Domestic Debt Restructuring – An Alternate View

By Romesh Bandaranaike, Ph.D.

There have been substantial and wide spread criticisms of the recently instituted Domestic Debt Restructuring (DDR) scheme carried out by Central Bank of Sri Lanka (CBSL), to reduce the Sri Lanka Government’s requirements for funding. 

In this article I argue that the scheme, as structured and carried out by CBSL, is appropriate, given the ground realities in the country, and that the critics have ignored a number of factors which forced CBSL to design the scheme as it did. 

I then suggest additional steps the Government could take to further improve its financial position in connection with past Bond issues.

The DDR Scheme

Faced with massive shortfalls in revenue, the Government recently carried out a “restructuring” of the debts it owed on Sri Lanka Rupee denominated Treasury Bills and Bonds in an effort to substantially improve Government finances, including the funds needed to service these Bills/Bonds. 

The two key elements of the DDR are a) Converting all Treasury Bills presently owned by CBSL to longer term Treasury Bonds, thereby substantially delaying the payment dates on these Bills; and b) Effectively ‘forcing’ the Employees Provident Fund (EPF) and other superannuation funds to exchange most of the Bonds they hold for 12 year Bonds with somewhat lower interest payments.

They did this by threatening to increase the tax rates that EPF pays on its annual income to 30% from the present 14%, if they did not accept the Bond exchange. 

Since such an increase was financially worse for EPF compared with the Bond exchange, EPF opted for the latter.

CBSL estimates that the scheme would reduce the Gross Financing Needs (GFN) of the Government by 1.5%, 1% by the CBSL Bill-Bond exchange and 0.5% by the EPF Bond exchange.

The Criticisms

The principal criticisms of the DDR, from the public, Trade Unions, Economic Think Tanks, and numerous ‘Experts’, is that the entire burden of the DDR is being placed on the backs of the retirement savings of ‘poor workers’ who are the members of EPF.

CBSL has justified the proposed increase in the tax rates for EPF and other superannuation funds to 30% from the present 14% on the basis that the banks have to pay the higher tax rate of 30% plus VAT.  

Dr. Wijewardena, the former CBSL Deputy Governor, has written several articles criticizing CBSL for not presenting the correct picture in this regard and not disclosing full information on the impacts of the restructuring on EPF members’ returns. 

Dr. W’s main argument is that, in the case of the banks, the tax rate applies to ‘net interest’, whereas, in the case of the EPF, it applies to ‘gross interest’. 

For the banks, net interest is interest earnings on its loans less the interest it pays to depositors. 

In the case of EPF, Dr. W points out that EPF is not allowed to subtract the interest it pays to EPF account holders to determine its income and tax liability, and that CBSL comparison of bank and EPF tax rates is therefore misleading. 

In a subsequent article by Dr. W, he criticizes CBSL/EPF for not fully disclosing the cost to EPF holders of accepting the proposed DDR compared with an increase in EPF’s tax rate to 30% and the justification for accepting the Bond exchange rather than the tax increase to 30%. 

He goes on to add that CBSL has a conflict of interest as both the developer of Government policy and as the administrator of EPF. 

My Responses to the Criticisms

The criticism that the entire burden of the DDR is on the backs of workers is factually incorrect. 

As stated above, two-third of the 1.5% reduction in GFN (1%) is being achieved by exchanging the Treasury Bills held by CBSL for long term Bonds. 

Only 0.5% of the reduction is from the exchange of Bonds held by the EPF and other superannuation funds. 

In other words, 67% of the restructuring cost is borne by CBSL (in effect by all citizens), while only 33% is borne by EPF account holders. 

Furthermore, only workers in the formal private sector contribute to the EPF/ETF, while workers in the informal private sector (e.g. farmers, fishermen, small transporters, traders and construction workers) and Government employees do not. 

As a result, only about 25-30% of the workers in the country have EPF accounts. 

Therefore, 70-75% of workers in the country will not suffer any burden due to the EPF bond restructuring. 

Finally, the workers with EPF accounts also include middle and senior management of companies who cannot be called ‘poor workers’ as referred to in the various criticisms of the present DDR scheme. 

It is true that this category may only be a very small percentage of those holding EPF accounts. 

