← Academy home
Lesson 4 of 12 · Reading the Market

Candles & Timeframes

Every chart your bot reads is made of candles — little four-number stories. Learn to read one candle, then learn how the same market looks completely different at different zoom levels.

4.1One candle, four numbers

A compresses everything that happened in one time slice — one hour on an H1 chart — into four prices: where it opened, the highest and lowest it reached, and where it closed.

HIGH CLOSE OPEN LOW bull — buyers won wick — visited, but rejected body — open→close, where it settled OPEN CLOSE bear — sellers won
Green (bull): closed above its open. Red (bear): closed below. The wicks show the full battlefield; the body shows who held the ground.

The colours tell you who won the hour; the shapes tell you how. That's the next stop.

4.2Four candle stories worth knowing

You don't need the fifty-name candlestick zoo. Four shapes cover most of what candles can honestly tell you:

conviction big body, tiny wicks — one side dominated indecision tiny body, long wicks — nobody won (doji) rejection long lower wick — sellers pushed down, got thrown out exhaustion long upper wick — rally tried and failed
Bodies = conviction. Wicks = rejection. Everything else is combinations of these two facts.

One honest warning: a single candle is a whisper, not a verdict. Strategies (and your bot) act on candles in context — several bars confirming, in the right regime. That's why IronGate supports multi-bar confirmation before it treats a signal as real.

⚙ Candle reader — type any four prices

—

4.3Timeframes — the same market at different zooms

A is just how much time each candle swallows. Nothing about the market changes — only your zoom level. One day is one candle on D1, 24 candles on H1, 96 on M15:

D1 one day, one candle H1 — the same day, 24 candles same day — up close, a bumpy climb full of red hours
Zoom out: smooth story. Zoom in: noise and counter-moves. Both are true; they answer different questions.

4.4The timeframe trade-off

There is no "best" timeframe — only a trade-off you should be able to recite:

The spectrum

Lower timeframes (M1–M15): many signals, fast feedback — but more noise, more false moves, and the eats a bigger share of each smaller win.

Higher timeframes (H4–D1): cleaner, more reliable moves and cheap-relative spreads — but few trades, slow feedback, and wider stops (which, from Lesson 1, just means smaller lots — not more risk).

H1 — where your bot's example card lived — is the classic middle ground: enough signals to learn from, calm enough to trust. But IronGate doesn't guess this: it measures which timeframe historically worked best per pair and strategy via testing, exactly as you saw on the Lesson 3 card ("H1 beat H4 and D1").

In your system

MT5 streams bars for M1–D1; the data loader (core/data_loader.py) fetches them, and strategies only ever act on closed candles — a candle still forming can change its mind, so acting on it is trading a rumour. This is also why the bot "waits for the next fully-closed bar" in its live loop.