Abstract
Purpose
The paper aims to investigate differences in non-professional and professional stock investors’ trust in and tendency to follow financial analysts’ buy and sell recommendations.
Design/methodology/approach
Online experiment conducted in Sweden in March 2022 comparing non-professional private investors (n = 80), professional investors (n = 33), and master students in finance (n = 28). Information was presented about four company stocks listed on the New York stock exchange. Two stocks were buy-recommended and two stocks sell-recommended by financial analysts. For one stock of each type, the recommendation was presented to participants. Dependent variables were predictions of the stock price after three months, ratings of confidence in the predictions and choices of holding, buying or selling the stock. Ratings were also made of the importance of presented stock-related information as well as trust in analysts’ skill and integrity.
Findings
More positive return predictions were made of buy-recommended than sell-recommended stocks. Non-professionals and to some degree finance students tended to trust financial analysts more than professional investors did and they were more influenced by the presentation of the buy recommendations. All groups made too optimistic return predictions, but the professionals were less confident in their predictions, more likely to sell the stocks and lost less on their investments.
Originality/value
A new finding is that non-professional stock investors are more likely than professional stock investors to trust financial analysts and follow their recommendations. It suggests that financial analysts’ recommendations influence non-professional investors to take unmotivated investment risks. Non-professionals in the stock market should hence be advised to exercise more caution in following analysts’ recommendations.
Keywords
Citation
Jansson, M., Michaelsen, P., Sonsino, D. and Gärling, T. (2024), "Non-professional versus professional investors’ trust in financial analysts’ recommendations and influences on investments", Review of Behavioral Finance, Vol. 16 No. 5, pp. 860-882. https://doi.org/10.1108/RBF-07-2023-0191
Publisher
:Emerald Publishing Limited
Copyright © 2024, Magnus Jansson, Patrik Michaelsen, Doron Sonsino and Tommy Gärling
License
Published by Emerald Publishing Limited. This article is published under the Creative Commons Attribution (CC BY 4.0) licence. Anyone may reproduce, distribute, translate and create derivative works of this article (for both commercial and non-commercial purposes), subject to full attribution to the original publication and authors. The full terms of this licence may be seen at http://creativecommons.org/licences/by/4.0/legalcode
1. Introduction
Financial markets are in academic research hypothesized to be efficient and impossible to predict (Fama, 1970; Malkiel, 2005). Hence, financial analysts’ forecasts and recommendations should have little if any importance for investors’ returns. Empirical research partly supports this assumption (e.g. Barber et al., 2001). Regardless of this, many retail investors appear to have a high trust in financial expertise (Huber et al., 2010; Peterson et al., 2015). Furthermore, although professional investors seem not in general to trust financial analysts’ advice and forecasts, they may value the analysts’ industrial expertise as well as their contacts with company managements (Brown et al., 2015, 2016, 2019). A vast research literature exists focusing on analysts’ important roles as financial intermediators, their work process, how they evaluate information, and the accuracy of their recommendations (Womack, 1996; Barber et al., 2001; Asquith et al., 2005; Ramnath et al., 2008; Twedt and Rees, 2012). Less research has investigated the problem raised in this study, whether financial analysts’ buy and sell recommendations are trusted and followed.
Professional investors employed by large institutional asset owners are the primary recipients of financial analysts’ forecasts and recommendations, although later these reports become public and available to small private retail investors. By making the forecasts and recommendations public, analysts can influence the prices of analyzed stocks in the direction of the forecasts and thus, at least in the short-term, confirm their correctness (Sant and Zamanb, 1996; Keasler and McNeil, 2010). It has also been argued that financial analysts make their forecasts and recommendations public with the purpose of benefitting themselves and their institutional clients (e.g. Michaely and Womack, 1999; Beyer et al., 2010; Malmendier and Shanthikumar, 2014). Trust in financial analysts’ integrity is hence possibly undermined by the suspicion that the analysts attempt to “boost” stock prices of affiliated companies. The implicated conflict of interest also presumably contributes to analysts’ inclination to issue significantly more buy than sell recommendations (Ramnath et al., 2008; Lin et al., 2013).
The observation that financial analysts’ forecasts and recommendations have significant influences on stock prices suggests that their ability to forecast the market is trusted. However, to what extent professional investors working for financial institutions and private retail investors differ in this respect is an open question. In previous research, a focus has been on understanding the analysts’ decision-making process as well as the statistical properties and accuracy of their earning forecasts (Ramnath et al., 2008). The aim of this study is to compare non-professional investors to professional investors, investigating whether the non-professionals differ from the professionals in trusting and following analysts’ buy and sell recommendations.
The paper proceeds as follows. The next section reviews previous research. The third section presents the research questions, followed by a section describing an experiment. The fifth section presents the results of the experiment, including the differences between professionals, finance students and non-professionals in a questionnaire measure of trust in financial analysts and the influences of the analysts’ buy and sell recommendations on investments, return predictions, and confidence. The sixth section is a discussion of the results and their interpretation. The seventh section concludes.
2. Previous research
Although it is not always true that advice by financial experts improves non-professional investors’ decision-making (Barber et al., 2001; Rahman et al., 2019), the non-professionals appear to have trust in expertise (Huber et al., 2010; Peterson et al., 2015). Perceiving that advice by financial experts is trustworthy may reduce uncertainty in assessments of risk among those taking advice. When uncertainty is high, the tendency to take advice furthermore increases (Bouteska and Mili, 2023). Trust in financial advice is hence derived partly from the inherent uncertainty that characterizes investments in financial markets in general and stock markets in particular.
In the finance context, trust can be decomposed into two primary dimensions (Ennew et al., 2011): (1) Trust in financial experts’ ability, skill, and experience to analyze and forecast the market; (2) Trust in that financial experts provide advice exclusively with the motive to benefit the advisees. The first aspect concerns the value of financial expertise for investment in the stock market. Previous research does not convincingly show that financial professionals are able to make more accurate forecasts and recommendations than non-professionals or chance (Törngren and Montgomery, 2004; Jansson, 2019). On the contrary, research has shown that stock-market forecasting is a challenging task in which professional investors rarely outperform market indexes (Barber et al., 2001; Rahman et al., 2019). Stock-market professionals as well as non-professionals are also prone to several judgmental biases that may cause bubbles in stock markets (Gärling, 2011). Prominent cases include under- and overreaction to news (De Bondt and Thaler, 1985) and the disposition effect, selling winning and keeping losing stocks (Shefrin and Statman, 1985). Investors are prone to these and other biases partly due to over-confidence in their ability to predict the market (Menkhoff et al., 2013; Sonsino and Regev, 2013). Other reasons are inherent cognitive limitations (Gärling et al., 2009). According to Shefrin (2007), judgmental biases are the main reason for stock investors’ irrationality and poor investment performance. Taken together, as also argued by Bodnaruk and Simonov (2015), the results of previous research suggest that non-professional investors would not benefit immensely from advice by financial experts.
The second aspect concerns financial analysts’ integrity, objectivity, and impartiality. Integrity threats arise due to the disparate incentives of the investment banks’ commercial clients (e.g. firms issuing stocks) and their brokerage clients (investors). The latter aim at unbiased research and forecasts, while the former typically benefit from optimistic forecasts of returns. Investment banks thus issue allegedly objective expert recommendations, while also gaining underwriting and brokerage commissions from promoting given stocks. Strategic distortion of analysts’ forecasts and recommendations has been documented in several studies (e.g. Lidén, 2004; Beyer et al., 2010; Malmendier and Shanthikumar, 2014). Michaely and Womack (1999) found that stocks that underwriter analysts recommend perform more poorly than stocks recommended by unaffiliated brokers prior to, at the time of, and after the recommendation date. Mola and Guidolin’s (2009) analysis of the 1994–2006 analysts’ recommendations for S&P 500 stocks reveals that sell-side analysts are likely to frequently assign more favorable ratings to a stock after that the analysts’ affiliated mutual funds invest in that stock. While analysts’ optimism about stocks held by affiliated investors declined from 2002, an upgrade from “buy” to “strong buy” still links with the investment of affiliated funds in the company. Studying the 2004–2013 stock ratings of Chinese analysts and the stock holdings of funds affiliated with these analysts, Yin (2018) showed that analysts make overoptimistic ratings of stocks held by affiliated funds, while the affiliated funds sell their holdings in the stocks realizing substantial net returns. The “affiliated” analysts appear to serve the interests of their associated funds rather than impartially advising investors. Hence, uncritically following sell-side financial analysts’ investment recommendations can be costly. Jegadeesh et al. (2005) showed that a consensus recommendation adds value only to stocks with favorable quantitative characteristics, for instance, value stocks and positive momentum stocks. Among stocks with unfavorable quantitative characteristics, higher consensus recommendations are associated with worse subsequent returns.
