Stated Preferences for Firm’s Characteristics and Asset Allocation Decisions

Introduction

There are several theoretical models on asset allocation. Tobin’s (1958) mutual fund separation model suggests asset allocation to different combination of risky and risk-free assets based on individual’s risk preferences. Sharpe (1964) and Lintner’s (1965) capital asset pricing model suggests asset allocation to all available assets in the market irrespective of risk associated with the assets. Markowitz’s (1959) men-variance model predicts that investors should care about the mean and variance of portfolio returns to achieve optimal asset allocation.

As opposed to optimal allocation, a particular strand of literature suggests that certain heuristics came into play when it comes to asset allocation, for example, 1/n heuristic, home bias, optimism, risk attitude, presentation format, future value projections, financial advice and, recently, peer effects and personal characteristics (Barasinska, Schäfer, & Stephan, 2012; Benartzi & Thaler, 1995, 2001; Ehm, Kaufmann, & Weber, 2014; Jacobsen, Lee, Marquering, & Zhang, 2014; Strong & Xu, 2003; Sundali & Guerrero, 2009; Tversky & Kahneman, 1974; Zhang, 2014; Zhang, Jacobsen, & Marshall, 2016). These studies conclude that such factors lead to hold under-diversified portfolio comprising excessive allocation to either risky or safe assets. On the competing argument between fundamental and behavioural, the empirical research on fundamental ground has received less attention as theoretical studies mostly assume and rarely examine the impact that fundamental factors may have on asset allocation.

Regarding fundamental factors, preferences for a firm’s characteristics could affect asset allocation decisions in several ways. First of all, investors’ stated preferences for firm’s characteristics can be regarded as relying on fundamental information for investments. Previous studies highlight that investors using fundamental analysis hold different combination of financial assets than others (Lease, Lewellen, & Schlarbaum, 1974). Therefore, it assumes that firm’s characteristics as a manifestation of fundamental analysis would affect asset allocation. Second, Elkinawy (2005) illustrated that a firm’s attributes affect an individual’s portfolio choice. Investors prefer to invest in firms which have greater accounting transparency (Aggarwal, Klapper, & Wysocki, 2005). Firm-level disclosure improves investors’ knowledge about a firm’s characteristics and thus likely to affect asset allocation. Third, a firm’s characteristics represent deeper insight into the type of stocks that investors like to hold in portfolio. Investors like to give higher priority to firms with the characteristics of higher profitability, efficient resource management, more liquidity, large market capitalization, low market risk and high leverage (Narang & Kaur, 2014). Duxbury, Hudson, Keasey, Yang, and Yao (2013) asserted that stock’s characteristics shape investors’ portfolio composition. They contend that stock’s characteristics do matter in portfolio holding. Fourth, individuals elicit risk–return expectations from a firm’s specific characteristics. They expect higher returns for firms that pay dividend, but higher risks for firms that do not pay dividend have low cash flows and are smaller firms (Kempf, Merkle, & Niessen-Ruenzi, 2014). Since a firm’s characteristics are perceived as a representation of risk and returns, based on this assessment, individuals can decide which assets to be included in portfolio. Given the aforementioned discussions, this study aims to examine the link between stated preferences for firm’s characteristics and asset allocation.

Malaysia is chosen as a context of this study due to its distinguished features. Malaysian households appear to less diversify their financial assets. They are excessively exposed to a single asset class, that is, safe assets (Bank Negara Malaysia, 2010, 2011). Hence, they fail to diversify portfolio with the combination of risky and safe assets. As a result, Malaysians forgo the opportunity to earn higher returns. In addition, regulators have identified excessive allocation to safe assets as one of the key challenges in the Malaysian stock market (Securities Commission Malaysia, 2011). Overconcentration to safe asset is considered dangerous anomalies in financial market, resulting in lower equity premium (Goetzmann & Kumar, 2008; Sunden & Surette, 1998). The aforementioned scenario grabs special attention to study asset allocation in Malaysia. The study finds that investors’ preferences for firm’s characteristics affect their ownership of financial assets, portfolio holdings, the number of financial asset holding and hypothetical asset allocation.

This study makes several contributions pertaining to asset allocation. First, the study focuses on stated preferences for firm’s characteristics as potential determinants of asset allocation decisions. To the best of author’s knowledge, there are few studies that empirically look at the potential role of preferences for firm’s characteristics in asset allocation. Second, the study explores asset allocation of retail investors in Malaysia. There are not many studies in the literature that have explored the factors influencing asset allocation of Malaysian investors, in spite of placing greater emphasis on structural imbalance between safe and risky assets. Third, the study comprehensively measures investors’ asset allocation decisions such as a particular financial asset holding, types of portfolio holdings, the number of asset holding and hypothetical asset allocation.

