Financial Risk Analysis Project with Python

"""Calculate synthetic credit expected loss and concentration indicators."""

from __future__ import annotations

import numpy as np
import pandas as pd


def make_demo_exposures(n: int = 1_500, seed: int = 42) -> pd.DataFrame:
    rng = np.random.default_rng(seed)
    rating = rng.choice(["A", "B", "C", "D"], n, p=[0.28, 0.36, 0.25, 0.11])
    pd_map = {"A": 0.004, "B": 0.015, "C": 0.055, "D": 0.16}
    probability_default = pd.Series(rating).map(pd_map).to_numpy()
    return pd.DataFrame({"exposure_id": range(1, n + 1), "sector": rng.choice(["Retail", "Manufacturing", "Services", "Technology"], n),
                         "rating": rating, "ead": rng.lognormal(12.0, 0.9, n).round(2),
                         "pd": probability_default, "lgd": rng.uniform(0.25, 0.7, n)})


def analyse_risk(data: pd.DataFrame) -> tuple[dict[str, float], pd.DataFrame]:
    risk = data.copy()
    risk["expected_loss"] = risk["ead"] * risk["pd"] * risk["lgd"]
    risk["stressed_expected_loss"] = risk["ead"] * np.minimum(risk["pd"] * 1.5, 1) * np.minimum(risk["lgd"] + 0.1, 1)
    total_ead = risk["ead"].sum()
    sectors = risk.groupby("sector", as_index=False).agg(ead=("ead", "sum"), expected_loss=("expected_loss", "sum"), stressed_expected_loss=("stressed_expected_loss", "sum"))
    sectors["ead_share"] = sectors["ead"] / total_ead
    kpis = {"total_ead": float(total_ead), "expected_loss": float(risk["expected_loss"].sum()), "stressed_expected_loss": float(risk["stressed_expected_loss"].sum()), "largest_sector_share": float(sectors["ead_share"].max())}
    return kpis, sectors.sort_values("ead", ascending=False)


def main() -> None:
    kpis, sectors = analyse_risk(make_demo_exposures())
    print({k: round(v, 3) for k, v in kpis.items()})
    print(sectors.round(3).to_string(index=False))
    print("Synthetic educational assumptions; not a validated regulatory, lending, or investment model.")


if __name__ == "__main__":
    main()