62 THE IMPACT OF INVESTMENT REGULATION ON THE PERFORMANCE OF PRIVATE PENSION FUNDS IN JAMAICA Leonie Anderson-Neita University of the West Indies, Mona June, 2012 Abstract This paper answers the question of whether investment regulations compromise the performance of private pension funds in Jamaica. The Jamaican case is an area lacking research on this particular topic. Most empirical literature in this area place their focus on Latin American Countries. The main finding of this paper is that even though the regulations to some extent safeguard the assets of pension funds and secure the retirement benefits to be enjoyed by members, they tend to place a limit on the potential earnings of pension funds, which could yield even a much greater benefit for members. The performance of the funds becomes particularly important for members in Defined Contribution Plans in which members accept the risks, unlike a Defined Benefit Plan, in which the members are guaranteed a particular pension benefit irrespective of the performance of the pension fund. JEL Classifications: G23, G11, G18. Key Words: Pension Funds, Interest Rates, Regulation, Fixed Income Securities, Stock Market. 1. Introduction. A Pension Plan is a Trust established in order to provide post retirement income for its members. Decisions relating to the plan are made by a Board of Trustees. Trustees are usually appointed by both the Sponsor and the members of the Plan. Two types of pension plans are Defined Benefit and Defined Contributions. In a Defined Benefit arrangement, the retirement pension a member receives is guaranteed. Therefore, irrespective of the performance of the various Funds, a member s pension is defined (for example, annual pension = 2% X average salary over the last three years prior to retirement X years of membership in the plan). As such, the sponsor accepts all investment risks.
63 On the other hand, in a Defined Contribution plan, retirement benefit is not guaranteed by a formula. The member and the sponsor contribute a fixed percentage of the member s salary on a monthly basis. The amount contributed is invested into the various funds subject to investment limits. At retirement, the accumulated contributions of the sponsor and the member are used to purchase an annuity (the respective purchase rates are determined by an Actuary), payable for the lifetime of the member. The amount available to purchase a retirement pension in a Defined Contribution plan is therefore dependent on the investment performance of the various funds. During the last decade, a consensus has been reached about the need to devise solutions for the looming old-age crisis (Dowers, Fassina, Pettinato 2001). There are two sets of phenomena that drive this situation: on one hand is the gradual decrease in the share of the working population which is caused by a mix of declining fertility and increasing life expectancy; on the other, is the widespread mismanagement and/or lack thereof on the part of the majority in effectively saving for retirement. It is expected that by 2025, nearly fifteen percent (15%) of the world population will be over 60 years of age. With increasing life expectancy, the population of developing countries like Jamaica is aging much faster than that of developed countries. It is expected that by 2030, 80% of the total elderly population will reside in developing countries. The Jamaican Case Until September 2004, private pension funds in Jamaica were not formally regulated. Approval of private pension funds was granted by the Commissioner of Taxpayer Audit and Assessment. The income tax act which was born in 1954 provided a significant tax benefit for contributors towards pension funds. Pension deductions are made from salaries prior to the deduction of income tax. Also, investment returns for pension fund are exempted from tax, hence the need to be somewhat overseen by the Commissioner of Taxpayer Audit and Assessment. In September 2004, the Pensions (Superannuation Funds and Retirement Schemes) Act, was passed. The Financial Services Commission was now given the mandate of regulating private pension funds in Jamaica. Supporting regulations were passed in March 2006. This first phase of regulation focused on three areas:
64 1. Registration, Licensing and Reporting 2. Governance 3. Investment Regulation. The Pensions, (Superannuation Funds and Retirement Schemes) (Investment) Regulations, 2006 are: 1. Limits on investments allowed Maximum of 20% of pension funds to be invested in foreign currency The inability of funds to hold or purchase any combination of investments in property or security of any one person or associate of that person exceeding 5% of the fair value of the assets of the fund. The inability of a fund to acquire, hold or invest in mortgages for properties located outside of Jamaica. A fund may only invest in ordinary shares listed on a recognized stock exchange of Jamaica or a recognized stock exchange of Canada, the United States of America and the United Kingdom. 2. Regulation of pension funds performance Investment transactions shall be undertaken at arm s length and at rates or prices comparable to those available in the market for similar investment transactions. My focus for this paper will be the impact of the investment regulation on the investment performance of pension funds in Jamaica. Whereas the regulation of private pension fund was needed in Jamaica, we cannot overemphasize the fact that the many limitations placed on pension funds may have compromised their performance. If the funds are to be built up enough to maximize the retirement benefit of their members, we have to be careful not to put too many limits on their capabilities to do just that. 2. Literature Review Srinivas and Yermo (1997) assessed the impact of regulatory regimes on the market performance of private pension funds in Latin American Countries (LAC) that have undertaken reforms of their pension systems. They focused, in particular, on the effects of draconian regulations, a set of rules on the industry s structure, investment regime and performance. The paper developed an intuitive analytical methodology for
