merton.core.result

MertonResult — the frozen container returned by every fit.

The result captures the calibrated state (asset_value, asset_vol) plus derived quantities (DD, PD). Heavy derived outputs (Greeks, term structures, confidence intervals, summaries) are lazy methods so callers only pay for what they need.

Classes

MertonResult

The output of fitting a single firm.

Module Contents

class merton.core.result.MertonResult

The output of fitting a single firm.

firm

The input Firm.

asset_value, asset_vol

Calibrated firm asset value and asset volatility.

asset_drift

Calibrated drift (only set by MLE/time-series methods).

default_point

Resolved default threshold from the firm’s chosen formula.

dd

Distance-to-default at firm.horizon.

pd

Risk-neutral probability of default at firm.horizon.

method

Name of the calibrator used.

n_iter, converged

Diagnostics from the solver.

log_likelihood, covariance_, residuals_

Statistical output for MLE methods (None for others).

diagnostics_

Free-form bag of solver-specific telemetry.

firm: merton.core.firm.Firm

asset_value: float | merton._typing.FloatArray

asset_vol: float

asset_drift: float | None = None

default_point: float | merton._typing.FloatArray

dd: float

pd: float

dd_series: merton._typing.FloatArray | None = None

pd_series: merton._typing.FloatArray | None = None

asset_value_series: merton._typing.FloatArray | None = None

method: str = 'unknown'

n_iter: int = 0

converged: bool = True

log_likelihood: float | None = None

covariance_: numpy.ndarray | None = None

residuals_: numpy.ndarray | None = None

diagnostics_: dict[str, Any]

classmethod from_calibration(firm: merton.core.firm.Firm, asset_value: float | merton._typing.FloatArray, asset_vol: float, *, method: str, asset_drift: float | None = None, n_iter: int = 0, converged: bool = True, log_likelihood: float | None = None, covariance: numpy.ndarray | None = None, residuals: numpy.ndarray | None = None, diagnostics: dict[str, Any] | None = None) → MertonResult

pd_term_structure(horizons: merton._typing.ArrayLike = (1 / 12, 3 / 12, 6 / 12, 1.0, 3.0, 5.0)) → pandas.DataFrame

Return a DataFrame of (horizon, DD, PD) across multiple horizons.

implied_spread(lgd: float = 0.6, *, in_bps: bool = True) → float | merton._typing.FloatArray

Implied credit spread (bps by default) at firm.horizon.

physical_pd(sharpe_ratio: float) → float | merton._typing.FloatArray

Convert risk-neutral PD to physical PD using the supplied Sharpe ratio.

confidence_interval(level: float = 0.95, method: str = 'asymptotic', *, quantities: tuple[str, Ellipsis] = ('asset_vol', 'asset_drift', 'dd', 'pd')) → dict[str, Any]

Confidence intervals for fitted parameters and derived quantities.

Parameters:

• level – Confidence level (e.g. 0.95).

• method – "asymptotic" uses the MLE observed Fisher information (requires covariance_ to be set, e.g. from method='duan_mle'). "bootstrap" uses the BootstrapResult stashed in diagnostics_['bootstrap'] (requires MertonModel(n_bootstrap=...)).

• quantities – Names to report. Supported: "asset_vol", "asset_drift", "dd", "pd".

greeks() → merton.greeks.equity.GreeksResult

Compute all closed-form Greeks at the calibrated operating point.

to_dict() → dict[str, Any]

to_pandas() → pandas.DataFrame

Single-row DataFrame summarising the fit.

to_polars() → Any

Return the result as a Polars DataFrame (requires polars installed).

to_excel(path: str, *, sheet: str = 'Merton') → None

Write the result to an Excel workbook.

summary() → str