Disclosure Practices Among Unit Trust Funds in Malaysia

This study examined the level of transparency among unit trust funds in Malaysia. Unit trust funds were the focus of this study due to its nature: unique, heavily regulated and rapidly growing. The study investigated the extent of disclosure in relation to the various mandato y and volunta y requirements and the roles of the specific characteristics of trustfunds in explaining their level of disclosure in the annual reports. Thefindings showed that the level of disclosure increased rapidly within the period of study suggesting that trust funds were becoming more transparent. With regard to the funds’ specific characteristics, the study showed that the influence was not consistent. The model best explained the disclosure level during thepre-mandato y period, i.e. 1996. During this year, the variables ’sponsor of the fund’ and ‘auditors’ significantly influenced the disclosure level in the predicted direction. In the other years (1997 and 1998), the explanato y power of the model was very low as suggested by the significant reduction in the adjusted R2. One argument was that during this period, the revised Guidelines had been issued and thus the model could not explain much the variation in the disclosure level as a large amount of the items in the study were already mandated.


INTRODUCTION
The issue of corporate transparency has recently been a topic of interest particularly with the Asian financial crisis in 1997.
Corporate transparency is about an entity disclosing all relevant and material information not already known to the public.In accounting, the major issue involving transparency is the level of disclosure in annual reports particularly relating to items that are not required to be disclosed either by regulatory bodies or by law.The amount of disclosure and the degree of compliance with the approved accounting standards serves as an indicator of high quality financial reporting.Malaysia's seriousness in increasing the level of transparency is evident in the number of disclosure regulations and accounting standards that have been introduced by the relevant bodies in recent years.As suggested by Watts and Zimmerman (1986), these governmentally enforced regulations will improve the capital market's ability to allocate capital more appropriately.
Extant studies investigating transparency mainly focused on public companies.This study, on the other hand, attempts to examine the level of transparency among unit trust funds in Malaysia.Unit trust funds are the focus of this study due to its unique nature; a unit trust fund is neither a company which utilizes its assets to produce services or product, , nor a finance institution which accepts money from depositors which in turn lent to borrowers for profits.A unit trust fund accepts money from investors and issues units to the investors in return for the money.Unlike a finance institution, a trust fund does not lend money for profits but rather it uses the money to invest in various securities.Thus, being investment driven, the level of disclosure of a unit trust fund is expected to be high as all of its unit holders are investors.Furthermore, as a unit trust fund is an investor in nature, it would expect the entities in which it holds securities to be transparent.Therefore, a unit trust fund is also predicted to be transparent to promote high transparency level among companies.Thus, it is the purpose of this paper to investigate the level of transparency among unit trust funds in Malaysia.Additionally, unit trust funds have become a very popular type of investment among Malaysians, particularly in the 1980's and 1990's.The government strongly encourages the public to invest in unit trusts because these funds are managed by professionals.This is evidenced by the large number of unit trust funds that are sponsored by either the Federal government or the State governments.The level of transparency among unit trust funds is expected to improve as a result of the various changes since the early 1990's involving disclosure requirements to ensure that the interests of unit holders are protected.
This study focused on the financial disclosure provided by unit trust funds in their annual reports.Although various defimtions were given to the word 'disclosure' (see Azhar, 1998), we, nonetheless, restrict our definition of disclosure to all information disclosed by unit trust funds in their annual reports as per required by the Securities Commission (SC) as contained in the Guidelines and the proposed accounting standard issued by the Malaysian Accounting Standard Board (MASB).

