Equity Premium Puzzle in Cambricon: Evidence from Behavioural
Finance
Zhixuan Li
1
a
and Weifeng Tian
2
b
1
School of Management, Shandong University, 27 Shanda Road, Jinan, China
2
School of Management, China University of Mining and Technology (Beijing), 11 Xueyuan Road, Beijing, China
Keywords: Behavioural Finance, Stock Premium Effect, Herding Behaviour, Investor Sentiment, Liquidity Risk.
Abstract: The stock premium effect in high-tech industries often deviates from the market effectiveness and investor
rationality emphasized by traditional financial theories. Based on behavioural finance theory, this study
analyses the premium effect of Cambricon, a leading Chinese AI chip company listed on the Science and
Technology Innovation Board (STB). Using financial indicators (e.g., negative price-to-earnings ratio, 58.03
times issue price-to-sales ratio), investor sentiment indices (e.g., Baidu search index) and institutional trading
data (2020-2025), the study finds that irrational factors-such as the herd effect, sentiment bias and narrative-
driven speculation (e.g., the “AI chip localization") drive high stock premiums to some extent. The results
suggest that the concentration of institutional positions and retail followers' behaviour together drive this
phenomenon, and the sentiment resonance mechanism exacerbates price distortions. The study proposes
regulatory reforms (e.g., dynamic position disclosure thresholds, ETF rule optimization) and institutional
strategies (e.g., long-term performance assessment) to curb speculative bubbles. These findings reveal the
role of behavioural factors in shaping tech stock premiums and provide a practical framework for stabilizing
high-volatility markets.
1 INTRODUCTION
In finance, the stock premium effect has been the
focus of research. However, empirical studies have
found that traditional financial theories based on the
efficient market hypothesis have difficulty in
explaining valuation deviations in high-tech stocks
(Fama, 1970; Shiller, 2003). Behavioural finance
proposes that the herd effect and investor sentiment
push up premiums through psychological account
segregation, while liquidity stratification exacerbates
price volatility (Bikhchandani et al., 1992; Baker &
Wurgler, 2006; Calzadilla et al., 2021; Thaler, 1985;
Shen, 2023).As a benchmark company in China's AI
chip industry, Cambricon occupies 30% of China's
cloud-based Ai chip market share and has received
capital injection from the National IC Industry
Investment Fund. It has attracted much attention since
its listing, with its capital fervour and domestic
substitution narrative providing an opportunity for
premium pricing(Li, 2020). Founded in 2016,
a
https://orcid.org/0009-0003-9287-0811
b
https://orcid.org/0009-0008-2129-557X
Cambium, which focuses on AI chip research and
development, has seen its revenues surge from
7,843,300 yuan in 2017 to 444 million yuan in 2019,
and through multiple rounds of financing was valued
at 2.5 billion dollars in 2018, making it a unicorn
enterprise. The stock price jumped 229.86% on the
first day of listing on the Science and Technology
Board in 2020, and rose over 1,089% after 2023
driven by the ChatGPT concept. The high level of
innovation and uncertainty in the AI industry has led
to technological developments, capital flows, and
macroeconomic changes significantly affecting
investor behaviour , which in turn leads to stock price
volatility. Existing studies mostly focus on traditional
industries, with insufficient research on the
behavioural premium mechanism of high-tech stocks,
and the special characteristics of China's science and
innovation board market due to the high proportion of
retail investors and strong policy drivers(Baker &
Wurgler, 2006). The high growth of the Ai chip
industry further increases irrational investor
30
Li, Z. and Tian, W.
Equity Premium Puzzle in Cambricon: Evidence from Behavioural Finance.
DOI: 10.5220/0013832200004719
Paper published under CC license (CC BY-NC-ND 4.0)
In Proceedings of the 2nd International Conference on E-commerce and Modern Logistics (ICEML 2025), pages 30-37
ISBN: 978-989-758-775-7
Proceedings Copyright © 2025 by SCITEPRESS – Science and Technology Publications, Lda.

