Build a strategy

Describe a trading strategy in plain language and Alphio AI assembles the rules, scaffolding, and entry/exit logic for you.

Find it at: /chat (just ask Alphio AI) or /agentic-trading → New Trading Task

Available on Pro and Elite plans.

---

1. What to expect

• Plain-language input — describe entry rule, exit rule, ticker, and timeframe in one sentence
• Alphio AI generates a strategy spec — structured representation of indicators, conditions, sizing, and timeframes
• You review before deploying — the spec surfaces as a confirmation card with every parameter editable
• Paper-test first — deploy to paper mode for at least one week before live capital
• Same execution rails as Auto Trade — the resulting strategy runs through the standard Monitor / Event / Periodic task surfaces
• Account-level guardrails apply — Strategy Settings (risk tolerance, max position, daily drawdown) constrain the AI-generated rules

---

2. Example prompts

Open /chat and describe the strategy — Alphio AI translates each one into a structured spec:

• "Build a strategy that goes long BTC on the 1-hour chart when 9-EMA crosses above 21-EMA. Exit on RSI > 70 or 3% drop from entry."
• "Long NVDA on 4h breakout above 20-day high with above-average volume."
• "Short ETH on 15-min when MACD crosses below signal and price < 200-EMA."
• "Long QQQ on daily close above 50-EMA, exit on 5% gain or close below 50-EMA."
• "Mean reversion: long SPY on 1h RSI below 30, exit when RSI crosses 50."

Alphio AI fills in any missing parameters with sensible defaults and asks for confirmation on ambiguous ones (e.g. position size, stop-loss).

---

3. How it works

1. Describe the strategy in chat. Include at minimum: ticker, direction (long/short), entry condition, exit condition, and timeframe.
2. Alphio AI parses intent — extracts indicators (EMA, RSI, MACD, volume), thresholds, timeframes, and order direction.
3. Generates a strategy spec — a structured representation of the rule set, mapped onto the Monitor / Event / Periodic task model:
   • Indicator-driven entries / exits → Monitor task
   • Earnings / macro-event entries → Event task
   • Recurring buys (DCA-style) → Periodic task
4. Confirmation card — Alphio AI surfaces the parsed spec with every field editable: ticker, entry, exit, stop-loss, position size, max run duration.
5. You review and deploy — to paper mode first, then live.

The same backend (/v1/tradeagent/tasks/*) powers both AI-built strategies and hand-configured Auto Trade rules, so monitoring, execution log, and notifications behave identically.

---

4. Strategy review before deploy

Before tapping deploy, run through this checklist on the confirmation card:

• Ticker and venue match your intent — "BTC" could route to spot or perpetuals; confirm which
• Entry condition is unambiguous — "9-EMA crosses above 21-EMA" is clear; "when momentum is strong" is not
• Exit rule has a hard cap — combine indicator exits with a percentage stop-loss as a backstop
• Position size respects Strategy Settings — the AI defaults to a fraction of max per position from your strategy settings
• Run duration is finite — max_run_duration_days set (default 60) so the strategy self-expires
• Trigger count cap is set — max_trigger_count prevents an over-active rule from blowing through Credits

If anything looks off, edit the field on the card before deploying.

---

5. Backtesting before live

Deploy to paper mode first. Paper mode runs the AI-generated strategy against live market data without sending real orders. See Paper vs live trading for the full paper-vs-live checklist — including the 20-closed-trade minimum, drawdown coverage, and signs of rule misfire to watch for.

Common failure modes that paper mode surfaces:

• AI parsed the timeframe wrong (intraday rule running on daily, or vice versa)
• Entry fires far too often (rule is too loose)
• Entry never fires (rule is too tight)
• Stop-loss too tight relative to typical asset volatility — strategy gets stopped out repeatedly

Tune the prompt, re-deploy, and re-test until paper performance matches your expectation.

---

6. Deploying

Once paper-tested, flip the strategy to live mode. It joins your running book on the Active strategies monitor — same performance curve, execution log, and AI_TRADE notifications as any other Agentic Trading or Auto Trade strategy.

Re-size on the first live deployment: start at 10–25% of the paper allocation. Scale up only after live behavior matches paper across at least 10 closed trades.