Volatility is an important measure of stock market performance. Competing securities market makers keep abreast of the pace of volatility change by adjusting the bid-ask spreads and bid/ask quotes properly and efficiently. For intradaily high frequency transaction data, the observed volatility of stock returns can be decomposed into the sum of the two components - the realized volatility and the volatility due to microstructure noise. The quote adjustments of the market makers comprise part of the microstructure noise. In this study, we define the ratio of the realized integrated volatility to the observed squared returns as the proportion of realized integrated volatility (PIV). Time series models with generalized error distributed innovations are fitted to the PIV data based on 70-minute returns of NYSE tick-to-tick transaction data. Both retrospective and dynamic online control charts of the PIV data are established based on the fitted time series models. The McNemar test supports that the dynamic online control charts have the same power of detecting out of control events as the retrospective control charts. The Wilcoxon signedrank test is adopted to test the differences between the changes of the market maker volatility and the realized volatility for in-control and out-of-control periods, respectively. The results reveals that the points above the upper control limit are related to the situation when the market makers can not keep up with the realized integrated volatility, whereas the points below the lower control limit indicate excessive reaction of the the market makers.