behaviour, with Asia-Pacific ai spending projected to
reach $90.7 billion by 2027, with a CAGR of up to
28.9% from 2022 to 2027 (IDC, 2006), as high as
28.9% (IDC, 2023).
Based on this, this paper applies the theoretical
framework of behavioural finance and introduces
narrative economics in conjunction with the liquidity
layering model to reveal how the 'AI chip autonomy'
narrative forms a positive feedback loop through
institutional hugging and retail followers(Shiller,
2003). Clarify how the degree of premium is
calculated and invoke the sentiment index construct
(Baker & Wurgler, 2006). The study is conducted in
two ways: one, to quantify the degree of Cambricon
premium by comparing financial data with industry
averages; and two, to quantify the abnormal returns
on stock price from events such as the release of
ChatGPT's concept and its inclusion in the SSE 50
index using event analysis. This research enriches the
application scenarios of behavioural finance in the
stock premium of high-tech enterprises and extends
the 'cognitive bias-price feedback' model , which not
only analyses the impact of investors' behavioural
biases on stock prices and provides theoretical
support for their investment decisions, but also
provides government, It also provides suggestions for
government, institutions and individual investors to
help the government to better manage the market and
investors to make more rational decisions(Barberis &
Huang, 2008). It can also design 'sentiment warning
indicators' for retail investors to avoid the risk of
bubbles.
2 THEORETICAL
FOUNDATIONS
2.1 Relevant Theories of Behavioural
Finance
The traditional "efficient market hypothesis" holds
that investors are completely rational in making
investment decisions in financial markets. However,
research results in behavioural finance show that
investors are often influenced by the external
environment to create cognitive bias and deviate from
the efficient market when making investment
decisions (Tang et al., 2017). As a high volatility
technology stock, Cambricon's stock price may be
significantly stronger affected by behavioural finance
factors than traditional industries. Based on this, this
paper investigates four aspects: herd effect, loss
aversion, sentiment influence, and information effect.
Herding effect refers to the tendency of investors
to mimic others' decisions rather than analyse them
independently when information asymmetry or
uncertainty is high, resulting in excessive volatility as
a result of blindly following the herd (Bikhchandani
et al., 1992). As an emerging technology company,
Cambricon's technological complexity and
uncertainty of future profitability may lead investors
to rely on the behaviour of others and form a herd
effect. For example, concentrated positions of
institutional investors may trigger retail investors to
follow suit, further amplifying stock price volatility
(Lakonishok et al., 1992).
Loss Aversion (Loss Aversion) is a core concept
of Prospect Theory , which refers to the fact that
investors are more sensitive to losses than to
equivalent gains, and this psychology leads investors
to be more conservative in making investment
decisions, demanding higher risk premiums, and
leading to irrational position-taking behaviour
(Kahneman & Tversky, 1979 ;Gupta & Shrivastava,
2022). Sentiment Effect (Sentiment Effect) refers to
the significant impact of investor sentiment on stock
prices through media rendering, social media
communication and other channels, when investment
sentiment is high, over-optimism affects investor
valuation of stocks, which leads to risky decisions
(Baker & Wurgler, 2007). With the iteration of
ChatGPT, investors show overconfidence in AI-
related stocks, and as a tech hotspot company,
Cambricon, its stock price is vulnerable to market
sentiment-driven, leading to short-term overshooting
of the stock price.
2.2 Theories Related to the Premium
Effect
Liquidity premiums and risk premiums have an
important place in finance. The liquidity premium
reflects the link between transaction costs and asset
prices, with investors demanding a higher rate of
return to compensate for the risk of holding less liquid
assets. The risk premium, on the other hand, is the
return that investors expect to receive over and above
the risk-free rate for taking on additional risk.
