In this paper, we study the overall statistical properties and multifractal characteristics of CSI 100 index and SCI 500 index which represent largecapstock and small-cap-stock respectively using statistical methods and multifractal detrending fluctuation analysis (MFDFA) based on the 1 minute high frequency data from January 15, 2007 to April 18, 2008. A new market risk measure based on two main parameters of multifractal spectrum is constructed to compare the volatility and risk of SCI 100 index with SCI 500 index. The results show that the distribution of the two indexes have peak fattail as a whole, exponential form in the center and powerlaw tails, however, SCI 500 index poses fatter tail, stronger multifractal, and the new risk measure(MFV) implies SCI 500 has more risk, which makes up the inadequacies of traditional risk measures in the complex nonlinear financial system.