Improving the Fama-French Model Based on Technical Indicators:

Evidence from the Financial Industry

Zhuoqun Lei

Department of Financial Engineering, Tianjin University of Science and Technology, Tianjin, China

Keywords: Technical Indicators, Fama-French Model, Financial Industry.

Abstract: As a matter of fact, Fama-French models are commonly applied in financial fields, which is an upgrading

model for CAPM. Contemporarily, plenty of proposals for adding indicators to improve the performances for

the model are demonstrated. With this in mind, this paper deeply and comprehensively discusses the

improvement of the Fama-French model based on technical indicators. To be specific, it elaborates on the

principles and methods of improvement in detail and systematically expounds its application in the financial

industry. By sorting out and analysing relevant research, it reveals the significant advantages and potential

problems of the improved model and conducts a forward-looking outlook on future research directions.

According to the analysis, the current limitations and prospects are proposed at the same time. Overall, this

provides a new perspective for the expansion of financial asset pricing theory and also provides a valuable

reference basis for practical operations in the financial industry.

1 INTRODUCTION

With the increasing complexity and uncertainty of

financial markets, asset pricing is facing severe

challenges. As an important cornerstone of asset

pricing, the Fama-French model gradually exposes

limitations in the new environment, especially in

dealing with short-term fluctuations, emerging

industry characteristics, and investor sentiment

(Carhart, 1997; Fama & French, 1993; Fama &

French, 2015). This study aims to use technical

indicators to improve the Fama-French model to

enhance the accuracy and practical value of asset

pricing.

First, this study will deeply analyse the principle

and limitations of the Fama-French model, and clarify

its way of explaining stock return differences and the

current problems it faces. Then this study will sort out

technical indicators suitable for improvement, and

analyse the calculation methods, data requirements,

application scenarios, and advantages of common

technical indicators such as moving averages, relative

strength indicators, and Bollinger Bands. Then

explore the integration method of technical indicators

and the Fama-French model, including incorporating

technical indicators as supplementary factors into the

original model and dynamically adjusting the factor

weights of the original model according to the real-

time signals of technical indicators. Finally, through

application cases in the financial industry, such as

portfolio management, asset allocation decisions, and

risk management, show the application process of the

improved model, and use quantitative analysis to

compare with traditional models to verify its

superiority in asset pricing, risk measurement, and

investment performance.

This study has important theoretical and practical

significance. In theory, it injects new vitality into

asset pricing theory, promotes method integration and

innovation, fills the gap in research on the

combination of technical indicators and traditional

models, and perfects the methodological system. In

practice, it provides more accurate decision support

for investors, helping them choose assets, seize

opportunities, and optimize portfolios; it provides

powerful tools for financial institutions in risk

management, product development, etc., and

enhances competitiveness and service levels.

Lei, Z.

Improving the Fama-French Model Based on Technical Indicators: Evidence from the Financial Industry.

DOI: 10.5220/0013264700004568

In Proceedings of the 1st International Conference on E-commerce and Artiﬁcial Intelligence (ECAI 2024), pages 423-428

ISBN: 978-989-758-726-9

423

2 DESCRIPTIONS OF THE

IMPROVED FAMA-FRENCH

MODEL

As an important theoretical framework in the field of

asset pricing, the Fama-French model is constructed

based on market factors, size factors, and value

factors. The market factor occupies a fundamental

position in the model. It mainly reflects the extensive

influence of the systemic risk of the entire financial

market on stock returns. In practical calculations,

usually the part of a representative market index, such

as the S&P 500 index exceeding the risk-free interest

rate, is used to quantify the market factor (Jegadeesh

& Titman, 1993; Barberis et al., 1998). This indicator

aims to capture the overall market trend of rise and

fall and its general influence on the returns of various

stocks.

The size factor focuses on capturing the return

differences caused by the size of companies. By

carefully constructing small-cap stock portfolios and

large-cap stock portfolios and calculating the return

difference between the two, the size factor can reveal

the significant differences in risk and return

characteristics between small companies and large

companies (Cochrane, 2001). This difference not

only reflects the unique challenges and opportunities

faced by small companies in market competition but

also reflects the differences in risk preferences and

expected returns of investors for companies of

different sizes (Hong & Stein, 1999).

The construction of the value factor is based on

the book-to-market ratio of companies. By rigorously

comparing the stock return differences between

companies with high book-to-market ratios and

companies with low book-to-market ratios, the value

factor can effectively reflect the key influence of the

company's intrinsic value attributes on stock returns

(Lo & MacKinlay, 1999). This factor helps investors

identify companies that are undervalued or

overvalued by the market, thereby making more

informed investment decisions (Campbell, 2000).

