How this app works
TraderPal turns public market data into plain-language analysis — and shows its work at every step. Nothing here is a black box. Here's the whole pipeline, from raw data to the report you read.
The pipeline, step by step
1 · We collect public data
On a schedule, background tasks pull prices and history from free, public sources — crypto exchanges and aggregators, currency and metals APIs, oil data, and Iran's bazaar rates. Nothing here requires you to log in or share anything.
2 · We store it — we never fetch on your click
Everything is saved to our own database first. When you open a page, you're reading from that database, not triggering a live call to anyone. That's why the app is fast and keeps working even if a data source is briefly down.
3 · We compute indicators and a signal
From the price history we calculate standard technical indicators (trend, momentum, overbought/oversold, and more) and blend them into one signal. We always show which indicator pushed the signal which way — the reasoning is never hidden.
4 · We measure the risk
For each asset we work out how much it typically swings, a likely bad day, and its worst drop over the last year, then combine them into a 1–10 risk score. Every formula is published on the risk-methods page.
5 · We test the signal honestly
We replay our own signal on past data — with trading costs included — and compare it to simply buying and holding. When our signal would have lost money, we say so, in the same place and with the same weight as a win.
6 · We read the news (optional AI)
When an AI model is switched on, it reads recent financial headlines, scores the mood, and links them to the right assets. If no model is available, the app still works — it just skips the mood score. The AI never invents numbers: every figure it writes is checked against our data first.
7 · We write it up in plain language
Finally we produce a short report that ties the signal, risk, backtest, and mood together in everyday words — always ending with the reminder that this is education, not advice.
Where the data comes from
These are the public sources and open-source tools behind the numbers. We use free tiers and respect each source's limits.
The open-source tools
Plain-language glossary
- Signal
- Our single read on an asset — leaning up, leaning down, or no clear direction — blended from several indicators. It's a summary of the charts, not a recommendation to buy or sell.
- Likely bad day (VaR)
- A rough size for a normal bad day. On about 5 days out of 100, a one-day loss bigger than this is expected. It doesn't tell you how bad the rare, really bad days get.
- Worst drop (drawdown)
- The biggest fall from a high point down to the following low over a period. It's the number that best captures how much pain holding the asset could have caused.
- Daily swing (ATR)
- How much an asset typically moves in a single day, on average. Bigger swings mean bigger risk and bigger potential moves.
- RSI
- A 0–100 gauge of whether something has been bought or sold hard recently. High can mean overbought, low can mean oversold — but neither is a guarantee of a turn.
- MACD
- A momentum indicator that compares a faster and a slower average to show whether upward or downward pressure is building.
- ADX
- A measure of how strong a trend is, regardless of direction. When it's low, the trend is weak — so we hold our signal at neutral.
- Price band position
- Where today's price sits inside its recent normal range (Bollinger Bands). Near the top or bottom can hint the move is stretched.
- Return vs. risk (Sharpe)
- How much return a strategy earned for the amount it bounced around. Higher is better — it means steadier gains, not just bigger ones.
- The gap (spread)
- The distance between two prices for the same thing — here, Iran's free-market dollar versus the official Central Bank rate. A wide gap signals pressure on the rial.
- News mood
- A score from about −1 (very negative) to +1 (very positive) that an AI model gives recent headlines, so you can see whether the news is leaning bad or good.
- Backtest
- Replaying a strategy on past data, with trading costs included, to see how it would have done. Past results don't guarantee the future — but a strategy that failed the past is a real warning.
Our honesty rules
No trade is ever placed here — this app can't buy or sell anything. We don't give personalized advice. Every number can be traced to a source or a published formula. And when we test our own signal and it loses, we show the loss just as plainly as a win. That honesty is the point of the whole project.