Stock item selection method and portfolio management method using trend strength index based on open, high, low and close price of stock

Abstract

The present invention relates to a method for selecting stock items and constructing a portfolio using a trend strength indicator based on OHLC data, comprising: (a) a step of calculating the central price of each stock item from the stock price fluctuating during the day (S10); (b) a step of converting the central price into a converted price for comparison between stock items (S20); (c) a step of determining a straight line connecting the stock prices on the observation start date and end date, and a straight line connecting the maximum and minimum points during the observation period (S30); (d) a step of determining an ideal trend line which is a straight line in which the stock price moves with an ideal slope during the observation period (S40); (e) a step of calculating the opening / closing energy, the high / low energy, and the peak energy by utilizing the energy concept (S50); and (f) a step of calculating the trend strength based on the opening / closing energy, the high / low energy, and the peak energy (S60), and is characterized by selecting promising stock items using the trend strength.

Description

Technology Field

[0001] The present invention relates to a method for selecting stock items and constructing a portfolio using an OHLC-based trend strength indicator, and more specifically, to a method for utilizing all stock opening, high, low, and closing price data, calculating trend strength by combining the directionality and path efficiency of the stock price, and then selecting promising stock items and constructing a portfolio based on the same. Background Technology

[0002] Recently, Home Trading Systems (HTS), which allow individuals to directly buy or sell stocks, are becoming increasingly popular.

[0003] In particular, interest in stock investment is gradually increasing as stock trading via personal mobile devices becomes possible.

[0004] When investing in stocks, it is important to refer to various stock price indicators to predict whether prices will rise or fall, and to determine the appropriate timing for buying and selling.

[0005] Examples of such stock price indicators include stock price fluctuation, the Price-Earnings Ratio (PER), the Price-to-Book Ratio (PBR), return on capital, and earnings yield.

[0006] However, when evaluating a company's value using earnings per share and the price-to-earnings ratio, only the most recent performance is reflected, so stock market trends and supply and demand conditions are not considered, which can lead to incorrect judgments.

[0007] In particular, trading in stocks where a downtrend or highly volatile market has formed, or where prices may skyrocket or plummet abnormally, can cause confusion among individual investors.

[0008] In addition, since stock prices fluctuate due to numerous factors, it is difficult to accurately predict the future stock market, even for experts with extensive information and experience.

[0009] Moreover, individual investors have difficulty selecting investment stocks by utilizing various stock price indicators.

[0010] Accordingly, various financial institutions and securities analysts analyze corporate operating performance, asset fluctuations, and stock price trends to recommend promising stocks, and individuals often refer to this information.

[0011] Meanwhile, recently, solutions that allow for the selection of investment priorities using various stock price indicators are also being provided.

[0012] However, even these solutions have limitations in providing accurate corporate information or indicators of investment value.

[0013] Accordingly, the inventors intend to calculate a new trend strength indicator by utilizing all of the stock's opening, high, low, and closing price data, and to propose a method for selecting stock items using this indicator.

[0014] Meanwhile, a search for prior art related to the present invention yielded numerous prior art records, some of which are introduced below.

[0015] Patent Document 1 discloses a stock information provision method capable of selecting undervalued stocks, characterized by comprising: a collection step in which corporate information and stock price indicators are collected; an auxiliary indicator calculation step in which auxiliary indicators are calculated using the stock price indicators as variables; an auxiliary indicator ranking calculation step in which rankings for each auxiliary indicator are calculated; a value indicator calculation step in which auxiliary indicators are combined for each stock and the rankings of the corresponding auxiliary indicators are summed to calculate a value indicator; and a value indicator ranking selection step in which rankings are selected for each value indicator combined with each auxiliary indicator in order of lowest value of the value indicator.

[0016] Patent Document 2 relates to a stock ranking analysis system and a stock selection method using the same. In a stock asset management system in which a plurality of user terminals and a server are connected via a network, the method for selecting stocks comprises the steps of: the server calculating score and ranking information through net profit analysis for individual stocks, calculating score and ranking information through sentiment analysis, and calculating a comprehensive score and ranking information by integrating the net profit score and the sentiment score; when there is a request for stock ranking analysis from a user terminal connected via a network to the server, the server providing the calculated score and ranking information to the user terminal; the server updating the score and ranking information according to a set period and providing the score and ranking information updated in real time to the user terminal; and the server also providing the trend of the score and ranking information to the user terminal according to a set period. Prior art literature

[0017] Korean Published Patent No. 10-2014-0134877 (Published Nov. 25, 2014) Korean Published Patent No. 10-2013-0123663 (Published Nov. 13, 2013) The problem to be solved

[0018] The present invention is proposed to solve the aforementioned conventional problems and aims to present a new trend strength indicator utilizing all stock opening, high, low, and closing price data, and to provide a method for selecting stock items and constructing a portfolio using this indicator.

[0019] Another objective of the present invention is to maximize the profitability of stock investment by selecting stock items based on trend strength according to the present invention.

[0020] Another objective of the present invention is to improve the accuracy of stock price prediction by combining the directionality of the stock price and path efficiency when calculating trend strength.