The Fama-French model has obvious limitations.

First, it is insufficient in responding to short-term

market fluctuations and investor sentiment. It mainly

relies on the company's long-term fundamental data

and historical returns to construct the factor system

(Ferson & Harvey, 1991; Banz, 1981). When facing

short-term impacts such as sudden major market news,

macro policy adjustments, and investor panic, it lacks

a rapid incorporation and response mechanism and is

difficult to accurately reflect the instantaneous drastic

changes in stock prices in a timely manner (Basu,

1977). Second, its explanatory power is limited in

specific market environments and emerging

industries. In extreme markets such as a frenzied bull

market or a desperate bear market, traditional size and

value factors are difficult to capture extreme changes

in stock returns and are beyond the scope of

interpretation of conventional market assumptions

(Piotroski, 2000). In emerging industries, which rely

on non-traditional assets to create value and are

different from traditional industries, the original

model is difficult to fully explain its business model

and market valuation. Third, there is subjectivity in

factor definition and selection. The definitions and

calculation methods of size and value factors are not

uniform (Novy-Marx, 2013). Different researchers

will adjust according to their goals, data, and

preferences, leading to inconsistent applications and

differences in results. At the same time, factor

selection is restricted by data availability and

computational complexity, and valuable factors may

be ignored, affecting the interpretation and prediction

accuracy of specific market phenomena.

3 SELECTION AND

CHARACTERISTICS OF

TECHNICAL INDICATORS

The moving average is a commonly used basic tool in

financial technical analysis. By calculating the

arithmetic average of security prices within a specific

time period, it smooths short-term random

fluctuations and shows long-term trends. It can

reduce short-term price noise and make the trend

clearer and more coherent, but it has a lag. There is a

delay in responding to real-time changes, which may

cause missed short-term opportunities. However, it is

very valuable in confirming the main price trend. If

the price is above the average, it is an uptrend.

Conversely, it is a downtrend, which can point the

direction for long-term investment (Asness et al.,

2019).

The relative strength index is used to measure the

relative strength of security price fluctuations. By

comparing the sum of rising and falling closing prices

within a specific time period, a relative strength value

between 0 and 100 is obtained to judge the

overbought and oversold states of the market. It is

extremely sensitive to short-term price changes and

can quickly capture the subtle changes in buying and

selling forces, providing timely signals for short-term

traders to predict short-term reversals and price

adjustments. By setting thresholds, such as when the

ECAI 2024 - International Conference on E-commerce and Artiﬁcial Intelligence

424

RSI value exceeds 70, it is overbought, and when it is

below 30, it is oversold, which can prompt investors

when prices may reverse and the timing for taking

profits or setting stop losses. Moreover, the RSI value

fluctuates between 0 and 100, making it convenient

and intuitive to compare the relative strength of the

market for different securities or time periods.

Bollinger Bands is a technical analysis tool

constructed based on statistical principles. It consists

of three track lines. The upper and lower limits of

price fluctuations and the middle moving average are

determined by calculating the standard deviation of

security prices. It can intuitively show the volatility

of prices. The upper track is usually a resistance level,

and the lower track is usually a support level.

Observing the relative position of prices to the upper

and lower tracks can judge whether the fluctuation is

abnormal (Zhang, 2005). When the price touches the

upper and lower tracks, it is often an important signal

of a possible reversal. If it touches multiple times and

then moves in the opposite direction, the signal is

more reliable, bringing potential trading opportunities

to investors. Besides, the bandwidth of Bollinger

Bands will automatically adjust according to the

actual market fluctuation. When the market fluctuates

violently, the bandwidth expands, and when it is

stable, it contracts, maintaining effectiveness and

applicability in different market environments.

4 INTEGRATION STRATEGY OF

TECHNICAL INDICATORS

AND THE FAMA-FRENCH

MODEL

4.1 Supplementary Factor

The calculation results of technical indicators are

integrated into the Fama-French model as

independent factors. For example, the slope or cross

signal of the moving average. When it is upward

sloping or in a golden cross, the new factor takes a

positive value. When it is downward sloping or in a

dead cross, it takes a negative value. This introduces

price trend and short-term momentum information

and enhances the capture of short-term market

dynamics. For the relative strength index (RSI), set a

threshold. When it is overbought (such as when the

RSI value exceeds 70), it is included in the model as

an overbought factor and takes a positive value. When

it is oversold (such as when it is lower than 30), it is

included in the model as an oversold factor and takes

a negative value, reflecting the short-term buying and

selling status of the market more sensitively and

providing timely signals for decision-making.

