Back to results
US 8,712,897 B2Granted
Stock analysis method, computer program product, and computer-readable recording medium
OwnerHwey-Chyi Lee, Sheng-Shyr Chenginventors
Life of this patent
Estimating the term…
Abstract
In a stock analysis method for performing an analysis on stocks to select target ones to be bought/sold from the stocks, each stock is grouped into a corresponding group based on stock return data thereof, market return data and industry return data of each corresponding classified industry. Clustering data for each stock corresponding to each time interval and associated with the groups is obtained based a clustering mode. Analysis data for each stock corresponding to a coming time interval is estimated based on the corresponding clustering data. Any ones of the stocks, whose analysis data matches predetermined selection criteria, are determined as the target stocks.
Claims14
A stock analysis method for performing an analysis on a plurality of stocks to select target ones to be bought/sold from the stocks, the company of each of the stocks belonging to a corresponding classified industry, said stock analysis method comprising the steps of:
- a) calculating stock return data of each of the stocks, market return data, and classified industry return data of each of the corresponding classified industries based on historical stock price information within a historical trading period that includes a current trading period and that consists of a number (N) of consecutive time intervals;
- b) according to the stock return data, the market return data and the classified industry return data obtained in step a), determining
- whether a stock return of each of the stocks in each of the N time intervals is greater than a classified industry return of the corresponding classified industry in a corresponding one of the N time intervals,
- whether the stock return of each of the stocks in the corresponding one of the N time intervals is greater than a market return in the corresponding one of the N time intervals, and
- whether the classified industry return of the corresponding one of the classified industries in the corresponding one of the N time intervals is greater than the market return in the corresponding one of the N time intervals;
- c) based on results determined in step b), grouping the stocks so that each of the stocks in each of the N time intervals is grouped into a corresponding one of a number (G) of different groups;
- d) obtaining clustering data of each of the stocks corresponding to each of the N time intervals and associated with the groups based on a specific one of the groups using a clustering mode;
- e) estimating by a computer, analysis data of each of the stocks corresponding to a coming time interval based on at least the clustering data obtained in step d);
- and
- f) determining any ones of the stocks whose analysis data estimated in step e) matches predetermined selection criteria as the target ones of the stocks;
- wherein, in step d):
- the clustering mode is one of a combination clustering mode and a permutation clustering mode;
- in the combination clustering mode, the clustering data of each of the stocks corresponding to an ith one of the N time intervals includes the corresponding one of the groups in the ith one of the N time intervals, and a number (Si) of the time intervals in a reference period from [i−(Q−1)]th to ith ones of the N time intervals, where 1≦i≦N and 2≦Q≦i, wherein a corresponding one of the stocks is grouped into said specific one of the groups in the number (Si) of the time intervals;
- and
- in the permutation clustering mode, the clustering data of each of the stocks corresponding to the ith one of the N time intervals includes the corresponding one of the groups in the ith one of the N time intervals, and group permutation pattern consisting of the corresponding ones of the groups that correspond respectively to [i−(Q−1)]th to ith ones of the N time intervals.