BACKGROUND
As at December 1998, there were a total of ninety-two approved funds or schemes managed by thirty-two management companies (http: / /www.sc.com.my).The 1990's alone saw a very rapid growth in the unit trust industry in Malaysia where sixtythree funds were set up during the period 1990 to 1998.The growth pattern continued in the year 2000 where another seventeen new trust funds were launched (The New Straits Times, 14 February 2001).
Prior to the establishment of the SC in 1993, the unit trust fund industry was under the purview of the Guidelines on Unit Trust Funds, which was issued in 1991 by the Capital Issue Committee (CIC).Compliance to the Guidelines was not compulsory.In 1994, following its establishment, the SC adopted the Guidelines, known as Guidelines on Unit Trust Funds 2994 (hereafter referred to as the Guidelines).Compliance to these Guidelines was mandatory.Following a revision, the revised Guidelines were issued by the SC in 1997 known as Guidelines on Unit Trust Funds (here after referred to as the revised Guidelines).These revised Guidelines were mandatory and superseded the earlier Guidelines.The revised Guidelines imposed various additional disclosures which were considered as important in evaluating the performance of a trust fund.Contents of the annual report of the trust funds were outlined in Schedule VI of the Revised Guidelines.
Variations in the items disclosed in annual reports of unit trusts still prevail despite the reporting requirements imposed by the SC (MASB, 1998).The issuance of the Draft Statement of Principles 2 (DSOP 2), Financial Reporting by Unit Trusts, in 1998 by MASB was an attempt to address this issue of variation in reporting.DSOP 2 specifies items to be disclosed in the financial statements and the methods of measuring certain items for disclosure, which were not previously covered in the revised Guidelines.Thus, DSOP 2 addresses the issue of disclosure and measurement not dealt with by the revised Guidelines.However, unlike the revised Guidelines, DSOP 2 is not mandatory.
The scope of this study was limited to the disclosure practice due to data unavailability at the time of the study.This study did not deal with the measurement and recognition aspects of financial reporting.The reason was that these issues of recognition and measurement were only addressed in 1998, following the issuance of DSOP 2.
No research evidence on disclosure practices by unit trust funds in Malaysia is currently available.Nevertheless, several studies on disclosure practices by companies in other industries are available.For example, Ku Nor Izah and Shamsul Nahar (1998) conducted a study on mandatory disclosure practice among general insurers and Hashanah (1999) examined disclosure practices of banks in Malaysia.On more general industries, Hossain, Lin and Adams (1994), Normah and Abu Bakar (1995) and Azhar (1998) studied the disclosure practices by companies listed on the Kuala Lumpur Stock Exchange (KLSE).Thus, the present study will add further to our knowledge about disclosure practices in Malaysia by focusing on the unit trust industry.
Given the foregoing background, this study attempts to investigate the reporting trend among Malaysian unit trust funds, starting from the period before the issuance of the revised Guidelines (i.e. 1996)until shortly after the DSOP was released for comment in 1998.The objectives of the study are as follows: 1. To investigate the extent of disclosure in the annual reports of unit trust funds following the issuance of Guidelines in 1994, after the issuance of the revised Guidelines in 1997, and after the issuance of the DSOP 2 in 1998, and 2. To exam&e the extent to which disclosure is influenced by size of a fund, auditor and sponsor of the fund.

HYPOTHESIS DEVELOPMENT
Despite the growing criticism that annual reports provide outdated information and are thus not considered useful in helping make decisions, at least in Malaysia, evidence shows that annual reports are among the most impor-tant sources of information analysts (Ku Nor Izah and used by financial Zuaini, 1995; and Azhar , 1998).This evidence was further supported by a comment from a financial analyst whoindicatedthattheannualreportwasamong the major formal sources of information that analysts usually relied upon (Razani, 2000).
Research on the extent of disclosure has been conducted extensively throughout the world, particularly in the US (e.g.Cerf, 1961; Firer and Meth, 1986;Chow and Wong-Boren, 1987;and Chew and Lee, 1990).However, none of extant studies on disclosures focussed on unit trust fund industry, whether in Malaysia or in other parts of the world.
A low level of disclosure was observed worldwide in the 1970's (Buzby, 1974;Barrett, 1976;Firth, 1979;and Azhar, 1998).In his comparative study, Barrett (1976) found that there were differences in the level of disclosure among companies in the US and those of other countries.Companies in the US were found to have a higher disclosure level than those in the UK, Japan, France, Germany, Sweden and the Netherlands.In his study on the UK companies, Firth (1979) found that the disclosure levels were very low.He argued that one possible reason for the lack of disclosure was that the finance director, and to some extent their auditors, might not have been be aware of the importance of some of the items to the users of the financial reports.
Studies conducted during the 1980's showed that the disclosure level remained unsatisfactory.For instance, the studies on voluntary disclosures in South Africa (Firer  and Meth, 1986) and Singapore (Chew and Lee, 1990) showed that the level of disclosures was not up to the level financial analysts were expecting.Wallace (1988), in a study on the mandatory disclosure among Nigerian companies, concluded that there was a poor compliance with the disclosure requirements.
In Malaysia, a series of research studies had been carried out in the 1990's to examine the disclosure practices by Malaysian companies.For instance, studies by Hossain et al. (1994), andNormah andAbu Bakar (1995) on voluntary disclosure by the Kuala Lumpur Stock Exchange (KLSE) listed companies suggested that the disclosure levels were unsatisfactory.Recent evidence employing the longitudinal approach offered by Hashanah (1999) and Azhar (1998) revealed that disclosure levels did improve over time.They attributed that the improvement was due to the increased mandatory disclosure requirements being imposed upon companies throughout the period of their study as evidenced by the sudden increase in the level of disclosures on mandatory items.In another related work, Shaari (1997) found that the Asian region countries were rapidly progressing towards developing a useful and comprehensive disclosure package.This development should thus, be reflective of the growing importance of capital markets in'this region as a source of financing of a Various firm-specific characteristics had been examined as an attempt to explain a firm's disclosure level (see, for example, Singhvi (1968), Buzby (1974), Stanga (1976) ' Firth (1979), Chow andWong-Boren (1987), Cooke (1989and 1991)' Ahmed and Nicholls (1994) ' Hossainet al. (1994), Wallace andNaser (1995), and Ku Nor Izah and Shamsul Nahar (1998)).Among the major characteristics studied which have indicated some positive rela-firm.
\ tionships were size of a reporting entity and auditor size.As this study attempts to extend previous studies into the unit trust industry, the previously-empirically-tested variables together with a new variable, that is the sponsor of the fund, will be utilised in this study as they are expected to affect the level of disclosure of unit trust funds.The type of sponsor is expected to be important as it is very common in Malaysia and it provides a varying degree of security in the investment.Each of the variables is explained in the following sections.