However, their account balances are likely to be very much higher than other EPF members and the impact of the restructuring on them will be proportional to these balances. 

It would be ideal if the EPF could release statistics in this regard which will allow an assessment of this element.  For example, the fraction of the total EPF funds held by those with EPF balances of over Rs 5 million (say), since such persons cannot be classified as ‘poor workers’.

Workers who have balances in EPF/ETF, built up as a result of deductions from their salaries and additional higher contributions from their employers, at least have a retirement fund they can turn to, even if it is somewhat less because of the DDR. 

It could conceivably be argued that the 70-75% of workers who do not have such balances, mostly working in the informal sector, are worse off or poorer than those with EPF balances. 

This would be a justification for placing the burden of part of the DDR on EPF account holders rather than on other, even ‘poorer’, workers.

In response to Dr. W’s criticisms that EPF and bank tax rates are not comparable the way CBSL has done; each year, EPF determines a percentage it will pay/accrue to the account of each EPF account holder based on the earnings by the EPF that year. 

This percentage is not an interest similar to that paid by banks to its depositors and is not a cost that is deductible by EPF to calculate its tax liability each year. 

It is simply a percentage decided by the EPF administrators to allocate the profits after tax earned by the EPF during the year.  Calling this percentage an ‘interest’ is a misnomer. 

Account holders in EPF are akin to shareholders in banks and not depositors. 

The amounts credited by EPF to a member’s account each year, based on EPF’s earnings during the year, is not a cost incurred by EPF in generating these earnings. 

It is something closer to a dividend paid by a bank to its shareholders in the form of additional shares. 

If EPF is allowed to determine its tax liabilities by subtracting these accrued amounts, banks should be able to deduct dividend costs in determining their tax liabilities. 

Dr. W’s criticism of the non-comparability of bank and EPF tax rates cannot be sustained.

Dr. W uses CBSL/EPF’s own numbers on the differences in returns under the two scenarios EPF has been offered and simply multiplies it by the total EPF Bond holding value to arrive at the cost to EPF of accepting the Bond exchange. 

This is something that could have readily been done by anyone and Dr. W’s implication that CBSL/EPF is hiding/not fully disclosing something in its statement cannot be sustained. 

The CBSL/EPF analysis is simply to show that the return to EPF is better under the scenario where it accepts the DDR option, compared to rejecting the DDR option and being subject to a 30% tax rate. 

This information is more than sufficient for EPF to recommend to its Board that it should accept the DDR. 

Dr. W than goes into the past history of taxation rates applicable to the EPF and points out that as originally envisaged EPF earnings were to be tax exempt. 

He shows calculations of the losses to EPF holders of the present DDR, compared with a situation if EPF earnings were tax exempt. 

This is a straw man put up by Dr. W to be knocked down as part of his criticism of CBSL/EPF. 

It is the tax rates that are applicable to EPF today, before the DDR, that are relevant and it would have been nonsensical for CBSL to show calculations based on what if the tax rates were those that existed many years ago. 

With respect to the criticism that CBSL has a conflict of interest in both being the administrator of EPF and the policy advisor recommending the DDR, I do not see any conflict. 

CBSL, as advisor to the Government, has made the DDR proposal which allows EPF to choose between two options, accept the proposed DDR Bond exchange or reject it and be subject to a 30% tax rate. 

As the administrator of the EPF, CBSL has simply analysed these two options and clearly shown that the EPF is better off accepting the Bond exchange compared with rejecting it and being subject to a 30% tax rate.

Other Relevant Issues

It is telling that none of those criticizing the present process have offered any viable alternative DDR arrangement to achieve the same objectives as the present exercise. 

In saying this, I am ignoring the suggestions by some parties who say there would be no need for the exercise if ‘The money stolen by the Rajapaksa’s is recovered’ or ‘The large corruption in Government is reduced’, and so on. 

Such actions, even if they were possible, are not alternatives, because they cannot be achieved in the short or medium term, which is one of the key objectives of the DDR. 

Verite Research did, some time ago before the announcement of the present DDR, present some analysis on a possible DDR which included sharing the cut across all Bond holders, but, for the reasons I refer to below, this is not a viable arrangement.