The structure of this study is as follows: the first section is an introduction of the article; the second section reviews the relevant literature; the third section stated the objective of this study; the fourth section discusses the rational of this study; the fifth section describes methodology, the sixth and seventh sections describe analysis and discussion, respectively and finally the conclusion is given in the eighth section following implications given in the ninth section.

Review of Literature

This study argues that preferences for firm’s characteristics affect investors’ asset allocation decisions. One of the arguments is that firm’s specific characteristics affect investors’ risk and return expectations, which in turn influence their asset allocation decisions. Along this line, Kempf, Merkle, and Niessen-Ruenzi (2014) reported that depending on a firm’s specific characteristics, investors form risk and return expectations. A firm’s profitability characteristics, such as dividend payments, profit margin, return on equity, earnings and return on assets (ROA), can be associated with investors’ risk and return expectations. Based on these expectations, a firm’s profitability characteristics can be an important determinant of a particular type of financial asset holding.

An individual’s decisions may be influenced not only by profitability characteristics but also by a firm’s value characteristics. Doran and Wright (2010) found that in a survey of finance professors in the USA, the use of a firm’s value-related characteristics, such as market capitalization, book-to-market ratio and price-to-earnings, has a stronger influence on an individual’s buying and selling decisions. Kempf, Merkle, and Niessen-Ruenzi (2014) argued that investors expect relatively higher returns from a firm with high price-to-earnings ratio, lower book values, debt ratio and larger firms. Stock selection, based on a firm’s price-to-equity ratio, price-to-book value ratio, debt-to-equity ratio, is found to be related to portfolio construction (Roy, 2016). Also, a firm’s value characteristics, like high price-to-earnings ratio and price-to-book ratio, are perceived as glamour firms. The glamour firms are generally attracted by sentiment-prone investors (Baker & Wurgler, 2006).

Besides value characteristics, a firm’s liquidity and trading volume characteristics can be regarded as potential determinants for asset allocation. The works of Ng and Wu (2006) and Khan, Tan, and Chong (2016) have shown that not only revealed preferences but also stated preferences for a firm’s liquidity and trading volume lead to a high degree of trading activity. Elkinawy (2005) found that a firm’s liquidity characteristic becomes more desirable for fund manager’s stock ownership decisions in financial crisis period than in tranquil period. Similarly, the demand for liquidity and firm’s greater trading volume would influence investors to own a particular financial asset and hold portfolio accordingly to meet their demand.

A firm’s management and product-related characteristics, and long-term value apart from its financial ratios, can lead to asset allocation decision. Kempf, Merkle, and Niessen-Ruenzi (2014) showed that, in a laboratory experiment, individuals display affective reaction to a firm’s marketing expenditure and its brand’s strength. Individuals’ confidence increases with renowned brand, higher marketing expenditures, higher media coverage and lower information asymmetry. Referring to Kempf, Merkle, and Niessen-Ruenzi’s (2014) study, it can be argued that a firm’s product and management-related characteristics carry significant information about firms and, as a result, this resourceful information can increase investors’ confidence. Confidence about a firm’s management and product-related characteristics could influence their asset allocation such as which assets to own, portfolio to hold and the number of financial asset holding.

As investors typically concern about risk–return concept, they can regard a firm’s risky characteristics in their asset allocation. Firm’s risky characteristics refer to leverage, beta and idiosyncratic volatility. In assessing these characteristics, investors are likely to expect higher return and lower risk (Kempf, Merkle, and Niessen-Ruenzi, 2014). Investors’ preferences for a firm’s risky characteristics are driven by past positive outcome, which in turn shape their portfolio composition (Duxbury, Hudson, Keasey, Yang, & Yao, 2013). A firm’s risky characteristics itself and its underlying information can influence asset allocation (Aggarwal, Klapper, & Wysocki, 2005).

Given the aforementioned discussion, the study expects that investors’ stated preferences for a firm’s profitability, value, liquidity and trading volume, management and product attributes, and risky characteristics affect their asset allocation decisions. This study is related to the study of Khan, Tan, and Chong (2016) where they investigate investors’ stated preferences not only for firm’s characteristics but also for trading.

Objective

This study aims to investigate the relationship between stated preferences for firm’s characteristics and asset allocation decisions in Malaysia. Stated preferences for firm’s characteristics are represented by investors’ preferences for a firm’s profitability, value, liquidity and trading volume, management and product attributes, and risky characteristics. Asset allocation decisions indicate a particular financial asset holding, types of portfolio holdings, the number of asset holding and hypothetical asset allocation.