12 more elements
A computer program product comprising a non-transitory computer readable medium storing instructions for causing a computer to perform consecutive steps of:
- a stock analysis method for performing an analysis on a plurality of stocks to select target ones to be bought/sold from the stocks, the company of each of the stocks belonging to a corresponding classified industry, the stock analysis method including the steps of:
- a) calculating stock return data of each of the stocks, market return data, and classified industry return data of each of the corresponding classified industries based on historical stock price information within a historical trading period that includes a current trading period and that consists of a number (N) of consecutive time intervals;
- b) according to the stock return data, the market return data and the classified industry return data obtained in step a), determining
- whether a stock return of each of the stocks in each of the N time intervals is greater than a classified industry return of the corresponding classified industry in a corresponding one of the N time intervals,
- whether the stock return of each of the stocks in the corresponding one of the N time intervals is greater than a market return in the corresponding one of the N time intervals, and
- whether the classified industry return of the corresponding one of the classified industries in the corresponding one of the N time intervals is greater than the market return in the corresponding one of the N time intervals;
- c) based on results determined in step b), grouping the stocks so that each of the stocks in each of the N time intervals is grouped into a corresponding one of a number (G) of different groups;
- d) obtaining clustering data of each of the stocks corresponding to each of the N time intervals and associated with the groups based on a specific one of the groups using a clustering mode;
- e) estimating analysis data of each of the stocks corresponding to a coming time interval based on at least the clustering data obtained in step d);
- and
- f) determining any ones of the stocks whose analysis data estimated in step e) matches predetermined selection criteria as the target ones of the stocks;
- wherein, in step d):
- the clustering mode is one of a combination clustering mode and a permutation clustering mode;
- in the combination clustering mode, the clustering data of each of the stocks corresponding to an ith one of the N time intervals includes the corresponding one of the groups in the ith one of the N time intervals, and a number (Si) of the time intervals in a reference period from [i−(Q−1)]th to ith ones of the N time intervals, where 1≦i≦N and 2≦Q≦i, wherein a corresponding one of the stocks is grouped into said specific one of the groups in the number (Si) of the time intervals;
- and
- in the permutation clustering mode, the clustering data of each of the stocks corresponding to the ith one of the N time intervals includes the corresponding one of the groups in the ith one of the N time intervals, and group permutation pattern consisting of the corresponding ones of the groups that correspond respectively to [i−(Q−1)]th to ith ones of the N time intervals.
13 more elements
A system for performing an analysis on a plurality of stocks to select target ones to be bought/sold from the stocks, the company of each of the stocks belonging to a corresponding classified industry, the system comprising:
- one processor and a non-transitory computer readable medium, wherein the non-transitory computer readable medium stores instructions that when executed by the processor, causes the processor to perform the steps of:
- a) calculating stock return data of each of the stocks, market return data, and classified industry return data of each of the corresponding classified industries based on historical stock price information within a historical trading period that includes a current trading period and that consists of a number (N) of consecutive time intervals;
- b) according to the stock return data, the market return data and the classified industry return data obtained in step a), determining
- whether a stock return of each of the stocks in each of the N time intervals is greater than a classified industry return of the corresponding classified industry in a corresponding one of the N time intervals,
- whether the stock return of each of the stocks in the corresponding one of the N time intervals is greater than a market return in the corresponding one of the N time intervals, and
- whether the classified industry return of the corresponding one of the classified industries in the corresponding one of the N time intervals is greater than the market return in the corresponding one of the N time intervals;
- c) based on results determined in step b), grouping the stocks so that each of the stocks in each of the N time intervals is grouped into a corresponding one of a number (G) of different groups;
- d) obtaining clustering data of each of the stocks corresponding to each of the N time intervals and associated with the groups based on a specific one of the groups using a clustering mode;
- e) estimating analysis data of each of the stocks corresponding to a coming time interval based on at least the clustering data obtained in step d);
- and
- f) determining any ones of the stocks whose analysis data estimated in step e) matches predetermined selection criteria as the target ones of the stocks;
- wherein, in step d):
- the clustering mode is one of a combination clustering mode and a permutation clustering mode;
- in the combination clustering mode, the clustering data of each of the stocks corresponding to an ith one of the N time intervals includes the corresponding one of the groups in the ith one of the N time intervals, and a number (Si) of the time intervals in a reference period from [i−(Q−1)]th to ith ones of the N time intervals, where 1≦i≦N and 2≦Q≦i, wherein a corresponding one of the stocks is grouped into said specific one of the groups in the number (Si) of the time intervals;
- and
- in the permutation clustering mode, the clustering data of each of the stocks corresponding to the ith one of the N time intervals includes the corresponding one of the groups in the ith one of the N time intervals, and group permutation pattern consisting of the corresponding ones of the groups that correspond respectively to [i−(Q−1)]th to ith ones of the N time intervals.
13 more elements
Dependencies are inferred from the claim text.