In terms of liquidity premium, for example the
scale has expanded rapidly in recent years(Shen,
2023) . The literature takes the high-frequency
trading data of SSE 50ETF options and CSI 300ETF
options market as the research samples, and
concludes that there is a significant liquidity premium
phenomenon in the market of all moneyness
categories of SSE 50ETF options as well as OTM Put
and ATM Put categories of CSI 300ETF options
Equity Premium Puzzle in Cambricon: Evidence from Behavioural Finance
31

(Jiang, 2022). It is mentioned that liquidity premium
represents the extra rate of return that investors expect
to get to compensate for the risk of holding illiquid
assets, and in stock investment, the liquidity factor
captures the excess return from illiquid stocks, and it
is also suggested that investors should consider the
liquidity risk when pursuing the liquidity premium
and realize the balanced portfolio allocation to the
liquidity risk factors, and four measures of liquidity
in the stock market are also proposed. Indicators.
On risk premium, the uncertainty affecting asset
prices is categorized into fundamental uncertainty,
market-level and firm-level external factor
uncertainty (Yang, 2011). It is argued that investors
form heterogeneous beliefs about fundamentals and
external factors, respectively, on the basis of which a
model of consumer capital asset pricing based on
investors' heterogeneous beliefs is established, which
theoretically proves that in addition to the
fundamental risk from total consumption/endowment,
the differences in investors' beliefs about market-
level and firm-level external factors also have a
significant influence on the risk premium of
idiosyncratic volatility. In addition to the fundamental
risk from total consumption/endowment, it is
theoretically demonstrated that the differences in
investors' beliefs about market-level and firm-level
external factors are also risk factors affecting stock
prices Based on the perspective of institutional
investors, the study of China's financial risky
investment and bond risk premium empirically
examines the impact of institutional investors'
motivation of stockholding behaviour on bond risk
premium as well as the transmission paths of its
action mechanism, and points out that the different
risky investment behaviours of institutional investors
can be identified and manifested in bond risk
premiums(Guan, 2020).
Liquidity premiums and risk premiums play a
crucial role in financial markets and have an
important impact on investors' decisions and asset
pricing. An in-depth study of liquidity premiums and
risk premiums can help to better understand the
operating mechanism of financial markets and
provide investors with a more accurate basis for
investment decisions.
3 CASE DESCRIPTIONS
Founded in 2016, it is a leading AI chip enterprise in
China, focusing on the research and development of
AI chip products and technological innovation, and
providing a series of intelligent chip products and
platform-based system software with a unified
ecology, such as the integration of cloud, hardware
and software, and the integration of training and
reasoning.
Cambium-U, on June 23, 2020, received approval
from the China Securities Regulatory Commission
(CSRC) for the registration of its initial public
offering, and on July 20, it was listed on the Kechuan
Board of the Shanghai Stock Exchange. Initial public
offering of RMB 40.1 million ordinary shares to the
public, with a total share capital of 360 million shares
before issuance and 400.1 million shares after
issuance. On the first day of listing, the stock price
soared 229.86%, but since then the stock price fell
continuously, and once fell to 46.59 yuan/share in
2022. 2023 ushered in the rise by the concept of
Chatgpt, and rose 1089% since the beginning of 2023
to 2024. 2024 December 23, the intraday rose to 700
yuan/share, and then fell 4.02% on the same day to
close at 648.75 yuan / shares, January 24, 2025
closing share price of 612.98 yuan, compared with the
previous trading day fell 8.02 yuan, or 1.29%.
4 EFFECTS ANALYSES
4.1 Existence of Premium Issue
As shown in Fig. 1, Cambricon landed on the S&T
board in 2020 with an offering price of $64.39. From
the information provided in its prospectus, there is a
possibility that Cambricon's stock offering will be at
a premium. During the period 2017-2019, Cambricon
is in a continuous loss. Among them, the net profit
was -380,704,000 yuan in 2017, -410,465,000 yuan
in 2018, and -117,898,560,000 yuan in 2019. In the
case that the company has not yet achieved
profitability for the stock issue, its issue price is
difficult to obtain strong support from the current
earnings data, and is based more on the expectation
of future earnings. Analysed from the traditional
perspective of earnings valuation, the issue price
determined in this earnings situation is likely to be
overestimated, which also increases the possibility of
Cambricon stock premium issuance to a certain extent.