Size of Fund
Research evidence involving companies has consistently shown that size has a positive relationship with the disclosure level.Several arguments have been offered to support the positive relationship.First, collecting and disseminating information is a costly exercise, and thus, it is always the larger firms that can afford such expenses (Atiase, Babber and Freeman, 1988).Additional costs that will be incurred by large firms are considered insignificant since the information systems are designed to fulfil the various reporting requirements.Therefore, with the information that is readily available, it is just a matter of determining what is to be disclosed.For a small firm, most of the information is collected on the specific need basis and a comprehensive information system rarely exists.Another reason that has been put forward is that fuller disclosure by small firms will put them in a competitively disadvantageous position in comparison with larger firms in the industry (Hossainand Adarns, 1995).Hence, small firms are likely to disclose less information than big firms.Large companies are always regarded as entities of economic significance so that there may be greater demands on them to provide information for their customers, suppliers, analysts, government as well as the public.Since unit trust funds are also economic entities, and are to some extent similar to companies, the foregoing arguments would also apply to the former.
Size of an entity appeared to be a common characteristic selected as an independent variable in most studies.Except for studies undertaken by Ahmed and Nicholls (1994) in Bangladesh, and Ku Nor Izah and Shamsul Nahar (1998) in Malaysia, other studies e.g.Singhvi (1968), Firth (1979), Chow and Wong-Boren (1987), Cooke (1989 and1991), Gibbins et al. (1990), Frost and Pawnall (1994), Hossain et al. (1994), Wallace and Naser (1995), and Marston and Robson (1997)) showed that the extent of disclosure increases with firm size.However, in a study conducted on Hong Kong firms, big and small companies were found to disclose more mandatory items than the medium-sized firms (Tai, Au-Yeung and  Lau, 1990).
Given the foregoing discussion, the following hypothesis is proposed:

23, :
The level of disclosure in the annual reports of unit trustfunds is positively related to the size of t h e f i n d .
Previous studies involving public-listed companies employed various proxies to measure variable size.The most commonly used measurements for size are total assets and market capitalization.Of these two, market capitalization reflects the market valuation of a firm's total wealth.In the case of a unit trust fund in Malaysia, the market does not value its units in circulation; rather, the value of its unit in circulation is determined by the market value of its portfolio (i.e. investment in securities).The market value of a unit trust fund is the net assets value (NAV).Thus, being influenced by the market, the use of a fund's NAV as a measure of size was appropriate for this study.Furthermore, NAV is an important indicator for a unit trust fund performance as well as an important factor in determining the selling and buying price of the units.Thus, NAV is like market capitalization for a listed company which is reflective of the market valuation of the company.

