Multifractal Analysis of Stock Market Prices
| Název práce v češtině: | Analýza multifraktality akciových trhů |
|---|---|
| Název v anglickém jazyce: | Multifractal Analysis of Stock Market Prices |
| Klíčová slova: | multifraktalita, Hurstův exponent, dlouhá paměť, analýza časových řad |
| Klíčová slova anglicky: | multifractality, Hurst exponent, long memory, time series analysis |
| Akademický rok vypsání: | 2010/2011 |
| Typ práce: | diplomová práce |
| Jazyk práce: | angličtina |
| Ústav: | Institut ekonomických studií (23-IES) |
| Vedoucí / školitel: | prof. PhDr. Ladislav Krištoufek, Ph.D. |
| Řešitel: | skrytý - zadáno vedoucím/školitelem |
| Datum přihlášení: | 04.06.2011 |
| Datum zadání: | 08.06.2011 |
| Datum a čas obhajoby: | 11.09.2013 00:00 |
| Místo konání obhajoby: | IES |
| Datum odevzdání elektronické podoby: | 18.07.2013 |
| Datum proběhlé obhajoby: | 11.09.2013 |
| Oponenti: | prof. Ing. Miloslav Vošvrda, CSc. |
| Kontrola URKUND: | |
| Seznam odborné literatury |
| 1. BARUNÍK, Josef; KRIŠTOUFEK, Ladislav. On Hurst exponent estimation under heavy-tailed distriubution. Physica A. 2010, 389. p. 3844-3855
2. COUILLARD, Michel; DAVISON, Matt. A comment on measuring the Hurst exponent of financial time series. Physica A. (2005), 348. p. 404-418 3. KRIŠTOUFEK, Ladislav. On spurious anti-persistence in the US Stock Indeces. Chaos, Solitons&Fractals. (2010), 43. p. 68-78 4. LO, Andrew W. Long-Term Memory in Stock Market Prices. Econometrica. (Semptember 1991), 59. p.1279-1313 5. MATTEO, T.Di; ASTE, T.; DACORAGNA, M.M. Scaling behaviours in differently developed makrets. Physica A. (2003), 324. 183-188 6. MISHRA, Ritesh Kumar; SEHGAL, Sanja; BHANUMURTHY, N.R. A search for longe-range dependence and chaotic structure in Indian stock market. Review of Financial Economics. (2011), 20, p. 96-104 7. ZUNINO, L., et al. A multifractal approach for stock market inefficiency. Physica A. (2008), 387. p.6558-6566 |
| Předběžná náplň práce |
| The aim of my thesis is to apply multifractal analysis on group of stock market returns. The assumption of no long-term memory is present in asset pricing theories but it has been questioned by many researchers. Several studies tested for long-range dependence in stock indices, especially in the very important indices. Nevertheless, some of them have been criticized because their analyses were influenced by sampling properties of their methods under heavy-tailed data. They therefore provided biased results. Studies concerning returns of individual stock are less common and these are therefore the main goal of my thesis. Moreover, both the data from developed and developing stock markets are going to be used and the results are going to be compared to verify the hypothesis that developing markets are persistent while developed markets are not. With the increase in trading volume on the stock markets, they are believed to be more efficient and any long-term memory should disappear. Therefore I will also look on development in time. Data on stock returns are available on yahoo finance.
1. Hypothesis #1: There are long-range dependencies in stock market returns. 2. Hypothesis #2: There are differences in persistence between developing and developed stock markets. 3. Hypothesis #3: Long-term memory has been disappearing in time. Concerning the analysis itself, three methods are going to be applied in order to analyze multifractality of the stock market data. These are the generalized Hurst exponent, detrended fluctuation analysis and detrended moving average. The advantage of generalized Hurst exponent is that it is suitable for processes with heavy tails. The null hypothesis is that there is no long memory present. The logarithmic transformed weakly and daily data are going to be used. The results for developing and developed markets are going to be compared and development in time is going to be checked. 1. Literature Review: a. Description of methods with their advantages and disadvantages b. Overview of empirical studies 2. Data description 3. Multifractal analysis: a. Generalized Hurst exponent b. Modified rescaled range analysis 4. Discussion of results: a. Developing versus developer stock markets b. Changes in results in time |
| Předběžná náplň práce v anglickém jazyce |
| The aim of my thesis is to apply multifractal analysis on group of stock market returns. The assumption of no long-term memory is present in asset pricing theories but it has been questioned by many researchers. Several studies tested for long-range dependence in stock indices, especially in the very important indices. Nevertheless, some of them have been criticized because their analyses were influenced by sampling properties of their methods under heavy-tailed data. They therefore provided biased results. Studies concerning returns of individual stock are less common and these are therefore the main goal of my thesis. Moreover, both the data from developed and developing stock markets are going to be used and the results are going to be compared to verify the hypothesis that developing markets are persistent while developed markets are not. With the increase in trading volume on the stock markets, they are believed to be more efficient and any long-term memory should disappear. Therefore I will also look on development in time. Data on stock returns are available on yahoo finance.
1. Hypothesis #1: There are long-range dependencies in stock market returns. 2. Hypothesis #2: There are differences in persistence between developing and developed stock markets. 3. Hypothesis #3: Long-term memory has been disappearing in time. Concerning the analysis itself, three methods are going to be applied in order to analyze multifractality of the stock market data. These are the generalized Hurst exponent, detrended fluctuation analysis and detrended moving average. The advantage of generalized Hurst exponent is that it is suitable for processes with heavy tails. The null hypothesis is that there is no long memory present. The logarithmic transformed weakly and daily data are going to be used. The results for developing and developed markets are going to be compared and development in time is going to be checked. 1. Literature Review: a. Description of methods with their advantages and disadvantages b. Overview of empirical studies 2. Data description 3. Multifractal analysis: a. Generalized Hurst exponent b. Modified rescaled range analysis 4. Discussion of results: a. Developing versus developer stock markets b. Changes in results in time |