The Macro Model uses Random Forest machine learning to predict 12-month market outcomes. It combines technical indicators (like RSI, MACD, ATR) with macro-economic factors (Policy Rate, Inflation, Yield Curve, Consumer Sentiment, Credit Spread) to provide probability-weighted forecasts for strategic positioning.

The model analyzes historical price data over a 252-trading-day horizon (approximately 1 year). It uses sliding windows with monthly steps to create training samples, then predicts one of four outcome classes: Strong Bullish (+2), Mild Bullish (+1), Mild Bearish (-1), or Strong Bearish (-2) based on forward returns.

The Net Score shows overall market sentiment based on the model’s probability-weighted analysis. Positive values indicate bullish sentiment, while negative values indicate bearish sentiment. Higher absolute values indicate stronger signals.

The model outputs four probabilities: P(-2) for Strong Bearish, P(-1) for Mild Bearish, P(1) for Mild Bullish, and P(2) for Strong Bullish. These probabilities sum to 100% and show the likelihood of each outcome over the 12-month horizon. The chart displays these as a visual probability distribution.

The model can incorporate Policy Rate (Federal Funds Rate), Inflation Rate (CPI), Yield Curve (10Y-2Y spread), Consumer Sentiment, and Credit Spread. These are optional features you can toggle on or off. The model also always uses MA10, ATR, and MACD (locked features) along with other technical indicators like RSI, Bollinger Bands, Stochastic RSI, ADX, OBV, Williams %R, Ichimoku Cloud, CCI, ROC, and PSAR.

Expected Value combines the probability of each outcome with average win and loss rates. It shows the anticipated return percentage over the 12-month horizon. The calculation depends on the Net Score’s sign: when Net Score is positive, Expected Value represents anticipated gains from a long position (bullish outcomes); when Net Score is negative, Expected Value represents anticipated gains from a short position (bearish outcomes). Therefore, a negative Net Score with a positive Expected Value indicates the model expects downward movement, and the Expected Value shows the anticipated profit from taking a short position.

Model Confidence shows how certain the model is about its predictions. It’s calculated as the maximum probability from the probability distribution. Higher confidence (closer to 100%) means the model is more certain about a particular outcome, while lower confidence suggests more uncertainty or mixed signals.

The Macro Model supports stocks (e.g., AAPL), ETFs (e.g., SPY), cryptocurrencies (e.g., BTC-USD), and forex pairs (e.g., EURUSD=X). Simply enter the symbol and the model will analyze it using both technical and macro factors.

Yes, TSL is designed for educational purposes. All tools are for learning and simulation. We do not provide financial advice, and past performance does not guarantee future results.