65 benchmarking performance in defined contribution systems. Expected replacement rates for workers of different demographic characteristics were calculated based on historical rates of return of pension funds and alternative investments. It was shown that the poor past performance of the Chilean pension funds relative to market alternatives will create significant differences between the expected replacement rates of the pension system and the replacement rates that could have been expected under a more liberal investment regime. The following conclusions were made: 1. While the rules have achieved their basic objectives of safeguarding workers retirement savings from financial systems that lack transparency and solidity, the rules are not without costs. They create distortions in asset management, limit opportunities for diversification, and, as a consequence, hamper the performance of pension funds. 2. The existing regulatory regimes should be liberalized as soon as possible to allow pension fund investments in a wide array of financial instruments, and that regulations should require evaluation of pension fund performance against market benchmarks as opposed to exclusive focus on comparison with industry average. According to Vittas (1998) and Queisser (1998) The main adverse effect of performance regulation is to exacerbate herding behavior with small fund managers behaving like Stackelberg Followers (Tirole, 1988), choosing portfolios similar to larger funds, which have greater weight in the industry average return. 3. Data Data on net annual return of twenty private pension funds in Jamaica for the period 1997 to 2007 was obtained from a Pension Fund Manager and Administrator in Jamaica. The annual returns for the selected pension funds were obtained prior to 2006 (the effective date of the investment regulation) and subsequent to the effective date of investment regulation. Information on the Benchmark used for the different asset classes were also obtained from the annual report of the Pension Fund Manager. Demographic information such as age and sex was obtained for the active members of the funds examined. Similar studies in the area such as Srinivas and Yermo (1999) use gross monthly returns by a number of fund managers from the inception of each regulation until May 1998. They focused on three countries, Argentina, Chile and Peru, as opposed to one country used in this study.
66 4. Methodology The Investment Manager has two types of Investment Funds, the Self Directed Fund and a Pooled Fund. For the purpose of this paper, focus is placed on the Pooled Investment Fund. The Pooled Fund consists of the following asset classes: 1. Equity Fund - This fund invests in listed securities on the local and recognized regional exchanges (Barbados and Trinidad). The stocks that make up this fund are diversified across the insurance, manufacturing, finance, trading, communications and conglomerate sectors. 2. Fixed Income Fund - This fund invests mainly in Government of Jamaica (GOJ) Local Registered Stocks, Global Bonds and US Dollar Indexed Bonds and Repos. 67% of the instruments are fixed rate and 33% variable rate. 3. Mortgage and Real Estate Fund - This fund invests in hotels, commercial and warehousing properties, retail complexes and mortgage loans. 4. Foreign Currency Fund - This fund invests mainly in GOJ US and Euro dollar denominated fixed income instruments. 5. Money Market Fund - This fund invests in GOJ fixed income instruments with average tenure not exceeding two years. 6. Diversified Investment Fund - This is a balanced fund which invests in listed equities, GOJ securities, Corporate Bonds and Real Estate. An examination of the demographic characteristic of the selected pension plans was done and the number of years to retirement for each member calculated (using a retirement age of 65 years for males and females). This was done in order to determine the liability profile of the plans. The percentage of the combined Net Asset Value (NAV) of the selected pension funds that was invested in the various asset classes (equity, fixed income, foreign currency etc.) in the Pooled Investment Fund was then analyzed in order to determine if the fund allocation matches the liability profile of the plans. In determining the impact of the investment restrictions on the performance (measured in terms of the average annual return over the period of study) of these pension plans, two tests were conducted:
67 1. The average return earned by the pension funds from investments in the various asset classes over the period 1997 to 2005 (prior to the investment regulation) was compared with the average returns across the funds for 2006 and 2007 (after the effective date of the regulation). This is to determine if the annual returns deteriorated after the implementation of the investment regulation. 2. A Panel Model with fixed effects was used, since the data consists of a combination of cross section and time series data. Let M= Number of units (In this case the 20 pension funds to be assessed). Suppose that the data set consists of time series observed at n time moments (t= 1,.,n), since the analysis covers a ten (10) year period, n= 10. The dependent variable Y it = the average annual returns. The explanatory variable is the investment restriction placed on the pension fund. Restrictions are given the value 0 for periods prior to 2006 and 1 for periods after. The assumption is that the marginal effects of the explanatory variables on the dependent variable are the same for all plans, hence the restriction of the slope, γ i = γ. In order to account for the differences between the pension funds, for example, the investment knowledge of Trustees or how proactive Trustees are in overseeing the performance of the Funds, α i is allowed to vary. It is also assumed that ε it is homoskedastic and uncorrelated, both overtime and across pension funds. The model will therefore be: Y it = α i + x it γ + ε it Attempts are then made to answer the following questions: 1. What is the impact of the investment restrictions on the average annual return of the pension funds over the period? 2. Is the impact of investment restrictions significant? 3. Have the funds outperformed the market benchmark on a risk adjusted basis? 4. Is the performance of private pension funds compromised by the investment regulations in Jamaica?