[0021] Another objective of the present invention is to minimize losses associated with stock investment by constructing a portfolio using the trend strength indicator according to the present invention. means of solving the problem

[0022] To solve the above problem, the stock selection method according to the present invention comprises a stock analysis device comprising a computer program for analyzing stock prices and hardware for implementing said program, and a method for selecting promising stocks using a trend strength indicator, wherein the method comprises: (a) a step of calculating the central price of each stock from the stock price fluctuating during the day in said stock analysis device (S10); (b) a step of converting said central price into a converted price for comparison between stocks (S20); (c) a step of determining an opening-closing line, which is a straight line connecting the stock prices on the observation start date and end date, and a high-low line, which is a straight line connecting the maximum and minimum points during the observation period (S30); (d) a step of determining an ideal trend line, which is a straight line in which the stock price moves with an ideal slope during the observation period (S40); (e) a step of calculating opening-closing energy, high-low energy, and peak energy by utilizing the concept of energy (S50); and (f) a step of calculating the trend strength based on said opening-closing energy, high-low energy, and peak energy (S60), and using said trend strength It is characterized by the selection of promising stock items.

[0023] In addition, the above S10 step is characterized by calculating the central price by considering all data regarding the Open Price, High Price, Low Price, and Closing Price for the day.

[0024] In addition, in the above S20 step, when calculating the conversion price, the above central price is converted to a level of 10,000 won for comparison between stocks.

[0025] In addition, it is characterized by including a process of determining the above-mentioned straight line and high-low line in the form of a linear function having a slope and an intercept, respectively.

[0026] In addition, it is characterized by including a process of determining a bisector that bisects the angle between the two lines, the above-mentioned straight line and the high-low line.

[0027] In addition, the above trend strength is characterized by being calculated as follows.

[0028]

[0029] In addition, after the above S60 step, it is characterized by further including a step of calculating a bisector divergence rate, which indicates the divergence between the actual price path and the bisector line, and a trend divergence rate, which indicates the divergence between the actual price path and the trend line.

[0030] In addition, the above-mentioned bisection divergence rate and trend divergence rate are characterized by being calculated as the average of the distances between the actual center price and the ideal trend line at each point in time.

[0031] In addition, it is characterized by further including a step of deriving the trend strength in the form of a score from 0 to 100 by combining the above-mentioned divergence rate and trend strength.

[0032] And the method for constructing a stock portfolio using trend strength according to the present invention is characterized by calculating the trend strength of each stock at the end of each month, and dividing the portfolio into an upper group of 20%, a middle group of 60%, and a lower group of 20% according to the calculated trend strength ranking. Effects of the invention

[0033] According to the present invention, stock items can be selected and a portfolio constructed using a trend strength indicator calculated by utilizing all of the stock's opening, high, low, and closing price data.

[0034] This has the effect of maximizing the profitability of stock investments.

[0035] In addition, by combining the directionality of the stock price and path efficiency when calculating trend strength, it is possible to improve the accuracy of stock price prediction.

[0036] In addition, by constructing a portfolio using the trend strength indicator according to the present invention, there is an effect of minimizing losses in stock investment.

[0037] In addition, the trend strength indicator according to the present invention has the effect of being usable in various ways, such as as a practical filtering tool that complements existing momentum strategies, a risk management device that reduces risk during sudden market changes, and a complementary signal applicable to asset allocation and sector strategies. Brief explanation of the drawing

[0038] FIG. 1 is a flowchart illustrating a method for calculating trend strength according to the present invention. Specific details for implementing the invention

[0039] The following describes a method for selecting stock items and constructing a portfolio using a trend strength indicator based on OHLC data according to the present invention.

[0040] First, the concept of the "Trend Strength Index" proposed in this invention will be explained.

[0041] The trend strength proposed in the present invention is calculated by considering all data on the Open Price, High Price, Low Price, and Closing Price for one day (hereinafter simply referred to as 'OHLC data').

[0042] In other words, the trend strength of the present invention is a concept further expanded from the conventional momentum signal, which simply indicates the direction of the price, and aims to quantify how consistently and efficiently the price has moved over a certain period.

[0043] Traditional technical analysis of stock prices focused on capturing directional reversals using methods such as moving averages or breakout rules, but recently it has been evolving to statistically define patterns and emphasize the reproducibility of signals.

[0044] In particular, while time-series momentum has been universally observed across various asset classes, even with the same rise, future performance and risk can differ depending on whether the price path fluctuates wildly or rises smoothly.

[0045] The trend strength indicator of the present invention, stemming from this awareness of the problem, can serve as a new measurement framework that reflects not only simple directionality but also path characteristics (reversal, irregularity, etc.).

[0046] In other words, the trend strength indicator according to the present invention can reduce signal noise and increase the reproducibility of predictive power by simultaneously reflecting directionality and path efficiency.

[0047] Figure 1 below shows the average trend strength from 2015 to 2024.

[0048] [Figure 1] Average trend strength 2015–2024

[0049]

[0050] The present invention utilizes all OHLC data information to design a path inefficiency penalty based on volatility estimates derived from high and low ranges and opening and closing price positions, and calculates a synthetic score by combining this with existing directional indicators (e.g., 3-, 6-, and 12-month cumulative returns).

[0051] In addition, for portfolio construction, trend strength scores for all stocks are calculated at the end of each month and classified into upper, middle, and lower groups.

[0052] Below, we will examine the method for calculating the trend strength indicator based on OHLC according to the present invention and how the trend strength indicator affects actual investment performance.

[0053] The subject of analysis was the Korean stock market, and the sample was limited to companies listed on the KOSPI and KOSDAQ during the period from January 2015 to December 2024.

