This Bloomberg article brings to our attention that

Shares in the yuan-denominated CSI 300 Index traded at 16.2 times earnings this month, compared with 8.6 times for 43 mainland companies in Hong Kong.

and that

The average gain in China is 23 percent this year, while the same companies are down 4.8 percent in Hong Kong, … China is among three of the four so-called BRICs economies where local shares are providing bigger returns than are available to foreigners … Russia’s Micex index, up 21 percent in rubles since Dec. 31, gained 2.5 percent when measured in dollars. A 9.2 percent decline in India’s Sensitive Index widens to 14 percent in dollars. The exception is Brazil, where the Bovespa Index has risen 5.7 percent, versus a 3.9 percent gain in reals.

The A-shares have always been overvalued compared to the H-shares, when measured against any benchmark currency and people have puzzled over this for a long time. The article itself offers two explanations

“Valuations are more appropriate in the H-share market [in Hong Kong] because more foreigners are paying attention.”

and

Victoria Mio at Robeco Group says mainland investors are quicker to anticipate changes in the local economy and have an incentive to spend their savings on stocks after the central bank cut interest rates five times since September.

Those may very well be the correct explanations, but don’t seem like the whole picture. A cryptic comment that Bloomberg extracted here may give a hint to an alternate explanation

Paul Chow, chief executive officer of Hong Kong Exchanges & Clearing Ltd., said in an interview last week that the valuation gap between the A and H shares is “determined by the market.”

Which is to say, the discrepancy is rational. This says to me that, in analogy to the other countries in which stock markets gained more in domestic currencies than in dollars (their currencies depreciated), China’s seemingly much excessive gains in the domestic currency would not be excessive if expectation about currency valuations is taken into account. If people believe that CNY is undervalued compared to the USD (and in the past year it has been pegged to the USD by policy rather than rising as it had for some time), they would drive up the price of the A-shares in anticipation of currency gain to the tune of the disparity between A-shares and H-shares (HKD is a USD proxy). That would indeed be a market determined phenomenon.

To be more concrete, market gain in local currency + currency appreciation against dollar = market gain in dollar. For example, Russian stocks gained 21% in roubles this year and 2.5% in dollars. So the rouble depreciated 18.5% against the dollar this year. Indeed this is true.

Now, a basket of Chinese stocks gained 23% this year in CNY and lost 4.8% in HKD (which is basically USD). The same argument says that CNY has depreciated 27.8% against the USD since the beginning of the year. But in fact the CNY-USD pair, which isn’t fully tradable, has been constant. So this is unrealized depreciation, or expectation about appreciation were full trading allowed.

One can only wonder about the actual market gain in CNY were market differences allowed to be arbitraged away. But suppose the H-shares market is “correct” at this moment and that A-shares gain in CNY should actually be -4.8% with tradable currencies, then we inevitably come to the conclusion that a one-time CNY gain of 27.8% against the USD would offset the -27.8% difference to reset the A-shares to H-share values as measured at the beginning of the year. In fact, the persistent valuation differences between A- and H-shares are perhaps not a bad measure of relative currency values in lieu of actual trading.