Rationale of the Study

To understand Malaysian investors’ asset allocation choices, Bursa Malaysia conducted a survey in 2009 which revealed that investors tend to hold on, priority basis, saving deposits, insurance, real estate assets and unit trusts more than shares (StarBiz, 2010; The Malaysian Insider, 2010). Although there was a tremendous growth in financial assets of Malaysians, they mainly allocate their assets to bank’s saving assets, representing overconcentration on a single asset class (Bank Negara Malaysia, 2011; Economic Report, 2011/2012). The trend has been observed for a period of time. Households’ savings in banks and other financial institutions have increased from 31 per cent in 2010 to 42 per cent in 2013, whereas their allocation to risky equity asset has declined from 32 per cent to 16 per cent between 2007 and 2011 (Bank Negara Malaysia, 2011, 2013). This suggests that households mainly composite their portfolios by excessively focusing on safe assets while avoiding risky assets. Such undiversified portfolio composition has been highlighted in the Capital Market Masterplan 2 published by Securities Commission (SC) Malaysia. They highlighted it as a structural gap between safe and risky assets. The SC emphasized more on reducing this structural gap between safe and risky assets as it is necessary for the acceleration of stock market and the economic growth. This sets the motivation to investigate the determinants of asset allocation.

Methodology

Operationalization of Variables

The independent variables in the study are investors’ stated preferences for various firm-specific characteristics. Profitability characteristics’ preferences were measured by multi-item psychometric scales. Participants were asked to express their degree of preferences for a firm’s profitability characteristics (profit margin, ROE, dividend, earnings and ROA) using a five-point Likert scale. The same explanation holds for value characteristics’ preferences (BM ratio, market capitalization, PE ratio and stock price), liquidity and trading volume preferences (liquidity and trading volume), quality of management and product preferences (quality of management, quality of product and services, long-term value) and risky characteristics’ preferences (leverage, beta and idiosyncratic volatility). The significant mean value of each firm’s characteristic indicates that investors exhibit preferences for firm’s characteristics. Finally, the independent variables were constructed by aggregating respective items of each construct.

Investors’ tendency to own a particular type of financial asset and the type of portfolio holdings were measured by asking investors about the combinations of financial assets they like to include in their portfolio. The riskiness of asset classes is classified as low risk (i.e., saving deposits), moderate risk (i.e., Government/Islamic bonds, insurance, real estate/property) and high risk (i.e., unit trust/mutual fund, stocks/equities) (Barasinska et al., 2012). Of seven types of portfolios, portfolios 1–3 are defined as undiversified portfolios which consist of only one type of assets such as low-risk assets, or moderately risky assets or high-risk assets. Portfolios 4–6 are defined as quite diversified portfolios which consist of maximum two types of assets. Portfolio 7 is defined as a fully diversified portfolio which is a combination of three types of assets (low-, moderate and high-risk assets). The number of financial asset holding is assessed by asking investors to indicate how many different types of financial assets they are currently holding in their portfolio. Asset allocation decision is also measured as hypothetical allocation to safe, moderately risky and risky assets (Guiso & Paiella, 2008; Nosić & Weber, 2010). The main highlights of the survey questionnaire are presented in Table 1.

OPEN IN VIEWER

Table 1.

Main Highlights of the Survey Questionnaire

Panel A: Stated Preferences for Firms’ Characteristics
Following are certain characteristics about companies that are traded on the stock market. As an investor which characteristics do you prefer in making investment decisions on a 5-point Likert scale (1: not preferred, 2: least preferred, 3: neutral, 4: generally preferred, 5: highly preferred).
Profitability characteristics’ preferences
1. Profit margin (profitability of a firm)
2. Return on equity (ROE)
3. Dividend yield (payment in the form of dividends)
4. Earnings/sales growth
5. Return on assets (ROA)
Value characteristics’ preferences
6. Book-to-market ratio (value of a firm by comparing book value to its market value)
7. Market capitalization (outstanding shares multiplied by market value per share)
8. Price-to-earnings ratio (how much earnings of the firm reflects in the stock price?)
9. Price of a stock (affordable or expensive stock)
Liquidity and trading volume preferences
10. Liquidity (how fast you can sell?)
11. Trading volume (volume of trades)
Quality of management and product preferences
12. Quality of management of a firm
13. Quality of products or services of a firm
14. Long-term investment value of a firm
Risky characteristics’ preferences
15. Leverage (debt-to-equity ratio)
16. Stock’s beta (systematic risk)
17. Idiosyncratic volatility (stock-specific volatility)
Panel B: Asset Allocation
1. What combinations of financial assets would you like to include in your portfolio? May select more than one.
□ Saving deposits	□ Insurance	□ Unit trusts/Mutual fund
□ Government or Islamic bonds	□ Real estate/Property	□ Stocks/Equities
2. How many different types of financial assets you are currently holding in your portfolio?
□ 0	□ 2–6	□ More than 10
□ 1	□ 7–10
3. Assume you have RM100 to invest into different types of assets, how would you allocate this money to the following assets? Please allocate 100 per cent
□ ___% in safe assets	□ ___% in moderately risky assets	□ ___% in risky assets
□ ___% in other assets

Source: Khan et al. (2016) and Barasinska et al. (2012).

