Dear Adam, thanks a lot for your good questions and comments. In 2008 and 2009, profitability of SOEs declined significantly amid the global crisis while non-SOEs held up pretty robustly. I think this is partly due to the sectorial distribution of the SOEs, which are heavily concentrated in the heavy industrial sectors. As we know, capital expenditure was hard hit in the global crisis, leading to significant contraction in the heavy industrial sectors and therefore weaker performance of enterprises in these sectors. Regarding figure 2, it is true that excluding the oil extraction sector would lead to a much lower R^2 in the regression. However, like I said in the post, there are other highly state monopolizing sectors such as utility supply suffering from low administrative pricing. Excluding these sectors from the regression would lead to higher R^2. Hence, to fully remove the impact of the oil extraction and sectors like utility supply, a regression of SOE ROA on SOE share is carried out for those sectors with SOE share lower than 50%. The R^2 of this regression is over 0.3, larger than that the R^2 of the same regression carried out for non-SOEs.