The Indian stock markets: can technical analysis and economic indicators be combined to develop a more efficient portfolio decision-making model?

Abstract

The perceived limitation of technical Analysis was that patterns (as distinguished from prices) needed a time lag to adsorb all information. This study aimed to test empirically a model which attempted to overcome this limitation by supplementing Technical Analysis Decisions by supplementing it with those economic indicators which did not get sufficient time to manifest themselves in the Patterns(i.e. the investor took a decision before the pattern could reflect the same). Prices of 5 scripts for 3 years were subjected to this analysis. Economic indicators were noted from the announcements in the Economic Times. The study was further classified into four sub studies – testing the effectiveness of Individual Technical Analysis Secondary Indicators, testing the efficiency of a composite model which used the signals of all the indicators, testing the assumption that economic indicators indeed affect investors(and finding which indicators do) and then finally testing the model, which incorporated the results of the third study wherever there was no time for the announcement to get reflected in the patters, by comparing the results with the second study. The results of the study were as follows: Sub study 1.1: There were no statistically significant differences between the efficiency of different indicators, as shown by the chi-square test. There were no significant differences across buy or sell decisions either. However for a few scripts, one or two trends came across – the 5-day stochastic and the 7 & 15 day moving average were significantly more efficient indicators for GSFC and TISCO, while the 15-day ROC was more efficient only for GSFC. The sell decisions were stronger for Reliance. However these decisions could not be generalized and extended to all scripts. Sub study 1.2: The composite model taking into consideration all signals of indicators was significantly superior and had an overall predictability of 68%. Sub study 2.1: Indian Investors definitely did not react to: -general announcements relating to the policy approach of the government. -Macroeconomics Variables as GDP growth rate, savings rate, Plan size and composition etc. -Inflation, WPL indices, etc. -Monetary policy excluding those directly concerning the interest rates and the call money rates. -The scope and scale of the budget and the budget deficit. However Indian Investors definitely reacted to , -Specific announcements related to any privatization and decontrol. -any information related to the monsoons -Any measure affecting the pricing or the cost of raw materials. -Expectations regarding the budget(soft of harsh) and the actual budget announcements. -Availing of foreign aid, and to factors affecting the flow of foreign funds in India. Sub study 2.2: The hypothesis that by incorporating the impact of those economic indicators which did not get time to get reflected in the patterns, one can improve the effectiveness of technical Analysis, could not: be proved statistically. However, in general, the prediction efficiency was always higher, hence, a rational investor who does follow this method cannot be said to be worse off from the one who relies on the composite technical analysis model devised by us. Of course, both the models developed by us were about three times more efficient than pure technical analysis.