Empirical Model

The functional form of the model that estimates the ownership of a particular financial asset holding is as follows:

U i = f ( Y i , Z i )

(1)

where U _i represents the ownership of a financial asset by investor i (i ϵ 1, 2, 3, …., 454), Y _i is a vector of the explanatory variables representing preferences for firm’s characteristics and Z _i is a vector of the control variables. On the left side of the distribution, the dependent variable, the ownership of a particular financial asset, is grouped into six dummy variables, that is, if the investors hold saving deposits, it takes the value 1 and 0 otherwise; if the investors hold government/Islamic bonds, it takes the value 1 and 0 otherwise; if the investors hold insurance, it takes the value 1 and 0 otherwise; if the investors hold real estate/property, it takes the value 1 and 0 otherwise; if the investors hold unit trusts/mutual funds, it takes the value 1 and 0 otherwise; if the investors hold stocks/equities, it takes the value 1 and 0 otherwise. In the model, the regressors are profitability characteristics, value characteristics, liquidity and trading volume, quality of management and product characteristics, and risky characteristics’ preferences. The control variables include an investor’s gender, age, marital status, income, ethnicity, education, investment experience and occupation. Given the nature of dependent variables, the binary logistic regression models were estimated for model (1).

The functional form of the model that estimates the types of portfolio holdings is as follows:

V i = f ( Y i , Z i )

(2)

where V _i represents the types of portfolio holdings by investor i (i ϵ 1, 2, 3, …., 454), Y _i is a vector of the explanatory variables representing an investor’s preferences for firm’s characteristics and Z _i is a vector of the control variables. The dependent variable, on the left-hand side of the equation, is denoted by seven dummy variables. ‘Portfolio 1’ assigns the value 1, if investors tend to hold only low-risk assets and 0 otherwise; ‘portfolio 2’ assigns the value 1, if investors tend to hold only moderately risky assets and 0 otherwise; ‘portfolio 3’ assigns the value 1, if investors tend to hold only high-risk assets and 0 otherwise. ‘Portfolio 1’, ‘portfolio 2’ and ‘portfolio 3’ are categorized as undiversified portfolios. ‘Portfolio 4’ denotes the value 1, if investors tend to hold ‘low and moderately risky assets’ and 0 otherwise; ‘portfolio 5’ denotes the value 1, if investors tend to hold ‘low- and high-risk assets’ and 0 otherwise; ‘portfolio 6’ denotes the value 1, if investors tend to hold ‘moderately risky and high-risk assets’. ‘Portfolio 4’, ‘portfolio 5’ and ‘portfolio 6’ are categorized as quite diversified portfolios. ‘Portfolio 7’ is categorized as fully diversified portfolio and assigned the value 1, if investors tend to hold a combination of low-risk, moderate-risk and high-risk assets and zero otherwise. Model (2) was estimated by binary logistic regressions.

To estimate the number of financial asset holding, the following functional form is estimated:

W i = f ( Y i , Z i )

(3)

In functional form (3), W _i represents the number of financial asset holdings by investor i (i ϵ 1, 2, 3, …, 454), Y _i is a vector of the explanatory variables representing an investor’s preferences for firm’s characteristics and Z _i is a vector of the control variables. The dependent variable in the functional form (3) takes the ordered-ranked values, where low value indicates few financial asset holdings and high value indicates more financial asset holdings. The independent and control variables remain as it is. The aforementioned model was estimated using an ordered logistic model.

The functional form of the model that estimates hypothetical asset allocation decisions is as follows:

X i = f ( Y i , Z i )

(4)

where X _i represents hypothetical asset allocation by investor i (i ϵ 1, 2, 3, …., 454), Y _i is a vector of the explanatory variables indicating an investor’s preferences for firm’s characteristics and Z _i is a vector of the control variables. Hypothetical asset allocation is denoted by three variables where an investor’s allocation of a certain amount to safe assets indicates safe asset allocation; the amount allocated to moderately risky assets indicates moderately risky asset allocation and the amount allocated to risky assets indicates risky asset allocation. The independent and control variables remain as before. This model was estimated using ordered logistic regression, given the econometric nature of the dependent variable.

