2004RePEc: Research Papers in EconomicsOpen access

Non linear behaviour of stock market volatility

R. Bartiromo

Open full text 0 citations

Abstract

We exploit a continuous time random walk description of stock prices to obtain a fast and accurate evaluation of their volatility from intraday data. We show that financial markets are usefully described as open physical systems. Indeed we find that the process determining market volatility is not stationary while the market response to external volatility shocks stays constant over the time period of more than two years covered by our experimental data. Furthermore the autocorrelation function of volatility increments yields a value of about -0.4 at one-day time lag that is nearly equal for all stocks we analyze. Conditioning the evaluation of the autocorrelation function, we show that the market response is non-linear and strongly stabilizing when external shocks push for higher volatility. This market behavior can be explained by the action of participants with different time horizon.

Open-access reader

About this research paper

What this paper is about

We exploit a continuous time random walk description of stock prices to obtain a fast and accurate evaluation of their volatility from intraday data. We show that financial markets are usefully described as open physical systems. Indeed we find that the process determining market volatility is not stationary while the market response to external volatility shocks stays constant over the time period of more than two years covered by our experimental data. Furthermore the autocorrelation function of volatility increments yields a value of about -0.4 at one-day time lag that is nearly equal for all stocks we analyze. Conditioning the evaluation of the autocorrelation function, we show that the market response is non-linear and strongly stabilizing when external shocks push for higher volatility. This market behavior can be explained by the action of participants with different time horizon.

Why it matters

A significance statement is not available in the OpenAlex record.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

We exploit a continuous time random walk description of stock prices to obtain a fast and accurate evaluation of their volatility from intraday data. We show that financial markets are usefully described as open physical systems. Indeed we find that the process determining market volatility is not stationary while the market response to external volatility shocks stays constant over the time period of more than two years covered by our experimental data. Furthermore the autocorrelation function of volatility increments yields a value of about -0.4 at one-day time lag that is nearly equal for all stocks we analyze. Conditioning the evaluation of the autocorrelation function, we show that the market response is non-linear and strongly stabilizing when external shocks push for higher volatility. This market behavior can be explained by the action of participants with different time horizon.

Key concepts: Volatility (finance), Autocorrelation, Econometrics, Stock market, Continuous-time random walk, Economics, Lag, Stock (firearms)

Related papers

Back to paper searchBrowse research topicsOriginal source
Non linear behaviour of stock market volatility — Research Paper | ScholarLens