initial commit

financial_tools.py

@@ -0,0 +1,131 @@
+import pandas as pd
+import numpy as np
+
+class FinancialTools:
+    @staticmethod
+    def calculate_returns(prices_series, in_percent=True): # <-- NEUER PARAMETER HIER!
+        """Berechnet die täglichen Renditen."""
+        if prices_series.empty:
+            return pd.Series(dtype='float64')
+        if isinstance(prices_series, pd.DataFrame):
+            if 'adj_close' in prices_series.columns:
+                prices_series = prices_series['adj_close']
+            elif 'close' in prices_series.columns:
+                prices_series = prices_series['close']
+            else:
+                return pd.Series(dtype='float64')
+        
+        returns = prices_series.pct_change().dropna()
+        if in_percent:
+            return returns * 100 # Für die Anzeige in Prozent
+        return returns # Für interne Berechnungen (Dezimalwerte)
+
+    @staticmethod
+    def calculate_cumulative_returns(prices_series):
+        """Berechnet die kumulativen Renditen."""
+        if prices_series.empty:
+            return pd.Series(dtype='float64')
+        # HIER: Rufe calculate_returns mit in_percent=False auf, um Dezimalwerte zu bekommen
+        daily_returns_raw = FinancialTools.calculate_returns(prices_series, in_percent=False)
+        if daily_returns_raw.empty:
+            return pd.Series(dtype='float64')
+        return (1 + daily_returns_raw).cumprod() - 1
+
+    @staticmethod
+    def calculate_moving_average(prices_series, window=20):
+        """Berechnet den gleitenden Durchschnitt."""
+        if prices_series.empty:
+            return pd.Series(dtype='float64')
+        if isinstance(prices_series, pd.DataFrame):
+            if 'adj_close' in prices_series.columns:
+                prices_series = prices_series['adj_close']
+            elif 'close' in prices_series.columns:
+                prices_series = prices_series['close']
+            else:
+                return pd.Series(dtype='float64')
+        
+        return prices_series.rolling(window=window).mean()
+
+
+    @staticmethod
+    def calculate_volatility(prices_series, window=20):
+        """Berechnet die rollierende Volatilität (Standardabweichung der Renditen)."""
+        if prices_series.empty:
+            return pd.Series(dtype='float64')
+        
+        # HIER: Rufe calculate_returns mit in_percent=False auf, um Dezimalwerte zu bekommen
+        daily_returns_for_vol = FinancialTools.calculate_returns(prices_series, in_percent=False)
+        
+        if daily_returns_for_vol.empty:
+            return pd.Series(dtype='float64')
+
+        # Annualisiere die Volatilität
+        return daily_returns_for_vol.rolling(window=window).std() * np.sqrt(252)
+
+
+    @staticmethod
+    def calculate_beta(stock_prices, market_prices, window=60):
+        """
+        Berechnet das rollierende Beta eines Wertpapiers relativ zu einem Marktindex.
+        stock_prices: Pandas Series der Aktie
+        market_prices: Pandas Series des Marktindex
+        """
+        if stock_prices.empty or market_prices.empty:
+            return pd.Series(dtype='float64')
+        
+        # Sicherstellen, dass es Series sind, falls doch DataFrames übergeben werden
+        if isinstance(stock_prices, pd.DataFrame):
+            if 'adj_close' in stock_prices.columns:
+                stock_prices = stock_prices['adj_close']
+            elif 'close' in stock_prices.columns:
+                stock_prices = stock_prices['close']
+            else:
+                return pd.Series(dtype='float64')
+        
+        if isinstance(market_prices, pd.DataFrame):
+            if 'adj_close' in market_prices.columns:
+                market_prices = market_prices['adj_close']
+            elif 'close' in market_prices.columns:
+                market_prices = market_prices['close']
+            else:
+                return pd.Series(dtype='float64')
+
+        # HIER: Rufe calculate_returns mit in_percent=False auf, um Dezimalwerte zu bekommen
+        returns_stock = FinancialTools.calculate_returns(stock_prices, in_percent=False)
+        returns_market = FinancialTools.calculate_returns(market_prices, in_percent=False)
+
+        # Kombiniere die Renditen und richte sie an den Daten aus
+        combined_returns = pd.concat([returns_stock.rename('stock'), returns_market.rename('market')], axis=1).dropna()
+
+        if combined_returns.empty:
+            return pd.Series(dtype='float64')
+
+        # Berechne die rollierende Kovarianz und Varianz
+        rolling_covariance = combined_returns['stock'].rolling(window=window).cov(combined_returns['market'])
+        rolling_variance = combined_returns['market'].rolling(window=window).var()
+
+        # Berechne Beta
+        beta = rolling_covariance / rolling_variance
+        return beta.dropna()
+
+# Beispielnutzung (für Tests) - Muss an die neuen Funktionssignaturen angepasst werden
+if __name__ == "__main__":
+    # Erzeuge fiktive Daten für Tests
+    dates = pd.to_datetime(pd.date_range(start='2023-01-01', periods=100))
+    stock_prices_series = pd.Series(np.random.rand(100) * 100 + 50, index=dates) # direkt als Series
+    market_prices_series = pd.Series(np.random.rand(100) * 10 + 200, index=dates) # direkt als Series
+
+    print("Tägliche Renditen (in Prozent):")
+    print(FinancialTools.calculate_returns(stock_prices_series, in_percent=True).head()) # Für Anzeige
+
+    print("\nKumulative Renditen (Dezimal):")
+    print(FinancialTools.calculate_cumulative_returns(stock_prices_series).head())
+
+    print("\nGleitender 20-Tage-Durchschnitt:")
+    print(FinancialTools.calculate_moving_average(stock_prices_series, window=20).head())
+
+    print("\nRollierende 20-Tage-Volatilität (annualisiert, Dezimal):")
+    print(FinancialTools.calculate_volatility(stock_prices_series, window=20).head())
+
+    print("\nRollierendes 60-Tage-Beta (Aktie vs. Markt, Dezimal):")
+    print(FinancialTools.calculate_beta(stock_prices_series, market_prices_series, window=60).head())