Analysis

To briefly describe financial asset holding of respondents, more than 80 per cent of investors in the sample hold equities. More than half of the investors hold saving deposits and real estate/property in the portfolio. A total of 34.1 per cent of investors own unit trust or mutual fund, while 29.3 per cent of investors own insurance. Government or Islamic bonds are held by only 16.3 per cent of investors. Of the six different types of financial assets, more than a quarter of investors hold one type of asset, whereas only 13.2 per cent of investors hold four or more financial assets. Constructing a portfolio based on the riskiness of assets, 26 per cent of investors hold fully diversified portfolios comprising low-, moderate and high-risk assets. A total of 36.1 per cent of investors hold quite diversified portfolios comprising either the set of low and moderately risky assets or the set of low- and high-risk assets or the set of moderate and high-risk assets. About 37.8 per cent of investors hold undiversified portfolio which consists of either low-, or moderate- or high-risk assets. The subsequent sections estimate the role of preferences for various firms’ characteristics in asset allocation.

In the multivariate analyses as presented in Table 2, the results show that preferences for a firm’s value characteristics reduce the probability of holding saving deposits, while preferences for a firm’s quality of management and product, and risky characteristics increase the probability of owning saving deposits. For the likelihood of holding government/Islamic bonds, the firm’s value characteristics turn positive along with the quality of management and product, and liquidity and trading volume preferences. The inclination to hold real estate/property is significantly increased by preferences for a firm’s profitability characteristics but decreased by a firm’s risky characteristics. Surprisingly, none of the firm-specific characteristics is significant to explain the inclusion of unit trust/mutual fund and stock/equity in portfolio.

OPEN IN VIEWER

Table 2.

Preferences for Firm’s Characteristics and Financial Assets Owned by Investors

	Dependent Variable
	Saving Deposit	Govt./Islamic Bond	Insurance	Real Estate/Property	Unit Trust/Mutual Fund	Stock/Equity
Profitability characteristics’ preferences	−0.195(0.369)	−0.186(0.534)	−0.203(0.388)	0.457**(0.037)	−0.253(0.264)	0.059(0.844)
Value characteristics’ preferences	−0.430*(0.060)	0.560*(0.081)	−0.307(0.213)	0.248(0.275)	−0.035(0.882)	0.457(0.135)
Liquidity and trading volume preferences	−0.077(0.610)	0.494**(0.020)	0.158(0.341)	0.235(0.122)	0.014(0.929)	0.104(0.617)
Quality of management and product preferences	0.340**(0.028)	0.545**(0.011)	−0.004(0.980)	−0.126(0.415)	0.001(0.994)	0.237(0.262)
Risky characteristics’ preferences	0.444***(0.009)	−0.314(0.161)	−0.126(0.492)	−0.295*(0.082)	−0.074(0.673)	−0.169(0.481)
Male	0.085(0.684)	−0.482*(0.083)	−0.098(0.665)	0.294(0.163)	−0.053(0.807)	0.306(0.285)
Middle-aged	0.694**(0.026)	0.168(0.696)	0.902**(0.020)	0.270(0.387)	0.411(0.215)	−0.041(0.922)
Older	0.045(0.924)	0.025(0.969)	0.880(0.103)	0.167(0.725)	0.114(0.818)	0.240(0.734)
Married	0.053(0.795)	−0.393(0.153)	−0.133(0.549)	−0.559***(0.007)	−0.055(0.796)	−0.087(0.760)
High income	−0.156(0.507)	0.140(0.664)	0.268(0.287)	0.374(0.118)	0.030(0.902)	0.074(0.819)
Malay	−0.526(0.255)	−0.652(0.278)	0.061(0.907)	0.500(0.272)	−1.466***(0.002)	0.356(0.576)
Chinese	−0.389(0.348)	−0.277(0.600)	0.365(0.435)	0.272(0.502)	−1.002**(0.012)	−0.054(0.921)
Degree	−0.324(0.479)	−0.616(0.261)	0.076(0.877)	−0.007(0.987)	0.116(0.805)	0.574(0.320)
Postgraduate	−0.528(0.320)	−0.380(0.548)	−0.053(0.927)	−0.596(0.263)	−0.579(0.314)	0.994(0.194)
Experience (4–10 years)	−0.007(0.977)	−0.428(0.257)	−0.147(0.610)	−0.393(0.142)	−0.149(0.593)	−0.097(0.790)
Experience (>10 years)	0.254(0.342)	−0.274(0.432)	−0.102(0.719)	−0.311(0.249)	0.159(0.562)	−0.013(0.972)
Employed	−0.271(0.463)	0.911(0.157)	−0.867**(0.020)	0.541(0.145)	0.004(0.992)	−0.393(0.490)
Finance-related sector	−0.193(0.616)	0.519(0.434)	−0.760*(0.051)	0.608(0.115)	−0.292(0.461)	−0.387(0.512)
Log-likelihood	597.156	374.803	526.419	588.123	560.744	362.957
Pseudo-R2	0.084	0.105	0.070	0.100	0.066	0.052

Source: The authors.