4.2 Adjusting Factor Weights

Dynamically adjust the weights of each factor in the

Fama-French model according to the real-time

signals of technical indicators. For example, when the

RSI enters the overbought region, reduce the weights

of the market and size factors and increase the weight

of the value factor to be vigilant against short-term

risks and prefer value assets. Conversely, when the

RSI enters the oversold region, increase the weights

of the market and size factors and reduce the weight

of the value factor to capture rebound opportunities.

5 MATHEMATICAL

EXPRESSION AND

THEORETICAL BASIS OF THE

IMPROVED MODEL

Supposing that the moving average (MA) and relative

strength index (RSI) are introduced. The improved

Fama-French model can be expressed as:

𝑅



−𝑅



= 𝛼 + 𝛽



𝑀𝐾𝑇 + 𝛽



𝑆𝑀𝐵 + 𝛽



𝐻𝑀𝐿 + 𝛽



𝑀𝐴 + 𝛽



𝑅𝑆𝐼 + 𝜖

(1)

Here ， 𝑅



−𝑅



represents the excess return of an

asset i; 𝛼 is the intercept term, which reflects the

fixed part of the return not explained by other factors

in the model. The 𝛽 series are the coefficients of each

factor, and their magnitudes represent the degree of

contribution of each factor to the excess return. MKT

represents the market factor and reflects the impact of

overall market fluctuations on asset returns; SMB is

the size factor and reflects the role of company size

on returns; HML is the value factor and measures the

relationship between the company's value

characteristics and returns; MA is the moving average

factor and reflects price trend information; RSI is the

relative strength index factor and reflects the

comparison of market buying and selling forces and

short-term momentum; 𝜖 is the error term and

represents the random part that the model cannot

explain. The theoretical basis lies in that the improved

model combines the advantages of fundamental

analysis (Fama-French model) and technical analysis

(technical indicators). The Fama-French model starts

from the basic financial characteristics of a company,

such as size and value, to explain the differences in

long-term stock returns, reflecting the intrinsic value

of assets and long-term investment logic. This

Improving the Fama-French Model Based on Technical Indicators: Evidence from the Financial Industry

425

fundamental-based analysis provides a stable and

reliable basis for asset pricing, especially suitable for

the evaluation and prediction of a company's long-

term value.

Technical indicators such as moving averages and

relative strength indicators focus on capturing

dynamic information such as short-term price trends,

investor sentiment, and short-term supply and

demand changes in the market. They can quickly

respond to immediate changes in the market and

provide signals about short-term market momentum

and reversal possibilities.

By combining these two analysis methods, the

improved model takes into account both the long-

term value driving factors of assets and the influence

of short-term market dynamics and investor

behaviour. This enables the model to explain the

formation mechanism of asset prices more

comprehensively and accurately and improve the

accuracy and adaptability of asset pricing. Under

different market conditions, whether it is a long-term

trend change or a short-term violent fluctuation, the

improved model is expected to provide more reliable

pricing and investment decision-making basis.

6 APPLICATION OF THE

IMPROVED MODEL IN THE

FINANCIAL INDUSTRY

6.1 Investment Strategy Formulation

The improved model provides a more accurate and

comprehensive evaluation framework for

constructing investment portfolios. By analyzing a

large amount of asset historical data and calculating

expected returns and risk levels, it can distinguish the

potential performance of similar stocks. For example,

for stocks in the same industry with similar size and

book-to-market ratio, the model can give

differentiated evaluations based on moving averages

and relative strength indicators, helping investors

choose stocks with high short-term upside potential

and low risk. At the same time, the model helps to

discover undervalued or overvalued assets and

provides support for contrarian investment (Hou et al.,

2015; Ang et al., 2006).

The integration of technical indicators makes the

improved model more sensitive and accurate in

investment timing judgment. When the moving

average forms a golden cross and the relative strength

indicator is oversold, it is usually a buying

opportunity; when there is a dead cross and the

indicator is overbought, it may be a time to sell or

reduce positions. Combined with the Bollinger Bands

indicator, when the price breaks through the upper

track and trading volume expands, it is time to take

profits; when it touches the lower track and stabilizes

after a decline, it is time to build positions or add

positions.

6.2 Risk Management

Traditional risk measurement methods have

limitations. The improved model incorporates

technical indicators and can timely reflect short-term

abnormal market fluctuations and changes in investor

sentiment. For example, changes in relative strength

indicators and moving averages can provide early

warning signals. Comprehensive analysis can more

accurately assess potential losses and provide a more

comprehensive and dynamic perspective for risk

measurement, helping investors respond in advance.