Research also shows that analysts affiliated with the analyzed company by banking ties tend to be slower in downgrading buy and hold recommendations and faster in upgrading from hold recommendations (O’Brien et al., 2005). Professional institutional investors and non-professional investors furthermore respond differently to downgrading of recommendations (Malmendier and Shanthikumar, 2007; Mikhail et al., 2007). Non-professional investors do not fully consider the effects of analysts’ incentives on the credibility of their reports. They buy stocks more than professionals following positive recommendation revisions, and they do not alter their positions in response to hold recommendations (Malmendier and Shanthikumar, 2007). Professional large investors on the contrary sell stocks classified as hold and make profitable trades based on negative analysts’ reports (Mikhail et al., 2007). In an international study Balboa et al. (2009) showed that by controlling for the analysts’ optimism bias in each of eight developed countries, the predictive power of their recommendations improved. The bias in analysts’ forecasts and recommendations still varies internationally, due to differences in regulatory environment (Rahman et al., 2019) and culture (Clement et al., 2003).
Previous research has also shown that professional investors focus less on financial analysts’ recommendations than on their industrial knowledge and access to company management (Brown et al., 2015, 2016, 2019; Spence et al., 2019). Buy and sell recommendations issued by financial analysts thus seem to be less valuable to professional investors who tend to perceive that the recommendations express biased beliefs. Financial analysts are instead valued for offering substantive value-adding interactions with companies (Brown et al., 2015, 2016). Contrary to professional investors, it is plausible to assume that non-professional investors are more reliant on financial analysts’ recommendations because their lack of time and knowledge makes it more difficult for them to utilize other sources of information.
To summarize, research shows that the stock market is volatile and difficult to forecast, and that financial analysts’ judgments are frequently inaccurate and biased, partly due to their conflicting interests. Regardless of this, financial analysts’ expertise is demanded by stock-market investors, both professional and non-professional (see, e.g. recent discussion by Boulland et al., 2022), although these groups of investors may perceive different benefits of financial analysts’ expertise. Professionals have been shown to attach little importance to analysts’ recommendations and forecasts, but they value analysts’ industrial expertise and that they offer a “back-channel” to company managers (Brown et al., 2016, 2019). In contrast, non-professional investors appear to have trust in financial expertise (Peterson et al., 2015) because it reduces uncertainty and enables assessments of risk (Huber et al., 2010).
3. Research questions
Based on previous research suggesting that professional and non-professional investors differ in how much they trust and follow financial analysts’ advice and recommendations, we conducted an experiment to investigate the influence of financial analysts’ buy and sell recommendations on the performance of non-professional compared to the performance of professional stock-market investors. The experiment was conducted online with actual stocks which were either buy or sell recommended by financial analysts. For each stock the participants were informed or not informed about the analysts’ recommendation before they made predictions of the future price, rated confidence in the predictions, and made fictitious investments. A scale was developed to measure trust in financial analysts' skill and integrity with the aim of investigating to which degree the influence of financial analysts’ recommendations was moderated by trust. The following specific research questions were posed:
To what degree do non-professional compared to professional stock-market investors follow financial analysts’ buy and sell recommendations?
To what degree are the different investors’ predictions of future stock prices influenced by financial analysts’ recommendations and do these influences improve the accuracy of the predictions?
To what degree do financial analysts’ recommendations increase the different investors’ confidence in their predictions?
How does the different investors’ trust in financial analysts’ skill and integrity influence their predictions and investments?
4. Method
4.1 Participants
A total of 141 Swedish participants recruited to three groups answered an online questionnaire programmed in Qualtrics (www.Qualtrics.com). The questionnaire was administered between March 15 and March 27, 2022. Participants in the different groups answered the questionnaire in a shuffled order. They were asked to predict future prices of and make fictitious investments in four stocks. In compensation for participation SEK 1000 (appr. USD 100) were paid to each of 15 participants who made the most accurate predictions of the stock prices after three months. Participants did not receive any other compensation.
In one group (henceforth referred to as “non-professionals”) participants (n = 80) were invited through an electronic newsletter published by the Swedish Shareholder Association. This organization is open to anyone to join at a modest annual membership fee. Through different channels it gives members advice on investments in the Swedish fund and stock markets. Approximately 70,000 members received the newsletter with the invitation to participate. Initially 185 agreed to participate.
A second group of participants referred to as “professionals” (n = 33) consisted of employees of Swedish investment firms in the roles of investors. Contacts were made with them through phone or email, using information listed on the firms’ web pages. In total approximately 135 employees were invited, of which 105 initially agreed to participate.
A third group of participants referred to as “finance students” (n = 28) was recruited to be compared to the other groups [1]. Approximately 349 students enrolled in the master program in finance at University of Gothenburg were invited by email. Of those invited, 291 initially agreed to participate.
Age, sex, higher finance education, and self-rated stock-market knowledge reported in the questionnaire are given in Table 1 for each group. Men dominates in all three groups. Compared to the other groups, more professionals have higher finance education and rate that they have more stock-market knowledge. Non-professionals are older and fewer have higher finance education.
4.2 Questionnaire
In the first part of the questionnaire, participants were given information about four stocks listed on the New York Stock Exchange in the Dow Jones large cap list. Two of the selected stocks were in the consumables sector (1 and 3) and two in the energy sector (2 and 4). Based on WSJ-aggregated data from financial analysts, the stocks were either buy-recommended (1 and 2) or sell-recommended (3 and 4).
The stocks were presented separately to the participants in individually randomized orders. For each stock the following information was given: (1) Company information (name of company, sector, and major products); (2) Financial key indicators (current p/e ratio as well as dividend yield, revenues, results, net income, EBITDA, EPS, and debt ratio for each year from 2017 to 2021); (3) Price history (a chart of the stock prices over the last three years); (4) Analysts’ buy and sell recommendation whose presence or absence was experimentally manipulated as described below; and (5) Current stock prices (continually updated), 1-day percentage change, and average prices for the last 50 and 200 days. Participants were asked to click a button to show (1)–(4), whereas (5) was mandatory shown. A sample display of the information for one stock is found in Appendix 1. After having read the information about the stock, participants made an investment choice, a prediction of the stock’s price after three months, and a rating of confidence in their prediction. For each stock participants also made a rating of each company on a seven-point scale ranging from not at all familiar (1) to very familiar (7).
In the second part of the questionnaire, participants rated the importance of each type of stock-related information for their four investment decisions considered together. The ratings were made on five-point scales ranging from “Not at all important” (1) to “Very important” (5).
The third part of the questionnaire consisted of measures of trust in the skill and integrity of financial analysts. The participants used five-point scales ranging from “Do not agree at all” (1) to “Agree completely” (5) to rate their degree of agreement with eight statements about financial analysts (see Appendix 2). In addition, two separate ratings were made on 5-point scales (1-to-5) of trust in analysts´ buy and sell recommendations, respectively. Three of the statements (e. g., “Analysts´ recommendations for major companies are usually accurate”) and the separate ratings were assumed to be related to trust in analysts’ skill and five of the statements (e. g., “Analysts’ recommendations for a company are influenced by whether they have the same company as a client”) were assumed to be related to trust in analysts’ integrity.
Finally, in the fourth part of the questionnaire, participants answered questions about sex, age, and higher finance education, and rated their knowledge of stock markets. Control questions were also asked about participants’ work role and experience in the finance industry.
4.3 Experimental treatments
For all participants in each of the three groups (non-professionals, finance students, and professionals), one of the stocks 1 and 2 (buy-recommended stocks) were presented with the buy recommendation disclosed, and one of the stocks 3 and 4 (sell-recommended stocks) with the sell recommendation disclosed (see Figure 1). Which stock was presented with or without the recommendation was counterbalanced such that across participants in each group the recommendations were presented about equally often for both stocks.
4.4 Measures
Investments. For each of the four stocks, participants imagined having $100,000 to invest with $50,000 already invested in the stock. Their task was to choose how much to decrease investment (sell) in the stock or how much of the remaining $50,000 to invest (buy) in the stock. Responses were made by moving a slider on a scale with marked increments of $1,000 from 0 (selling all owned stocks) to 100 (buying stocks for all money) through 50 (neither buying nor selling stocks).