Notes: Among controls, younger, Indian, primary or secondary, experience of less than 4 years and retired are used as reference categories. *, ** and *** denote levels of significance at 10 per cent, 5 per cent and 1 per cent, respectively.

This section demonstrates the role of preferences for a firm’s characteristics in portfolio composition consisting of undiversified, quite diversified and fully diversified portfolios. The results are presented in Table 3. Preferences for a firm’s value characteristics decrease the propensity to composite undiversified portfolio (i.e., ‘portfolio 1’). Similarly, preferences for a firm’s quality of management and product characteristics reduce the probability to composite undiversified portfolio (i.e., ‘portfolio 1’ or ‘portfolio 2’). Turning to the likelihood of holding quite diversified portfolio, preferences for quality of management and product characteristics are positively significant (i.e., ‘portfolio 4’ or ‘portfolio 5’). Risky characteristics’ preferences also increase the likelihood of holding quite diversified portfolio but only ‘portfolio 5’. Preferences for a firm’s liquidity and trading volume characteristics are significantly and positively related to holding quite diversified portfolio (i.e., ‘portfolio 6’). In the specification of fully diversified portfolio, none of the firm’s characteristics is statistically significant.

OPEN IN VIEWER

Table 3.

Preferences for Firm’s Characteristics and Undiversified, Quite Diversified and Fully Diversified Portfolio

	Undiversified			Quite Diversified			Fully Diversified
	(I)	(II)	(III)	(IV)	(V)	(VI)	(VII)
	Portfolio 1	Portfolio 2	Portfolio 3	Portfolio 4	Portfolio 5	Portfolio 6	Portfolio 7
Profitability characteristics’ preferences	0.958(0.115)	−0.115(0.871)	0.250(0.285)	0.998(0.358)	−0.243(0.336)	0.054(0.872)	−0.260(0.282)
Value characteristics’ preferences	−1.280**(0.029)	0.911(0.269)	0.146(0.545)	−0.384(0.699)	−0.054(0.836)	0.536(0.139)	−0.233(0.359)
Liquidity and trading volume preferences	−0.407(0.279)	−0.162(0.770)	−0.194(0.228)	−0.272(0.733)	−0.092(0.607)	0.581**(0.018)	0.113(0.507)
Quality of management and product preferences	−0.785*(0.052)	−0.354(0.494)	−0.377**(0.021)	1.683*(0.058)	0.406**(0.027)	0.048(0.845)	0.139(0.417)
Risky characteristics’ preferences	0.724(0.135)	−0.142(0.797)	−0.260(0.147)	0.148(0.870)	0.536***(0.009)	−0.417(0.100)	−0.064(0.732)
Male	0.301(0.592)	−0.474(0.491)	−0.049(0.828)	−0.963(0.331)	0.138(0.574)	−0.120(0.705)	0.028(0.905)
Middle-aged	0.086(0.915)	0.520(0.663)	−0.857***(0.007)	−1.187(0.435)	0.254(0.492)	0.255(0.604)	0.689*(0.078)
Older	−0.150(0.915)	−16.881(0.998)	−0.501(0.303)	−17.656(0.997)	−0.194(0.739)	1.066(0.109)	0.372(0.504)
Married	0.545(0.327)	0.317(0.648)	−0.026(0.906)	−1.349(0.167)	0.164(0.495)	−0.033(0.915)	−0.082(0.719)
High income	−0.607(0.337)	0.097(0.905)	0.157(0.539)	0.811(0.477)	−0.316(0.247)	0.076(0.834)	0.148(0.568)
Malay	−0.079(0.953)	16.252(0.998)	1.061**(0.037)	17.408(0.998)	−0.664(0.171)	−1.360*(0.086)	−0.107(0.831)
Chinese	0.612(0.596)	17.295(0.998)	0.382(0.412)	15.919(0.998)	−0.584(0.169)	−0.133(0.824)	−0.043(0.923)
Degree	−1.060(0.248)	16.779(0.998)	0.085(0.859)	−3.498**(0.047)	0.091(0.864)	1.105(0.311)	−0.098(0.838)
Postgraduate	−1.374(0.309)	18.098(0.998)	−0.088(0.878)	−20.846(0.996)	0.033(0.957)	1.437(0.210)	−0.169(0.765)
Experience (4–10 years)	−0.242(0.757)	0.147(0.864)	−0.129(0.653)	1.630(0.191)	0.258(0.399)	0.229(0.552)	−0.337(0.269)
Experience (>10 years)	0.472(0.495)	0.154(0.869)	0.006(0.984)	−0.351(0.818)	0.165(0.599)	−0.519(0.225)	0.033(0.910)
Employed	17.300(0.998)	0.237(0.836)	0.657(0.131)	17.518(0.997)	−0.130(0.752)	−0.603(0.244)	−0.457(0.232)
Finance-related sector	18.674(0.998)	−1.688(0.259)	0.652(0.147)	16.963(0.997)	−0.116(0.787)	−0.448(0.405)	−0.575(0.154)
Log-likelihood	123.161	80.923	540.642	44.681	468.637	305.530	507.607
Pseudo-R2	0.172	0.172	0.084	0.315	0.065	0.106	0.040