From the point of view of valuation indicators,
according to the prospectus and other public
information, although the price-earnings ratio was not
disclosed at the time of the issuance, the 2019 post-
issuance market-to-sales ratio was as high as 58.03
times, far exceeding the average static market-to-
sales ratio of comparable companies, which means
that relative to the level of the industry in the same
period, the market pricing or its market pricing is
ICEML 2025 - International Conference on E-commerce and Modern Logistics
32

overestimated. The market environment also has an
impact. 2020 Cambricon listing, the science and
technology innovation board for unprofitable
technology enterprises to open the listing channel, the
market for artificial intelligence and other emerging
technology fields enthusiasm. The opening price of
250 yuan on the first day of listing, the initial peak of
297.77 yuan, followed by significant fluctuations.
The dramatic ups and downs of the stock price reflect
the market's disagreement over the reasonableness of
its offering price, side by side indicating that there
may be a premium at the time of issuance. When
Cambricon went public in 2020, the Science and
Technology Innovation Board happened to open up a
listing channel for unprofitable technology
companies, which made the market unprecedentedly
enthusiastic about artificial intelligence and other
emerging technology fields. In this market
environment, investors generally have higher
expectations for emerging tech companies and are
more willing to give them higher valuations.
Cambricon's offering price is likely to be driven by
market sentiment and investor expectations, resulting
in a premium to some extent. This leads to the
conclusion that there is a degree of premium issuance
when Cambium issues in 2020.
Figure 1: Closing price of Cambricon stock from 2023 to 2025 on a line graph (Photo/Picture credit: Original).
Figure 2: Line graph of closing price of NVIDIA stock from 2023 to 2025 (Photo/Picture credit: Original).
0
50
100
150
200
250
300
350
400
450
500
550
600
650
700
750
800
$0,00
$20,00
$40,00
$60,00
$80,00
$100,00
$120,00
$140,00
$160,00
Equity Premium Puzzle in Cambricon: Evidence from Behavioural Finance
33

4.2 Select Companies in the Same
Industry for Comparison to
Determine the Extent of Premiums
According to Dolphin Wealth's February 14, 2025
data, Cambricon has a negative TTM P/E ratio and is
in the red. Tiger International's February 13, 2025
data shows that NVIDIA has a P/E TTM of 53.42.
Finance's February 6, 2025 data shows that the
semiconductor industry has a P/E average of 123.31x
and a median of 73.16x (seen from Fig. 2).
Cambricon as a newcomer in the field of AI chips,
in the expansion period, a large number of resources
into research and development, resulting in the
current unprofitable, negative price-earnings ratio.
Nevertheless, the market is optimistic about its
potential in the domestic AI chip market, giving it a
high market value. NVIDIA, with its leading position
in the global AI chip and graphics processing field,
has achieved steady growth in revenue and profit by
virtue of its mature product line, extensive customer
base and stable market share, and its price-earnings
ratio reasonably reflects its operation and market
value. Cambricon cannot be directly compared with
the industry average due to losses, deviation is
obvious, reflecting market pricing based on future
expectations; NVIDIA's P/E ratio is lower than the
industry average but higher than the median, if only
compared with the industry's high-quality enterprises,
the level reflects its leading advantage.
Cambricon although losses but high market value,
indicating that the market for its future earnings
expectations are extremely high, the stock price or
there is an overdraft on future performance, contains
a large potential premium, if the future earnings are
less than expected, the stock price adjustment
pressure. Compared with the industry, the degree of
deviation reflects the market's high expectations for it
to become the domestic AI chip leader. For NVIDIA,
assuming that future earnings growth is sustained,
when your former price-earnings ratio is in a
reasonably high range, there is a certain premium,
which stems from the market's recognition of its
leading position, technological and market
advantages and future profitability. Compared to
industry averages, NVIDIA's price-earnings ratio is
low, but compared to high-quality companies in the
industry, the premium reflects the market's
confidence in its future earnings growth.
