Feature
Backtesting Lab: Learn Why Strategies Fail
The TradingLead Backtesting Lab at /lab is a set of short, interactive notebook lessons that show why trading strategies fail, covering Indian trading costs, overfitting, drawdown, sample size, stop-losses and walk-forward testing. Lessons run on offline sample data or live NIFTY 50 daily candles. Practice Labs at /lab/practice then guide you through the real Backtesting panel and Indicator Studio on synthetic data.
What it does
The Backtesting Lab is TradingLead's hands-on learning area for strategy testing. Instead of reading theory, you run a strategy yourself and watch each common mistake change the numbers: chasing a perfect screenshot, ignoring costs, over-tuning settings, trusting win rate, or skipping out-of-sample checks.
It is aimed at beginner and intermediate traders in India who want to understand backtests before relying on one. The examples use NIFTY and Indian cost structures.
How it works
Each lab is a short notebook: a sequence of steps with controls (dropdowns, sliders, toggles), a result panel and a side chart showing price with trades or the account equity. Steps unlock in order, and some include quick-check questions. You can switch between:
- Sample data: built-in NIFTY-like candles that work offline, so everyone sees the same numbers.
- Live NIFTY 50 daily: roughly the last 600 days from TradingLead's data (may need sign-in).
Progress is saved in your browser, and Start over clears it.
Labs available
| Lab | Level | Approx. time | What you test |
|---|---|---|---|
| Why your strategy loses money | Beginner | ~20 min | A 9/21 EMA crossover from a cherry-picked window to every candle, Indian costs and slippage, trending vs sideways markets, the overfitting trap and position sizing. |
| Read a backtest report | Beginner | ~15 min | Why win rate alone misleads, expectancy and profit factor in rupees, max drawdown and recovery, sample size and beating buy-and-hold. |
| Stop-loss and exits, tested | Intermediate | ~20 min | No stop vs fixed stops, how far winners dip (MAE), too-tight vs too-wide stops, exit types and parameter plateaus. |
| Can you trust this result? | Intermediate | ~20 min | Locked rules on hidden data, walk-forward tuning, good-neighbour settings, other markets and stress tests with missed trades and extra slippage. |
Trading psychology and Technical analysis, tested are listed as coming soon.
Practice Labs
At /lab/practice, guided tasks run inside the real TradingLead chart, Backtesting panel and Indicator Studio, with a small coach alongside:
- Use the Backtesting panel: set up, run, inspect a trade and compare costs.
- Backtest your idea with free AI: turn your rules into a prompt, review the answer, backtest it and compare one change.
- Use Indicator Studio: compile, apply, change an input, save and reopen.
- Build with free AI: generate a prompt, review the answer, compile and apply.
Practice Labs use synthetic daily prices, not live NSE data. They place no broker orders and don't change your normal workspace.
Key capabilities
- Interactive, step-by-step notebooks with live-updating controls.
- Indian equity cost modelling: delivery vs intraday, discount, flat ₹20 and full-service broker plans, STT, stamp duty, exchange and SEBI fees, GST and slippage.
- Out-of-sample, walk-forward, plateau and stress-test lessons that mirror the Strategy Optimizer.
- A link from the Lab into the real Backtesting panel on a NIFTY chart, with the strategy preselected (never auto-run).
How to use it
- Go to
tradinglead.in/lab. - Choose a lab; start with Why your strategy loses money if you are new.
- Pick sample or live data, then run each step and try the controls.
- Finish with a Practice Lab to repeat the workflow in the real app, then follow Backtest a strategy.
Limitations
- Everything in the Lab is an educational simulation. It is not investment advice, and past results do not guarantee future returns.
- Cost figures are approximate published rates for learning, not exact broker charges.
- Live data is limited to NIFTY 50 daily candles. Practice Labs use synthetic prices.
- Progress lives in your browser storage only.
Steps
- Open https://tradinglead.in/lab and choose a lab.
- Pick sample (offline) or live NIFTY 50 data.
- Run each step in order, adjust its controls and answer the quick checks.
- Use Practice labs to try the same ideas in the real Backtesting panel.
Limitations
- Educational simulations only; not investment advice, and past results do not guarantee future returns.
- Lesson cost models are approximate Indian equity charges for learning, not contract-note accurate.
- Live lab data covers NIFTY 50 daily candles only and may need sign-in; Practice Labs use synthetic prices.
- Progress is stored in your browser; clearing it or switching browsers loses progress.
- Trading psychology and Technical analysis labs are listed as coming soon.
Frequently asked questions
- Do I need to sign in to use the Backtesting Lab?
- The /lab lessons open in the browser without the sign-in guard, and the offline sample data works without an account. The live NIFTY 50 data option may need sign-in.
- What will I learn in the Backtesting Lab?
- How a cherry-picked screenshot differs from a full-history test, how Indian brokerage, STT and slippage eat into returns, why over-tuned settings fail, how to read win rate, expectancy, drawdown and sample size, how to test stop-losses, and how to judge robustness with hidden data and walk-forward windows.
- Does the Lab use real market data?
- Each notebook can use built-in sample NIFTY-like data that works offline, or live NIFTY 50 daily data (about the last 600 days) from TradingLead. Practice Labs use synthetic daily prices, not live NSE data.
- Are the Lab's Indian trading costs exact?
- No. They use approximate published NSE equity charges (brokerage plan, STT, stamp duty, exchange and SEBI fees, GST) for learning. They are not meant to reconcile with a broker's contract note.
- Can I move a Lab strategy into the real backtester?
- Yes. The Lab links into the real Backtesting panel on a NIFTY chart, where the strategy is preselected for you to run. It never runs automatically.
Related
Last updated: