An Agent-Based Model of a Pricing Process with Power Law, Volatility Clustering, and Jumps

Figure 1

The empirical analysis of the SSE 50 Index's 1-minute price series. (a) 1-minute price series; (b) return series that eliminates the overnight effect (in percentage); (c) tail distribution of the absolute return in the log-log scale and a least-square fit for values with a slope of three; (d) autocorrelation function of return and absolute return, where the autocorrelation function of absolute return decays slowly, indicating that the volatility has long-memory characteristics.

(a)

(b)

(c)

(d)