4.3 Analysis of the Reasons for the
Premium
Investor sentiment amplifies irrational market
volatility through cognitive biases and group
contagion mechanisms, and when market sentiment is
in a state of exuberance, the valuation of emerging
technology assets may systematically deviate from
the fundamentals, creating a persistence premium
(Baker & Wurgler, 2006). Second, investors tend to
categorize risky assets in separate mental accounts
and relax risk constraints, resulting in significantly
higher allocation weights to Cambricon-related assets
than predicted by traditional asset pricing
models(Thaler, 1985) . Further, the overconfidence
bias prompts investors to focus excessively on
optimistic signals and ignore tail risks during
information processing, thus reinforcing the positive
feedback loop of premium formation (Odean, 1998) .
As an AI chip company at the forefront of technology,
Cambricon has a high level of industry attention.
Cambricon's stock price has risen sharply in 2024,
and the Baidu Information Index has crested more
and more frequently, indicating that investors
subconsciously amplify Cambricon's development
prospects and ignore some of the potential risks in
their decision-making, thus buying a large number of
Cambricon shares and pushing the stock price to
continue to rise (seen from Fig. 3).
Figure 3: Baidu Search Index (Photo/Picture credit:
Original).
Figure 4: Cambricon Share Price Trend (Photo/Picture
credit: Original).
0
200
400
600
800
1000
ICEML 2025 - International Conference on E-commerce and Modern Logistics
34

From the stock price volatility data (illustrated in
Fig. 4), Cambricon in 2023 at the end of September
stock price of about 230 yuan, to 2025 January rose
to more than 600 yuan, January 10 once rushed to
777.77 yuan, or 238%, more than the average rate of
increase of the artificial intelligence chip industry in
the same period. Seen from Table 1, on November 29,
2024, Cambricon was added to the SSE 50 index, and
its share price rose by +15% in the month before its
inclusion in the SSE 50, and +32% in the week after
its inclusion, with the turnover rate rising from 3.5%
to 12.8%. The impact of early leakage and official
release of information on the market should not be
underestimated, investors will adjust their investment
decisions based on this information, and the positive
nature of the information (e.g., inclusion in important
indices) makes investors more optimistic about the
future expectations of Cambricon, which increases
the demand for buying, and pushes the stock price up.
Table 1: Stock Price, Average Daily Volume, Turnover
Ratio.
Time Period Stock
Price
Increase
Average
daily
turnover
(billion yuan)
Turnover
Rate
One month
before
inclusion in
SSE
+15% 8.2 3.5%
One week
after inclusion
+32% 25.6 12.8%
Day after
brokerage
ratin
g
release
+12% 18.3 9.1%
Figure 5 Baidu Information Index (Photo/Picture credit:
Original).
As can be seen from the Baidu Information Index
given in Fig. 5, in order to meet the demand of
investors for relevant information, the organization
has increased its research investment. Positive ratings
and research reports from a number of investment
institutions have kept Cambricon hot. According to
the data of Securities Star, a total of 3 institutions
gave ratings, including 2 buy ratings and 1 hold rating.
This concentrated release of information and positive
guidance has intensified market demand for
Cambricon stock, driving the stock price ever higher.
From the data of the day following the release of the
brokerage ratings, the stock price rose by +12%, with
an average daily volume of 1.83 billion yuan and a
turnover rate of 9.1%, indicating that investors
reacted positively to this information, further pushing
the stock price away from the reasonable range and
causing the stock price to overflow (shown in Fig. 6).
The synergy of these three types of behavioural
mechanisms may trigger the phenomenon of
"cognitive resonance" , in which emotionally driven
attentional biases, the segregation effect of mental
accounts, and misattributions due to overconfidence
combine to drive up pricing distortions in specific
asset classes (Barberis & Huang, 2008).
Herd Behavior is an irrational behavior in which
investors abandon their own information analysis and
blindly imitate others' decisions in an environment of
information asymmetry or uncertainty (Banerjee,
1992). When individuals believe that others' behavior
implies private information, they may choose to
follow it even if it contradicts their own judgments,
thus forming an information waterfall (Bikhchandani
et al., 1992). In financial markets, the herd effect is
often manifested as convergent trading by investors
influenced by the information waterfall, resulting in
asset prices deviating from fundamental value
Figure 6: Cambricon Institutional Positions and Shares
(Photo/Picture credit: Original).