[0054] This period encompasses the stabilization phase following the global financial crisis, the COVID-19 pandemic, and the recent high-interest rate environment, making it a time to evaluate the durability of strategies amidst various shocks in the market environment.

[0055] In addition, survivorship bias was minimized during the sample selection process by excluding companies with incomplete data due to delisting, mergers, or spin-offs.

[0056] In addition, considering transaction liquidity, only companies that recorded an average transaction value above a certain level were included.

[0057] In addition, companies with an average daily transaction value of 100 million won or more were selected as a sample to increase the likelihood of actual transactions and prevent excessive illiquidity effects from distorting the results.

[0058] In addition, to ensure industry representation, all sectors were evenly included based on the KRX industry classification.

[0059] This not only ensured the comprehensiveness of the sectors but also minimized potential factors that could distort the performance of the trend strength indicator.

[0060] Ultimately, approximately 1,200 stocks were included in the analysis sample, which can be considered a sufficiently comprehensive scope compared to existing domestic and international momentum and technical analysis studies.

[0061] In addition, the above OHLC data was collected from the Korea Exchange (KRX) and financial information providers (FnGuide, Bloomberg).

[0062] The collected price data consists of daily Open Price, High Price, Low Price, and Close Price, and trend strength was calculated based on this.

[0063] In addition, all prices were adjusted to reflect corporate events such as dividends, stock splits, and paid-in capital increases.

[0064] If the above adjusted stock price is not used, the trend strength value may be distorted due to factors such as ex-dividends or stock splits.

[0065] In addition, to reflect transaction costs and market friction, actual measures of return were used that take into account liquidity conditions at the time of buying and selling.

[0066] In addition, outliers and missing values ​​that may occur during the data pre-processing process were handled using Winsorizing and linear interpolation.

[0067] In addition, the normality of the time series analysis was enhanced by using the log return when calculating the rate of return.

[0068] The aforementioned logarithmic return has time-additive characteristics compared to simple returns, which possesses essential advantages in long-term cumulative performance analysis.

[0069] The process of calculating trend strength according to the present invention will be explained in detail below with reference to FIG. 1.

[0070] The trend strength calculation method according to the present invention summarizes OHLC (opening price, high price, low price, closing price) data information for one day into a single Central Price, and then calculates the trend line and divergence based on this to finally derive a trend strength score.

[0071] First, the central price, which is the stock price that fluctuates over the course of a day. It is calculated as follows.

[0072]

[0073] Next, for comparison between stocks, the converted price obtained by converting the above-mentioned center price to a level of 10,000 won. Calculate.

[0074]

[0075] Here represents the central price on the observation start date.

[0076] Let the observation start and end dates be x-start and x-end, respectively, and the stock prices on the start and end dates be y-start and y-end, respectively. Let the days on which the maximum and minimum stock prices appear during the observation period be x-max and x-mini, respectively, and the maximum and minimum values ​​be y-max and y-mini, respectively. Then, two straight lines can be found.

[0077] In other words, if we define the straight line connecting the stock price on the observation start date and the stock price on the observation end date as the start-end line, and the straight line connecting the maximum and minimum points during the observation period as the high-low line, then the line connecting the start and end class, Connecting high and low prices You can obtain.

[0078] The above-mentioned straight line and high-low line can each be represented by the following linear functions.

[0079]

[0080] Here, a represents the slope of the starting line, b represents the intercept of the starting line, c represents the slope of the rising line, and d represents the intercept of the rising line.

[0081] Bisects the angle between the two lines, the above-mentioned straight line and the straight line. It can be represented as a linear function in the form of a straight line as follows.

[0082]

[0083] Here, e represents the slope of the bisector calculated by combining the slope of the above-mentioned straight line (a) and the slope of the high-low line (c).

[0084] And f represents the bisector intercept calculated by combining the intercept (b) of the above-mentioned straight line and the intercept (d) of the vertical line.

[0085] The aforementioned bisector refers to a straight line that exactly divides in half the angle between the start-end line, which indicates directionality connecting the beginning and the end, and the high-low line, which indicates variability connecting the highest and lowest points.

[0086] And using the angle bisector theorem, e and f can be found as follows.

[0087]

[0088] For reference, the bisector above is also a line that collects all points equidistant from the starting and ending lines, and this property can be used to find e and f.

[0089] And the ideal trend line, which is a straight line assuming the stock price moves daily with an optimal slope during the observation period, can be expressed in the form of a linear function as follows.

[0090]

[0091] Here, g represents the slope of the ideal trend line, indicating the total range of price fluctuation (y-max-y-minimum) over the entire observation period (x-end-y-start), and signifies the slope of maximum efficiency through which the price can move within the period.

[0092] And h is the intercept representing the central price (start of y) on the observation start date, signifying the price point where the ideal trend line begins.

[0093] That is, the above ideal trend line represents the path assuming that the stock price rose or fell most efficiently and evenly during the observation period.

[0094] Trend strength is calculated using the concept of energy, and the initial energy ( ), high and low energy( ), supreme energy( ) is calculated as follows.

[0095]

[0096] Here, E-start / end refers to energy based on the start / end line, E-high / low refers to energy based on the high / low line, and E-maximum refers to energy based on the maximum fluctuation range.

[0097] Based on the aforementioned opening / closing energy, high / low energy, and peak energy, the strength of the upward trend is calculated as follows.

[0098]

[0099] In other words, the strength of the upward trend is calculated by averaging the ratio of the opening and closing energies and the high and low energies relative to the peak energy.

