Deputy Minister of Labour Mahinda Jayasinghe said on January 6 that the government is considering a new system that would allow private sector employees eligible for the Employees’ Provident Fund (EPF) to receive their savings as a pension instead of a one-off lump sum payment, but longstanding concerns about the EPF’s management remain unaddressed.
Established under the EPF Act No. 15 of 1958, the EPF is the largest social security scheme, valued at Rs4.4 trillion by the end of 2024.
The fund is meant to help private and semi-government sector workers save money for when they retire. While they work, 8% of an employee’s salary is taken each month and added to the fund, and the employer also contributes another 12% of the employee’s salary.
Echelon sat down with Dr Nishan de Mel, Executive Director of Verité Research, to understand how long-running governance failures at the EPF have severely hurt returns and public trust, why a pension-style payout could work only if it is voluntary and competitive, and what reforms are needed, including moving EPF management to an independent, professional body focused on maximising returns.
Do you think offering an EPF pension option would be better for members than receiving their savings as a lump sum?
A pension is understood as a regular monthly income that will be received for the rest of one’s lifetime. Everyone likes to have a pension. But when you offer to take a person’s life savings in the EPF and give them a pension instead, the details matter.
The EPF is a “defined contribution” scheme, which means members get back what they and their employers paid in, plus the investment returns earned by the fund. At retirement, members can currently withdraw this as a lump sum.
Dr Nishan de Mel, Executive Director, at Verité Research
If a person feels that they would like to convert this lump sum into a pension — known as a lifetime annuity in finance jargon — that type of option already exists in the private sector financial markets, especially through insurance companies. A variety of pension-like products have emerged over time, many of them offering monthly payments for a fixed number of years, coupled with life and health insurance.
So, from a public interest perspective, there is no particular reason for the EPF to get into the pension business. If more attractive pension payments are possible on lump-sum investments, the state banks and state-owned insurance companies can be encouraged to offer better products. Financial institutions have the relevant experience in structuring such instruments, which the EPF does not.
“For instance, when I look at the numbers at hand between 2017 and 2024, the EPF made an average return of 10% on its member balances, but 5-year government bonds returned an average of 13%.”
What would be rather detrimental to people is if the government robbed people of choice. That is, if members are told they cannot take a lump sum and can only access their lifelong savings through a pension given by the government. That would be a forced expropriation of their savings. If choice is taken away, it would likely be another compulsory, low-quality scheme, similar to what the EPF has become today. In the last few decades, Verité has evidence from forensic audits of serious mismanagement and abuse of the EPF. Even today, mutual funds available in the market can easily outperform the returns on the EPF. For instance, when I look at the numbers at hand between 2017 and 2024, the EPF made an average return of 10% on its member balances, but 5-year government bonds returned an average of 13%.
These differences in returns can seem small in any one year. But the compounded consequence of even a 3% difference is such that over a period of about 35 years of working, it would lead to the interest accrued in the EPF account being only half of what would have been accrued by investing it outside. So it is a serious loss of welfare and security for retirement that is created by the poor performance of the EPF.
This development of the EPF being used exploitatively by the government is something that has arisen more in the last two decades, I believe. From the numbers, it seems that prior to the turn of the century, the EPF performed relatively better. This exploitation is done not only through poor management of the EPF, but also through the structure of taxation. Verité submitted several proposals for Budget 2026, including a recommendation to remove the 14% tax on EPF investment returns. This is because we found that over 92% of EPF members earn so little that they are not liable to pay any tax on their income. Yet their EPF returns are still taxed at 14%. This is unfair. People are getting taxed more than they would be if they had been able to keep the savings and invest them in banks.
“Irregular bond trading is estimated to have cost the fund nearly Rs9.8 billion in losses between 2002 and 2015″
It also goes against the international norm of retirement schemes. Almost 80% of the countries don’t place a tax on the investment returns of provident funds. Others have reduced taxes. All because it is universally agreed that retirement savings should be encouraged. Sri Lanka is very unusual for having higher taxes on the EPF, discouraging retirement savings rather than encouraging them.