68 5. Estimation Results and Analysis Based on the age distribution of the pension funds shown in Figures 1 and 2, a minimum of 4% of the members will retire within the next five years and approximately 50% will retire within the next 25 to 35 years. Based on this result, there is a fairly young membership in the plans examined. They can therefore tolerate some amount of risk. The equity and mortgage and real estate funds are considered to be the more risky of the funds that make up the Pooled Investment Fund. Figure 1 shows the percentage of the pension funds invested in the various asset classes. From examining Figure 3, approximately 30% of the funds under management are invested in equities, 30% in fixed income, 20% in mortgage and real estate, 6% in money market and 8% in the foreign currency and balanced funds. This distribution of the funds is in line with the age distribution, hence matches the liability profile of the funds. Since about 50% of the membership will retire within the next 25 to 30 years, having 30% of the fund values invested in equities and fixed income is essential. This is because they are both of a long term nature and younger members are able to tolerate the risk, brought about by the equity fund. Figure 1 AGE DISTRIBUTION 30% 4% 6% 9% 0-5 5.1-10 10.1-15 14% 15.1-20 20.1-25 25.1-30 Over 30 21% 16%
69 Figure 2 AGE TO RETIREMENT 7000 6000 5000 NUMBER 4000 3000 2000 1000 0 0-5 5.1-10 10.1-15 15.1-20 20.1-25 25.1-30 Over 30 YEARS TO RETIREMENT Figure 3 DISTRIBUTION OF POOLED PENSION INVESTMENT FUND 8% 8% 30% 6% 19% EQUITY FIXED INCOME MORT. & REAL ESTATES MONEY MARKET FOREIGN CURRENCY BALANCED FUND 29%
70 Figures 4 and 5 show that even though equity is very volatile, it does not only outperform the other funds over the long term but also inflation (measured by the CPI). Outperforming inflation is a major goal of pension fund Managers, since the retirement benefit of members must be preserved. Figure 1 shows that only 4% of members will retire within the next 5 years and 6% in the next 10 years. The money market fund provides liquidity for the fund, therefore, 6% of the value of the funds is invested in the money market fund. This is an adequate amount in order to meet the short term obligation. Figure 4 FIVE YEAR FUND PERFORMANCE 35% 30% 25% 20% 15% 10% 5% 0% Equity Fixed Income Mort. & RE Foreign Currency DIF CPI Figure 5 5 10 YEAR PERFORMANCE CPI Mort. & RE Fixed Income Equity 0% 5% 10% 15% 20% 25% 30%
71 Result 1 - Comparison of Average Returns Before and After Effective Date of the Regulation Figure 6 shows the average returns of the Pooled Investment Fund prior to the implementation of the investment regulation and the average returns of the Benchmark over the period. Figure 7 shows the average returns of the Pooled Investment Fund and the corresponding benchmark performance since the implementation of the regulation. Firstly, the foreign currency fund averaged 21% in the pre regulation period and 16% in the post regulation period. The returns of the benchmark (10 year GOJ Global Bond Yield) for both periods were 14% and 13% respectively. On the other hand, the equity fund averaged 32% in the pre-regulation period compared to 8% in the post regulation period. The average returns of the benchmark for the equity fund (the Jamaica Stock Exchange main index) for both periods were 18% and -2% respectively. Mortgage and Real Estate Fund averaged 22% in the pre-regulation period compared to 17% in the post-retirement period. The benchmark (inflation) for the periods averaged 22% and 13% respectively. The fixed income fund averaged 18% in the preregulation period compared to 15% in the post retirement period. From the empirical evidence above, it can be clearly seen that all the funds yielded a much higher return in periods prior to the imposition of the investment regulation that after its implementation. It is also worth pointing out that all the funds outperformed their respective benchmarks both before the implementation of the investment regulation and after its implementation. Figure 6 Performance of Pension funds Relative to Benchmark- 1997-2005 ( Pre Regulation) Average Returns 35% 30% 25% 20% 15% 10% 5% 0% EQUITY FIXED INCOME MORT. &RE FOREIGN XC AVERAGE RETURNS 1997 TO 2005 BENCHMARK
72 Figure 7 Performance of Pension Funds Relative to Benckmark- 2006 to 2007 (Post Regulation) 20% Average Returns 15% 10% 5% 0% -5% EQUITY FIXED INCOME MORT. &RE FOREIGN XC MONEY MARKET AVERAGE RETURN 2006 TO 2007 BENCHMARK Result 2 - Panel Model Estimated Annual Return Equation: Annual Return = 18.87706-15.7759 Investment Standard Independent Variables Coefficient Error t Statistic Pvalue Constant 18.87706 2.035061 9.275917 0.0000 Investment Restriction -15.77579 5.422152-2.909507 0.0041 R-squared 0.426826 S.E. of Regression 7.473921 From the panel regression results, If gross investment restriction is set to zero, the Annual Return = 18.87706. There is a negative correlation between investment restrictions and annual returns of the pension fund. T-Statistic This is the ratio of the estimated coefficient to its standard error.
