Wavelet analysis of the international markets: A look at the next eleven (N11)

In this paper, we investigated theco-movement of the Nigerian stock market with that of the N11 (Next 11) countries Bangladesh(Dhakha), Egypt(MSCIEgy), Indonesia(EIDO), Iran(RANI3), Mexico(MSCIMex), Nigeria(NGE), Pakistan(PAK), the Philippines(PHS), Turkey(MSCITurk), South Korea(EWY), and Vietnam(FTSE).The stock returns is denoted by ‘-r’ where ‘-’ denote the stock variable. For instance, the return for Nigeria is denoted by ‘NGEr’.The denoised stock returns is denoted by ‘-rd’ where ‘-’ denote the stock variable. For instance, the denoised return for Nigeria is denoted by ‘NGErd’.We also looked at the effect of noise on the dynamics of the market. We used some wavelet based measures of co-movement to analyze the market dynamics. The data cover the period from 5th September 2017 to 24th of May 2019 a length of 512. This period falls in the period Nigeria experienced her second worst recession in history. The summary statistics shows that the Iranian stock (RANI3) has some quite interesting properties. The standard deviation of its price is the least with a value of 0.43 and highest positive skewness of 0.62. Its returns however has the highest standard deviation of 0.03, highest skewness of 2.08 and the highest kurtosis of 31.7. The wavelet power spectrum shows that most of the power and variability are concentrated at the lower scales (higher frequencies). It can be seen in the stock returns that most of the power as well as the variability is seen in the lower scales 0–16 (higher frequencies) while lower power is mostly seen at higher scale 32–128 (lower frequencies). The 1D scalogram, showed that for the return series. The energy peaks between 0 and 4 scale and decreases almost linearly. The denoised series has a very contrasting scenario; the energy is at its lowest at scale 0–32 and then rises sharply to its peak(s) and then descends again. We conclude that there is a very strong co-movement among the N11 countries and that the presence of noise has almost the same effect in the time frequency domain for all the N11 countries. It can be said that most of the market dynamics of the N11 countries in the lower scales (high frequency) are driven by noise while the dynamics of the higher scale (low frequency) are driven by market fundamentals. Finally, from our results, we deduced that the skewness and the kurtosis can be used to determine the energy distribution in a wavelet decomposition.