Political landscape and liquidity of non-U.S. stocks from emerging markets

Abstract

We investigate the empirical relation between country governance quality and stock market liquidity, as well as information asymmetry, using a sample of non-U.S. stocks from 17 emerging markets listed on the NYSE between 2004 and 2019. We find that non-U.S. stocks from emerging markets with higher democracy quality tend to have narrower spreads and larger depth, suggesting improved liquidity. Higher autocracy levels, on the other hand, are associated with wider spreads and lower depth, indicating poorer liquidity. Additionally, stronger democracy and polity qualities are linked to reduced price impact, while heightened autocracy levels are associated with increased price impact and a higher probability of informed trading. Moreover, we show that changes in our liquidity and information asymmetry measures significantly relate to changes in the country governance index over time. Our results remain remarkably robust across regions and when using different measures of liquidity and information-based trading.

Appendix: Calculation of the probability of information-based trading (PIN)

The EKOP model of the trade process for firm i over trading day j is represented by the following likelihood function:

$$\begin{aligned} & {\text{L}}_{{\text{i}}} ({\text{B}}_{{{\text{i}},{\text{j}}}} ,{\text{S}}_{{{\text{i}},{\text{j}}}} |\theta _{{\text{i}}} ) = (1 - \alpha _{{\text{i}}} ){\text{e}}^{{ - \upvarepsilon _{{\text{i}}} {\text{T}}_{{{\text{i}},{\text{j}}}} }} \frac{{(\upvarepsilon _{{\text{i}}} {\text{T}}_{{{\text{i,j}}}} )^{{{\text{B}}_{{{\text{i,j}}}} }} }}{{{\text{B}}_{{{\text{i,j}}}} !}}{\text{e}}^{{ - \upvarepsilon _{{\text{i}}} {\text{T}}_{{{\text{i,j}}}} }} \frac{{(\upvarepsilon _{{\text{i}}} {\text{T}}_{{{\text{i,j}}}} )^{{{\text{S}}_{{{\text{i,j}}}} }} }}{{{\text{S}}_{{{\text{i,j}}}} !}} \\ & \quad\quad\quad\quad\quad\quad\quad + \upalpha _{{\text{i}}} \updelta _{{\text{i}}} e^{{ - \varepsilon _{{\text{i}}} {\text{T}}_{{{\text{i,j}}}} }} \frac{{(\upvarepsilon _{{\text{i}}} {\text{T}}_{{{\text{i,j}}}} )^{{{\text{B}}_{{{\text{i,j}}}} }} }}{{\text{B}_{{{\text{i,j}}}} !}}e^{{ - (\upmu _{{\text{i}}} + \upvarepsilon _{{\text{i}}} ){\text{T}}_{{{\text{i,j}}}} }} \frac{{[(\upmu _{{\text{i}}} + \upvarepsilon _{{\text{i}}} )\text{T}_{{{\text{i,j}}}} ]^{{{\text{S}}_{{{\text{i,j}}}} }} }}{{\text{S}_{\text{i,j}} !}} \\ & \quad\quad\quad\quad\quad\quad\quad + \upalpha _{{\text{i}}} (1 - \updelta _{{\text{i}}} ){\text{e}}^{{ - (\upmu _{{\text{i}}} + \upvarepsilon _{{\text{i}}} ){\text{T}}_{{{\text{i,j}}}} }} \frac{{[(\upmu _{{\text{i}}} + \upvarepsilon _{{\text{i}}} )\text{T}_{{{\text{i,j}}}} ]^{{{\text{B}}_{{{\text{i,j}}}} }} }}{{\text{B}_{{{\text{i,j}}}} !}}e^{{ - \upvarepsilon _{{\text{i}}} {\text{T}}_{{{\text{i,j}}}} }} \frac{{(\upvarepsilon _{{\text{i}}} {\text{T}}_{{{\text{i,j}}}} )^{{{\text{S}}_{{{\text{i,j}}}} }} }}{{{\text{S}}_{{{\text{i,j}}}} !}}. \\ \end{aligned}$$

where B_i,j is the number of buyer-initiated trades for the day, S_i,j is the number of seller-initiated trades for the day, α_i is the probability that an information event has occurred, δ_i is the probability of a low signal given an event has occurred, µ_i is the probability that a trade comes from an informed trader given an event has occurred,⁴$\varepsilon_{i}$ is the probability that the uninformed traders will actually trade, T_i,j is total trading time for the day, and θ_i = (α_i, δ_i, ε_i, μ_i) represents the vector of parameters to be estimated.

We estimate these parameters θ_i for firm i for each year by maximizing the joint likelihood over the J observed trading days in a calendar year:

$$L_{i} (M_{i} |\theta_{i} ) = \prod\limits_{j = 1}^{J} {L_{i} (B_{i,j} ,S_{i,j} |\theta_{i} ).}$$

We then estimate the probability of information-based trading (PIN) for firm i for each year as

$$PIN_{i} = \frac{{\hat{\alpha }_{i} \hat{\mu }_{i} }}{{\hat{\alpha }_{i} \hat{\mu }_{i} + 2\hat{\varepsilon }}}_{i} .$$