To summarise, there is nothing wrong if the government were to offer a competitive and attractive pension scheme through its banks and insurance companies that people could choose from the marketplace of pension options. But it would be rather problematic if the government were to force EPF members to put their under-performing and over-taxed savings into a pension scheme that would then continue that exploitation for the rest of their lives.
If the EPF introduces a pension option, what new governance risks could arise beyond the issues that already exist today?
Running a long-term fund that must pay out an agreed amount in the future is inherently risky. When finance and insurance firms offer similar products, they are subject to strict regulation to ensure the money is protected and the promised payments can be met when they fall due.
One advantage of having private sector providers manage pension-style payouts is that the market is already regulated. The EPF, however, is run by the Central Bank of Sri Lanka (CBSL), which makes all investment decisions. What we have seen is that while the CBSL is effective at regulating others, it has struggled to hold itself to the same standard.
The forensic audits of the EPF tell a clear story. Irregular bond trading is estimated to have cost the fund nearly Rs9.8 billion in losses between 2002 and 2015. The EPF also recorded losses of about Rs9.8 billion from equity investments between 1998 and 2017. In 2007 alone, it suffered a net loss of Rs389 million from investments in unlisted companies that were not publicly traded.
If the EPF was privately managed, the CBSL or another regulator would likely not have allowed it to be run this poorly. The bigger risk is that even if a pension-style option is introduced, weak oversight could persist, because that is what we are already seeing today.
If oversight fails, the money can run out. A past example is the farmers’ pension scheme launched in 1987, which was later halted due to a lack of funds. The scale of the problem is reflected in the Auditor General’s remarks in the Agricultural and Agrarian Insurance Board’s 2013 annual report, which stated:
“The Scheme had a negative Pension Fund account balance of Rs53,968.5 million by the end of the year under review and it had been increased continuously since the year 2010 as compared with a positive balance of Rs663.8 million in the year 2009 due to making the pension payments by obtaining loans from other existing funds of the Board instead of collecting the adequate contribution from the members of the Scheme. Thus, the ability of the Scheme to continue as a going concern without the financial assistance from the General Treasury and other Funds is doubtful.”
“My primary recommendation is to set up a professional retirement fund management office, similar to the professional debt management office that has been created. Its sole job should be to maximise returns for EPF members”
So the cart cannot come before the horse. The government must first prove it can regulate the system properly and build strong oversight before offering new payout options to members.
In the public sector, retirement benefits come in the form of unfunded pensions. How does that system differ from the EPF, and what risks does it create for pensioners and for the government?
Last year, the Department of Pensions estimated that public sector pensions cost the government Rs491 billion.
In the public sector, a key risk is that politicians can make promises without setting aside money to pay for them. These pensions are “unfunded”, unlike the EPF in the private sector, where workers and employers contribute while the employee is working. With an unfunded pension, the government promises to pay a retirement benefit in the future but does not have funds set aside today to do so. That can create serious fiscal pressure over time and, if not managed carefully, become unsustainable.
When the government commits to large future pension payments without setting money aside, it is taking on a long-term liability that is effectively a debt obligation, even though it is not recorded as debt in the way fiscal accounting is done.
Between fully funded and unfunded pensions, there is another accounting approach called “notional funding”. This means the government calculates its future pension bill and records it clearly in official documents, such as budget papers, even if it does not set aside money in the budget, just like it is expected to record contingent liabilities. Doing this improves transparency of the obligation and can discourage decisions that make the system unaffordable later. Notional funding of pensions is a well-accepted practice that Sri Lanka could adopt, as it focuses on sustainable recovery from the recent economic crisis and tries to manage public finances more professionally.
Taking a professional approach to managing pension debt obligations is important because Sri Lanka has repeatedly seen today’s actions turn into tomorrow’s crisis. If we ignore the long-term consequences, it becomes easy to make decisions now that cause a crisis later.