Auditor
An Auditor is entrusted to give an opinion about the "true and fair" view of the financial position of a firm.The type of opinion issued by an auditor about the financial position of a company is determined by audit quality.The issue of audit quality becomes important due the nature of an audit work, which is subjective and inevitably varies between auditors.The extent of audit investigation determines the auditor's independence (O'Sullivan, 2000) so as to enable the auditor to arrive at the right opinion.
The opinion of "true and fair" depends largely upon the audit investigation, a firm's compliance with the stipulated requirements, and the guidelines pertinent to the firm to reflect its underlying financial position.It has been argued and empirically shown that the quality of audit varies between auditors (Becker, Defond, Jiambalvo and Subramayam, 1998).DeAngelo (1981) defined audit quality as the joint probability of detecting a breach in the client's accounting system and subsequent reporting of the breach.In addition, Collier and Gregory (1996) noted that an auditor's ability to discover irregularities depends largely on the freedom to determine the suitable audit procedures and the extent of their implementation.
Hence, an auditor who is perceived to be more independent will carry out a detailed audit investigation to ensure that the information contained in the published annual reports is reflective of the firm's economic underlying.In the literature, audit independence is measured by its size.A large audit firm is argued to maintain its reputation as a provider of quality audit than its smaller counterparts (Hossain et al., 1995).Inability to discover errors and low disclosure level would have direct negative consequences on the reputation of a large audit firm, as perceived by the markets.The associated hypothesis is stated as follows: H,: The level of disclosure in the annual reports of unit trust f i n d s is positively influenced by auditor independence.
, Extant evidence on the influence of auditor size, a measure of auditor independence on the level of discloshre, is mixed.While Ahmed and Nicholls (1994), found that the extent of mandatory disclosure varies with auditor size in Bangladesh, Wallace and Naser (1995) documented a reverse relationship among Hong Kong firms.However, in an earlier study conducted in Hong Kong, Tai et al. (1990) showed that there was no relationship between auditor size and firms' level of disclosure, and this finding is consistent with those of Firth (1979), Hossain et al. (1994) and Ku Nor Izah and Shamsul Nahar (1998).The fact that those studies were conducted in different countries with different degrees of market sophistication and extensiveness of regulations might explain the conflicts in the findings.

A Fund's Sponsor
In Malaysia, unit trust funds can be divided into two: government funds (i.e.established by the Federal or State Governments) and private funds.A government fund is established with the objective of upgrading the status of the people and it usually allows a small investment (i.e.minimum of RM10).Moreover, in the case of Permodalan Nasional Berhad, managing the largest trust funds in Malaysia, unit holders ranging between R M l O and RM5, OOO of investment make up the largest portion of the investors in its unit trust funds (Utusan Malaysia, 27 April 2002).A total of 6.7 million investors fall within this range of investment.These investors are considered unsophisticated and are expected to place less reliance on published information in making their investment decisions.Moreover, a great majority of the unit holders in government-sponsored funds hold small amounts of investment.Thus, the reporting practices of a government fund are predicted to be different from those of a private fund.Thus, given the users' low dependence on the annual reports, it is expected that the level of disclosure among unit trust funds sponsored by the government is low.
A private fund is established with the primary objective of saving and securing stable future dividend flows.It always works as an alternative to a fixed deposit as it generally offers attractive returns.The amount of in-vestment by individual unit holders in private trust funds is generally large with a minimum investment of RM500.Thus, given the significant amount being invested in the fund, unit holders are predicted to be very concerned with their investments and their primary source of information is contained in the annual reports.Thus, it is expected that their dependence on published information is high.
Unit holders of a private fund are also different compared with those of a state fund with respect to their ability to make a sound investment analysis.Given the fact that their investment is not backed by the government and would usually involve a big amount of money, unit holders of a private fund are predicted to be more careful when making their investment-related decisions.Hence, their reliance on the accounting data is predicted to be high.Therefore, given this expectation, themanagers of private funds are predicted to disclose more information and to comply with all the relevant standardsbetter thana state fund does.
The foregoing discussion thus leads to the following hypothesis.

H3: The t p e of the sponsor of a f i n d does not
influenc'e the level of disclosure in the annual reports.

Sample Selection
Data were collected from the annual reports of unit trust funds for a period covering 1996 to 1998.The period from 1996 to 1998 was chosen because the revised Guidelines was made effective in May 1997 and the DSOP was issued in July 1998; hence this would enable researchers to compare disclosure practices during the premandatory period (i.e.1996); after the revised Guidelines was made mandatory (i.e.1997) and after the issuance of DSOP 2 (i.e.1998).
Only trust funds that issued complete annual reports during this period were included in the analysis of this study.Ninetytwo funds were operating as of December 1998.However, only sixty-four issued their annual reports between 1996-1998.Requests for annual reports were made to respective fund managers through the mail.Annual reports were also gathered from the library.From these two sources, the authors managed to gather complete annual reports from fiftyseven funds.This represented about sixty-two percent of the total population.Table 1 presents the distribution of the fifty-seven unit trust funds by sponsorship and auditor size.