Predictions. Participants were asked to predict each stock’s price three months after the day of answering the questionnaire. The predicted stock price in USD was typed in a designated field. Predicted percent return was obtained as the signed difference between the prediction and the stock price at the time of prediction divided by the stock price and multiplied by 100. Accuracy in % was calculated as the difference between the signed predicted percent difference and the signed percent difference between the stock price after three months and the stock price at the time of prediction.
Confidence ratings. For each stock participants used a percent scale to indicate their level of confidence that the stock price after three months would be within ±5% of their prediction. On a following page in the questionnaire separated from the prediction task of each stock, participant’s prediction was presented with the calculated ±5% limits displayed. Responses were made by means of moving a slider ranging from 0% (no confidence) to 100% (full confidence) in steps of 1%.
5. Results
5.1 Data preparation
Participants who had not answered all questions were excluded. Almost all of them quitted directly after starting to answer the online questionnaire without providing any replies. Five participants had when quitted answered half or more of the questionnaire.
Two participants originally recruited to the non-professional group and three to the finance student group were reassigned to the professional group since they indicated in the questionnaire that they worked as professional investors.
Some outlying values were observed in the predictions of future stock prices. Most of them were attributed to a few participants who appeared to have made estimates in their native currency SEK. Assuming this was the mistake they had made, 68 of 564 values (12.0%) were corrected by conversion to USD [2]. One finance student was discarded because of making predictions and investments markedly different from the others.
5.2 Manipulation checks
The check of how familiar the companies were to participants showed that except for stock 2 for which the means of the familiarity ratings were at the midpoint of the 7-point scale, the ratings were close to the lowest value in each condition and group. The results of the checks of whether participants attended to the information about the stocks showed that more than 90% of professionals and with one exception more than 90% of non-professionals but fewer finance students viewed the stock-related information. In all three groups with two exceptions (both 67.9%), between 80 and 90% viewed the analysts’ recommendations. The subsequent analyses were performed on all data assuming that the minority who did not view the information considered it nonessential to the tasks they were requested to perform.
5.3 Experimental treatment effects
Predictions. The results presented in Table 2 show that the return predictions are higher in all groups for buy-recommended than sell-recommended stocks. This effect was significant in the mixed effects linear regression analysis [3] reported in Table 3. Another significant effect in the regression analysis was associated with the higher return predictions in all three groups when the buy recommendation is presented. It should furthermore be noted that the stock prices decreased substantially during the three-month period, thus resulting in large over-predictions for particularly the buy-recommended stocks.
Confidence ratings. Table 4 presents the average ratings of confidence in the three-month return predictions made on the scale from 0% confident to 100% confident. As shown, confidence in the predictions of the buy-recommended stocks exceeds confidence in the predictions of the sell-recommended stocks. Furthermore, non-professionals are overall more confident in their predictions than professionals are. These observations are substantiated by significant effects in the regression analysis (see Table 5). It is also suggested that buy-recommended stocks have stronger effects on confidence than sell-recommended stocks for non-professionals and finance students and that professionals are more confident than non-professionals when being presented the sell recommendation.
Investments. The means in Table 6 show that more buy-recommended stocks are bought and less sold compared to sell-recommended stocks, and that non-professionals buy more and sell less than professionals. Both effects are significant in the regression analysis (see Table 7). The difference between non-professionals and professionals in how much they buy buy-recommended stocks and sell sell-recommended stocks was marginally significant. Although less pronounced for the professionals, all average returns were negative due to the substantial decrease in stock prices during the three-month period.
Table 8 presents the ratings of the importance of the stock-related information for the investment. One-way analyses of variance (ANOVA) [4] yielded significant differences between the groups for company information, price history, and analysts’ recommendations. Information about financial key indicators was rated equally important by the groups. Together with company information it was rated highest. Price history was rated third highest and analysts’ recommendations lowest. Both were rated higher by non-professionals than professionals with finance students in between. The ratings of importance of the specific financial key indicators did not differ significantly between the groups. The Pearson correlation between the ratings by non-professionals and professionals was 0.27 (p = 0.563), between non-professionals and finance students 0.80 (p = 0.031), and between professionals and finance students 0.55 (p = 0.205). Net income was rated as the most important to non-professionals and finance students, whereas EBITDA was rated as the most important to professionals. Importance of dividend yields was rated lowest by all three groups.
5.4 Trust effects
Means, standard deviations and Pearson correlations between the ratings of trust in financial analysts’ skill and integrity are shown in Appendix 2. The ratings were submitted to principal component analysis (PCA) extracting two orthogonal factors accounting for 48.0% of the total variance. After varimax rotation the first factor interpreted as trust in skill accounted for 23.3%, and the second factor interpreted as trust in integrity accounted for 24.7% of the total variance. Ratings of two of the statements and both separate ratings of trust had high loadings (ranging from 0.65 to 0.81) on only the first factor and ratings of three statements had high loadings (ranging from −0.64 to −0.79) on only the second factor. Ratings of the two remaining statements had lower loadings (−0.38, −0.46) on the second factor but loaded only on this factor. Uncorrelated standardized factor scores (M = 0; SD = 1) were estimated and used in the following analyses. Table 9 shows that the average factor scores are higher for the non-professionals, but only the difference in trust in analysts’ integrity was significant due to non-professionals’ higher factor score compared to finance students’ factor score.
Table 10 shows the results of the mixed effects linear regression analysis of investment repeated with both trust factor scores entered. A significant interaction substantiates that trust in integrity increased the difference between buy-recommended and sell-recommended stocks. Trust in skill tended to increase the effect of presenting the buy recommendation and more in non-professionals than professionals. Trust in integrity likewise tended to increase the effect of presenting the buy recommendation more in non-professionals than professionals.
6. Discussion
Previous research has indicated that buy versus sell recommendations have different influences on professional than non-professional investors (Lidén, 2004; Mikhail et al., 2007). While the professionals tend to have more trust in financial analysts’ sell than buy recommendations, the opposite is true of the non-professionals. In our experiment we found that three-month return predictions, confidence in the predictions, and investments were all higher for buy-recommended than sell-recommended stocks. By experimentally presenting or not presenting the buy or sell recommendations, the effect of the recommendations was assessed for the same stocks. The predictions were more positive when the buy recommendations were presented for the buy-recommended stocks, but presenting the sell recommendations for the sell-recommended stocks made no difference. A similar tendency was found for investments. The market outlook was pessimistic during the period when the experiment was conducted (March 2022). On average all the return predictions overshoot the prices. The professionals more than the non-professionals and finance students appeared to take the market outlook into consideration when making their investments. Thus, on average, they sold the stocks and were less confident in their return predictions than the non-professionals and finance students. As all the stock prices were lower after three months, average returns on investments were negative but the professionals’ returns less negative than those of the other groups. Taken together, the results suggest in line with previous research (Jansson, 2019) that having professional experience and expertise do not place professional investors in a much better position to predict the short-term returns of specific stocks, although it may make investors less confident in their predictions and possibly thus also more aware of the investment risk.
The non-professionals as well as the finance students differed from the professionals most clearly with respect to investments and confidence in the return predictions. Furthermore, the non-professionals tended to have more trust than the professionals in financial analysts’ skill and integrity. A tendency was also observed that both trust in the analysts’ skill and integrity increased the influence on investments of the presentation of the analysts’ recommendations. A stronger trust in financial analysts and their recommendations is therefore one possible explanation of why non-professionals’ investment differs from professionals’ investment.
An interesting and potentially important finding of the experiment is the major impact on investors’ return predictions of the buy-recommended versus sell-recommended stocks that was independent of the presentation of the recommendations. Thus, without knowing the consensus recommendations made by the financial analysts, on average the participants in all three groups made return predictions aligned with the recommendations. Evidence shows that herding or clustering is frequent in stock markets (Spyrou, 2013). Common knowledge that has been proposed as one explanation of herding or clustering (Andersson et al., 2014; Li et al., 2017) may account for these effects of buy-recommended versus sell-recommended stocks despite that no information was disseminated about the recommendations. This raises the question of what common knowledge is utilized by the investors and analysts alike.
Since most participants chose to view all the presented stock-related information, differences in viewing tendency do not shed light on the question of what common knowledge was utilized. The self-ratings of the importance of the different types of information for the investments offer another possibility. Comparing the different groups, the rank order (company information > financial key indicators > price history > analysts’ recommendations) was identical except that finance students ranked financial key indicators higher than company information. Yet, there were differences in importance placed on the four types of stock-related information. Professionals tended to place more importance on company information than did non-professionals and finance students who placed more importance on price history. A common feature of the buy-recommended stocks was that they had a long positive price trend, while the sell-recommended stocks had a long negative price trend. Previous research has shown that inexperienced investors tend to be more trend-chasing than experienced investors (Greenwood and Nagel, 2009), and that non-professional investors herd more and are more sensitive to public information (Spyrou, 2013). Professional investors are also slower in responding to price movements (Li et al., 2017). The influence of buy-recommended and sell-recommended stocks among the non-professionals and finance students in our experiment may thus reflect that they placed high importance on the momentum and price trend. It appears likely that the professionals were less influenced by the same information. Previous research (Brown et al., 2015, 2016, 2019; Spence et al., 2019) has shown that professional investors tend to distrust financial analysts´ recommendation but rely on their expertise in analyzing company information. These observations are consistent with the present findings.
Additional support is provided by the correlations reported in Table 11 between investment and the experimental-treatment variables, return prediction, and confidence. The first two rows showing the correlations with buy-recommended versus sell-recommended stocks and presentation versus no presentation of the recommendations only repeat the results already presented. Of more interest is the significant positive correlations of investment with return prediction and confidence. As may be seen, the professionals’ investments are only influenced by their return prediction and confidence. In contrast the non-professionals and finance students are also influenced by the buy/sell recommended stocks. A possibility is still that all investors extrapolated the price trends when making their return prediction. The differences between the non-professionals and professionals may then be the difference in the confidence in the return predictions. In a companion paper, Sonsino et al. (2023) explored the current data by constructing a competence index having the maximal correlation with the participants' returns across the four investment tasks of the experiment. Consistent with the results presented here, professional occupation had the strongest positive loading on the competence index while confidence in the accuracy of the predictions loaded negatively.
Investor sentiment is a phenomenon believed to result from interpersonal communication (Nofsinger, 2005). Examples include collective optimism versus pessimism. The downturn of the stock markets during the March–June period of 2022 possibly sparked a pessimistic sentiment, faster among the professional than the non-professional investors and finance students. Even though the professionals still made return predictions as the non-professionals and finance students did, they expressed a lower confidence in their predictions. The lower confidence, presumably contributing to the choice of selling the stocks, may be the outcome of a pessimistic investor sentiment not yet spread to the non-professionals and finance students. Selling stocks was the right decision, even though still not enough to avoid losses.
More research is needed of how different types of investors, not only retail and professional investors but also asset owning organizations and separate demographic groups, perceive financial analysts’ forecasts and recommendations. When the usefulness of sell-side analysts’ recommendations has been studied in previous research, some that question the value of financial analysts’ advice argue that their primary role is not to provide accurate investment advice but to offer a backchannel to company management and to generate banking businesses (Brown et al., 2015, 2016). For examples studies by Millo et al. (2023) and Spence and Carter (2014) have shown how path dependency and social inertia conserve what is perceived to be analysts’ role but also force analyst to converge to consensus estimates in their predictions and thus undermine the investment value of their analysis. This study stresses the split perception between professional and non-professional investors of what analysts’ role is and the value of their advice. Professionals hold a more skeptical view of the validity of analysts’ recommendations, while analysts’ recommendations seem to influence non-professionals to take unmotivated risks. Future research should further investigate the relation between financial analysts’ recommendations and non-professionals' risk taking.
7. Conclusions
It is concluded that more positive return predictions are made of stocks that are buy-recommended than stocks that are sell-recommended by financial analysts. Whether this depends on common knowledge of the different price trends or dissemination of the analysts’ recommendations cannot be definitely settled. The evidence still suggests that non-professionals and to some degree finance students tend to trust financial analysts more than professional investors do. Their investments are also more influenced by knowing the financial analysts’ buy recommendations. Professional investors like non-professional investors make too optimistic return predictions but are less confident in their predictions and more likely to sell stocks. A possible explanation is that professionals have stronger awareness of the risk associated with investing in stocks. A supplementary explanation is that in the times of the downturn stock markets in March 2022, a pessimistic sentiment spread faster among professional investors than among other investors.
Figures
Figure 1
Presentation to participants of financial analysts’ buy and sell recommendations
Figure A1
Historical prices/price charts
Descriptives of participants
| Descriptives | Non-professionals n = 80 | Finance students n = 28 | Professionals n = 33 | Significance tests |
|---|---|---|---|---|
| Age in years [M, (SD)] | 61.92a (12.72) | 24.21b (3.37) | 46.73c 12.69) | F(2,138) = 113.76, p < 0.001 |
| Sex (% men) | 87.5 | 85.7 | 93.9 | χ2(2) = 1.26, p = 0.532 |
| Higher finance education (%) | 20.0a | 50.0 b | 78.8c | χ2(2) = 35.22, p < 0.001 |
| Self-rated stock-market knowledge (1–7) [M, (SD)] | 4.60a (1.39) | 4.50a (0.92) | 6.15b (1.25) | F(2,138) = 19.23, p < 0.001 |
Note(s): Values with different subscripts differ significantly at p < 0.05 in Bonferroni-corrected post hoc tests
Source(s): Authors' own creation/work
Means (M) and 95% confidence intervals (CI) of return predictions (%), stock price differences in percent, and accuracy of return predictions
| Non-professionals n = 80 | Finance students n = 28 | Professionals n = 33 | ||
|---|---|---|---|---|
| Dependent variable | Condition | M 95%CI | M 95%CI | M 95%CI |
| Return prediction (%)a | Presentation of buy recommendation (Stock 1, 2) | 7.01 [4.48; 9.55] | 4.13 [−0.15; 8.42] | 5.39 [1.44; 9.34] |
| No presentation of buy recommendation (Stock 1, 2) | 3.82 [0.69; 6.95] | 1.77 [−3.52; 7.05] | −1.44 [−6.32; 3.42] | |
| Presentation of sell recommendation (Stock 3, 4) | −1.31 [−4.82; 2.20] | −2.14 [−8.07; 3.78] | 0.04 [−5.42; 5.50] | |
| No presentation of sell recommendation (Stock 3, 4) | −0.62 [−5.26; 4.02] | −4.81 [−12.65; 3.04] | −1.21 [−8.43; 6.01] | |
| Price difference (%)b | Presentation of buy recommendation (Stock 1, 2) | −7.17 [−9.19; −5.15] | −8.71 [−12.13; −5.29] | −7.70 [−10.85; −4.55] |
| No presentation of buy recommendation (Stick 1, 2) | −8.25 [−10.26; −6.25] | −6.67 [−10.06; −3.28] | −8.31 [−11.43; 5.19] | |
| Presentation of sell recommendation (Stock 3, 4) | −2.79 [−3.44; −2.13] | −3.20 [−4.30; −2.08] | −2.98 [−4.00; −1.95] | |
| No presentation of sell recommendation (Stock 3, 4) | −2.70 [−3.38; −2.02] | −3.06 [−4.21; −1.92] | −3.74 [−4.80; −2.69] | |
| Accuracy of return predictions (difference in %)c | Presentation of buy recommendation (Stock 1, 2) | 14.18 [10.63; 17.74] | 12.84 [6.83; 18.85] | 13.09 [7.56; 18.63] |
| No presentation of buy recommendation (Stick 1, 2) | 12.07 [8.04; 16.10] | 8.44 [1.63; 15.24] | 6.86 [0.59; 13.14] | |
| Presentation of sell recommendation (Stock 3, 4) | 1.48 [−1.91; 4.87] | 1.05 [−4.69; 6.78] | 3.01 [−2.70; 8.30] | |
| No presentation of sell recommendation (Stock 3, 4) | 2.08 [−2.50; 6.65] | −1.74 [−9.48; 5.99] | 2.53 [−4.60; 9.66] |
Note(s): a100[(Prediction in USD ˗ Stock price in USD at t0)/Stock price in USD at t0]. Time t0 refers to when predictions were made. At this time the prices varied across participants for the buy-recommended stocks from USD 43 to USD 45 and from USD 72 to USD 76 and for the sell-recommended stocks from USD 88 to USD 93 and from USD 128 to USD 137
b100[(Stock price in USD at t1 ˗ Stock price in USD at t0)/Stock price in USD at t0]. Time t1 refers to three months after the predictions were made
