Western observers commonly compare China’s stock markets to a casino, a wild gambling house where stock picking requires the same amount of skill as a turn at the roulette wheel. While it may be true that in China, ‘the house’ always wins, I’d like to challenge the underlying assumption that China’s markets operate without rhyme or reason. In fact, there exist several clearly observable patterns governing the behavior of China’s stock markets.

Using DataYes’s dataset of 400 factors for A-share stocks, I analyzed the performance of a few well-known factors for predicting future returns. Each one clearly illuminates patterns in China’s stock markets, even if they are unrecognizable at first glance.

To evaluate factor performance, I used a monthly rebalancing decile ranking methodology of factors and returns from January 2011 through May 2017. I calculated the decile returns for three factors: the size effect, where small stocks outperform large stocks, the momentum effect, where recent winners continue to outperform, and the beta effect, where high-beta stocks outperform low-beta stocks. The DataYes factor definitions are respectively the natural log of total assets, current stock price divided by average price over the past year, and beta relative to the CSI 300 index.

I find evidence for dramatic size and momentum effects; that is, small stocks and recent winners are the top performers in China’s stock market. Additionally, I find that high-beta stocks modestly underperform low-beta stocks.

Although Chinese investors worship Warren Buffett, there isn’t an obvious case to be made for value investing in China’s public equity markets. In fact, the most expensive stocks by measures such as price-to-book or price-to-sales outperform their cheaper counterparts. I don’t believe that this implies an “inverse” value effect, but rather that the momentum and size effects swamp out the value effect.

Also, the price to earnings factor produces a return structure with a wrinkle on the low end; that is, the cheapest decile of stocks by this measure outperforms each of the next five deciles. This pattern appears curious, but following closer inspection is simply a predictable distortion in stock market behavior due to government intervention in capital markets.

Specifically, the CSRC will delist firms posting three consecutive years of losses. This requirement incentivizes firm management to book multiple years of losses in one accounting year. Thus, the lowest decile of price-to-earnings ratio is populated with firms employing accounting legerdemain to take on big-bath losses. In fact, 95% of firms in the bottom decile are loss-making firms. Additionally, firms in the bottom decile posted positive earnings in the two years previous to and following their appearance in the bottom decile – which explains why the news of massive losses doesn’t commensurately impact the stock price.

In sum, although China’s stock markets behave differently from developed markets, they aren’t casinos where stock prices lurch with neither rhyme nor reason. Clear patterns have governed the behavior of stock prices in China, even if investors may struggle to take advantage of them.