for a long time and still never open a trade.
What you have when you only have the number
A moving average, a momentum oscillator, or a comparison between two averages is a function of candles. It answers "what is the value now?" It does not answer "should a position open, and how large should it be?"
That missing half is why two people can share one indicator and hold opposite positions. One treats a cross above as an entry. Another treats the same cross as a place to exit. The calculation did not choose. The rule did.
A named indicator also leaves operating choices unspecified. Which candle is allowed to update the value. What happens while the candle is still forming. Where the stop sits. What cost would erase the edge. A forming candle is not a closed candle is the timing half of that list. This page is the decision half.
What Tradee will store in a component
A component is a versioned piece of graph you can drop into more than one strategy. The palette matches the calculation tools on a strategy board: data, indicators, math, comparisons, and candlestick tests. It stops before actions. A component cannot enter, exit, set a stop, set a target, or contain another component.
The component declares inputs and outputs. The strategy pins a specific component version and wires those ports. If you delete the component, a strategy that still pins it will not save. The pin is the point: shared arithmetic stays in one place, and the decision stays on the strategy.
This is stricter than a notebook where a plotted line and an order sit in the same cell. The split is visible. Zero action nodes means the fragment cannot trade, even if the reading looks like a signal.
The entry that makes it a strategy
A strategy has boards for entry, exit, and protection. Structural save allows a board that is still unfinished, so you can leave work in progress. A board that never reaches an entry action still cannot open a position. Tradee treats enter_long or enter_short as that action. Exit and stop logic can be rich and still leave the strategy inert if nothing is allowed to enter.
What is automated trading? puts this in the wider system. A signal is not an order, and an order is not custody. An indicator is one step earlier than the signal. It is material for a signal.
A worked example with one reading
The figures in this article's images are a worked example, not a recorded backtest. The reading is defined here so the count is checkable. It is not a claim about a published indicator's formula.
Both sides "use the averages." Only the strategy can open. The component can be reused by a second strategy that fades the same cross, with the opposite entry. One reading, two rules, two different bots if you bind them that way. Strategy, bot, and execution is the next split, after the rule exists.
The averages also need a warmup. Until 30 candles exist, the long average is not defined. A strategy that enters on an undefined reading is a different bug from a strategy that has no entry node. Count the lookback when you test. Do not treat the first bars as evidence.
What to write down before you test
A testable strategy states every choice the indicator left open.
- The reading, including periods and the price input.
- The entry condition, in words a second person can apply.
- The exit condition, separate from the entry.
- Whether stops and targets are part of the rule.
- The candle that is allowed to update the decision.
- The size rule and the cost assumptions.
If item 2 is missing, you have a chart study. If items 2 through 6 are present, you have something a backtest can accept or reject. Why win rate is not enough is about reading the result after that test exists.
FAQ about technical indicator
Is a crossover a strategy?
Only after you add the actions. "Short average above long average" is a reading. "Enter long when that reading turns on, and exit when it turns off" is a strategy. The second sentence can still be a weak strategy. It is at least a complete one.
Should every indicator become its own component?
No. A component pays off when more than one strategy must share the same calculation and stay in sync. A one-off comparison can live on the strategy board. Moving it into a component does not make it a better signal.
Methodology and sources
The component boundary is Tradee's authoring rule: calculation nodes are allowed, action nodes are not, and components do not nest. The 10-candle and 30-candle averages are a constructed reading with outputs 0 and 1. This article does not evaluate that cross. It shows where the calculation ends and the decision starts. This article is educational and does not provide personalized investment or trading advice.
This article is for education. It is not financial advice. Trading can lose money. Past results do not guarantee future results.