Skip to content
All library documents

Using GARCH Conditional Volatility to Simulate and Forecast Returns

Article Quant Q&A · Author: TAN YONG SHENG
Read the original

Summary

The document explains how to pursue GARCH-based simulation and forecasting in Python, distinguishing conditional volatility modeling from implied volatility. GARCH models describe the changing conditional variance of asset returns; forecasts and simulated returns can then be used to study possible price paths, though a volatility forecast alone does not determine a unique future price.

The answer points readers to Python’s arch package for fitting, simulating, and forecasting ARCH and GARCH models, and mentions examples for implementing a GJR-GARCH model from scratch and estimating a GARCH model with the package. The cited implementation example also covers sandwich covariance estimates for standard errors. These are learning resources rather than a worked forecast or performance evaluation. The document reports no empirical validation, and any simulated price path depends on the chosen model, estimated parameters, and return assumptions.

Key ideas

  • GARCH models represent conditional volatility in returns, distinct from implied volatility.
  • A conditional volatility model can support return simulation and forecasting in Python.
  • The arch package includes model estimation, simulation, and forecast examples.
  • A from-scratch GJR-GARCH example illustrates estimation and robust standard errors.
  • Simulated price paths depend on model specification and assumptions.

Tags

EquitiesVolatilityStatisticsMachine learning
Full text
# How to use conditional volatility under GARCH model to forecast price?


# How to use conditional volatility under GARCH model to forecast price?












I have come across videos on youtube about GARCH model in stimulating and forecasting stock price, however, it is programmed in R language. Is there any tutorials teach the similar as the videos shown below, but programmed in python?

(Aim: using conditional volatility under GARCH model to stimulate stock price.)

Thanks in advance.

https://m.youtube.com/playlist?list=PL34t5iLfZdduGEuSXYrleeBdvfQcak0Ov

## Answer by Pleb (score 1, accepted)

https://quant.stackexchange.com/a/66077

Be aware that GARCH models are used to model the conditional volatility of asset returns, and in this sense, it has nothing to do with implied volatility (eg. see the wiki page).

#### Python package & example for GARCH modeling:

Within the `Python` framework you can find the well-known `arch` package developed by Kevin Sheppard. The package have many different ARCH & GARCH models, which can be viewed in this list. The author also include ways of simulating and forecasting asset returns from the models, which is helpful in your scenario. If you read the documentation you will see that he has provided an abundance of examples, that will help you implement and understand how the package works.

If you want to implement a GARCH model from scratch in `Python`, then you can follow his example of implementing and estimating a GJR-GARCH(1,1) model. In this example, he also teaches you how to find the standard errors using the sandwich covariance estimator.

Alternatively, there's a YouTube video, that shows you how to simulate (from scratch) a GARCH(2,2) model and furthermore how to estimate & forecast the model using the `arch` package.

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.