cReturn prediction (%) – Price difference (%)
Source(s): Authors' own creation/work
Mixed effects linear regression analysis of return predictions (%)
| Independent variable | b | SE | df | t | p |
|---|---|---|---|---|---|
| Intercept | 0.88 | 0.88 | 138.00 | 1.00 | 0.318 |
| Buy-recommended stock (1, 2) vs sell-recommended stock (3, 4) | 2.55 | 0.70 | 414.00 | 3.66 | <0.001 |
| Presentation vs no presentation of buy recommendation (stock 1, 2) | 2.05 | 0.99 | 414.00 | 2.08 | 0.038 |
| Presentation vs no presentation of sell recommendation (stock 3, 4) | 0.54 | 0.99 | 414.00 | 0.54 | 0.586 |
| Non-professionals vs Professionals | 0.19 | 1.29 | 138.00 | 0.14 | 0.886 |
| Non-professionals vs Finance students | 1.16 | 1.35 | 138.00 | 0.86 | 0.392 |
| Buy-recommended stock (1, 2) vs sell-recommended stock (3.4) | 1.28 | 1.02 | 414.00 | 1.24 | 0.214 |
| X Non-professionals vs Professionals | |||||
| Buy-recommended stock (1, 2) vs sell-recommended stock (3.4) | −0.64 | 1.07 | 414.00 | −0.60 | 0.552 |
| X Non-professionals vs Finance students | |||||
| Presentation vs no presentation of buy recommendation (stock 1, 2) | −1.36 | 1.45 | 414.00 | −0.94 | 0.347 |
| X Non-professionals vs Professionals | |||||
| Presentation vs no presentation of buy recommendation (stock 1, 2) | 0.91 | 1.52 | 414.00 | 0.60 | 0.550 |
| X Non-professionals vs Finance students | |||||
| Presentation vs no presentation of sell recommendation (stocks 3, 4) X | −0.09 | 1.45 | 414.00 | 0.06 | 0.953 |
| X Non-professionals vs Professionals | |||||
| Presentation vs no presentation of sell recommendation (stock 3, 4) X | −0.79 | 1.52 | 414.00 | −0.55 | 0.602 |
| X Non-professionals vs Finance students |
Note(s): Pseudo R2Fixed-effects = 0.04, Pseudo R2total = 0.17
Source(s): Authors' own creation/work
Mean (M) confidence ratings (0–100%) and 95% confidence intervals (CI)
| Non-professionals n = 80 | Finance students n = 28 | Professionals n = 33 | |
|---|---|---|---|
| Condition | M 95%CI | M 95%CI | M 95%CI |
| Presentation of buy recommendation (Stock 1, 2) | 51.92 [46.10; 57.75] | 48.79 [38.94; 58.63] | 38.21 [29.14; 47.28] |
| No presentation of buy recommendation (Stock 1, 2) | 50.41 [45.23; 55.60] | 53.11 [44.34; 61.87] | 37.00 [28.93; 45.07] |
| Presentation of sell recommendation (Stock 3, 4) | 44.14 [38.74; 49.53] | 39.61 [30.49; 48.73] | 37.49 [29.08; 45.88] |
| No presentation of sell recommendation (Stock 3, 4) | 48.71 [43.16; 54.26] | 44.61 [35.22; 53.99] | 34.27 [25.63; 42.92] |
Source(s): Authors' own creation/work
Mixed effects linear regression analysis of confidence ratings
| Independent variable | b | SE | df | t | p |
|---|---|---|---|---|---|
| Intercept | 44.02 | 2.02 | 139.00 | 21.84 | <0.001 |
| Buy-recommended stock (1, 2) vs sell-recommended stock (3, 4) | 2.55 | 0.66 | 414.00 | 3.84 | <0.001 |
| Presentation vs no presentation of buy recommendation (stock 1, 2) | −0.27 | 0.94 | 414.00 | −0.28 | 0.777 |
| Presentation vs no presentation of sell recommendation (stock 3, 4) | −1.06 | 0.94 | 414.00 | −1.13 | 0.260 |
| Non-professionals vs Professionals | 7.28 | 2.96 | 138.00 | 2.70 | 0.015 |
| Non-professionals vs Finance students | −2.50 | 3.10 | 138.00 | −0.81 | 0.420 |
| Buy-recommended stock (1, 2) vs sell-recommended stock (3.4) | 1.69 | 0.98 | 414.00 | 1.73 | 0.084 |
| X Non-professionals vs Professionals | |||||
| Buy-recommended stock (1, 2) vs sell-recommended stock (3.4) | 1.87 | 1.02 | 414.00 | 1.83 | 0.068 |
| X Non-professionals vs Finance students | |||||
| Presentation vs no presentation of buy recommendation (stock 1, 2) | −0.87 | 1.38 | 414.00 | −0.63 | 0.528 |
| X Non-professionals vs Professionals | |||||
| Presentation vs no presentation of buy recommendation (stock 1, 2) | 1.89 | 1.46 | 414.00 | 1.31 | 0.191 |
| X Non-professionals vs Finance students | |||||
| Presentation vs no presentation of sell recommendation (stock 1, 2) | −2.67 | 1.38 | 414.00 | −1.93 | 0.054 |
| X Non-professionals vs Professionals | |||||
| Presentation vs no presentation of sell recommendation (stock 1, 2) | −1.44 | 1.45 | 414.00 | −0.99 | 0.320 |
| X Non-professionals vs Finance students |
Note(s): Pseudo R2Fixed-effects = 0.05, Pseudo R2Total = 0.69
Source(s): Authors' own creation/work
Mean (M) investments (0–100), returns, and 95% confidence intervals (CI)
| Non-professionals n = 80 | Finance students n = 28 | Professionals n = 33 | ||
|---|---|---|---|---|
| Condition | M 95%CI | M 95%CI | M 95%CI | |
| Investment (103 USD) | Presentation of buy recommendation (Stock 1, 2) | 57.12 [52.24; 62.01] | 47.43 [39.47; 55.39] | 34.61 [25.06; 44.15] |
| No presentation of buy recommendation (Stock 1, 2) | 51.32 [46.13; 56.52] | 47.29 [38.23; 56.34] | 30.39 [22.11; 38.68] | |
| Presentation of sell recommendation (Stock 3, 4) | 41.90 [35.76; 48.04] | 38.50 [30.18; 46.82] | 31.70 [21.58; 41.82] | |
| No presentation of sell recommendation (Stock 3, 4) | 42.52 [36.82; 48.24] | 36.43 [28.20; 44.66] | 27.61 [18.78; 36.42] | |
| Return (103 USD)a | Presentation of buy recommendation (Stock 1, 2) | −4.09 [−3.75; −4.45] | −4.13 [−3.44; −4.82] | −2.67 [−1.93; −3.40] |
| No presentation of buy recommendation (Stock 1, 2) | −4.24 [−3.81; −4.67] | −3.15 [−0.97; −3.76] | −2.52 [−1.84; −3.21] | |
| Presentation of sell recommendation (Stock 3, 4) | −1.17 [−1.00; −1.34] | −1.23 [−0.97; −1.50] | −0.95 [−0.64; −1.24] | |
| No presentation of sell recommendation (Stock 3, 4) | −1.15 [−0.99; −1.30] | −1.11 [−0.86; −1.37] | −1.03 [−0.70; −1.36] |
Note(s): a(Investment at t0) (% price difference at t1)/100 (see Table 4). Time t0 refers to when predictions were made; time t1 refers to three months after the investments were made
Source(s): Authors' own creation/work
Mixed effects linear regression analysis of investment
| Independent variable | b | SE | df | t | p |
|---|---|---|---|---|---|
| Intercept | 40.57 | 1.52 | 138.00 | 26.62 | <0.001 |
| Buy-recommended stock (1, 2) vs sell-recommended stock (3, 4) | 4.13 | 0.98 | 414.00 | 4.19 | <0.001 |
| Presentation vs no presentation of buy recommendation (stock 1, 2) | 1.69 | 1.39 | 414.00 | 1.22 | 0.224 |
| Presentation vs no presentation of sell recommendation (stock 3, 4) | 0.92 | 1.39 | 414.00 | 0.66 | 0.507 |
| Non-professionals vs Professionals | 9.49 | 2.24 | 138.00 | 4.24 | <0.001 |
| Non-professionals vs Finance students | −1.84 | 2.34 | 138.00 | −0.79 | 0.430 |
| Buy-recommended stock (1, 2) vs sell-recommended stock (3.4) | 2.70 | 1.45 | 414.00 | 1.87 | 0.062 |
| X Non-professionals vs Professionals | |||||
| Buy-recommended stock (1, 2) vs sell-recommended stock (3.4) | −0.82 | 1.51 | 414.00 | −0.54 | 0.578 |
| X Non-professionals vs Finance students | |||||
| Presentation vs no presentation of buy recommendation (stock 1, 2) | −0.41 | 2.04 | 414.00 | −0.20 | 0.840 |
| X Non-professionals vs Professionals | |||||
| Presentation vs no presentation of buy recommendation (stock 1, 2) | 1.62 | 2.14 | 414.00 | 0.76 | 0.449 |
| X Non-professionals vs Finance students | |||||
| Presentation vs no presentation of sell recommendation (stock 1, 2) | −1.12 | 2.04 | 414.00 | −0.55 | 0.583 |
| X Non-professionals vs Professionals | |||||
| Presentation vs no presentation of sell recommendation (stock 1, 2) | −0.11 | 2.14 | 414.00 | −0.05 | 0.958 |
| X Non-professionals vs Finance students |
Note(s): Pseudo R2Fixed-effects = 0.12, Pseudo R2Total = 0.34
Source(s): Authors' own creation/work
Mean (M) self-report ratings of importance (1–6)
| Stock-related information | Non-professionals n = 80 | Finance students n = 28 | Professionals n = 33 | Significance tests |
|---|---|---|---|---|
| M | M | M | F(2, 141) | |
| Company information | 3.35ab (1) | 3.04a (2) | 3.70b (1) | 5.31, p = 0.006 |