Based on the risk assessment of the improved

model, financial institutions and investors can adjust

the structure of their investment portfolios. When

market risk rises, reduce the allocation of high-risk

assets and increase the proportion of low-risk assets,

and use financial derivatives for hedging. The

improved model can also help set reasonable stop-

loss and take-profit levels and send signals in time to

limit losses or lock in gains.

6.3 Case Analysis

E Fund Management Co., Ltd. has long relied on

classic asset pricing models to construct investment

portfolios in the field of stock investment. However,

with the development of the market and the

intensification of competition, they found that the

traditional Fama-French model performed poorly in

responding to rapid market changes and the rise of

emerging industries. For example, during the period

from 2019 to 2020, the new energy vehicle industry

rose rapidly. When the fund company evaluated

related enterprises, the evaluation results given by the

traditional model based on market factors, size factors,

and value factors failed to fully reflect the potential

growth value of these enterprises. Therefore, E Fund

Management Co., Ltd. decided to introduce technical

indicators to improve the Fama-French model (Bali et

al., 2011).

They selected technical indicators such as moving

average (MA), relative strength indicator (RSI), and

Bollinger Bands. When evaluating new energy

vehicle enterprises, combined with these technical

indicators, it was found that although the stock prices

ECAI 2024 - International Conference on E-commerce and Artiﬁcial Intelligence

426

of some enterprises fluctuated in the short term, the

moving average showed a clear upward trend, and the

relative strength indicator showed that it was in a

strong area. Based on the signals of these technical

indicators, E Fund Management Co., Ltd. increased

its investment in these enterprises. After a period of

practice, compared with the investment portfolio

using the traditional model in the same period, the

investment portfolio using the improved model

achieved significantly higher investment returns in

the field of new energy vehicles, bringing more

substantial returns to fund investors. This case shows

that the Fama-French model improved based on

technical indicators can help fund companies capture

market opportunities more sensitively, optimize

investment strategies, and improve investment

performance (Avramov et al., 2009).

7 LIMITATIONS AND

PROSPECTS

The effectiveness of technical indicators depends on

market conditions: The effectiveness of different

technical indicators varies significantly in different

market environments. For example, in a market with

obvious trends, trend-following indicators can

provide accurate buy and sell signals, but in a volatile

or trendless market, they may frequently make

mistakes. In addition, technical indicators are easily

affected by market manipulation, abnormal trading

behaviors, or sudden major news, leading to signal

deviations or failures. Increased model complexity:

Integrating technical indicators into the Fama-French

model increases complexity, which is reflected in the

mathematical expressions, the number of parameters,

and the increased requirements for data processing

and computing power. Complex models are difficult

to understand and explain, and are prone to overfitting,

losing the ability to predict and generalize to new data,

reducing practicability and reliability. Impact of data

quality and sample bias: low-quality data and

unrepresentative samples can seriously affect the

accuracy of model results. Financial market data

often has quality problems such as missing values,

incorrect records, and asynchronous trading, leading

to deviations in technical indicators and model

parameters. Sample bias is also common. If the data

sample cannot represent market diversity and

dynamic changes, the model conclusions may be

limited and cannot be widely applied.

In the future, more effective combinations of

technical indicators and parameter optimization can

be explored. Comprehensively use multiple

indicators, combine big data and machine learning

algorithms to screen out the optimal combination, and

use historical data backtesting to determine the

optimal weights and parameters to improve model

adaptability and accuracy. For example, fuse

common indicators with emerging tools and use

intelligent optimization algorithms to determine the

best parameter combination to adapt to market

changes.

Incorporating investor psychology and behavioral

factors into the model to improve the asset pricing

mechanism. Behavioral finance shows that investors'

irrational behaviors have an important impact on the

market. In the future, improved models can quantify

these behavioral factors and combine them with

technical indicators and Fama-French factors to

reflect market laws, explain abnormal phenomena

and bubble formation mechanisms, and provide

support for investment decisions. For example,

introduce relevant indicators, study their synergistic

relationship, and construct a pricing model based on

behavioral finance.

Real-time monitoring and adaptive adjustment:

Developing a dynamic model that responds to market

changes in real time is an important trend. As

financial market changes accelerate, traditional static

models are difficult to adapt. It is necessary to

construct a model with self-learning and adaptive

adjustment capabilities, monitor data and indicator

changes in real time, and adjust parameters and

weights. For example, use online learning and

reinforcement learning technologies, combined with

cloud computing and big data processing

technologies, to achieve real-time calculation, rapid

decision-making, and provide investment advice and

risk warnings.