The spike in Cambricon's stock price is essentially
a positive feedback loop triggered by the collective
irrational behaviour of market participants. Head
0
0,5
1
1,5
2
2,5
0
200
400
600
800
End of
2022
End of
2023
Mid 2024 Q3 2024
Institutional Position (number)
Number of Shares in Position (Billions)
Equity Premium Puzzle in Cambricon: Evidence from Behavioural Finance
35

institutions accelerated the concentration of positions,
institutional holdings soared from 146.6 million
shares in 2022 to 193.9 million shares in 2024Q3 (an
increase of 32.3%), but the number of institutions fell
from a peak of 739 to 302. Among the top ten
shareholders, Huaxia, Efontaine and other science
and technology ETFs combined position accounted
for more than 12% of the formation of passive capital
bottoming effect - index funds need to be configured
according to the weighting machinery, even if the
valuation of the abnormally high is still forced to buy
The relative return ranking mechanism used by
public equity funds has created a "prisoner's
dilemma" among fund managers. When technology
stocks like Cambricon show excess returns, even if
fund managers think the stock fundamentals may be
overvalued, but in order to avoid short-term
performance decline, they have to follow the buy,
which further pushes up the stock price. Vanguard
Mo Haibo has taken a long position in Cambricon for
six consecutive quarters since 2023, gaining a
demonstration effect of 37.63% excess returns,
triggering other fund managers to follow, which led
to the concentration of Cambricon's institutional
holdings soaring from 12.7% to 21.4% in three
months. As listed in Table 2, from the end of 2023 to
the third quarter of 2024, retail investors continued to
sell, while institutions absorbed chips through block
trading and other channels, the number of
shareholders fell from 33.18 million to 26.47 million.
2024, when the number of institutions reached 739,
the market is close to the "crowded trading" tipping
point. According to the game theory model, when
more than 70% of institutional positions are
concentrated in the same underlying, the individual
optimal strategy from "follow the group" to
"preemptive withdrawal". The number of institutions
slashed in the third quarter, but total holdings fell by
only 7.4%, suggesting that institutions still holding
shares are maintaining control by adding to their
positions, creating a new equilibrium.
Table 1: Concentration Quartile of Shareholding of Top 10
A-Shareholders (2000-2023)
Quartile 10% 25% 50% 75% 90%
Shareholdin
g
18% 24% 30% 33% 34%
The current shareholding of the top ten
shareholders has reached 37.2%, exceeding the 90%
quartile of A-share history. According to the liquidity
layering model, when the concentration of
institutional positions exceeds 30%, the market depth
(Market Depth) falls to the danger zone, when 1% of
the selling volume can trigger more than 15% of the
price retraction. The irrational prosperity of
Cambricon is the ultimate interpretation of narrative
economics. The grand narrative of "autonomous and
controllable AI chips" has made Cambricon a vehicle
for symbolic capital. Institutions have reinforced this
narrative by continuing to take positions, creating a
consensus premium similar to that of Bitcoin - prices
are no longer dependent on financial metrics, but
rather on the strength of participants' belief in the
narrative.
5 CONCLUSIONS
To sum up, the abnormal volatility of Cambricon's
stock price reveals the structural defects in the A-
share market in institutional design and behavioural
regulation. From the perspective of government
regulation, it is necessary to build a dynamic risk
prevention and control system: on the one hand, for
the risk of institutional position concentration
breaking through the historical quartile value, it is
recommended to introduce counter-cyclical
adjustment tools, dynamically adjust downward the
disclosure threshold for position concentration (e.g.,
mandatory disclosure of trading intent when the
shareholding ratio exceeds 25%), and correct
unilateral market volatility by expanding the
underlying securities financing and introducing
hedging tools such as individual stock options; on the
other hand, it is necessary to Optimize ETF
subscription and redemption rules, set valuation
deviation thresholds for index constituents (e.g., P/E
ratio deviating from the industry average ± 2 standard
deviations), and allow fund managers to suspend
subscriptions or adjust weightings to curb the bubble-
boosting effect of passive funds.