| Price history | 2.95a (3) | 2.93ab (3) | 2.52b (3) | 3.79, p = 0.025 |
| Analysts’ recommendations | 2.11a (4) | 1.96ab (4) | 1.48b (4) | 7.04, p = 0.001 |
| Financial key indicators | 3.21 (2) | 3.46 (1) | 3.36 (2) | 1.38, p = 0.255 |
| p/e number | 2.95 | 3.07 | 2.82 | 0.77, p = 0.466 |
| Revenues | 2.91 | 2.86 | 3.06 | 0.68, p = 0.506 |
| Dividend yields | 2.75 | 2.43 | 2.46 | 2.46, p = 0.089 |
| Net income | 3.26 | 3.14 | 3.03 | 1.50, p = 0.227 |
| EBITDA | 2.80 | 2.89 | 3.15 | 1.86, p = 0.159 |
| EPS | 2.84 | 2.75 | 3.09 | 1.43, p = 0.242 |
| Debt ratio | 2.78 | 2.64 | 2.97 | 0.89, p = 0.411 |
Note(s): Means in each row not sharing subscripts differ significantly in independent samples Bonferonni-corrected t-tests at p < 0.05
Source(s): Authors' own creation/work
Mean (M) factor scores and 95% confidence intervals (CI) based on ratings of statements about trust in financial analysts’ skill and integrity (reverse coded)
| Variable | Non-professionals n = 80 | Finance students n = 28 | Professionals n = 33 | Significance tests |
|---|---|---|---|---|
| Factor score | M [95%CI] | M [95%CI] | M [95%CI] | F(2,139) |
| Trust in skill | 0.09 [−0.12; 0.29] | −0.09 [−0.52; 0.33] | −0.13 [−0.51; 0.26] | 0.69, p = 0.505 |
| Trust in integrity | 0.17a [0.05; −0.39; ] | −0.38b [−0.01; 0.77] | −0.08a [−0.24; 0.42] | 3.48, p = 0.034 |
Note(s): Means in each row not sharing subscripts differ significantly in Bonferroni-corrected t-tests at p < 0.05
Source(s): Authors' own creation/work
Mixed effects linear regression analysis of investment with factor scores of trust in skill and integrity entered
| Independent variable | b | SE | df | t | p |
|---|---|---|---|---|---|
| Intercept | 43.05 | 1.50 | 138.00 | 28.77 | <0.001 |
| Trust in skill | 0.45 | 1.50 | 138.00 | 0.32 | 0.763 |
| Trust in integrity | 0.24 | 1.50 | 138.00 | 0.16 | 0.873 |
| Buy-recommended stock (1, 2) vs sell-recommended stock (3, 4) | 0.92 | 0.94 | 411.00 | 0.98 | 0.328 |
| X Trust in skill | |||||
| Buy-recommended stock (1, 2) vs sell-recommended stock (3, 4) | 2.25 | 1.01 | 411.00 | 2.22 | 0.027 |
| X Trust in integrity | |||||
| Presentation vs no presentation of buy recommendation (stock 1, 2) X Trust in skill | 2.34 | 1.33 | 411.00 | 1.76 | 0.080 |
| Presentation vs no presentation of buy recommendation (stock 1, 2) X Trust in integrity | 1.37 | 1.44 | 411.00 | 0.96 | 0.340 |
| Presentation vs no presentation of sell recommendation (stock 3, 4) | 1.38 | 1.33 | 411.00 | 1.04 | 0.299 |
| X Trust in skill | |||||
| Presentation vs no presentation of sell recommendation (stock 3, 4) | 0.29 | 1.44 | 411.00 | 0.20 | 0.840 |
| X Trust in integrity | |||||
| Buy-recommended stock (1, 2) vs sell-recommended stock (3, 4) | 0.44 | 1.07 | 411.00 | 0.41 | 0.680 |
| X Trust in skill X Non-professionals vs Professionals | |||||
| Buy-recommended stock (1, 2) vs sell-recommended stock (3, 4) | 0.62 | 1.15 | 411.00 | 0.54 | 0.592 |
| X Trust in integrity X Non-professionals vs Professionals | |||||
| Presentation vs no presentation of buy recommendation (stock 1, 2) X Trust in skill X Non-professionals vs Professionals | 2.18 | 1.52 | 411.00 | 1.44 | 0.151 |
| Presentation vs no presentation of buy recommendation (stock 1, 2) X Trust in integrity X Non-professionals vs Professionals | 2.66 | 1.62 | 411.00 | 1.64 | 0.103 |
| Presentation vs no presentation of sell recommendation (stock 3, 4) | 0.39 | 1.52 | 411.00 | 0.26 | 0.780 |
| X Trust in skill X Non-professionals vs Professionals | |||||
| Presentation vs no presentation of sell recommendation (stock 3, 4) | 0.67 | 1.62 | 411.00 | 0.41 | 0.681 |
| X Trust in integrity X Non-professionals vs Professionals |
Note(s): Pseudo R2Fixed-effects = 0.03, Pseudo R2Total = 0.32
Source(s): Authors' own creation/work
Pearson correlations in each group of investment with buy-recommended versus sell-recommended stocks, presentation vs no presentation of recommendation, return prediction and confidence in return prediction
| Variable | Non-professionals (n = 80) | Finance students (n = 28) | Professionals (n = 33) |
|---|---|---|---|
| Buy/sell recommended stocks | 0.30*** | 0.30** | 0.08 |
| Presentation of recommendation | 0.06 | 0.03 | 0.12 |
| Return prediction | 0.69*** (0.66***) | 0.66*** (0.63***) | 0.76*** (0.75***) |
| Confidence in return prediction | 0.28*** (0.24***) | 0.21* (0.13) | 0.26** (0.17) |
Note(s): Buy-recommended versus sell-recommended stocks are coded 1 versus −1, and presentation versus no presentation of recommendation is coded 1 versus −1. These variables are uncorrelated with each other. Their point-biserial correlations are reported with investment
Partial correlations are reported in parentheses controlling for the buy-recommended versus sell-recommended stocks. All variables are mean-centered
*p < 0.05: **p < 0.01; ***p < 0.001
Source(s): Authors' own creation/work
Financial measures 2017–2021
| 2017 | 2018 | 2019 | 2020 | 2021 | |
|---|---|---|---|---|---|
| Revenue | 15.454 | 15.544 | 15.693 | 16.471 | 17.421 |
| Result | 2.024 | 2.400 | 2.367 | 2.645 | 2.338 |
| EBITDA (earnings before interest, taxes, depreciation and amortization) | 4.182 | 4.361 | 4.281 | 4.450 | 4.442 |
| EPS (earnings per share) | 2.28 | 2.76 | 2.76 | 3.15 | 2.56 |
| Debt ratio | 52.0% | 52.2% | 48.8% | 46.1% | 47.2% |
Means (M), standard deviations (SD) and Pearson correlations between varimax rotated factor scores and ratings of trust in financial analysts (n = 141)
| Statements | M | SD | FSTS | FSTI | S9 | S10 | S2 | S1 | S6 | S8 | S7 | S5 | S4 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| S9 What trust do you have in analysts' buy recommendations? | 2.81 | 0.80 | 0.81*** | 0.29** | |||||||||
| S10 What trust do you have in analysts' sell recommendations? | 3.06 | 0.86 | 0.80*** | 0.08 | 0.69*** | ||||||||
| S2 Stocks with positive recommendations from several analysts will probably show a good return | 2.64 | 0.98 | 0.73*** | 0.01 | 0.48*** | 0.34*** | |||||||
| S1 Analysts' return predictions for large cap companies are usually accurate | 2.91 | 0.82 | 0.65*** | −0.06 | 0.33*** | 0.36*** | 0.38*** | ||||||
| S6 Analysts are on purpose overly optimistic in their analysis to boost stock returns | 2.78 | 1.08 | −0.13 | −0.79*** | −0.32*** | −0.20* | −0.03 | −0.18* | |||||
| S8 Analysts predicts purposely low short-term profits for companies to give the companies' management better opportunities to overperform expectations | 2.39 | 1.11 | −0.02 | −0.73*** | −0.20* | −0.18* | −0.04 | −0.04 | 0.51*** | ||||
| S7 Analysts' recommendations concerning a company are influenced if they also have the company as client | 3.57 | 1.16 | −0.01 | −0.70*** | −0.21* | −0.02 | −0.03 | −0.14 | 0.43*** | 0.41*** | |||
| S5 Analysts tend to refrain from giving recommendations that strongly deviates from consensus | 3.39 | 1.03 | −0.27** | −0.64*** | −0.40*** | −0.17* | −0.26 | −0.10 | 0.41*** | 0.30*** | 0.35*** | ||
| S4 Analysts rather give buy than sell recommendations | 3.57 | 1.17 | −0.01 | −0.46*** | −0.16 | 0.02 | −0.11 | 0.00 | 0.24** | 0.15 | 0.16 | 0.36*** | |
| S3 Analysts are not able to separate temporary from stable changes in companies’ profits | 2.67 | 0.91 | −0.10 | −0.38*** | −0.11 | −0.07 | −0.07 | −0.16 | 0.30*** | 0.22** | 0.12 | 0.14 | 0.03 |
Note(s): FS refers to standardized factor scores, subscript TS trust in skill and subscript TI trust in integrity. The numbering of statements corresponds to presentation order in the online questionnaire
*p < 0.05; **p < 0.01; ***p < 0.001
Source(s): Authors' own creation/work)
Notes
Studies in several other expertise domains (e.g. Camerer and Johnson, 1991) have shown that recent graduates may be more accurate in making judgments and decisions than those with the same education who also have experience.