[0100] And when the deviation from the bisector is defined as |y-(ex+f)| and the deviation from the trend line as |y-(gx+h)|, the deviation rate between the actual price path and the bisector and trend line is calculated as follows.

[0101]

[0102] Here, |y-(ex+f)| represents the distance between the actual price (y) and the bisector, and |y-(gx+h)| represents the distance between the actual price (y) and the ideal trend line.

[0103] And the final divergence rate is calculated as bisector divergence rate / trend divergence rate x 100(%).

[0104] In other words, the final deviation rate is calculated by using the difference between the stock price and the path when the stock price moves most ideally and evenly as the reference value.

[0105] In other words, the aforementioned final divergence rate represents the degree of 'divergence from the bisector' relative to the 'divergence from the ideal trend line' as a percentage, and can be considered an indicator that measures how well the actual price path tracks the bisector (center of the trend).

[0106] Then, the aforementioned trend strength and divergence rate are combined to derive a final trend strength score (range 0 to 100).

[0107] The trend strength calculated as described above is distinguished from conventional indicators in that it is an extended indicator that incorporates path efficiency, going beyond simple closing price-based momentum or existing volatility estimates.

[0108] Furthermore, since the aforementioned trend strength can be aggregated at the portfolio level, consistency and practical applicability can be simultaneously secured when applying investment strategies.

[0109] Next, a portfolio construction method using the aforementioned trend strength will be explained.

[0110] In the present invention, the trend strength of each stock is calculated at the end of each month, and a portfolio is constructed based on this.

[0111] The stocks were divided into a High group (20%), a Middle group (60%), and a Low group (20%) according to the trend strength ranking calculated as above.

[0112] In addition, portfolio performance was evaluated by comparing it to the market benchmark, the KOSPI index, and the basic performance measures calculated were Annualized Return and Excess Return.

[0113] Risk-adjusted performance evaluation used the Sharpe Ratio, M² (Modigliani-Modigliani Measure), and alpha (α).

[0114] Looking at the distribution of trend strength of the sample stocks used in the analysis, the overall distribution was evenly distributed between 0 and 100, and on average, it recorded a level of 52.4.

[0115] This means that more than half of the sample stocks are maintaining a moderate level of trend strength.

[0116] The top 20% group showed an average of 73.2, and the bottom 20% group showed an average of 31.5, demonstrating a clear distinction between the groups.

[0117] This suggests that it is not merely the magnitude of returns, but the smoothness and efficiency of the price path that actually create differentiation between stocks.

[0118] These distribution characteristics can be interpreted to mean that trend strength is a distinct signal different from simple closing price momentum.

[0119] Furthermore, the distribution of trend intensity reflected the volatility of different market phases. For example, the average trend intensity increased during bull markets, while conversely, it decreased significantly during periods of sharp decline.

[0120] In addition, the trend strength indicator according to the present invention indicates that it effectively captures differentiation between stocks.

[0121] Additionally, looking at the distribution by period, the median of the entire sample increased significantly during the years 2016-2017 and 2020-2021, when the bull market continued.

[0122] On the other hand, during sharp declines such as the early stages of the COVID-19 pandemic, the distribution shifted to the left and the average trend strength decreased.

[0123] This suggests that the trend strength indicator reflects investor sentiment and trading concentration by market phase.

[0124] In addition, an analysis of the distribution by industry revealed that the average trend intensity was relatively high for technology and biotech stocks, whereas relatively low values ​​were observed in the financial and traditional manufacturing sectors.

[0125] These differences can arise because information asymmetry and liquidity structures vary by industry, implying that trend strength can be linked not merely to technical signals but also to industrial characteristics.

[0126] And when comparing portfolio performance, the top group portfolio recorded an annual average return of 12.8%, significantly outperforming the market benchmark (KOSPI index) of 7.5%.

[0127] On the other hand, the lower group showed poor performance compared to the market, with an average annual return of only 3.1%.

[0128] This can be seen as proof that a strategy of concentrating investments in stocks with high trend strength provides substantial excess returns.

[0129] In addition, the middle group also recorded a return of 9.4%, outperforming the benchmark, which demonstrates that trend strength-based classification creates continuous and systematic performance differences.

[0130] [Table 1] Comparison Table of Cumulative Returns of Trend Strength-Based Portfolios

[0131]

[0132] [Figure 2] Graph comparing cumulative returns of trend strength-based portfolios

[0133]

[0134] The difference in annualized returns strengthens the evidence that trend-strength-based portfolios systematically provide outperformance compared to the market.

[0135] In particular, the fact that the performance of the top group was significantly higher than that of the middle and bottom groups implies that the ranking of trend strength can be directly linked to investment performance.

[0136] In addition, looking at the cumulative return converted to 100 as of January 2015, the top group reached 486.7 as of the end of 2024, the end of the analysis.

[0137] On the other hand, the sub-group remained at 135.2 during the same period, showing significantly lower performance compared to the market.

[0138] These results are also consistent with the practical observation that “strong trends become stronger, and weak trends become weaker.”

[0139] Therefore, the trend strength according to the present invention can be seen as serving as a meaningful signal for predicting investment performance.

[0140] [Table 2] Annual Cumulative Returns of Trend Strength-Based Portfolios

[0141]

[0149] [Figure 3] Graph comparing yearly cumulative returns of trend strength-based portfolios

[0150]

[0151] [Figure 4] Graph of the yearly cumulative return trend of the trend strength-based portfolio

[0152]

[0153] The gap in cumulative returns was more clearly revealed in long-term investments.

