Explaining discrepancy between H1 listed profits and total profits

Listed Chinese companies' profits fell 5.5 percent in H1 2016, according to the China Securities Journal, a better-than-expected performance in light of the rocky and depressing start to the year. This also compares favorably with the S&P 500, which is on pace for a nearly 9 percent year-on-year decline for the same period (nearly 90 percent of companies having reported Q2 earnings).

However, this contrasts sharply with the National Bureau of Statistics' (NBS) reported 6.9 percent increase in total industrial profits year-on-year for January-July 2016 - including 6.2 percent increase for H1 2016 (January-June). How to account for the nearly 12-percentage point gap between the two metrics, both by official agencies? Probably a combination of the following factors.

1. The NBS figure is a more generous figure of profits before interest and/or taxes.
2. The NBS figure allows for more aggressive revenue recognition, given that listed companies' income statements must be complemented by cash flow statements which make it harder to fudge marginal sales.
3. The listed companies' figures are generally more conservative and less prone to artificial inflation, given that these firms are under more scrutiny, including in many cases by international investors and analysts.
4. The less than 3,000 listed companies are still a small slice of the economy and don't include many highly profitable medium-size private firms; if anything, the role of the stock market in Chinese business finance has stagnated since last year's crash, with only selective high-tech IPOs somewhat picking up since Q1; since smaller and medium-size tech firms (or those claiming to be high-tech) are a significant portion of privately owned companies that have listed since 2013-14, and many of these are still loss-making and dependent on future prospects, that's been a drag on listed firms' profit growth over the past year.

Like other apparent inconsistencies between sets of Chinese economic data, these earnings spreads between listed and all firms is troubling but shouldn't be dismissed out of hand as evidence of statistical fabrication; over time they will necessarily reconcile.