Analyses after excluding the corrected values did not change the results in any systematic ways.
All reported mixed effects linear regression analyses were conducted using R version 4.2.1 and the lme4 package version 1.1–30. The lmerTest package version 3.1.3 (Kuznetsova et al., 2017) was used to approximate degrees of freedom for the t-statistics with the Satterthwaite technique. Pseudo R-squares were calculated with the MuMIn package version 1.47.1 according to the method proposed by Nakagawa et al. (2017).
Separate repeated-measures ANOVAs showed that the differences in rated importance and their interactions with group differed significantly for the general stock-related information (F(3,414) = 106.38, p < 0.001; F(6,414) = 6.24, p < 0.001), as well as for the specific financial key indicators (F(6,828) = 2.06, p < 0.001; F(12,828) = 2.06, p = 0.018). The overall differences between groups did not reach significance (F < 1).
Appendix 1 Sample display of presentation of the stock-related information
Information is shown by clicking on “Show me the information”.
Company information
[Show me the information]
Colgate-Palmolive is a consumer company. The company manufactures and distributes a wide range of consumer products, the most famous of which include toothpaste, shampoo, detergent, and other cleaning products. The company operates world-wide, and the best-known brands include Colgate, Palmolive and Softsoap. The company was founded in 1806 and has its headquarters in New York.
Financial measures
[Show me the information].
Current p/e-number: 35.6.
Dividend (previous years dividend divided by current stock price): 1.92%
[Show me the information]
Green (lower) line shows the stock price and yellow (higher) line the Dow Jones index.
Financial analysts’ recommendations
[Show me the information].
The aggregated recommendation from financial analysts for this stock is: Buy.
Share price data
The current price of the stock: $75.20.
Share price for one day: $75.64.
50-days average: $79.54.
200-days average: $74.02.
Updated at [date and time].
References
Andersson, M., Gärling, T. and Hedesström, M. (2014), “A social-psychological perspective on herding in stock markets”, Journal of Behavioral Finance, Vol. 15 No. 3, pp. 226-234, doi: 10.1080/15427560.2014.941062.
Asquith, P., Mikhail, M.B. and Au, A.S. (2005), “Information content of equity analyst reports”, Journal of Financial Economics, Vol. 75 No. 2, pp. 245-282, doi: 10.1016/j.jfineco.2004.01.002.
Balboa, M., Gómez‐Sala, J.C. and López‐Espinosa, G. (2009), “The value of adjusting the bias in recommendations: international evidence”, European Financial Management, Vol. 15 No. 1, pp. 208-230, doi: 10.1111/j.1468-036x.2007.00421.x.
Barber, B., Lehavy, R., McNichols, M. and Trueman, B. (2001), “Can investors profit from the prophets? Security analysts' recommendations and stock returns”, The Journal of Finance, Vol. 56 No. 2, pp. 531-563, doi: 10.1111/0022-1082.00336.
Beyer, A., Cohen, D.A., Lys, T.Z. and Walther, B.R. (2010), “The financial reporting environment: review of the recent literature”, Journal of Accounting and Economics, Vol. 50 Nos 2-3, pp. 296-343, doi: 10.1016/j.jacceco.2010.10.003.
Bodnaruk, A. and Simonov, A. (2015), “Do financial experts make better investment decisions?”, Journal of Financial Intermediation, Vol. 24 No. 4, pp. 514-536, doi: 10.1016/j.jfi.2014.09.001.
Boulland, R., Ornthanalai, C. and Womack, K.L. (2022), “Speed and expertise in stock picking: older, slower, and wiser?”, Journal of Financial and Quantitative Analysis, Vol. 58 No. 4, pp. 1612-1644, doi: 10.1017/s0022109022000199.
Bouteska, A. and Mili, M. (2023), “The role of investor sentiment and valuation uncertainty in the changes around analyst recommendations: evidence from U.S. firms”, Journal of Behavioral Finance, Vol. 24 No. 1, pp. 73-96, (In press), doi: 10.1080/15427560.2021.1948853.
Brown, L.D., Call, A.C., Clement, M.B. and Sharp, N.Y. (2015), “Inside the ‘black box’ of sell‐side financial analysts”, Journal of Accounting Research, Vol. 53 No. 1, pp. 1-47, doi: 10.1111/1475-679x.12067.
Brown, L.D., Call, A.C., Clement, M.B. and Sharp, N.Y. (2016), “The activities of buy-side analysts and the determinants of their stock recommendations”, Journal of Accounting and Economics, Vol. 62 No. 1, pp. 139-156, doi: 10.1016/j.jacceco.2016.06.002.
Brown, L.D., Call, A.C., Clement, M.B. and Sharp, N.Y. (2019), “Managing the narrative: investor relations officers and corporate disclosure”, Journal of Accounting and Economics, Vol. 67 No. 1, pp. 58-79, doi: 10.1016/j.jacceco.2018.08.014.
Camerer, C. and Johnson, E.J. (1991), “The process-performance paradox in expert judgment: how can experts know so much and predict so badly?”, in Ericsson, K.A. and Smith, J. (Eds), Toward a General Theory of Expertise: Prospects and Limits, Cambridge University Press, New York, pp. 195-217.
Clement, M.B., Rees, L. and Swanson, E.P. (2003), “The influence of culture and corporate governance on the characteristics that distinguish superior analysts”, Journal of Accounting, Auditing and Finance, Vol. 18 No. 4, pp. 593-618, doi: 10.1177/0148558x0301800410.
De Bondt, W.F. and Thaler, R. (1985), “Does the stock market overreact?”, The Journal of Finance, Vol. 40 No. 3, pp. 793-805, doi: 10.2307/2327804.
Ennew, C., Kharouf, H. and Sekhon, H. (2011), “Trust in UK financial services: a longitudinal analysis”, Journal of Financial Services Marketing, Vol. 16 No. 1, pp. 65-75, doi: 10.1057/fsm.2011.8.
Fama, E. (1970), “Efficient capital markets: a review of theory and empirical work”, The Journal of Finance, Vol. 25 No. 2, pp. 383-417, doi: 10.2307/2325486.
Gärling, T. (2011), “Heuristics and biases making people rich or poor: a psychological account of stock market anomalies”, in Brun, W., Keren, G., Kirkeboen, G. and Montgomery, H. (Eds), Perspectives on Thinking, Judging, and Decision Making, Universitetsforlaget, Oslo, pp. 146-156.
Gärling, T., Kirchler, E., Lewis, A. and van Raaij, F. (2009), “Psychology, financial decision making, and financial crises”, Psychological Science in the Public Interest, Vol. 10 No. 1, pp. 1-47, doi: 10.1177/1529100610378437.
