Two essays on China's stock markets

Abstract

China’s stock markets have become the second largest in the world after that

of the United States. Both the Chinese institutional setting and the behaviors of

the populous Chinese investors and listed firms provide novel opportunities to

explore the classical theories in the field of economics and finance. Using two

natural experiments, this thesis attempts to shed new light on these theories.

The local bias puzzle was originally proposed from the analysis of investors’

investment portfolios. In the first essay, I test and confirm the hypothesis that

local bias has already existed in investor attention subconsciously regardless of

their investment. In contrast to literature which focuses on investment accounts, I

examine local bias in investor attention by analyzing investor messages posted on

China’s Internet stock message boards. I find that individual investors pay more

attention to the stocks of local companies. This finding is strong and robust to

local-bias proxy variables. By examining factors that affect investor attention

local bias, I find that local bias is particularly strong in underdeveloped regions,

for SOEs, for small-investor base and low-turnover stocks, and for stocks with

name indicating locality. Furthermore, distance plays a significant role: the

marginal effect of local bias is much stronger for distances within 500 kilometers.

All these results are consistent with my explanation that local bias is affected by

factors which can attract investors’ attention. Thus, investment local bias is the

natural consequence of investor attention local bias, and I attribute the local bias

puzzle to limited investor attention.

Chinese stock market has plunged into an unlocking flood of non-tradable

shares since June 2006. This radical transition provides a unique natural

experimental setting to ascertain earnings management incentives. In the second

essay, I explore whether earnings management behavior exists in listed Chinese

firms during the unlocking process. I find that non-tradable shareholders

opportunistically manipulate earnings upward to offset price pressures for

subsequent selling. Firms have higher levels of accruals when unlocking incentive

is higher. Furthermore, actual selling incentive is higher in firms which have

higher levels of accruals. The results document a novel case that equity incentives

give rise to the incidence of earnings management.