Data Collection
The Data collection process involved preparing a scoring sheet listing the sections and items that needed to be disclosed in the annual reports according to Schedule VI of the revised Guidelines as well as those of DSOP 2. Items that required disclosure as contained in the revised Guidelines were considered mandatory in both 1997 and 1998.Items that were found in the DSOP 2 were considered voluntary throughout the period under study.The revised Guidelines specified eight sections to be reported in the annual reports, each with a different number of disclosure items totalling ninety-seven items, which were deemed relevant for constructing the disclosure index.Details on the items included in the index are listed in Appendix A. Out of the eight sections covered by the revised Guidelines, four were already contained in the 1994 Guidelines.These sections were balance sheet, income statement, audit report and management review.The cash flow statement, a section found in DSOP 2, was also included in the index of disclosure resulting in a total of nine sections.Consequently, this study examined each of the funds' annual reports to determine the frequency of reporting of each of the nine sections and the disclosure level involving the ninety-seven items.Table 2 reveals the sections and the distribution of the relevant disclosure items.

Disclosure Level
To determine the level of disclosure by a fund for each year under study, two measures of disclosure level were adopted.The first measure treats each of the nine sections in the annual reports as equally important without assigning weight.This means that a fund would be assigned 1 /9* of the total score for the disclosure of each section regardless of the comprehensiveness of the section.Thus the total score (labelled as E) would range between "0" (for nondisclosure) and "1" (for disclosing all the sections).
The second method to measure the disclosure level used in this study (i.e. total disclosure, TD) was to treat each of the disclosure items deemed to be relevant to the fund as equally important.This is known as a dichotomous procedure.This method is found in most studies investigating disclosure in annual reports [see, for example, Buzby (1974), Firth (1979), Firer and Meth (1986), Wallace (1988), Cooke (1989), Chew and Lee (1990), Azhar (1998) and Ku Nor Izah and Shamsul Nahar (1998)l.This method assigns a value "1" if it is disclosed, and "0" if it is not disclosed to a partiplar item.The TD score for each fund was calculated as follows: m TD = Zdi i=l where d = 1 if the item diis disclosed, 0 if the item di is not disclose, and m 5 n (see below) In cases involving funds which did not disclose irrelevant items, their non-disclosure of that item was not penalised.However, if the relevant item was not disclosed, a score "0" would be given.Decisions on whether an item was of relevance to the fund were arrived at upon a review of the annual report, as suggested by Cooke (1989).For instance, if it was mentioned in the annual reports that dividends were distributed by the fund, but the date of the distribution was not disclosed, then it was concluded that the item was relevant but not disclosed, and thus a score "0" would be given.Similar to the approach used by Cooke (1989), the additive model used was not weighted as each item of disclosure was treated as equally important.
Once the score on disclosure (i.e.TD) was computed, an index was computed to measure the relative level of disclosure by the fund.The index is the ratio of the total score of the fund to the score that the fund was expected to receive (if all relevant items were disclosed).
The maximum score (M) a fund could receive varies depending upon the number of relevant items: where d = expected item of disclosure, and n = the number of relevant items The total disclosure index (TDI) for each fund for each year was computed by dividing TD (score) over M (score).The index ranged between "0" and "1".A score of "1" indicated that the fund disclosed all the relevant items as required by the Guidelines and DSOP 2; while a score of "0" meant that the fund disclosed none of the relevant items.

Hypotheses Testing
A multiple regression analysis was conducted for each period to test the hypotheses: 1996, 1997and 1998.Each regression model included fund size, auditor, and a fund's sponsor as the independent variables.The influence of a fund's year end was included in the regression model as the effective date of revised Guidelines had different impact on a fund's disclosure depending on the fund's balance sheet date.Total disclosure index (i.e.TDI) was used as the dependent variable rather than total score (i.e.E) due to the comprehensive nature of the former.The ordinary least square (OLS) technique therefore suggests the following model (for each year): where: Total Disclosure Index; Natural log of net asset value; 1 for 'big 6' audit firm or '0' otherwise; 1 for a government-spon sored fund or 0 otherwise; 1996 and 1998 models: "1" i f k e balance sheet date falls after mid-year or "0" otherwise; 1997 model: "1" if the bal ance sheet date falls in May and after, or "0" otherwise; error term; and Size, being measured by NAV, was expressed in logarithm form.The transforma-tion was necessary due to the non-normal distribution of the data.Transformation of such continuous variables was found in Cooke (1989), Ahrned and Nicholls (1994) and Ku Nor Izah and Shamsul Nahar (1998).Classification of the balance sheet date for the 1997 model differs from those of 1996 and 1998.The difference in models was due to the fact that the revised Guidelines were made effective in May 1997.