[0154] Over the course of 10 years, the top group recorded performance approximately 3.5 times greater than the market, whereas the bottom group performed significantly poorly compared to the market.

[0155] These results demonstrate that the persistence of trend strength remains valid even in long-term investments.

[0156] In particular, the fact that the performance of the lower group has been sluggish for an extended period implies that defensive performance of the strategy can be expected simply by avoiding stocks with low trend strength.

[0157] In addition, when measuring risk-adjusted performance, the Sharpe ratio of the top group was 0.82, which is more than twice as high as the KOSPI index (0.39).

[0158] On the other hand, the lower group scored only 0.15, indicating poor performance relative to risk.

[0159] The Sharpe ratio mentioned above refers to a method in finance that adjusts for and reflects the risk of an investment when evaluating investment performance.

[0160] [Table 3] Cumulative Return and Sharpe Ratio Table of Trend Strength-Based Portfolios

[0161]

[0162] [Figure 5] Graph of cumulative return and Sharpe ratio of trend strength-based portfolio

[0163]

[0164] The difference in Sharpe ratios was clearly evident not only in differences in returns but also in differences in risk-adjusted efficiency.

[0165] The top group had returns per unit of risk more than twice as high as the market, which means that the trend strength is not just a simple return signal but also carries risk management significance.

[0166] The extremely low Sharpe ratio of the lower group confirms that stocks with low trend strength have lower investment efficiency.

[0167] Furthermore, examining the Sharpe ratios by market phase, the top group showed the greatest dominance during bull markets, while the middle group displayed a relatively stable Sharpe ratio during bear markets.

[0168] This suggests that the performance structure of trend strength-based strategies can operate differently depending on the market phase.

[0169] Looking at the M² measure, the top group recorded +6.39%, demonstrating a distinct outperformance compared to the KOSPI index at the same level of volatility.

[0170] On the other hand, the bottom group showed -4.1%, demonstrating that performance was inferior even after accounting for market volatility.

[0171] The M² measure can be seen as having significant practical meaning from an investor's perspective because it demonstrates excess performance at the same level of risk.

[0172] While the M² values ​​of the top group consistently recorded positive (+) performance, the bottom group showed negative (-) performance.

[0173] This means that even under the same risk conditions, stocks with stronger trends provide excess returns.

[0174] Meanwhile, the portfolio's excess performance was analyzed using MDD, Var, and Cvar.

[0175] The aforementioned MDD (Maximum Drawdown) is an indicator that measures how significantly the value of an investment asset has fallen from its peak.

[0176] Mean-variance-based indicators (Shape, M²) view “volatility” as a risk, but MDD represents the maximum loss actually experienced by investors, so it can be seen as reflecting psychological risk well.

[0177] The above VaR (Value at Risk) refers to the maximum expected loss under a certain confidence level.

[0178] For example, “a 1-day VaR of 5% at a 95% confidence level” means that there is a 5% probability of a loss of 5% or more occurring during the day.

[0179] The above CVaR (Conditional Value at Risk) represents the "average loss in the extreme loss region" exceeding VaR.

[0180] In other words, the Sharpe ratio indicates average efficiency and the MDD indicates maximum loss risk; VaR allows setting daily loss limits, and CVaR enables responding to stress scenarios.

[0181] Considering that the Sharpe ratios by trend strength group and M² reflect only mean-variance-based risk, the inventor evaluated the actual risk-adjusted performance by considering both the investor's perceived loss (maximum drawdown) and extreme loss (CVaR).

[0182] As a result, the maximum drop in the top group of trend strength was -18.2%, which was significantly smaller compared to the bottom group (-37.8%), and the CVaR at the 95% confidence level was also stable in the top group at -6.3% compared to the bottom group (-9.5%).

[0183] [Table 4] Comparison Table of MDD, Var, and CVaR by Trend Strength Group

[0184]

[0185] In addition, the excess performance of the portfolio was analyzed using the Fama-French 3-factor model and the Carhart 4-factor model.

[0186] As a result, the top group recorded a positive (+) alpha of 0.42% on a monthly basis.

[0187] And in the Carhart model, the alpha of the subgroup was 0.27%, maintaining significance, whereas the subgroup recorded a negative (-) alpha.

[0188] The fact that the alpha of the top group was statistically significant positive (+) in the analysis using the Fama-French 3-factor model indicates that the trend strength signal is not simply explained by existing factors, and that this suggests the trend strength contains an independent set of information.

[0189] [Table 5] Comparison Table of Alpha, Beta, SMB, HML, MOM, Adj R², and t(Alpha) by Trend Strength Group

[0190]

[0191] [Figure 6] Alpha by Trend Strength Group Comparison graph

[0192]

[0193] [Figure 7] Beta by Trend Strength Group Comparison graph

[0194]

[0196] [Figure 8] SMB by Trend Strength Group Comparison graph

[0197]

[0198] [Figure 9] HML by Trend Strength Group Comparison graph

[0199]

[0202] [Figure 10] MOM by Trend Strength Group Comparison graph

[0203]

[0204] [Figure 11] Adj R² by Trend Strength Group Comparison graph

[0205]

[0209] [Figure 12] Regression analysis results (Portfolio and Model Alpha)

[0210]

[0211] In addition, market phases were classified based on the long-term trend of the KOSPI index, and a bull market was defined as a period in which the KOSPI index rose by more than 20% from the previous low and the 12-month moving average line was in an upward phase.

