strategy-dev

Conditional Bot — your first regime-switching rule set

Worked example: BTC/USDT 1h reference → pause in Strong Downtrend → Grid Bot in Range → Smart Trade in Uptrend

10 min · intermediate

What you'll have when finished

  • Build a 4-rule conditional strategy that adapts to BTC regime
  • Understand the 6 regime classifications and which strategies fit each
  • Order rules correctly (most-restrictive first, fallback last)
  • Read the live regime panel and verify the classifier matches your view
  • Know when to pause vs delete a conditional orchestrator

Before you start

  • Conditional Bot does NOT trade directly. It activates/pauses OTHER bots. If those inner bots are losers, this just loses faster.
  • Get your inner strategies (Grid, DCA, Smart Trade) working in MANUAL mode for 2-4 weeks BEFORE wiring them into a Conditional Bot.
  • Auto-evaluation runs every 5 minutes — not instantaneous. Don't expect tick-level reactions.
  • High Volatility regime catches most retail traders off guard. Default to PAUSE in High Volatility unless you have a specific high-vol strategy.

Walkthrough

  1. Decide if Conditional Bot is actually right for you yet

    **Pre-flight check.** Conditional Bot is for traders who have ALREADY run individual bots successfully. If you can't answer these, come back later: 1. Have you deployed Grid Bot at least 3 times with profitable results? 2. Have you run Smart Trade with TP ladders at least 10 times and understand which setups work? 3. Have you completed at least 5 DCA Safety cycles? 4. Can you describe what "Strong Downtrend" looks like on a chart without thinking? If you said NO to any of these, master the inner bots FIRST. Use [Choose the right bot](/learning-library?guide=choose-the-right-bot) to pick your starter and run it for 2-4 weeks. Then come back here. **The most common Conditional Bot failure:** wiring up untested inner strategies and letting the orchestrator amplify their problems. Get the strategies right; then automate the switching.

    Success criteria: Already comfortable with 2-3 inner bot types · Have proven each inner bot profitable for its target regime

  2. Open Conditional Bots + start a new strategy

    Navigate to **CEX Trade → Conditional Bots**. You'll see two tabs: **List** (your existing strategies) and **Create New**. Click Create New. The form starts with: - **Name:** e.g. "BTC-Regime Master" - **Reference Exchange + Symbol:** Coinbase + BTC/USDT (the default; almost always the right choice) - **Timeframe:** 1h (the default; balanced between responsiveness and noise) **Why BTC/USDT 1h:** - BTC drives the broader crypto market — regime detection on the leader is more reliable than on followers - 1h smooths out the noise of 15m/30m while staying responsive enough to catch regime shifts within hours You can change these (BTC/USDC, ETH/USDT for ETH-led orchestrators, etc.), but for your first deploy, defaults work.

    Success criteria: Strategy name set · Reference symbol = BTC/USDT · Timeframe = 1h

  3. Understand the 6 regimes before writing rules

    The classifier outputs ONE of 6 regimes every 5 minutes: **🟢 Strong Uptrend** — high trend strength + RSI > 70 + momentum bullish. Best for trend strategies (Smart Trade with trailing TPs, momentum-following bots). **🟢 Uptrend** — moderate trend bullish but not extreme. Same strategies as Strong Uptrend but with tighter risk parameters. **🟡 Range / Sideways** — no clear trend, low-to-moderate volatility. **PERFECT FOR GRID BOT.** Range strategies and mean-reversion thrive here. **🔴 Downtrend** — moderate bearish bias. DCA Safety (averaging into the dip) or pause depending on conviction. **🔴 Strong Downtrend** — high trend strength bearish + RSI < 30 + bearish momentum. **PAUSE EVERYTHING** unless you specifically have a short strategy. **🟣 High Volatility** — high Bollinger Band width but unclear direction. Whipsawing markets where most strategies bleed. Default to pause unless you have a specific high-vol strategy. The detail panel will show **Trend Strength** (EMA slope), **Volatility %** (BB width), and **RSI** live so you can verify the classifier matches your visual read.

    Success criteria: Memorized the 6 regimes · Understand which strategies fit which regime

  4. Build Rule 1: PAUSE on Strong Downtrend (most-restrictive first)

    Rules evaluate **TOP TO BOTTOM. FIRST MATCH WINS.** Always order from most-restrictive to least-restrictive. **Rule 1: "BTC Strong Downtrend → Pause All"** - WHEN: regime = Strong Downtrend - THEN: pause all bots This catches the worst-case regime first. If BTC is in a confirmed strong downtrend, NOTHING you have running is going to work well. Pause everything, save capital for the next opportunity. **Why this rule comes first:** if Rule 4 (Range → Grid) and Rule 1 (Strong Downtrend → Pause) both somehow match, you want the Pause to win. Most-restrictive first ensures the safety nets always trigger.