At the same time, the herd behaviour of
institutional investors and appraisal mechanism flaws
urgently need systematic correction. It is
recommended to reconstruct the appraisal system
with three-year rolling returns as the core, break the
“prisoner's dilemma” of short-term relative return
ranking, and establish an internal AI public opinion
monitoring system to identify emotional bias in the
market narrative (e.g., the concept of “localization of
AI chips” speculation). At the level of liquidity risk
management, it is necessary to preset rules for
mandatory position reduction (e.g., initiate a gradient
reduction when the concentration of positions
exceeds 30%), so as to avoid the risk of stampede due
to insufficient market depth.
For individual investors, behavioural finance
interventions are needed to improve decision-making
rationality. Investor education should strengthen the
ICEML 2025 - International Conference on E-commerce and Modern Logistics
36

correction of cognitive biases such as mental accounts
and overconfidence, and be especially alert to the risk
accumulation effect of categorizing high-risk assets
as “dream accounts”. At the practical level, we can
establish a reverse layout framework, when the stock
search index and the stock price show a positive
correlation (such as the correlation coefficient > 0.8),
it will be regarded as an early warning signal of a
bubble to implement the position reduction operation.
The above multi-dimensional policy synergies can
provide institutional safeguards to improve the
pricing efficiency of high-volatility technology
stocks.
AUTHOR CONTRIBUTIONS
All the authors contributed equally and their names
were listed in alphabetical order.
REFERENCES
Baker, M., Wurgler, J., 2006. Investor sentiment and the
cross-section of stock returns. Journal of Finance,
61(4), 1645-1680.
Banerjee, A. V., 1992. A simple model of herd behavior.
The Quarterly Journal of Economics, 107(3), 797-817.
Barberis, N., Huang, M., 2008. The loss aversion/narrow
framing approach to the equity premium puzzle.
Handbook of the Equity Risk Premium, 199-212.
Bikhchandani, S., Hirshleifer, D., Welch, I., 1992. A theory
of fads, fashion, custom, and cultural change as
informational cascades. Journal of political
Economy, 100(5), 992-1026.
Calzadilla, J. F., Bordonado-Bermejo, M. J., González-
Rodrigo, E., 2021. A systematic review of ordinary
people, behavioural financial biases. Economic
research-Ekonomska istraživanja, 34(1), 2767-2789.
Fama, E. F., 1970. Efficient Capital Markets: A Review of
Theory and Empirical Work. Journal of Finance, 25(2),
383-417.
Guan, 2020. Venture Capital and Bond Risk Premium:
From the Perspective of Institutional Investors. Beijing:
Social Sciences Academic Press.
Gupta, S., Shrivastava, M., 2022. Herding and loss aversion
in stock markets: mediating role of fear of missing out
(FOMO) in retail investors. International Journal of
Emerging Markets, 17(7), 1720-1737.
Jiang, B., 2022. Liquidity Premium. in: Investment
Strategies. Palgrave Macmillan.
Lakonishok, J., Shleifer, A., Vishny, R. W., 1992. The
impact of institutional trading on stock prices. Journal
of financial economics, 32(1), 23-43.
Odean, T., 1998. Volume, volatility, price, and profit when
all traders are above average. Journal of Finance, 53(6),
1887-1934.
Shen, W., 2023. Research on the Liquidity Premium in
China's Stock Option Market. Zhejiang University.
Shiller, R. J., 2003. From efficient markets theory to
behavioral finance. Journal of economic
perspectives, 17(1), 83-104.
Tang, W., Wu, T., Xu, L., 2017. Skewness Preference and
IPO Anomalies in China. Annals of Economics &
Finance, 18(1).
Thaler, R. H., 1985. Mental accounting and consumer
choice. Marketing Science, 4(3), 199-214.
Yang, H., 2011. External Risk, Heterogeneous Beliefs and
Idiosyncratic Volatility Risk Premium. Journal of
Management Sciences in China, 11, 71-80.
Equity Premium Puzzle in Cambricon: Evidence from Behavioural Finance
37