[0212] A bear market was defined as a period in which the market falls by more than 20% from the previous high and the moving average line maintains a downward trend.

[0213] Based on the above criteria, each market phase was divided into a total of 5 bullish phases and 4 bearish phases during the period 2015-2024, and the monthly returns are as shown in [Table 6] below.

[0214] [Table 6] Comparison Table of Returns by Group by Market Phase

[0215]

[0216] Table 6 above shows the average monthly returns of portfolios by market phase. In a bull market, the returns of the top group were the highest, and there was a clear excess performance compared to the market.

[0217] On the other hand, in a bear market, losses in the top group widened, while the middle group showed defensive performance with relatively small losses (-0.18%).

[0218] This suggests that trend strength indicators do not rely solely on a single phase, but can be distinguished into aggressive (High) and defensive (Middle) signals depending on market volatility.

[0219] These results are also consistent with changes in market participants' risk preferences.

[0220] In a bull market, investor sentiment strengthens, leading to a concentration of funds in stocks with reinforcing trends; however, in a bear market, investors tend to mitigate relative losses by shifting to moderate, stable trends.

[0221] In addition, looking at performance by market phase, the top group recorded an average excess return of over 12% in a bull market, and while performance slowed in a bear market, it still maintained an advantage over the market.

[0222] This demonstrates that while the performance of trend strength strategies may vary depending on the market cycle, they are generally most effective in bull markets.

[0223] A more detailed analysis of performance by market phase reveals that the top group recorded this overwhelming excess performance during bull markets.

[0224] Conversely, in a bear market, the middle group showed relatively defensive performance, while the lower group recorded additional losses relative to the market.

[0225] Additionally, in highly volatile markets, stocks with stronger trends tended to show reduced performance variance, demonstrating that strong trends can provide relatively stable performance even in phases of high uncertainty.

[0226] Looking at performance by industry, IT and bio sectors showed outstanding results, while the financial and traditional manufacturing sectors showed relatively weak performance.

[0227] This shows that the persistence of trend strength may vary by industry.

[0228] In particular, a tendency for strong trends to persist over a long period was observed in the IT sector, which can be interpreted as a result reflecting growth expectations and investor interest in the industry.

[0229] Looking at performance by sector, technology and biotech stocks maintained strong trends over an extended period, with the top group showing particularly notable outperformance.

[0230] On the other hand, the effect of trend strength was relatively weak in the financial and traditional manufacturing sectors, which implies that trend strength signals can operate differently depending on the industry.

[0231] Furthermore, trend strength strategies recorded relatively strong performance in sectors with high liquidity, suggesting that market structural factors can influence strategy performance.

[0232] The results above demonstrate that trend strength indicators contain extended information compared to simple momentum and can simultaneously provide market outperformance and risk-adjusted performance.

[0233] In particular, trend strength reflecting path efficiency was effective in compensating for the weaknesses of existing momentum strategies, namely the risk of loss during sharp reversals.

[0234] In other words, the trend strength indicator according to the present invention provides information that is more extensive than simple momentum and can be seen as making a meaningful contribution to both investment performance and risk management.

[0235] The following describes actual application cases of the Trend Strength Index (TSI) proposed in the present invention to the KOSPI 200, KOSDAQ 150, and ETFs.

[0236] The trend strength indicator simultaneously quantifies the direction of short- and long-term trends and path efficiency by reflecting the geometric characteristics of the price path, such as the Start-End Line, High-Low Line, and Angle Bisector Line, with the Center Price as the axis.

[0237] The results were verified by dividing the cases into three categories.

[0238] The first case involved classifying KOSPI 200 stocks from November 2024 to September 2025 into 3-month and 6-month trend strengths and the intersection of the trend strengths.

[0239] In the second case, 10 stocks with strong trend strengths ranging from 1 to 6 months were selected from the KOSPI 200 stocks, and stocks from the intersection period of each period were also tested.

[0240] The third case involved verifying the performance through actual investments by three actual experimenters from July 1, 2025, to September 30, 2025. In addition to the first case, it was divided into KOSPI 200, KOSDAQ 150 stocks, and ETFs.

[0241] For actual data collection, the Yfinance (Yahoo Finance) library was used, and the KOSPI 200 stock list consisted of 100 stocks extracted from TradingView. The code was written in Python and output tables of monthly and cumulative returns when executed.

[0242] In addition, the case study was supplemented through Portfolio Construction Theory (MPT) from the perspective of diversifying unsystematic risk.

[0243] First, the first example is the result of measuring the trend strength of KOSPI 200 stocks by applying a trend strength indicator using OHLC information and utilizing it as an investment strategy.

[0244] The KOSPI 200 is the representative index of the Korean stock market and has the advantages of stability centered on large-cap stocks, a large market capitalization, large trading volume and liquidity, relatively low volatility, and low slippage.

[0245] Specifically, the first case compared investment performance of 3-month (short-term) and 6-month (medium-term) trends using the OHLC-based TSI formula (central price and energy-based strength) and evaluated the stability and excess performance of the intersection (3M∩6M) strategy.

[0246] The results of the first case are as shown in [Table 7] below.