    Success criteria: Rule 1 created · Action = pause all bots

  5. Build Rule 2: PAUSE on High Volatility (the second safety net)

    **Rule 2: "BTC High Volatility → Pause All"** - WHEN: regime = High Volatility - THEN: pause all bots High Volatility is the most dangerous regime for retail traders. Strategies designed for trends OR range BOTH fail in whipsaws. Even good strategies get stopped out repeatedly on noise. **Real example:** August 2024 SOL crash — Bollinger Band width hit 12% in a few hours. Trend was unclear (down then up then down). Grid Bot deployments got whipsawed, Smart Trades stopped out, DCA Safeties triggered safety orders that then fell more. Pausing in High Volatility would have saved most retail traders that week. If you have a specific high-volatility strategy (e.g., wide-range Grid Bot with very wide stops), you can route that here instead. For your first deploy, pause is the right default.

    Success criteria: Rule 2 created · Action = pause all bots (or wide-vol strategy if you have one)

  6. Build Rule 3: GRID BOT in Range, Rule 4: SMART TRADE in Uptrend

    **Rule 3: "BTC Range/Sideways → Grid Bot (tight range)"** - WHEN: regime = Range / Sideways - THEN: activate Grid Bot tight range strategy Range is where Grid Bot SHINES. The strategy of placing N buy/sell orders across a price range PROFITS from oscillation — and Range regimes are by definition oscillating. **Rule 4: "BTC Uptrend OR Strong Uptrend → Alpha Momentum Pro"** - WHEN: regime IN [Uptrend, Strong Uptrend] - THEN: activate Alpha Momentum Pro strategy In confirmed uptrends, momentum strategies catch the move. Smart Trade with trailing TPs is the classic fit. **Rule 5 (fallback): "Always → Pause All"** OR your default strategy This is the safety net. If somehow no regime matches (rare), the fallback kicks in. Pause is the safest default; or use a defensive strategy like DCA Guardian. You now have a 5-rule conditional strategy that adapts to BTC regime: pauses in dangerous markets, runs Grid in calm ones, runs Momentum in trending ones, pauses by default.

    Success criteria: Rule 3 (Range → Grid) created · Rule 4 (Uptrend → Smart Trade) created · Rule 5 (fallback → Pause) created

  7. Activate + verify the regime panel

    Click **Create** to save the rule set, then go to the **List** tab. Your new strategy appears. Click it. The detail panel shows: **Currently:** Range / Sideways (or whichever regime matches BTC right now) **→ Range/Sideways → Grid Bot tight range** (the matched rule + active strategy) **3 live stats:** - **Trend strength:** 0.23 (between -1 and +1; near zero = sideways) - **Volatility (BB %):** 2.4% (moderate) - **RSI:** 52 (neutral) **Click Activate.** The orchestrator is now live. It evaluates every 5 minutes and switches strategies as BTC regime changes. **Verify the classifier matches your visual read:** - Look at BTC/USDT 1h chart on TradingView (or in CEX Trading) - Does it look like a sideways range? Then "Range" should be the regime - Does it look like a clean uptrend? Then "Uptrend" should show - If the regime label disagrees with what you SEE, that's a debug signal — the classifier might be lagging or your view might be wrong. Use Evaluate Now to force a refresh.

    Success criteria: Strategy activated · Regime panel shows live values · Visual confirmation that classifier matches the chart

  8. What to watch for in the first week + when to tune

    **Healthy patterns (first week):** - Switches 1-3 times across the week (regime changes are normal but not constant) - Each switch is followed by a visible regime change on the chart (not just classifier noise) - The active inner bot accumulates fills consistent with its strategy **Warning signs:** - **Switches every few hours** — your regime thresholds are too sensitive. The classifier is reacting to noise, not real regime shifts. Widen the rule conditions. - **Never switches** — your thresholds are too tight or your rules don't cover the actual regime. Add more rules or relax conditions. - **The active inner bot is bleeding** — the regime classifier might be right, but the inner bot for that regime is wrong. Replace the strategy in Rule N. - **Fallback always firing** — your rule conditions don't cover the regimes that actually occur. Add explicit rules. **Review weekly.** The Switch History in the detail panel shows every regime → rule → strategy transition. After 1-2 weeks you'll see patterns — which inner bots actually fired profitably, which got switched away from before they could work, which regimes you over-reacted to. **Tuning is a 4-6 week process.** Conditional Bot rewards patience.

    Success criteria: Reviewed Switch History after 1 week · Identified at least 1 tuning opportunity

What's next

**The "BTC Master" pattern (what you just built):** one orchestrator with BTC/USDT 1h reference, covering the full regime cycle. Most retail users only ever need this. **The "Per-asset" pattern:** separate orchestrators per asset (one for BTC, one for ETH, one for SOL). Useful when assets decouple — ETH can be in Uptrend while BTC is Range. More complex; only worth it if you're actively trading 3+ assets. **The "Timeframe stack" pattern (advanced):** one orchestrator on 1h timeframe controlling Smart Trades, one on 4h timeframe controlling Grid Bots, one on 1d timeframe controlling DCA Safety. The 1d catches the strategic regime, 4h the tactical, 1h the operational. Hedge funds use this. Most retail traders don't need this complexity. **Pause your orchestrator before major economic events** (FOMC, CPI prints, regulatory news). Even good rules don't anticipate news-driven volatility spikes. For automated regime classification + alerts (without auto-switching strategies), use [Market Trends](/market-trend) instead — same engine, just informational.