Source: The authors.

Note: *, ** and *** denote levels of significance at 10 per cent, 5 per cent and 1 per cent, respectively.

OPEN IN VIEWER

Table 4.

Preferences for Firm’s Characteristics and the Number of Financial Asset Holding and Hypothetical Asset Allocation

	Number of Financial Asset Holding	Safe Asset Allocation	Moderately Risky Asset Allocation	Risky Asset Allocation
	(I)	(II)	(III)	(IV)
Profitability characteristics’ preferences	−0.281(0.140)	−0.038(0.837)	−0.116(0.525)	0.153(0.415)
Value characteristics’ preferences	−0.021(0.917)	0.496**(0.010)	0.359*(0.061)	−0.709***(0.000)
Liquidity and trading volume preferences	0.162(0.224)	−0.222*(0.082)	−0.028(0.825)	0.232*(0.077)
Quality of management and product preferences	0.327**(0.016)	0.086(0.506)	0.206(0.112)	−0.293**(0.027)
Risky characteristics’ preferences	0.008(0.957)	−0.135(0.338)	0.074(0.601)	0.032(0.825)
Male	0.044(0.809)	0.389**(0.029)	0.016(0.926)	−0.345*(0.058)
Middle-aged	−0.773***(0.006)	−0.672**(0.011)	0.178(0.497)	0.447*(0.098)
Older	−0.408(0.326)	−0.879**(0.028)	0.760*(0.057)	0.334(0.409)
Married	0.098(0.582)	0.118(0.492)	0.026(0.879)	−0.020(0.911)
High income	−0.089(0.667)	0.287(0.151)	−0.053(0.790)	−0.304(0.135)
Malay	1.010**(0.011)	0.369(0.335)	0.163(0.671)	−0.395(0.324)
Chinese	0.634*(0.071)	0.515(0.133)	0.065(0.848)	−0.514(0.153)
Degree	−0.075(0.848)	0.670*(0.080)	−1.023***(0.008)	−0.097(0.806)
Postgraduate	0.068(0.882)	1.098**(0.014)	−0.840*(0.061)	−0.668(0.144)
Experience (4–10 years)	0.116(0.618)	0.356(0.113)	0.222(0.322)	−0.520**(0.023)
Experience (>10 years)	−0.083(0.722)	0.405*(0.073)	−0.241(0.285)	−0.237(0.305)
Employed	0.477(0.135)	−0.182(0.555)	0.424(0.170)	−0.579*(0.074)
Finance-related sector	0.606*(0.069)	0.131(0.684)	0.180(0.576)	−0.655*(0.052)
Log-likelihood	1177.756	2056.829	2042.346	1884.787
Pseudo-R2	0.063	0.084	0.050	0.092

Source: The authors.

Note: *, ** and *** denote levels of significance at 10 per cent, 5 per cent and 1 per cent, respectively.

The regression results for the number of financial asset holding are presented in column (I) of Table 4. Given the same set of independent variables, the result in this specification provides interesting finding. Among the five groups of firm characteristics, preferences for a firm’s quality of management and product characteristics increase the likelihood of number of financial asset in the portfolio.

Investors are asked to allocate a certain amount to safe, moderately risky and risky assets in a hypothetical task. As shown in columns (II)–(IV) of Table 4, the results indicate that preferences for a firm’s value characteristics lead to a higher fraction of safe asset allocation, whereas preferences for liquidity and trading volume lower the fraction of safe asset allocation. Viewing moderately risky asset allocation, the significant positive effect is merely observed for a firm’s value characteristics. Considering risky asset, a firm’s value characteristics rather reduce risky asset allocation. Similarly, preferences for quality of management and product attributes lower the likelihood of risky asset allocation. As expected, the tendency for risky asset allocation increases with the preferences for liquidity and trading volume characteristics.

