Sample code for: Python for Stock Analysis: Build a Portfolio Analyzer

python-for-stock-analysis/README.md

@@ -0,0 +1,9 @@
+# Python for Stock Analysis: Build a Portfolio Analyzer
+
+This folder provides the code examples for the Real Python tutorial [Python for Stock Analysis: Build a Portfolio Analyzer](https://realpython.com/python-for-stock-analysis/).
+
+- `fetch_prices.py`: Downloads adjusted daily closing prices with yfinance and saves them to `prices.csv`
+- `prices.csv`: Snapshot of the prices used in the tutorial (2021-01-04 to 2025-12-31, downloaded in October 2026)
+- `portfolio.py`: The finished portfolio analyzer
+
+Install the dependencies with `python -m pip install -r requirements.txt`, then run `python portfolio.py` to print the summary table and save `growth.png`.

python-for-stock-analysis/fetch_prices.py

@@ -0,0 +1,15 @@
+import sys
+
+import yfinance as yf
+
+TICKERS = ["AAPL", "JPM", "XOM", "JNJ", "SPY"]
+
+prices = yf.download(
+    TICKERS, start="2021-01-01", end="2026-01-01", progress=False
+)["Close"]
+prices = prices.reindex(columns=TICKERS)
+failed = prices.columns[prices.isna().all()].tolist()
+if failed:
+    sys.exit(f"Download failed for {', '.join(failed)}")
+prices.to_csv("prices.csv")
+print(prices.tail())

python-for-stock-analysis/portfolio.py

@@ -0,0 +1,90 @@
+import matplotlib.pyplot as plt
+import pandas as pd
+
+TRADING_DAYS = 252
+WEIGHTS = {"AAPL": 0.4, "JPM": 0.2, "XOM": 0.2, "JNJ": 0.2}
+BENCHMARK = "SPY"
+
+
+def load_prices(path="prices.csv"):
+    return pd.read_csv(path, index_col="Date", parse_dates=True)
+
+
+def calculate_returns(prices):
+    if prices.isna().any().any():
+        raise ValueError("Found missing prices. Choose dates without gaps.")
+    if len(prices) < 3:
+        raise ValueError("You need at least three days of prices.")
+    return prices.pct_change().iloc[1:]
+
+
+def build_portfolio(returns, weights):
+    weights = pd.Series(weights)
+    if not (weights >= 0).all() or abs(weights.sum() - 1) > 1e-9:
+        raise ValueError("Weights must be non-negative and add up to 1.")
+    holdings = returns[weights.index]
+    return holdings @ weights
+
+
+def growth_of(returns, initial=10_000):
+    return initial * (1 + returns).cumprod()
+
+
+def annualized_return(returns):
+    years = len(returns) / TRADING_DAYS
+    return (1 + returns).prod() ** (1 / years) - 1
+
+
+def annualized_volatility(returns):
+    return returns.std() * TRADING_DAYS**0.5
+
+
+def drawdowns(returns):
+    growth = (1 + returns).cumprod()
+    peak = growth.cummax().clip(lower=1)
+    return growth / peak - 1
+
+
+def max_drawdown(returns):
+    return drawdowns(returns).min()
+
+
+def sharpe_ratio(returns, risk_free_rate=0.0):
+    daily_risk_free = (1 + risk_free_rate) ** (1 / TRADING_DAYS) - 1
+    excess = returns - daily_risk_free
+    if excess.std() == 0:
+        return float("nan")
+    return excess.mean() / excess.std() * TRADING_DAYS**0.5
+
+
+def summarize(returns, risk_free_rate=0.0):
+    sharpe = returns.apply(sharpe_ratio, risk_free_rate=risk_free_rate)
+    return pd.DataFrame(
+        {
+            "Return": returns.apply(annualized_return),
+            "Volatility": returns.apply(annualized_volatility),
+            "Max Drawdown": returns.apply(max_drawdown),
+            "Sharpe": sharpe,
+        }
+    )
+
+
+def plot_growth(returns, filename="growth.png"):
+    fig, (top, bottom) = plt.subplots(
+        2, 1, figsize=(10, 7), sharex=True, height_ratios=[3, 1]
+    )
+    growth_of(returns).plot(ax=top, title="Growth of $10,000")
+    drawdowns(returns).plot(ax=bottom, title="Drawdown", legend=False)
+    top.yaxis.set_major_formatter("${x:,.0f}")
+    bottom.yaxis.set_major_formatter("{x:.0%}")
+    fig.tight_layout()
+    fig.savefig(filename)
+
+
+if __name__ == "__main__":
+    prices = load_prices()
+    returns = calculate_returns(prices)
+    returns["Portfolio"] = build_portfolio(returns, WEIGHTS)
+    comparison = returns[["Portfolio", BENCHMARK]]
+    print(summarize(comparison, risk_free_rate=0.03).round(3))
+    plot_growth(comparison)