Extent of Disclosure
Table 3 presents the results of the analysis of the reporting of sections in the annual reports for the period 1996 to 1998.Overall, there was a steady and significant increase in the reporting of sections in the annual reports.The mean score for the overall sections reported in 1996 was 0.55, compared to 0.77 and 0.89 in 1997 and 1998 respectively.The increase was significant as shown by the paired t-tests of the years 1996-1997 and 1997-1998.The Findings also showed that both the management review and the income statement were disclosed by almost all funds.Audit reports were also disclosed by more than ninety-five percent of the funds, except for 1998.These findings were expected since these sections were already made mandatory before 1996.Nevertheless, the reporting of the balance sheet was relatively low in 1996 although it was also made mandatory in both the 1991 and 1994 Guidelines.The reporting frequency of the balance sheet increased sigruficantly in 1997 and 1998.
As shown in Table 3, the reporting of the cash flow statement and performance data were relatively low.This evidence could be attributed to the fact that the cash flow statement and the performance data have never been mandated by the revised Guidelines.
However, the steady and sigruficant increase in the scores of all sections suggests that mandatory requirements were the reasons for the funds' high disclosure level.Subsequently, results on the extent of disclosure using the TDI as the measure of disclosure level also showed that there was a significant increase in the disclosure levels during the three-year period.As depicted in Table 4, the mean TDI scores in 1996,1997 and 1998 were 0.4279,0.6429and 0.8252 respectively.Further analysis using t-tests revealed that theincrease between 1996 to 1997, and 1997 to 1998 were significant at the one percent and five percent levels respectively.Results of the correlation analysis of the variables are presented in Table 5.It shows that the size of a trust fund is positively associated with the size of the au&t firm.This is generally consistent with our expectation.However, one interesting finding is that the correlation between the type of sponsor and the size of the audit firm is negative and sigruficant.This finding suggested that government (private) sponsored funds are associated with (non) big-six audit firms.This implies that private funds rely big-six audit finns, which may be interpreted more on the services of as being consistent with thesignallingtheory.Thecorrelationbetween the type of a fund's sponsor and the TDI was consistently negative but it was only sigruficant in 1997.\ Thus, this negative correlation is consistent with our prediction.Finally, the direction of association between the auditors and the TDI was not consistent throughout the period of study.The association was positive and significant in 1997.The relation was positive in 1997 but it was not significant.In 1998, the association became negative and significant.One explanation is that the role of the auditor might be important in the absence of regulations (i.e.1996).In the presence of regulations (i.e.1998), big-six auditors' roles might have lagged behind that of the non big-six as all trust funds were subjected to the additional requirements which had complied with trust funds which employed big-six audit firms.The results in Table 5 suggested that the problem of multicollinearity was not serious.However, each of the explanatory variables was alternately regressed on all the other explanatory variables to determine the severity of the multicollinearity problem in the regression model used in testing the hypothesis.*According to Kleinbaum, Kupper and Muller (1988), a severe multicollinearity problem exists when the resulting R2 from each regression is more than ninety percent or when R is greater than ninety-five percent.
The results of this analysis showed that multicollinearity was not a problem in any of the regression analysis as the R2from each regression was far below the ninety percent level.