Greenwood, R. and Nagel, S. (2009), “Inexperienced investors and bubbles”, Journal of Financial Economics, Vol. 93 No. 2, pp. 239-258, doi: 10.1016/j.jfineco.2008.08.004.
Huber, J., Kirchler, M. and Stöckl, T. (2010), “The hot hand belief and the gambler's fallacy in investment decisions under risk”, Theory and Decision, Vol. 68 No. 4, pp. 445-462, doi: 10.1007/s11238-008-9106-2.
Jansson, M. (2019), “The contested significance of financial expertise in predicting short-and long-term risk and return on the stock market”, The Journal of Prediction Markets, Vol. 13 No. 2, pp. 15-31, doi: 10.5750/jpm.v13i2.1762.
Jegadeesh, N., Kim, J., Krische, S. and Lee, C. (2005), “Analyzing the analysts: when do recommendation add value?”, The Journal of Finance, Vol. 59 No. 3, pp. 1083-1124, doi: 10.1111/j.1540-6261.2004.00657.x.
Keasler, T.R. and McNeil, C.R. (2010), “Mad money stock recommendations: market reaction and performance”, Journal of Economics and Finance, Vol. 34 No. 1, pp. 1-22, doi: 10.1007/s12197-008-9033-7.
Kuznetsova, A., Brockhoff, P.B. and Christensen, R.H.B. (2017), “lmerTest package: tests in linear mixed effects models”, Journal of Statistical Software, Vol. 82 No. 13, pp. 1-26, doi: 10.18637/jss.v082.i13.
Li, W., Rhee, G. and Wang, S.S. (2017), “Differences in herding: individual vs institutional investors”, Pacific-Basin Finance Journal, Vol. 45, pp. 174-185, doi: 10.1016/j.pacfin.2016.11.005.
Lidén, E. (2004), “Essays on the information and conflict of interest in stock recommendations”, Unpublished paper, University of Gothenburg, Göteborg.
Lin, W.C., Chang, S.C., Chen, S.S. and Liao, T.L. (2013), “The over-optimism of financial analysts and the long-run performance of firms following private placements of equity”, Finance Research Letters, Vol. 10 No. 2, pp. 82-92, doi: 10.1016/j.frl.2012.12.001.
Malkiel, B.G. (2005), “Reflections on the efficient market hypothesis: 30 years later”, Financial Review, Vol. 40 No. 1, pp. 1-9, doi: 10.1111/j.0732-8516.2005.00090.x.
Malmendier, U. and Shanthikumar, D. (2007), “Are small investors naive about incentives?”, Journal of Financial Economics, Vol. 85 No. 2, pp. 457-489, doi: 10.1016/j.jfineco.2007.02.001.
Malmendier, U. and Shanthikumar, D. (2014), “Do security analysts speak in two tongues?”, Review of Financial Studies, Vol. 27 No. 5, pp. 1287-1322, doi: 10.1093/rfs/hhu009.
Menkhoff, L., Schmeling, M. and Schmidt, U. (2013), “Overconfidence, experience, and professionalism: an experimental study”, Journal of Economic Behavior and Organization, Vol. 86, pp. 92-101, doi: 10.1016/j.jebo.2012.12.022.
Michaely, R. and Womack, R. (1999), “Conflict of interest and the credibility of underwriter analyst recommendations”, Review of Financial Studies, Vol. 12 No. 4, pp. 653-686, doi: 10.1093/rfs/12.4.653.
Mikhail, M., Walther, B. and Willis, R. (2007), “When security analysts talk, who listens?”, The Accounting Review, Vol. 82 No. 5, pp. 1227-1253, doi: 10.2308/accr.2007.82.5.1227.
Millo, Y., Spence, C. and Valentine, J. (2023), “The field of investment advice: the social forces that govern equity analysts”, The Accounting Review, Vol. 98 No. 7, pp. 457-477, doi: 10.2139/ssrn.4440019.
Mola, S. and Guidolin, M. (2009), “Affiliated mutual funds and analyst optimism”, Journal of Financial Economics, Vol. 93 No. 1, pp. 108-137, doi: 10.1016/j.jfineco.2008.06.006.
Nakagawa, S., Johnson, P.C. and Schielzeth, H. (2017), “The coefficient of determination R2 and intra-class correlation coefficient from generalized linear mixed-effects models revisited and expanded”, Journal of the Royal Society Interface, Vol. 14 No. 134, 20170213, doi: 10.1098/rsif.2017.0213.
Nofsinger, J.R. (2005), “Social mood and financial economics”, Journal of Behavioral Finance, Vol. 6 No. 3, pp. 144-160, doi: 10.1207/s15427579jpfm0603_4.
O´Brien, P., McNichols, M. and Hsiou-Wei, L. (2005), “Analyst impartiality and investment banking relationship”, Journal of Accounting Research, Vol. 43 No. 4, pp. 623-650, doi: 10.1111/j.1475-679x.2005.00184.x.
Peterson, D., Carlander, A., Gamble, A., Gärling, T. and Holmen, M. (2015), “Lay people beliefs in professional and naive stock investors' proneness to judgmental biases”, Journal of Behavioral and Experimental Finance, Vol. 5, pp. 27-34, doi: 10.1016/j.jbef.2015.02.002.
Rahman, J.M., Zhang, J. and Dong, S. (2019), “Factors affecting the accuracy of analysts' forecasts: a review of the literature”, Academy of Accounting and Financial Studies Journal, available at: https://ssrn.com/abstract=3610334
Ramnath, S., Rock, S. and Shane, P. (2008), “A review of research related to financial analysts' forecasts and stock recommendations”, available at: http://ssrn.com/abstract=848248
Sant, R. and Zamanb, M. (1996), “Market reaction to business week ‘inside wall street’ column: a self-fulfilling prophecy”, Journal of Banking and Finance, Vol. 20 No. 4, pp. 617-643, doi: 10.1016/0378-4266(95)00025-9.
Shefrin, H. (2007), “How the disposition effect and momentum impact investment professionals”, Journal of Investment Consulting, Vol. 8 No. 2, pp. 68-79.
Shefrin, H. and Statman, M. (1985), “The disposition to sell winners too early and ride losers too long”, The Journal of Finance, Vol. 40 No. 3, pp. 777-790, doi: 10.2307/2327802.
Sonsino, D. and Regev, E. (2013), “Informational overconfidence in return prediction – more properties”, Journal of Economic Psychology, Vol. 39, pp. 72-84, doi: 10.1016/j.joep.2013.06.006.
Sonsino, D., Michaelsen, P., Gärling, T. and Jansson, M. (2023), “The sober outlook of competent investors – a framed-field experiment in times of volatile markets”, paper presented at the 15:th Nordic Conference on Behavioral and Experimental Economics, Lund, Lund University.
Spence, C. and Carter, C. (2014), “An exploration of professional habitus in Big 4 accounting firms”, Work Employment and Society, Vol. 28 No. 6, pp. 946-962, doi: 10.1177/0950017013510762.
Spence, C., Aleksanyan, M., Millo, Y., Imam, S. and Abhayawansa, S. (2019), “Earning the ‘write to speak’: sell‐side analysts and their struggle to be heard”, Contemporary Accounting Research, Vol. 36 No. 4, pp. 2635-2662, doi: 10.1111/1911-3846.12496.
Spyrou, S. (2013), “Herding in financial markets: a review of the literature”, Review of Behavioral Finance, Vol. 5 No. 2, pp. 371-398, doi: 10.1108/rbf-02-2013-0009.
Törngren, G. and Montgomery, H. (2004), “Worse than chance? Performance and confidence among professional and laypeople in the stock market”, Journal of Behavioral Finance, Vol. 5 No. 3, pp. 148-153, doi: 10.1207/s15427579jpfm0503_3.
Twedt, B. and Rees, L. (2012), “Reading between the lines: an empirical examination of qualitative attributes of financial analysts' reports”, Journal of Accounting and Public Policy, Vol. 31 No. 1, pp. 1-21, doi: 10.1016/j.jaccpubpol.2011.10.010.
Womack, K.L. (1996), “Do brokerage analysts' recommendations have investment value?”, The Journal of Finance, Vol. 51 No. 1, pp. 137-167, doi: 10.2307/2329305.
Yin, Y. (2018), “Fund portfolio holdings and affiliated analyst's objectivity: do equity and employment relationship matter?”, Journal of Behavioral Finance, Vol. 19 No. 1, pp. 16-29, doi: 10.1080/15427560.2017.1366493.
Acknowledgements
This research was funded by Handelsbankens research foundation, Sweden.