4. Market Cycles Study
4.3. Presentation of Results
國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
It is important to mention that the analysis is already broaden compared to what have been previously conducted in the United States as we do not work only with the main indices from France and Taiwan but we also include indices to take into account the market capitalization factor. Once this first analysis conducted, we will focus more on this factor by doing a
“comparative” analysis between the big and mid-caps for France and between the big and mid-caps for Taiwan. The purpose of this analysis will be to see whether it is possible to see significant differences between indices according to the market capitalization of the stocks listed on them. Another type of t-test will be conducted, to compare two means and see whether they are significantly different from each other or not. The formula for this t-test is:
With:
7- x1 and x2 the two sample means
8- s1 and s2 the standard deviation of the two samples 9- n1and n2 the sample size
10- Delta the specified value to be tested, in this case 0
4.3. Presentation of Results
We will first show our hypothesis based on the literature review we conducted and the previous findings that have been made in the United States. We will assume this is what we expect to find for our results (even if the French presidential term does not last 4 years like in the United States, we expect to find similarities). The hypotheses are presented in the following table:
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Year 1 Year 2 Year 3 Year 4
Worst YAR Second worst YAR Best YAR Second Best YAR
Market cycle study
France
Regarding the results we have found for France, we will analyze them in two parts for the two indices of our study. The first one is the CAC 40, the reference index in France which lists the 40 biggest market capitalization stocks. Let us focus first on the 7 years presidential term study.
It is possible to see a positive trend as the yearly returns are in average positive for the first 6 years. The highest value appears to be the year after following the presidential election with a return of 27,352% on the index. The only negative YAR (yearly average return) happens to be the year number 7 which is basically the year before the presidential election. It does not match with the findings made in the United States which showed that the year before the election normally has positive yearly returns in average. Except the findings made for the year 2, these results are not significant at 90% level of confidence.
Figure 1: YAR on the CAC 40 Stock Index: 5 Years Presidential Term
year 1 year 2 year 3 year 4 year 5 2007 ‐18.016% ‐34.766% 7.054% 13.978% ‐21.322%
2002 ‐31.455% 21.981% 11.939% 31.572% 17.831%
average ‐24.735% ‐6.393% 9.497% 22.775% ‐1.745%
sd 0.095029251 0.401266 0.034545 0.124409 0.276851
t‐test ‐3.68109017 ‐0.2253 3.887764 2.588893 ‐0.08915
‧
Figure 2: YAR on the CAC-Mid 60 Stock Index: 5 Years Presidential Term
When we focus at the 5 years presidential term, the results are not the same. In fact, a pattern closer to the one seen in the US can be found. The first two years show negative returns of respectively -24,735% and -6,393%. A trough seems to be reached during the second year like in the United States. In average, the worst yearly returns happen the year after the presidential election outcome. Again, it matches the results found for the United Stated. The next two years, 3 and 4 show higher positive YAR. Not like in the US, the third year does not show the highest YAR. The fourth year is ahead with 22,775% yearly returns in average. Nevertheless, if we look at this from another perspective, the fourth year in the French presidential term is the second year before the next election and, in this way, the results match the US-ones in the sense that in the US, the second year before the election shows the best YAR. Finally, the fifth year, which is the year previous to the new election, shows relatively small negative YAR, which is different from the US findings. It is important to mention that none of these results are significant at, at least 90% level of confidence. Nevertheless, they show a good trend and allow us to compare them with the US ones. The results seem different considering the 7-years and the 5-years presidential term. The majority of positive YAR for the 7-years terms may be explained by the fact that France faced better economic conditions during that period compared to the 5-years ones. The results regarding the 5-year term in France appear to be close to the US ones. Only
year 1 year 2 year 3 year 4 year 5
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
the last year YAR (the year before the election) does not seem to match the observation made in the US as they are (not significantly) negative.
The second is the CAC-mid 60 which lists the mid-capitalization stocks in France. Our findings show that negative yearly average returns happen on the year 1,2 and 5 while the positive ones happen to be the year 3 and 4. The worst year for this index returns is the year following the election day with a YAR of -18,361%. The best year for this index returns is the 4th year of the presidential election which is the penultimate year before the new election. The returns of the 3rd year (25,581%) appears to be the most relevant as the t-test value shows a result significant at a 10% level according to the t distribution table. As for the CAC 40, we can observe partly the same trend for the CAC-mid 60 as the one in the US as the worst year in returns is the first one and a trough is observable during the second year after the presidential election on this index but the year before the election does not show positive returns as the it is possible to see in the United States. Nevertheless a trend made of 3 successive “bear” years and 2 successive “bull”
years can be observed. In the US the period of “bear” market is much lower than the “bull” one (approximately 1 for 4 ratio). It is obviously not the case for this index, even the contrary as the
“bear” years are predominant (3 for 5 ratio).