Regress ion Res uI t s
Results of the regression analysis are shown in Table 6.Overall, the model showed a de-creasing trend as depicted by the decreasing adjusted R2, from fifty-seven percent in 1996 to justed R2, from fifty-seven percent in 1996 to only fifteen percent in 1998.Throughout the period, the findings were mixed.For variable size, the influence was only significant in the year 1998 in the predicted direction.Inconsisteent findings were also observed with regard to the status of the sponsor, wereby the influence was only significant in the predicted direction in 1996.Significant influence of audit firms was found only for 1996 and 1998, but the direction of the influence was not consistent for these two years.It was in the hypothesised direction in 1996 and in the opposite direction in 1998.
Based on the above findings, the extent of the influence of size of funds, auditors and status of the sponsor was mixed.However, the mixed influence was expected as the year 1996 was considered a pre-manda-tory period and year 1998 was considered a post-mandatory period.The year 1997, on the other hand, was considered a transitory period.Thus the crucial periods were 1996 and 1998.
In panel A of Table 6, two variables, namely sponsors and auditors, mfluence the extent of disclosure by unit trust funds, in the hypothesised directions.The regression model was statistically sigruficant with an adjusted R2 of 57 percent.The year, i.e. 1996, was considered a voluntary period as the revised Guidelines was not yet issued.Thus, a great amount of information disclosed was entirely determined by the management.Moreover, the arguments put forward in developing the hypothesis might have been relevant in a voluntary setting.
Panel B of Table 6 showed entirely different findings from those found in Panel A where the influence of sponsors and auditors became insignificant.The variable size remained statistically insignificant, as was found in 1996.However, the variable yearbecame statistically significant as predicted.Thus, companies whose year-ends fell after May demonstrated a high level of disclosure as opposed to companies whose year-ends fell before May.Therefore, the influence of year-end became more important than other variables in the model in explaining the level of disclosure for the year 1997.The reason is that the SC issued the revised Guidelines only in May 1997 and, therefore, the level of compliance increased accordingly for companies whose year-ends were after May 1997.The influence of auditor, size and sponsor could not be observed because the model contained the data of all companies without segregating those that had issued the annual reports before and those that had issued them after the revised Guidelines was in place.
As for the year 1998, the results shown in Panel C of Table 5 suggested that size of funds was positively and statistically significant in the predieted direction.The influence of size of funds was not significant for 1996 and 1997.One explanation is that when the SC issued the revised Guidelines, the variable fund size became important because larger trust funds might have been under close scrutiny by the SC and they were the ones that were expected to take the lead in complying with the SC Guidelines.Alternatively, the higher level of disclosure by larger trust funds might have been attributed to their ability to comply with the Guidelines given the resources they al-ready had and the infrastructure that existed in the firm.This could support the earlier argument of low information-gathering costs.
The insignificant influence in both 1996 and 1997 might have been attributed to their unwillingness to disclose information beyond what they deemed necessary.With the revised Guidelines and DSOP 2, they did not have the options but to comply.As for the negative influence of auditors on the level of disclosure in 1998, one explanation is possible.With the Guidelines being mandated to all funds, the level of disclosure by funds with smaller auditors was expected to improve in order to comply with all the necessary requirements.On the other hand, lesser initiatives were expected to be taken by the funds with larger auditors in order to improve the level of disclosure in 1998 since the previous disclosure had been reasonably satisfactory (a positive and sigruficant influence of auditors on the level of disclosure was observed in 1996).

CONCLUSIONS AND LIMITATIONS
Unit trust funds responded positively to the effort taken by both the SC and the MASB to strengthen the financial reporting practices of the industry.This was evidenced by the siccant increase in the reporting sections of the annual reports and accounts among unit trust funds as well as the disclosure level, at least, immediately after the issuance of the revised Guidelines.Nonetheless, the compliance with DSOP 2 was still very low.One explanation was that DSOP 2 was only introduced during the period of the study and it was not considered mandatory, as it has not been issued as a Malaysian Accounting Standard.Nonetheless, there was a clear upward trend for information disclosed in the annual reports over the years under study.The total disclosure level also showed a signlhcant increase from 0.4279 in 1996 to 0.8252 in 1998.This trend, at least, signals that the industry was moving towards the higher disclosure level.
Findings on the influence of size, auditors and sponsors on the disclosure level were mixed.Two reasons could be offered.First, the period of the study was not constant.In other words, the reporting requirements for each year of the study were different.Significant events took place in 1997 and in 1998, with 1997 being the most crucial with the issuance of the revised Guidelines.Though efforts had been taken to control these effects by including a variable known as year-end, the effects might stdl persist.Second, the study collapsed mandatory requirements (as in the Guidelines) together with the items considered voluntary (as in the DSOP 2).Thus, the differing effects might have contributed to the mixed findings.Furthermore, because the majority of the items were considered as mandatory (ninety out of ninety-seven), the mandatory items might have confounded the findings.Hence, regardless of the specific characteristics, the disclosure level was predicted to be high since the SC had mandated them.
Further study, however, needs to be carried out to discover the sections of the Guidelines and standards that are frequently complied with and to identify reasons why these certain sections were complied with.
Finally, this paper has several limitations.First, the findings do not extend beyond the unit trust industry.Second, the study examined only three aspects of the financial reporting for the unit trust industry: the Guidelines issued in 1991, the revised Guidelines and the DSOP 2. There are also other reporting requirements imposed on unit trust funds which were not covered in this paper.Examples of these include other accounting standards that are also applicable to a unit trust fund.With the inclusion of these other standards, the amount of voluntary disclosure would increase and the model could predict better the voluntary disclosure practices among unit trust funds in Malaysia.Perhaps, this could be addressed in future research.Third, the period of the study was mixed with various significant events: the pre-mandatory period (1996), mandatory (but considered transitory) period (1997) and the issuance of the DSOP 2 (1998).Perhaps, future research could concentrate on looking at the pre-mandatory period (prior to 1996) and post-mandatory period (post 1998) so that the effects could be measured better.