[0247] [Table 7] Performance Analysis of Case Study I (November 2024 – September 2025, 3M / 6M / Intersection)

[0248]

[0249] According to the interpretation from the MPT perspective, the 10-stock portfolio diversifies unsystematic risk by more than 90%, and the MDD of the intersection strategy (-0.85% monthly average) was lower than that of the single strategy (-1.56%), confirming the diversification effect.

[0250] In addition, the 6-month strategy showed stable growth (all months were good), the 3-month strategy performed well in high-volatility months (May 2025: 3.67%), and the intersection had a risk mitigation effect.

[0251] In addition, the 3-month strategy achieved high returns by capturing market momentum (strength in the semiconductor sector), but exhibited significant volatility (lowest -1.56%).

[0252] And the 6-month strategy recorded a 100% win rate due to long-term trends (energy-based stability) and was superior in Sharpe ratio (risk-adjusted return) (estimated 1.2 vs. 0.9).

[0253] In addition, the intersection effect showed that the overlap stocks were over 70%, indicating stability (volatility σ=0.85%) despite the lack of diversification.

[0254] Considering the above, a portfolio manager can use a 6-month strategy as a baseline and 3 months as a timing signal.

[0255] However, the first example above has limitations in that it did not include transaction costs and used synthetic data.

[0256] The second case involves comparing the Top 10 strategies based on 1 to 6 months of TSI and evaluating the risk-return profile when trend persistence is reinforced through intersection and union strategies.

[0257] That is, for the period of 5 years and 8 months from January 1, 2020 to August 31, 2025, 10 stocks with strong trend strengths in the 1-6 month range were selected from the KOSPI 200 stocks for the experiment, and stocks in the intersection period of each period were also tested.

[0258] As a result, the 3-month (M3) strategy performed best with an average monthly return of 1.84% and showed relative stability with an MDD of -35.17%.

[0259] This is interpreted as being due to the fact that the 3-month short-term trend effectively captures market momentum. On the other hand, the 1-month (M1) period showed high volatility, resulting in a high risk with an MDD of -41.71%.

[0260] Furthermore, multi-period intersections demonstrated the strongest performance; in particular, the 2∩3∩6 month intersection (Inter M2∩M3∩M6) outperformed the benchmark with an average monthly return of 4.31% and a CAGR of 55.73%, and with an MDD of -23.78%, the risk-adjusted return (Sharpe Ratio estimate) 2.35) was found to be excellent.

[0261] Furthermore, the Union showed a slight reduction in MDD (-30.04%) due to diversification effects compared to the single period, but fell short of the Intersection in CAGR (21.18%), which is interpreted as excessive diversification diluting the trend strength.

[0262] And the average monthly return was best when the trend strength was calculated using the intersection of the 3-month and 6-month periods, with a value of 2.69% per month.

[0263] The Compound Annual Growth Rate was best for the stock calculated using the trend strength of the intersection of the Top 10 trend strengths for 2, 3, and 6 months, with a value of 55.73%.

[0264] Tables 8 and 9 below show the results.

[0265] [Table 8] 1–6 Month Trend Strength Performance from January 1, 2020 to August 31, 2025

[0266]

[0271] [Table 9] Monthly Trend Strength Performance from January 1, 2020 to August 31, 2025

[0272]

[0273] The third case involved three experimenters utilizing a trend strength indicator based on OHLC information to actually invest in KOSPI 200, KOSDAQ 150, and ETF stocks from July 1, 2025, to September 30, 2025, and verify the results.

[0274] That is, from July 1, 2025 to September 30, 2025, the experimenter freely selected stocks using the trend strength indicator, and the holding period was set to one month.

[0275] In addition, the investment amount for each portfolio was set to 10 million won, and it was diversified into 10 stocks.

[0276] In addition, stocks with a high point close to the current time within a period of 3 to 6 months were extracted using heuristic information.

[0277] Figures 13 and 14 below show the KOSPI 200 stocks sorted in descending order of trend strength, and Figure 15 shows the trend strength of specific stocks selected as intersection stocks.

[0279] [Figure 13] KOSPI 200 stocks sorted in descending order by trend strength

[0280]

[0281] [Figure 14] KOSPI 200 stocks sorted in descending order by trend strength

[0282]

[0287] [Figure 15] Trend strength of specific items

[0288]

[0290] From July to September 2025, the average return of the three experimenters was 9.13%, the standard deviation was 5.67%, and the Sharpe Ratio It was found that when the intersection (e.g., 3∩6 months) is applied, the return on selected stocks improves to 12.45%.

[0291] In addition, it was confirmed that the rate of return improves when selecting and investing in portfolio stocks using heuristic information, specifically those whose peak is close to the current time within a period of 3 to 6 months.

[0292] Tables 10 to 12 below show the results of experimenters 1 to 3, and Figures 16 to 21 show the specific items extracted by the experimenters.

[0293] [Table 10] Experimenter 1's Performance

[0294]

[0295] [Table 11] Performance of Experimenter 2

[0296]

[0297] [Table 12] Performance of Experimenter 3

[0298]

[0299] [Figure 16] Extracted items of Experimenter 2

[0300]

[0301] [Figure 17] Extracted items of Experimenter 2

[0302]

[0304] [Figure 18] Extracted items of Experimenter 3

[0305]

[0306] [Figure 19] Extracted items of Experimenter 3

[0307]

[0308] [Figure 20] Extracted items of Experimenter 3

[0309]

[0310] [Figure 21] Extracted items of Experimenter 3

[0311]

[0312] As can be seen from the above example, the investment stock selection method using the trend strength indicator according to the present invention has demonstrated investment utility in the Korean stock market from 2015 to 2024.