73 Ho: β=0 H 1 : β 0 At the 5% level of significance, there is strong evidence to conclude that the estimated coefficient of investment is significantly different from 0. Therefore, we reject the null. P Value At the 5% level of significance(α) the P value is <.05- statistically significant- reject the null of a zero coefficient for Investment Restriction. R-squared This measures the success of the regression in predicting the independent variable. It is the fraction of the variance of the dependent variable, explained by the independent variables. 42.68% of the variation in Annual Return is accounted for by the model. 5. Conclusion From the results above, it is clear that investment restrictions have a negative impact on the performance of the pension funds holding all other factors constant. Whereas this information may prove meaningful, the regulation was implemented in 2006 and there were a few events that could have triggered such a result. A few examples include the poor performance of the stock market in 2005 into 2006, hurricane Dean in August 2007 and the rise in interest rates in the post regulation period. The rise in interest rates will particularly affect the performance of the fixed income fund due to the reduction in the value of the bonds. Despite that fact, it cannot be overemphasized that strict investment regulations compromise the performance of pension funds. Therefore, in answering the questions above: 1. Investment restrictions adversely affect the performance of private pension funds in Jamaica. 2. Despite the restrictions, pension funds have out-performed market benchmarks. 3. The cap of 20% of investment in foreign currency should be removed and funds be allowed to invest in properties overseas, once there is a creditable property management company.
74 REFERENCES D Arista, Jane, (2006). The implication of aging for the structure and stability of financial markets. 2006:1-30 Dowers, Kenroy, Stefano Fassina and Stefano Pettinato. Pension reform in small emerging economies, issues and challenges. 2001: 1-34 Heij, Christiaan, Paul De Boer, Philip Franses, Teun Kloek and Herman Van Dijk. Econometric Methods with Application in Business and Economics. New york: Oxford University Press, 2004 Srinivas, P.S. and Juan Yermo. Do Investment Regulations Compromise Pension Fund Performance? Evidence from Latin America. (1999):1-54 Srinivas, P.S., Edward Whitehouse and Juan Yermo. Regulating Private Pension Funds Structure, performance and investments: cross country evidence, Pension Reform Primer series Social Protection Discussion Paper, World Bank, forthcoming. (2000) Vittas, D. Regulation controversies of private pension funds, Policy Research Working Paper no.1791, World Bank. (1998) World Bank. Averting the Old Age Crisis. World Bank Policy Research Report. New York Oxford University Press Inc. (1994) Appendix 1: DATA SOURCES 1. Financial Services Commission- www.fscjamaica.org 2. Life of Jamaica (Now, Sagicor Jamaica ) Limited: Largest Investment Manager for Pension Funds 3. Employee Benefits Administrator Limited (Subsidiary of Life of Jamaica Limited)- Pension Fund Administrator
75 Appendix 2: FIXED EFFECTS- RESULT Dependent Variable: ANNUALRETURN Method: Panel Least Squares Date: 04/21/08 Time: 15:40 Sample: 1997 2007 Cross-sections included: 20 Total panel (balanced) observations: 220 Variable Coefficient Std. Error t-statistic Prob. C 18.87706 2.035061 9.275917 0.0000 RESTRICTION -15.77579 5.422152-2.909507 0.0041 Effects Specification Cross-section fixed (dummy variables) Period fixed (dummy variables) R-squared 0.426826 Mean dependent var 13.14041 Adjusted R-squared 0.335846 S.D. dependent var 9.170947 S.E. of regression 7.473921 Akaike info criterion 6.990654 Sum squared resid 10557.44 Schwarz criterion 7.468847 Log likelihood -737.9720 F-statistic 4.691432 Durbin-Watson stat 2.068646 Prob(F-statistic) 0.000000