Contents of Annual Report
A. Management Review 1. Description of the scheme.2. Investment objectives and policies.3. Explanation of the management company's strategies employed towards achieving objectives.4. The scheme's asset allocation as at the date of the report.5. Particulars of significant changes in the state of affairs of the scheme.

17.
Analysis of the performance of the scheme (on the basis of unit prices) as at the date of the report.Breakdown of its components, namely capital growth and income distribution.Review of the market(s) in which the scheme invested during the period.
Where the scheme invests in foreign section, the names of the countries.If the manager has not done so, the shortcomings that may have impact on investor's decision, and 2.

8.
Steps taken to address/prevent shortcoming in cases where management has not done so.

Balance Sheet
Investment not classified as current or non-current, but presented in an order that reflects each category's relative liquidity.

Statement of Changes in Net Asset Value
Net asset value (beginning).Net asset value (ending).
Net income for the period.Changes in unrealised investment gains or losses.Amounts received from units created.Amounts paid for units cancelled.Distribution to the unit holders. -.

Cash Flow Statement
Operating and investment activities.Financing activities.

Notes to the Financial Statements
Income recognition basis for the following investments.2. Income recogrution basis for the following dividends.3. Income recognition basis for the following interests.
Maximum units allowed for issue under the provision of the trust deed. 5. Units created/cancelled during the period (number of units created as additional distribution, if any).6. Broker transactions with the top 10 brokers by value trade.7. Aggregate amount of brokerage/commission fees.8. Fees paid to each broker expressed in value.
Fees paid to each broker expressed in percentage. 4.

9.
Portfolio composition of the scheme (e.g, distribution among industries sectors, market and category of investment).Total Net Asset Value of the scheme, breakdown into NAV per unit.Tot$ Net Asset Value of the scheme, broken into number of units in circulation.Selling price for units.Repurchase price for units.Highest/lowest selling price.Highest/lowest repurchase price Total return of the scheme.Breakdown into capital growth and income distribution.
non-financial company characteristics on mandatory disclosure compliance in developing countries : The case of Bangladesh.The International Journal of Accounting, 29(1): 62-77.

Table 1
Distribution of Funds by Type of Sponsor and Auditor I * The big-

Table 2
Distribution of the Items in the Disclosure Index ** Required merely under DSOP 2. * Sections to be disclosed since the 1994 Guidelines.Malaysia Management Journal 5 (1&2), 65 -87 (2001)

Table 3
Results on the Reporting of Sections in the Annual Reports during 1996

to 1998 Section in the Annual Report Overall Content (E)
** Sigruficant at 1% level.aMandated since the 1994 Guidelines.

Table 6
Results of Regression Tests 11. Number of units held by management 12.Number of units held by related company.13.Composition of investment portfolio of the scheme as at the date of the report grouped in the appropriate categories, showing quantity held.14.Composition of investment portfolio of the scheme as at the date of the report grouped in the appropriate categories, showing cost of investment.15.Composition of investment portfolio of the scheme as at the date of the report grouped in the appropriate categories, showing MV of investment.16.Composition of investment portfolio of the scheme as at the date of the report grouped in the appropriate categories, showing MV of each holding as a % of NAV.Composition of investment portfolio of 13.Portfolio turnover of the scheme.Average annual return measured over the periods since the launch to the date of report.18. Voluntary disclosure performance of the relevant benchmark indices.19.Voluntary disclosure clearly stated that past performance is not necessarily in- dicative of future performance.20.Voluntary disclosure unit prices and in- vestment returns may fluctuate.A comparative table covering the last 5 financial years or over the whole of the period in which it has been in existence, showing: company.Malaysia Management Journal 5 (1&2), 65 -87 (2001) 17.BIBLIOGRAPY Ahmed, K. & Nicholls,