[0313] In particular, the top group with high trend strength showed significant annualized excess returns and higher Sharpe ratios compared to the market benchmark.

[0314] Conversely, the subgroup with low trend strength consistently underperformed in cumulative returns and risk-adjusted performance, and it was confirmed that simply excluding weak-trending stocks improved the portfolio's defensive capabilities.

[0315] In other words, the trend strength indicator according to the present invention can be described as a new measurement frame that simultaneously reflects direction and quality of the path by mathematically integrating the concept of path efficiency, which was overlooked by existing closing price-centered signals, into an OHLC-based indicator.

[0316] In addition, the trend strength indicator according to the present invention can be applied to portfolio management using only simple ranking, grouping, and equal weighting rules.

[0317] In particular, it can be utilized as a practical filtering device to systematically exclude stocks showing weak trends.

[0318] For example, an asset management company can construct a portfolio by calculating the trend strength of each stock at the end of each month and including only the top 30%, thereby automatically removing stocks with frequent irregular price rebounds and securing a more stable profit curve.

[0319] This is expected to provide substantial operational improvements in investment environments requiring long-term stability, such as retirement pensions or insurance company accounts.

[0320] In addition, according to the trend strength indicator of the present invention, even among stocks with the same momentum, stocks with low path efficiency can be excluded.

[0321] In other words, since it is possible to filter out stocks that have moved erratically even if they have shown high past returns, the predictive power and reproducibility of the portfolio can be improved as a result.

[0322] In addition, the trend strength indicator according to the present invention can be applied in multi-strategy operation frameworks such as asset allocation, sector rotation, and style mixing strategies.

[0323] For example, when managing industry ETFs, a conditional asset allocation strategy can be implemented by increasing weight only in sectors with high trend strength and reducing weight in sectors with weak trend strength, instead of buying the entire sector.

[0324] This has the advantage of enhancing the efficiency of capital allocation and reflecting heterogeneity across economic phases, compared to a method that simply tracks the entire market.

[0325] In addition, transaction costs can be minimized by excluding stocks with significant size and liquidity constraints, so the effect of improved cost-effectiveness can be maintained even in real-world operating environments.

[0326] That is, the trend strength indicator according to the present invention can be utilized in various ways, such as a practical filtering tool that complements existing momentum strategies, a risk management device that reduces risk during sudden market changes, and a complementary signal applicable to asset allocation and sector strategies.

Claims

Claim 1 A stock analysis device comprising a computer program for analyzing stock prices and hardware for implementing said program, wherein a method for selecting promising stock items using a trend strength indicator comprises: (a) a step of calculating the central price of each stock item from the stock price fluctuating during the day in said stock analysis device (S10); (b) a step of converting said central price into a converted price for comparison between stock items in said stock analysis device (S20); (c) a step of determining a straight line connecting the stock prices of the observation start date and end date, and a straight line connecting the maximum and minimum points during the observation period in said stock analysis device (S30); (d) a step of determining an ideal trend line in which the stock price moves with an ideal slope during the observation period in said stock analysis device (S40); (e) a step of calculating the opening / closing energy, the high / low energy, and the peak energy by utilizing the concept of energy in said stock analysis device (S50); and (f) a step of calculating the trend strength based on said opening / closing energy, the high / low energy, and the peak energy in said stock analysis device. A method for selecting promising stock items using a trend strength indicator, characterized by including step (S60). Claim 2 A method for selecting promising stock items using a trend strength indicator, characterized in that, in step S10, the central price is calculated by considering all data regarding the Open Price, High Price, Low Price, and Closing Price for one day in the stock analysis device. Claim 3 A method for selecting promising stock items using a trend strength indicator, characterized in that, in step S20, the central price is converted to a level of 10,000 won for comparison between stocks when calculating the converted price in the stock analysis device. Claim 4 A method for selecting promising stock items using a trend strength indicator, characterized in that, in step S30, the stock analysis device determines the start-end line and the high-low line in the form of a linear function having a slope and an intercept, respectively. Claim 5 A method for selecting promising stock items using a trend strength indicator, characterized in that, in the stock analysis device of claim 4, the process of determining a bisector that bisects the angle between two straight lines, the opening and closing line and the high and low line. Claim 6 A method for selecting promising stock items using a trend strength indicator, characterized in that, in step S60 above, the trend strength in the stock analysis device is calculated as follows. (Here, E-start / end represents energy based on the start / end line, E-high / low represents energy based on the high / low line, and E-max represents energy based on the maximum fluctuation range). Claim 7 A method for selecting promising stock items using a trend strength indicator, characterized in that, in the stock analysis device after step S60, the method further includes a step of calculating a bisector deviation rate indicating the deviation between the actual price path and the bisector line, and a trend deviation rate indicating the deviation between the actual price path and the trend line. Claim 8 A method for selecting promising stock items using a trend strength indicator, characterized in that, in the stock analysis device of claim 7, the bisection deviation rate and the trend deviation rate are calculated as the average of the distances between the actual center price and the ideal trend line at each point in time. Claim 9 A method for selecting promising stock items using a trend strength indicator, characterized in that, in the stock analysis device of claim 8, the method further includes the step of deriving the trend strength in the form of a score from 0 to 100 by combining the divergence rate and the trend strength. Claim 10 delete

Images