Economic Integration of the Guangdong-Hong Kong-Macao Greater Bay Area: an Empirical Assessment
Abstract. 9 Guangdong (G) cities, Hong Kong (H) and Macao (M) were strategized as an integrated zone, called the Greater Bay Area (GBA) since 2017. The GBA has encountered both opportunities and challenges. There are many trade ties, but border effects such as post-COVID travel restrictions remain. The regions have different exchange rate regimes – HKD and Pataca are pegged, but Renminbi semi-floats. These observations motivate us to evaluate the degrees of real and financial integration in the GBA. With price index, interest rate and exchange rate data from 2016M1 to 2022M7, we computed purchasing power differential (PPD), uncovered interest rate differential (UID) and real interest rate differential (RID). Only two differentials have statistically significant means. All series have means within 1.13%, indicating high degrees of real, financial and overall economic integration in the GBA. GM has the strongest real integration and HM has the strongest financial and overall economic integration. The unit root tests find most differentials to be mean-reverting in the long run. HM UID approaches a small but positive mean in the long run. GM’s RID is non-stationary, but both their PPD and UID have close-to-zero means and are mean-reverting, so economic integration in GM is still anticipated. Keywords: Economic integration; Financial Integration; Economic development policy; Greater Bay Area; GBA.
Steps to reproduce
We built the dataset mainly based on interest rate, price index and exchange rate data compiled by IFS of IMF (2022). When the database was updated more slowly than the local statistics agencies, namely NBS (2022) of mainland China and DSEC (2022) of Macao, we calculated the latest data points with local data. Besides, only national data of mainland China are available in IFS. To compute the CPI of Guangdong province, we used mainland Chinese CPI of December 2015 available in IFS as a starting point and adjusted it with monthly inflation rates downloaded from NBS (2022). The preprocessed dataset covers the period from January 2016 to July 2022, so the sample has 79 periods. Afterwards, the UID, PPD and RID series were computed according to identity (1). All the differential series were expressed as annualized percentage points for easy interpretation.