Discussion

The findings regarding the probability of holding a particular asset show that preferences for a firm’s various characteristics are significant in holding saving deposits, government/Islamic bonds and real estate/property. However, the likelihood of holding unit trust/mutual fund and stock/equity is unaffected by a firm’s characteristics. The finding can be explained by the fact that financial assets that are characterized by higher risk may be irrelevant to a firm’s characteristics. In the probability of holding a financial asset, a firm’s quality of management and product-related attributes have a stronger influence. This is consistent with the argument of Kempf, Merkle, and Niessen-Ruenzi (2014) that investors show an affective reaction to a firm’s product-related attributes.

With respect to type of portfolio holdings, preferences for a firm’s value characteristics, and quality of management and product characteristics reduce the probability of undiversified portfolio holding. Preferences for such characteristics indicate preferences for fundamental valuation technique (Doran & Wright, 2010), which lower the propensity to hold undiversified portfolio.

Investors’ tendency to hold quite diversified portfolios (a mix of low and moderately risk assets; or a mix of low- and high-risk assets) is caused by preferences for a firm’s quality of management and product characteristics. When it comes to portfolio composition, investors highly regard firm’s external product-related attributes on which they derive risk–return expectations. A similar finding is found by Khan, Tan, and Chong, (2016), but for trading decision. Also, preferences for a firm’s risky characteristics increase the probability of holding a quite diversified portfolio (a mix of low- and high-risk assets). Risky characteristics’ preferences lead to hold low-risk assets to overcome the downside of excessive risk and high-risk assets to meet the desire for risk. This finding is relevant to the studies conducted by Duxbury, Hudson, Keasey, Yang, and Yao (2013) and Khan, Tan, and Chong, (2016). A quite diversified portfolio (a mix of moderate- and high-risk assets) is further increased by investors’ preferences for liquidity and trading volume characteristics, which can be explained by investors’ preferences for easily tradable characteristics of a firm (Khan, Tan, & Chong, 2016; Ng & Wu, 2006).

When asset allocation is measured as the number of financial asset holding, a firm’s quality of management and product characteristics have significant impact. The visibility of this characteristic due to its external attributes conveys important information and thus influences number of asset holding. To explore further, investors are asked to allocate a certain amount to safe, moderately risky, risky and other assets, hypothetically. The evidence shows that value characteristics’ preference increases the probability of safe asset allocation and ensures investors’ safety net (Doran & Wright, 2010). Investors’ preferences for liquidity and trading volume characteristics lower the probability of safe asset allocation but increase moderately risky and risky asset allocation. The nature of this firm characteristic may increase investors’ risk-seeking propensity.

Value characteristics’ preferences also increase the probability of moderately risky asset allocation. Investors’ expectations of higher returns from a firm’s value characteristics such as market capitalization, BM ratio, PE ratio and stock price may cause asset allocation to moderately risky assets. Preferences for this characteristic, however, reduce the probability of risky asset allocation along with quality of management and product characteristics.

All in all, preferences for a group of firm’s characteristics affect investors’ asset allocation.

Conclusion

This study explores the relationship between stated preferences for firm’s characteristics and asset allocation. The results show that in holding a particular type of financial asset such as saving deposits, government/Islamic bond and real estate/property, preferences for a firm’s value characteristics, risky characteristics, and quality of management and product characteristics have significant impact. Regarding portfolio holdings, value characteristics and quality of management and product characteristics reduce the probability of under-diversification, while quite diversified portfolio holding increases with a firm’s quality of management and product, risky characteristics and the liquidity and trading volume characteristics. A firm’s quality of management and product characteristics affect number of financial asset holding. In a hypothetical task, preferences for a firm’s value characteristics, quality of management and product characteristics, and liquidity and trading volume affect asset allocation. While these findings support behavioural explanation, it contrasts conventional mean-variance assumption (Markowitz, 1959).

Implications

The findings imply that a firm’s quality of management and product-related attributes as well as liquidity and trading volume characteristics play a significant role in holding quite diversified portfolio and risky assets. Thus, preferences for firm-specific characteristics can help to design policies to reduce overinvestment in safe assets and the structural imbalance in the Malaysian capital market. SC and regulators may give more priority to investors’ preferences for a firm characteristic besides their own initiatives to minimize the structural difference. For investors, it implies that investors base their asset allocation on their preferences for firm’s characteristics. This will enhance investors’ understanding that different groups of firm’s characteristics have distinct impacts on asset allocation.

Acknowledgement

The authors are grateful to the anonymous referees of the journal for their extremely useful suggestions to improve the quality of the article. Usual disclaimers apply.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The authors received no financial support for the research, authorship and/or publication of this article.