‧
Figure 3: YAR on the TAIEX Stock Index: 4 Years Presidential Term
Focusing on the results for Taiwan, we will start with the main index: the TAIEX. A trend can be identified with the results we found and it does not look like what can be observed in the United States. In fact, the year 1,2,3,4 of the presidential term are alternatively showing, in average for our study period, positive and negative yearly returns. Negative values happen to be for the year 1 and 3 with very similar results of -4.823% and -5,645% while positive values happen to be for the year 2 and 4, again with very close results of 16,452% and 16,085%. None of these values appear to be significant at, at least 10% of significance. Indeed the results of the t-test do not overpass the critical values of significance which interest us. Nevertheless a pattern can be identified and this one is different from the one observed in the United States. In fact, the second year YAR is strongly positive (not significantly though) while in the US, this is the year when the market bottoms (negative abnormal returns are found). The recovery observed during the third year of the presidential term in the US (“best” year in term of returns) does not happen in Taiwan, the YAR being negative.
year 1 year 2 year 3 year 4
‧
Figure 4: YAR on the FTSE TWSE Taiwan 50 Stock Index: 4 Years Presidential Term
Figure 5: YAR on the FTSE TWSE Mid-Cap 100 Stock Index: 4 Years Presidential Term
Regarding the results obtained for The FTSE TWSE Taiwan 50 index and FTSE TWSE Mid-Cap 100 index , studied for the big and mid-caps indices in Taiwan, it is possible to see that a similar pattern can be established between the two indices. The YAR appear to be negative the year 1 and 4 for both indices while they are positive at the middle of the presidential term, for the year 2 and 3. Due to the low degree of freedom for this study (df=N-1=1, due to our small size sample), none of the results found are significant at the 10% significance level or more.
Nevertheless, it seems for these indices that the yearly returns are in average negative and lower the year before and after the election (the lowest value being the year after the election). The uncertainty of the election outcome and the confidence in the newly elected president might be factors affecting the market in this way. On the other way, the middle of the presidential term shows higher and positive yearly returns in average, the most important values being the second
2008 ‐47% 63% 7% ‐14%
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
year for both indices. Again, our findings do not match the research conducted in the United States. The possible intervention of the president on the US market as we discussed before does not seem to be an explanation here for Taiwan as the last year (year 4) of the presidential term shows yearly returns negative in average, which does not rationally fit this possibility.
Nevertheless, as for the United States the year after the election shows negative abnormal returns. We can reach the same conclusion that the one previously made in the United States, that is to say, the year after the presidential election is the one that shows the “worst” results on the stock market for both the indices FTSE TWSE Taiwan 50 and FTSE TWSE Mid-Cap 100.
Comparative study
As mentioned previously, the comparative study aims to study whether or not we can establish significant differences between the big and mid-caps indices, respectively for France and Taiwan, analyzing the average yearly returns for each year of the presidential term. We would then be able to know for France and Taiwan which index (big or mid-caps one) reacts significantly more strongly than the other during the presidential cycle. This study does not aim to explain the reason of the differences if some are found, but instead identify possible differences between different market-capitalization indices, the market capitalization being, we believe, a potential source of differences for the YAR and by extension, for the presidential cycles. We will then be able to see if different market-capitalization indices react significantly differently during the presidential cycle, a study never conducted before, which will give us more understanding of the presidential cycles of the two countries of our study, and possibly, prospective for further research.The hypotheses for the study are the following:
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
H0: The YAR for the big and mid-caps indices are identical
H1: The YAR for the big and mid-caps indices are significantly different
Regarding France, our results show that for the four years after the presidential election the YAR seems to be higher for the mid-caps index in comparison to the big-caps one. The contrary can be observed for the 5th year, the year before the presidential election. Nevertheless, none of these results are significant at 90+ % level of significance. For Taiwan, the YAR values are alternatively higher for the big-caps index and the mid-caps one. The YAR seems higher for the big-caps index for the year 1 and 3 and higher for the mid-caps index for the year 2 and 4. Again the t-test we conducted does not show any significance for those results and show a trend interesting to take into consideration. It is not possible then, for both France and Taiwan to reject the hypothesis H0 stating that, the YAR for the big and mid-caps indices are identical. It means we could not identify significant differences in the Yearly Average Returns for the big-caps and mid-caps indices defined for this study for both countries.
‧
國立 政 治 大 學