Financial volatility obeys two well-established empirical properties: it is fat-tailed (power-law distributed) and it tends to be clustered in time. Many interesting models have been proposed to account for these regularities, notably agent-based computational models, which typically invoke complicated mechanisms, however. It can be shown that trend-following speculation generates the power law in an intrinsic way. But this model cannot exaplain clustered volatility. This paper extends the model and offers a simple explanation for clustered volatility: the impact of exogenous news on traders’ expectations. Owing to the famous no-trade results, rational expectations, the dominant model of news-driven expectations, is hard to reconcile with the incessant high-frequency trading behind the volatility clustering. The simplest alternative model of news-driven expectations is to assume that traders have prior views about the market (an asset’s future price change or its present value) and then modify their views with the advent of a news. This simple news-driven random walk of traders’ expectations explains volatility clustering in a generic way. Liquidity plays a crucial role in this dynamics of volatility, which is emphasized in a discussions section.
Inoua, S. M. (2020). News-driven expectations and volatility clustering. ESI Working Paper 19-33. https://digitalcommons.chapman.edu/esi_working_papers/294/