Most trading bots fail for the same reason: they pick one strategy and run it in all market conditions. Trend-following in a choppy market. Mean reversion in a trending market. The strategy isn't wrong — the context is.
Here's how to think about combining proven algorithmic strategies with a top-down sector model so the bot knows not just how to trade, but where and when.
The Strategies That Actually Hold Up
Trend-Following
Moving average crossovers (50/200 EMA), MACD (12/26), momentum ranking. Simple and robust across decades of data. Works best in trending markets, bleeds in chop. Every serious systematic fund has some version of this.
Mean Reversion
Bollinger Bands, RSI extremes, pairs trading. Bets that prices snap back to their average. Negatively correlated with trend-following — which is exactly why you want both. When one is losing, the other is usually winning.
Breakout / Range Expansion
Entry on volume-confirmed breaks of consolidation zones. Captures the start of new trends early. ATR-based position sizing keeps the risk clean regardless of how volatile the name is.
Volatility Regime Filtering
VIX or realized volatility thresholds to switch between aggressive and defensive posture. Not a strategy by itself, but it makes every other strategy better. When VIX spikes above 25, reduce size. When it drops below 15, lean in.
Factor-Based Sector Rotation
Rank sectors by momentum, fundamentals, and macro regime. Rotate into leaders, exit laggards. This is where the VisionBoard sector model lives — and it is the layer that tells you where to deploy the other strategies.
The Architecture: How They Stack
The mistake most people make is treating these as competing strategies. They are not. They operate at different timeframes and answer different questions.
Think of it as four layers:
Layer 1 — Macro Regime (monthly): What is the economic environment? Rising rates, falling rates, inflation, deflation, growth, recession? The VisionBoard sector model processes this and outputs a sector ranking. This tells you which sectors to be in and which to avoid.
Layer 2 — Sector Confirmation (weekly): Does price agree with the macro thesis? Use trend-following (50/200 EMA on sector ETFs like XLE, XLF, XLK) to confirm the sector is actually moving in the direction the macro model predicts. Do not rotate into a sector just because the model says so — wait for price confirmation.
Layer 3 — Entry Timing (daily): Once you are in the right sector, use breakout signals or mean reversion to time individual stock entries. Breakout for momentum names, mean reversion for beaten-down names in a recovering sector.
Layer 4 — Position Sizing (always): ATR-based sizing at the individual stock level. Volatility regime (VIX) at the portfolio level. Size down when the environment is uncertain, size up when conditions are clear.
What This Looks Like in Practice Right Now
The current macro regime: elevated oil, sticky inflation, geopolitical risk, Fed on hold. The VisionBoard model outputs: overweight Energy (XLE), Financials (XLF), Industrials (XLI). Underweight Tech (XLK), Consumer Discretionary (XLY).
Layer 2 check: XLE is in a clear uptrend above its 50 EMA. Confirmed. XLF holding. Confirmed. XLK below its 200 EMA — the macro thesis and price agree.
Layer 3: Within Energy, look for breakout setups on pullbacks to the 20 EMA. Within Financials, regional banks showing mean reversion bounces off support.
Layer 4: VIX elevated — keep individual position sizes at 60-70% of normal. The macro is right; the volatility says do not max out.
The Point
The sector model handles the top-down. The bot strategies handle execution and risk. They do not compete — they stack. A bot without a macro filter is flying blind. A macro model without execution discipline leaves money on the table.
The VisionBoard sector model is built to be the top layer of exactly this kind of system. Learn more at VisionBoard Finance.