Font Size: a A A

Hidden Markov Algorithm In Timing Detecting

Posted on:2020-09-16Degree:MasterType:Thesis
Country:ChinaCandidate:R Z JiangFull Text:PDF
GTID:2370330626964688Subject:Applied statistics
Abstract/Summary:PDF Full Text Request
With the continuous development of China's capital market and financial investment theories,investing in index-based financial products is increasingly favored by people because of its low management costs,and relevant theories and methods are increasingly researched by more and more people.The main investment theories can be divided into two categories,the passive and the active and management.The main work of this paper is to combine these two types of strategies and apply them to the ETF300 index.This paper focuses on two aspects,selecting factor construction index tracking method and using the hidden Markov model to model the numerical sequence,and then use implicit least squares algorithm to solve the timing problem of index tracking.The main research goal of this paper is to combine passive management with quantitative investment methods,whose management cost can be low and also capture market information by active investment,and explore the combination of them.In the construction of the algorithm,exponential tracking is used as the first layer strategy,and the timing is used as the second layer strategy.By using the quantitative method to fine-tune the passive investment,we can significantly raise the goal of combining excess returns.In the empirical analysis,the ETF300 index-related daily data will be chose.The time span is from the index structure to the present,the data set is divided into training set and test set,the training set data is used to determine the appropriate factor,and the hidden Markov is run in real time on the test set.The algorithm and the implicit minimum algorithm will be used to determine the timing of the appropriate timing.Selecting the maximum retracement and excess return as the performance indicator,and the static classic algorithm is used as the benchmark.Through the empirical analysis of this paper,we can see that the combined algorithm does achieve the expected results.At the same time,the algorithm is simple in construction,and can change the specific configuration in the same framework to add many basic methods.
Keywords/Search Tags:Hidden Markov Model, Timing, Stock Index, Factor, Utility Maximum Portfolio
PDF Full Text Request
Related items