Usage guide for alphabench
Your guide to AI-powered strategy development, backtesting, and portfolio management on Indian markets
Usage guide for alphabench
Your guide to AI-powered strategy development, backtesting, and portfolio management on Indian markets
alphabench is an AI-powered quantitative trading platform for Indian markets (NSE/BSE). Describe a trading idea in plain language, and the platform will develop, backtest, and refine it into a deployable strategy. This guide covers the key workflows from first chat to live paper trading.
Strategy Development via Chat
1. Start a Conversation
Describe your trading idea in the chat input. You can be as specific or as broad as you like. Examples:
- "Build a momentum strategy on Nifty 50 stocks using RSI crossover"
- "I think banking sector will outperform in a rising rate environment"
- "Backtest a mean-reversion strategy on RELIANCE with Bollinger Bands"
- "Create an intraday options strategy on Bank Nifty using straddles"
2. Planning Phase
The AI planner will engage you in a conversation to refine your idea. It can look up instruments, fetch live market quotes, and research sector data before building a strategy. For broad macro theses (e.g. "rate-sensitive sectors will rally"), it automatically decomposes your idea into multiple instrument-level strategies.
3. Execution Phase
Once the plan is ready, the system hands off to the execution engine which backtests the strategy across historical data. You'll see real-time progress as each tool runs — data fetching, signal computation, backtesting, and result analysis.
4. Review and Iterate
Results appear with equity curves, trade tables, and performance metrics (Sharpe, CAGR, max drawdown, win rate). Ask follow-up questions to tweak parameters, change instruments, or try different approaches — the AI retains context from your conversation and previous strategy results.
Equity Strategies
Instruments and Universes
The platform covers NSE-listed equities and indices. Strategies can target individual instruments or entire universes (Nifty 50, Nifty Next 50, Nifty Bank, Nifty Midcap 150, and more). Universe-wide strategies sweep across all constituents and rank results.
Strategy DSL
Behind the scenes, the AI writes strategies in a compact expression language. You don't need to learn it — the AI handles translation from your natural language description. Supported indicators include SMA, EMA, RSI, MACD, Bollinger Bands, ATR, ADX, Supertrend, VWAP, and many more.
Research Jobs
For broad exploration, you can run research jobs that sweep an entire universe across multiple strategy templates (momentum, mean-reversion, breakout, volatility) with parameter variations. These run in the background and notify you when complete.
Options Strategies
Intraday Options
Build and backtest intraday options strategies on Nifty and Bank Nifty. The platform supports straddles, strangles, spreads, and custom multi-leg positions with time-based or signal-based entry/exit rules.
Adaptive Strategies
Adaptive strategies dynamically switch between spread types based on intraday signals (e.g. switch from iron condor to directional spread when volatility spikes). The engine reconstructs historical option chains for accurate backtesting.
Strike Selection
Define strike selection rules by delta, moneyness (ATM, OTM, ITM), or fixed offset. The platform handles expiry calendar logic, chain snapshots, and proper symbol resolution automatically.
Portfolio Management
Building a Portfolio
Combine multiple strategies into a portfolio using the portfolio wizard. Add equity strategies, options strategies, or manual holdings (buy-and-hold positions you already have). Set target allocations and the platform will track combined performance.
Goal-Based Recommendations
Not sure what strategies to include? Use the goal-based mode — specify your risk tolerance (conservative, moderate, aggressive) and time horizon (short, medium, long), and the platform recommends strategies from its memory of past backtests.
Optimization and Analytics
Optimize portfolio weights using Monte Carlo simulation or mathematical optimization (max Sharpe, efficient frontier). View risk analytics, benchmark comparisons, factor exposure, and return attribution. The platform runs Monte Carlo projections to visualize potential future outcomes.
Monitoring
Active portfolios are tracked daily with NAV updates and drift alerts. Strategy health monitoring scores each strategy on Sharpe, returns, drawdown, and win rate, flagging any that need attention.
Paper Trading
Deploying a Strategy
Once you're satisfied with a backtest, deploy it to paper trading directly from the chat. The platform connects to live market tick data and executes your strategy signals in real-time with simulated orders — no real capital at risk.
Live Monitoring
Monitor deployed strategies via the paper trading dashboard. Track open positions, P&L, order fills, and signal triggers as they happen during market hours. Pause or stop strategies at any time.
Supported Modes
Paper trading supports both daily equity strategies and intraday options strategies. Intraday strategies evaluate signals on 1-minute or 5-minute bars constructed from live ticks.
Tips for Best Results
Be Specific
Instead of "make me a strategy," try "build a momentum strategy on Nifty Bank stocks using 20-day vs 50-day EMA crossover with RSI confirmation." Specific queries produce better strategies on the first iteration.
Iterate Progressively
Start with a simple strategy, review the backtest, then refine. Ask to tighten stop losses, add filters, change timeframes, or test on different instruments. The AI remembers your conversation context and previous results.
Use Research Jobs for Exploration
When you want to scan broadly (e.g. "which Nifty 50 stocks respond best to mean-reversion?"), use a research job rather than testing stocks one by one. It runs all combinations in the background and returns ranked results.
Paper Trade Before Committing
A strong backtest doesn't guarantee future performance. Deploy to paper trading for at least a few sessions to verify that the strategy behaves as expected with live data before considering real capital.
Important Disclaimer
All analysis, strategies, and results on alphabench are for educational and research purposes only. They do not constitute financial advice.
Past performance and backtesting results are not indicative of future returns. Trading in securities and derivatives involves substantial risk of loss. Always conduct your own due diligence, understand the risks involved, and consult with a qualified financial advisor before making investment decisions.