The Sinhala song Sumihiri Pane captures this playfully and rather well: “Play the devil with the bottle, freedom in the bar. When you go broke, don’t worry, sell your lovely car.” This is what happens when you don’t take a long-term sustainability approach to your financial management. We must avoid the government doing a Sumihiri Pane with the EPF or government pensions. We don’t want the country or the retired elderly to pay the price of “selling their lovely car” in the future because of the government’s failure to act professionally in the present.
“The Sinhala song Sumihiri Pane captures this playfully and rather well: “Play the devil with the bottle, freedom in the bar. When you go broke, don’t worry, sell your lovely car.” This is what happens when you don’t take a long-term sustainability approach to your financial management”
So I think fully funding the public sector pension scheme over the long term is the best way to avoid a future economic crisis caused by pension promises the country cannot afford. But funding it notionally is a sensible way to start, which I think is important and necessary for professional governance of the pension obligation.
Current government servants can stand to lose in the future if we fail to do that.
What are some clear examples from the past that show EPF funds were mismanaged, or that raise serious concerns about how the fund has been run?
In an old insight Verité published titled, Flawed Rationale Behind EPF Re-enterth entered the Stock Market, we found that through 2009–2010, EPF returns from equity investments were less than 5% while the All Share Price Index (ASPI) achieved a return of approximately 100%.
Then there is the infamous bond scam, which actually happened over a period; it was not just a single event at the end of February. Unfortunately, the scope of the forensic audit carried out by the Central Bank was carefully designed to avoid looking at trading during the bond scam period, but it only looked at anomalous bond trades from 2002 up to 28 February 2015. Even then, it recorded significant bond-scam-like activity where windfall profits were generated for private parties by the abuse and manipulation of the EPF and the bond market.
These limited forensic audits showed that irregular bond trading cost the EPF approximately Rs9.8 billion between January 2002 and February 2015.
In a publication Verité made last year, Forensic Audit of Central Bank 2019: Assessment of Losses to the EPF, we showed some of the more scandalous actions revealed by the forensic audits. For instance, in a classic pump-and-dump scenario, the EPF purchased 2.5 million shares of Brown and Company when the price of a share had risen from Rs17.5 in April 2009 and peaked at about Rs290 in July 2011. After the purchase, the share price subsequently plummeted back to its original levels. The forensic audits show that the EPF recognised an impairment loss of Rs1.3 billion on this investment, losing 69% of the investment value.
Finally, during domestic debt restructuring, Verité wrote on its PublicFinance.lk platform that Sri Lanka is the only country where restructuring targeted social security funds while excluding banks and individual investors. Simply, debt restructuring is meant to ease pressure on the government by reducing what it has to repay.
“EPF returns from equity investments were less than 5% while the All Share Price Index (ASPI) achieved a return of approximately 100%”
In Sri Lanka’s case, three groups could have been affected: individual holders of government debt, banks, and provident funds. Individuals and banks did not see a cut in their returns and will be repaid in full, while the EPF saw its returns reduced.
If you could give key recommendations to the EPF and its governing body, the Central Bank of Sri Lanka, what would they be?
The EPF’s management should be moved into a separate, professionally managed institution. Even if it remains state-owned, it should be treated like any other non-bank financial institution and regulated in the same way by the CBSL. The CBSL generally does a reasonable job regulating private non-bank financial institutions, but the problem here is that it is not effectively regulating itself.
My primary recommendation is to set up a professional retirement fund management office, similar to the professional debt management office that has been created. Its sole job should be to maximise returns for EPF members.
Secondly, a similar office can be set up to cost and manage notional and actual funding set aside for sustainably financing government pension promises for the future.
Thirdly, until that happens, we need a fresh set of forensic audits on the EPF immediately and more transparency and public disclosure on how the EPF is managed and its performance is reported, because right now the old Latin maxim holds: Quis custodiet ipsos custodes? The EPF is being abused because the guardians are not guarding themselves.