8 CONCLUSIONS

To sum up, this study focuses on the Fama-French

model improved based on technical indicators and

conducts a comprehensive and in-depth discussion. It

elaborates on its theoretical basis, integration strategy,

application fields, potential limitations, and future

development directions. The research results clearly

show that this model successfully combines the

advantages of fundamental analysis and technical

analysis. Not only does it build a more comprehensive

and in-depth explanatory framework for the asset

price formation mechanism at the theoretical level,

but also in practical applications, it plays a significant

role in key fields such as investment strategy

Improving the Fama-French Model Based on Technical Indicators: Evidence from the Financial Industry

427

formulation and risk management, providing more

accurate and flexible decision-making tools for

investors and financial institutions. Looking to the

future, it is necessary to further deepen the integration

mechanism of technical indicators and the model,

actively explore more advanced data analysis

methods and optimize the model structure, so as to

significantly improve the model's predictive ability

and adaptability. At the same time, it is necessary to

strengthen empirical research, fully verify the

effectiveness of the model with the help of large-scale

actual data, and continuously expand the wide

application range of the model in different financial

markets and asset categories. In addition, closely

monitor the innovative development trends of the

financial market and changes in regulatory policies,

and timely adjust and improve the model to ensure

that it always maintains a high degree of practicability

and excellent guiding value in the financial field. This

research provides a new and extremely valuable

perspective and method for financial practitioners,

and strongly promotes them to re-examine and

optimize existing asset pricing and risk management

strategies, which is of great significance for

improving the decision-making level and risk

management ability of the financial industry.

REFERENCES

Ang, A., Hodrick, R. J., Xing, Y., Zhang, X., 2006. The

cross-section of volatility and expected returns. The

Journal of Finance, 611, 259–299.

Asness, C. S., Frazzini, A., Pedersen, L. H., 2019. Quality

minus junk. The Review of Accounting Studies, 241,

34–112.

Avramov, D., Chordia, T., Jostova, G., Philipov, A., 2009.

Credit ratings and the cross-section of stock returns.

The Journal of Finance, 643, 1099–1137.

Bali, T. G., Cakici, N., Whitelaw, R. F., 2011. Maxing out:

Stocks as lotteries and the cross-section of expected

returns. The Journal of Financial Economics, 992, 427–

446.

Banz, R. W., 1981. The relationship between return and

market value of common stocks. Journal of Financial

Economics, 91, 3–18.

Barberis, N., Shleifer, A., Vishny, R., 1998. A model of

investor sentiment. Journal of Financial Economics,

493, 307–343.

Basu, S., 1977. Investment performance of common stocks

in relation to their price-earnings ratios: A test of the

efficient market hypothesis. The Journal of Finance, 323,

663–682.

Campbell, J. Y. 2000. Asset pricing at the millennium. The

Journal of Finance, 554, 1515–1567.

Carhart, M. M., 1997. On persistence in mutual fund

performance. The Journal of Finance, 521, 57–82.

Cochrane, J. H., 2001. Asset Pricing. Princeton University

Press.

Fama, E. F., French, K. R., 1993. Common risk factors in

the returns on stocks and bonds. Journal of Financial

Economics, 331, 3–56.

Fama, E. F., French, K. R., 2015. A five-factor asset pricing

model. Journal of Financial Economics, 1161, 1–22.

Ferson, W. E., Harvey, C. R., 1991. The variation of

economic risk premiums. The Journal of Political

Economy, 992, 385–415.

Hong, H., Stein, J. C., 1999. A unified theory of

underreaction, momentum trading, and overreaction in

asset markets. The Journal of Finance, 546, 2143–2184.

Hou, K., Xue, C., Zhang, L., 2015. Digesting anomalies:

An investment approach. The Review of Financial

Studies, 283, 650–705.

Jegadeesh, N., Titman, S., 1993. Returns to buying winners

and selling losers: Implications for stock market

efficiency. The Journal of Finance, 481, 65–91.

Lo, A. W., MacKinlay, A. C., 1990. An econometric

analysis of nonsynchronous trading. The Journal of

Econometrics, 451-2, 181–211.

Novy-Marx, R., 2013. The other side of value: The gross

profitability premium. Journal of Financial Economics,

1081, 1–28.

Piotroski, J. D., 2000. Value investing: The use of historical

financial statement information to separate winners

from losers. Journal of Accounting Research, 381, 1–

41.

Zhang, L., 2005. The value premium. The Journal of

Finance, 601, 67–103.

ECAI 2024 - International Conference on E-commerce and Artiﬁcial Intelligence

428