Methodology
Methodology
The AnyMarket Algorithm runs automatically every evening after the market closes. The sequence is the same every day, without exception.
1. Market data is downloaded. At around 4:30 PM EST, the model pulls the latest S&P 500 closing price and appends it to the full historical dataset.
2. Indicators are recalculated. Every technical indicator is updated using the new closing price: momentum readings, trend measurements, volatility metrics, and short-term price behavior.
3. Signals are evaluated. The model checks whether any buy or sell conditions have been met and determines whether the current allocation should change.
4. Allocation is recorded. If a change is warranted, the model logs the trade and updates the portfolio state. If nothing has changed, no action is taken.
5. Subscribers are notified. When a rotation occurs, subscribers receive an automated email alert that same evening, before the next trading day opens. See the Support page to learn more.
6. The public site is updated. Performance data is published here on a 30-day delay for transparency and accuracy verification.
Every decision is fully deterministic. The same inputs always produce the same output, with no discretion or manual override.
The Signals
The model evaluates four categories of signals. No single one drives the decision. They work together within a structured set of rules designed to respond to meaningful shifts in market conditions while filtering out everyday noise.
Momentum
The primary momentum indicator is the Relative Strength Index (RSI), a widely used measure of how quickly prices have been rising or falling over the past couple of weeks. High RSI readings suggest a market that may be overextended after a strong rally. Low readings often appear during heavy selling pressure.
The model watches both extreme single-day readings and sustained clusters of signals over time. A persistent run of unusually high readings can indicate that buying pressure has become unsustainable. A persistent run of weak readings may point to exhaustion in selling pressure. Depending on broader context, either pattern can contribute to a buy or sell signal.
Trend
Momentum can shift quickly. To capture longer-term direction, the model compares short-term and long-term moving averages of market prices.
When the short-term average strengthens relative to the long-term average, it suggests building upward momentum. The reverse can signal the beginning of a broader weakening. These crossovers are among the most widely used tools by institutional investors to identify major directional shifts.
In this model, trend signals are never acted on alone. They must align with other conditions before triggering a trade.
Volatility
Volatility tends to rise before large market declines. The model monitors price stability continuously and uses it as a filter on certain signals.
Some bearish signals are only permitted to trigger when volatility is already elevated. This keeps the model from rotating defensively during otherwise calm conditions and focuses its protective moves on the periods when a meaningful decline is statistically much more likely.
Sharp Declines
Markets occasionally experience sudden, steep selloffs driven more by fear than fundamentals. These moves can unfold in just a few trading days.
The model includes a rule designed to detect these unusually steep short-term drops. When one is identified and other guardrails allow it, the signal can trigger a buy. The logic is that panic may have pushed prices well below where they should be. Additional filters prevent this from treating every sharp decline as an automatic entry point.
Guardrails
Signals alone do not trigger trades. Several guardrails layer on top to reduce unnecessary activity and limit downside risk.
The model prevents rapid reversals. Markets regularly produce short-term conflicting signals, and the system requires meaningful confirmation before reversing a recent decision. This keeps it focused on sustained shifts rather than noise.
When the model is in defensive assets, it monitors whether the market has begun recovering. If enough time has passed and prices have climbed sufficiently from the level where the defensive rotation began, the model may re-enter equities even if some indicators remain cautious.
One Clear Output
Each trading day, all signals and guardrails are evaluated together. The result is intentionally binary: stay in equities or rotate to safe haven assets.
Because the rules are fixed and the data is objective, the model is fully transparent and fully auditable. There is no judgment call, no override, and no reaction to the news cycle. It simply does what the data says, every evening, the same way it always has.
