pyLPA

This is a python repository implementing the Local Parametric Approach from Spokoiny, V. (2009a).

Scripts

You can find a list of scripts located in the scripts folder.

Scripts parameters:

• model: dict, model parameters:
  • name: str, ex: "garch"
  • params: dict, ex: {"p": 1, "q": 1}
• data: dict, data parameters:
  • path: str, path to the data file
  • feature: str, name of the feature, ex: "log_returns"
  • preprocessing: dict, preprocessing configuration, ex: {"name": "StandardScaler"}
• bootstrap: dict, parameters for the multiplier bootstratp:
  • generate: str, name of the distribution from which to generate multiplier bootstrap weights, ex: normal
  • num_sim: int, number of simulation
  • njobs: int, number of parallel jobs
• min_steps: int, mininum distance between two break point test
• max_trial: int, maximum attempts to estimate MLE
• maxiter: int, maximum number of iterations to perform to estimate MLE

References

Spokoiny, V. (1998). Estimation of a function with discontinuities via local polynomial fit with an adaptive window choice, The Annals of Statistics 26: 1356–78.

Spokoiny, V. (2009a). Multiscale local change point detection with applications to value-at-risk, The Annals of Statistics 37: 1405–1436.

Spokoiny, V. and Zhilova, M. (2015). Bootstrap confidence sets under model misspecification, The Annals of Statistics 43(1): 2653–2675

Spilak, Bruno and Härdle, Wolfgang Karl, Tail-Risk Protection: Machine Learning Meets Modern Econometrics (October 7, 2020). Spilak, B., Härdle, W. K. (2021). In: Lee, CF., Lee, A.C. (eds) Encyclopedia of Finance. Springer, Cham. https://doi.org/10.1007/978-3-030-73443-5_94-1