The “Economic Conditions Snapshot, September 2014″ McKinsey Global Survey
results presented two interesting sets of conclusions concerning the global economic conditions expected to dominate the next six months around the world. The first insight gained from the survey is that for the second quarter in a row, geopolitical risks are seen as the dominant risks to economic growth worldwide. Some 80 percent of the top executives surveyed now believe geopolitical instability in the Middle East and North Africa (MENA) is very likely or extremely likely to shock the global economy in the coming year. The second insight is that concerns with geopolitical instability in the MENA region are dominant amongst the executives from North American, European and Developing Markets.
Whether we like it or not, political and institutional risks are shape-makers in the global economic environment and financial markets.
Geopolitical and institutional risks relating to political environments are notoriously hard to measure and capture in quantitative models of asset returns. There are several good reasons for this. Firstly, these risks evolve over longer periods of time and require careful selection of time horizons for analysis. Choosing a wrong time frame for modelling asset returns in the presence of these risks can lead to strongly biased and inaccurate results. Secondly, measuring such risks requires somewhat subjective interpretations of risk parameters. For example, political instability drives up financial risks. But political stability, especially in the cases of less democratic, less transparent institutions found in a number of countries around the world can temporarily mask build up of substantial risks in the system. Thirdly, some geopolitical risks manifest themselves in incremental changes in the political environment. Others tend to be ‘black swan’ events: unforeseen until they materialise, rare in occurrences, but exceptionally large in their impact on the financial markets.
In a study, completed this year, Prarthana Ravi Kumar and myself looked at the asymmetric impacts of political risk metrics on stock prices in the MENA countries. To determine if political risks have an impact on stock markets in the region, we used the data from the Euromoney Country Risk surveys of economic analysts since the beginning of 2005 and through H1 2014 covering Egypt, Jordan, Kuwait, Oman, Qatar, Saudi Arabia, UAE and Bahrain.
In the first stage of our analysis, we deployed a linear model linking political risk ratings for the above countries to the long-term performance of their core equity indices. We found that linear models fail to capture any significant impact of political risk on equity valuations.
As the result, we investigated whether Momentum Threshold Autoregressive model (M-TAR) provides a better insight into the potential relationship between political environment quality and the stocks returns. In traditional Threshold Autoregressive model setting, the dynamic process driving the variable of interest (in our case, stock prices, yt) is assumed to follow the autoregressive process that changes in relation to a specific threshold, ‘tau’. For example, as stock prices rise above specific level of ‘tau’, the future dynamics of stock prices (yt) can become more or less dependent on past realisations of stock prices (yt-1, yt-2…) than prior to cross the threshold.
Standard specification of the threshold autoregressive model is given by:
