國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
2.7 Conclusion
In this paper, we examine whether the information of market volatility is able to improve the profitability of technical analysis. In the previous literature, market volatility acts as a key input in many financial issues. It is, however, not clear whether market volatility has any value in technical analysis. That is, could the information of market volatility enhance the profitability of technical trading rules? If it does, why and how does it work?
We adopt the VMA rule proposed by Chande (1992) as the representative of technical anal-ysis comprising the information of market volatility. Market volatility in the VMA rule is built to detect whether the market price makes big moves in up or down direction or whether it moves in a narrow range. If the market volatility does help the technical trading rules to detect market trend timelier, the VMA rule should be more profitable than other rules.
Using the Superior Predictive Ability test proposed by Hansen (2005), we find that the VMA rule outperforms others with higher profitability. Then we carry out the market timing ability test of Cumby and Modest (1987), and compare the predictive ability for upward and downward trends of the best VMA rule with that of the best MA and best MSV rule. The results that the best VMA rule enjoys better market timing ability may provide evidence to support the value of market volatility in better trend detecting ability.
Finally, we also investigate whether the best VMA rule has differential market timing ability in different market conditions. Empirical results suggest that in bull markets, the best VMA, best MA and best MSV rule do possess predictive ability for future upward trends, but the best VMA rule earns more daily profit than the others. On the other hand, the average loss per day the best VMA rule suffers is less than that of the best MA and best MSV rule in downward trend forecasts. In bear markets, the best VMA rule as well as the best MA and best MSV rule is able to forecast a price decrease well; however, the best VMA rule has less daily loss than the others in the upward trend forecast. As a whole, the best VMA rule outperforms the best MA and best MSV rule both in bull and bear markets.
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
3 Exploring the Information Content of Market Volatility in Technical Analysis
3.1 Introduction
In last chapter, the information of market volatility improving the profitability of technical anal-ysis is proved. We also explain its better performance possibly stemming from having better timing ability in generating profitable trading signals. Higher profitability and better timing ability display what investors can profit from the information of market volatility considered in forecasting. However, how the market volatility ratio is informationally relevant for the deter-mination of trading signals in technical analysis is still unknown. Thus, the intent of this chapter is to investigate how the information of market volatility affects the generation of trading signals in technical analysis based on the case of the variable moving average (VMA) rule.
The information of market volatility is used to measure whether there is trading opportunity through examining more recent price movements are big or small. The theoretical mechanism of the information of market volatility on affecting the trading signals in the VMA rule appears to be nonlinear and it is usually difficult to empirically capture such nonlinear relationship.
However, due to the decision mechanism in the moving average, buying (or selling) when the moving average rises above (or falls below) the current price, using the time-varying-transition-probability (TVTP) Markov switching model can account for their nonlinear relationship.6 We use a two-state Markov switching model for the gap between the price line and the VMA line in which the transition probabilities of regime-switching is allowed to respond to changes in the information of market volatility. As a result, studying how the information of market volatility influences the regime switching between ”price-above-VMA” and ”price-below-VMA” states is as well as investigating how it affects the generation of trading signals.
Our estimation results indicate that the increase of the change in market volatility will leads to a higher probability of generating signals in the VMA rule. Furthermore, its effect on trading signals is asymmetric across bull and bear markets. The effect that the increase of market
6For the detail of the TVTP Markov switching model, see the study of Filardo (1994) and Diebold et al. (1994).
‧
volatility producing higher probability of a selling signal generated is stronger in bear markets than in bull markets. While for a buying signal, its effect in bull markets is larger than in bear markets. These empirical results coincide with the theoretical purpose of the information of market volatility in the VMA rule, taking better advantage of the market price movements.
In this chapter, we also re-explore the value of the information of market volatility for a particular simple MA rule. From the study in last chapter, how the technical trading rule adjusts its time the signal generated to the information of market volatility cannot be revealed directly no matter in the SPA tests or the market timing tests, since the time that the chosen trading rules signal to buy or sell are totally different. It is resulted from the best VMA rule and other trading rules (i.e., the best MA and best MSV rules) we chose to compare are based on the different parameter settings (i.e., the best VMA and best MA rule) or different trading mechanisms (i.e., the best VMA and best MSV rule).
In order to clearly figure out the value of market volatility in technical analysis, we add the information of market volatility to a particular n-day simple MA rule. Through using the Fixed-transition-probability (FTP) and the TVTP Markov-switching models, we compare how the time, the signal generated, changes and how the profit varies after considering the information of market volatility in detecting price movements. For the sake of checking the robustness of the empirical results, six simple MA rules with different settings for period n, 5, 20, 40, 75, 100 and 250, are studied, and we categorize these trading signals into four types according to the time point the trading signals generated between the FTP and the TVTP Markov-switching models. Our results reveal that the same information of market volatility has similar effect on the generation of trading signals for all simple MA rules. We find that the signal time change due to the information of market volatility will benefit investors with higher profit. This result again verifies the information of market volatility is the valuable information in future price movement prediction for technical analysis.
The chapter is organized as follows. Section 2 introduces the moving average trading sys-tems and discusses the theoretical design of market volatility in moving averages. Section 3 presents the TVTP Markov-switching model, describes the data and reports the empirical
re-‧
sults in the case the VMA rule. Section 4 contains the estimation results in the TVTP Markov-switching model, data description and the profitability analysis for six simple MA rules. The analysis for four types of trades is also presented. Finally, Section 5 presents conclusions.