CBSL in its original presentation to the Cabinet, and subsequently to the public, argued that it would be prudent to exclude the Banks from the DDR exercise, because these institutions were already under stress as a result of COVID related business failures and because the banks would also be taking a hit from the future restructuring of USD Bonds, some of which are held by them. 

CBSL was of the view that such an exclusion was essential to ensure financial system stability. 

None of those criticising the present DDR arrangements have objected to this and I concur with that view. 

Even if the banks are able to bear the burden of some restructuring of the domestic bonds they hold, bank stability is also dependent on public perception, and excluding them from the DDR has certainly had a positive impact on such perception.

There are two unstated conditions applicable to the DDR exercise which have received no mention in the ongoing discussions. 

First, the DDR should be concluded in a short time frame since it will be a pre-condition to the restructuring of foreign currency debt. 

Second, it must be carried out in a legally valid manner. 

A Government Bond is a legal contract between the Government and the holder of the Bond. 

The Government is legally obligated to pay the interest coupon and the principal of the Bond on specified dates according to this legal contract. 

The Government has no legal authority to change the conditions of this Bond. 

It could, of course, pass a new law in Parliament giving itself the authority to make changes to existing Bonds. 

In doing so, however, if all Bond holders are not treated equally (in particular, if the bank holdings of Bonds are excluded), there are bound to be legal challenges in the Courts to any such changes, and the changes may well be struck down by the Courts. 

Even if such changes are not struck down, this process can take considerable time to be determined and would not be achievable in the time frame required for the DDR process to be concluded as discussed previously. 

The present DDR has finessed the issue of different treatments of Bond holders by making it ‘voluntary’, albeit by holding a gun to the head of the EPF and superannuation funds in the manner detailed earlier. 

This is draconian, but effective.

As per CBSL statistics presented in connection with the original DDR proposal, the total outstanding amount of Treasury Bonds at that time was Rs. 8,700 billion. 

Of this amount Rs 1,644 billion is held by ‘Others’, after excluding the banks and EPF and superannuation funds being subject to the DDR. 

Even if the banks are excluded from the DDR to ensure ‘financial system stability’ reasons mentioned previously, it would have been ideal if the ‘Others’ holding the Rs 1,644 billion in outstanding Bonds were subject to some ‘restructuring’, in the form of cuts in coupon rates and/or face value, or an extension of the tenor of these Bonds. 

If this was done, the benefit of these cuts could have been passed on in the form of a reduction in the burden passed onto EPF and superannuation funds. 

However, such an unequal treatment where some Bond holders (the banks) are excluded, even if the necessary legislation is passed, would certainly have resulted in legal challenges which would, at the least, have delayed the process as I have pointed out earlier, or even been found unconstitutional by the Courts.

Further Actions in Respect of Bonds

Very high yield rate Bonds (over 20% yield) were issued by the Government during the period prior to the announcement of the DDR (April 8, 2022 to March 13, 2023) as a consequence of severe financial shortages faced by the Government. 

Bidders for these Bonds added a premium to the bid yield rates, expecting a future restructuring of these Bonds as part of the DDR they knew was coming. 

By being excluded from the DDR, these Bond holders have enjoyed a ‘windfall profit’.  

It is not possible to specifically target these Bonds for a cut in coupons or face value, because many of the original holders may have sold some of these Bonds and have already earned the windfall profits. 

[The buyers of the Bonds would only receive a ‘normal’ profit in the form of coupon payments and final redemption.] 

Furthermore, in the case of individual Bond holders, their windfall profit would be in the form of a capital gain, which only attracts a tax rate of 10%. 

Corporates/ banks in the same situation would be paying a tax of 30% on their capital gains.  

I propose that the Government should ‘claw back’ some of these profits by imposing a special windfall tax rate of 50% applicable to all profits derived from these Bonds. 

If the original purchaser has not sold the Bonds, the coupon interest and the final face value redemption should also be taxed at 50%. 

There would be some complications in structuring this tax, since some high yield Bonds may have changed hands within the period that high yield Bonds were still being issued, which means that the initial purchaser would only have made a small capital gain and the next buyer would still be enjoying the high coupon rate or subsequently selling the Bond for a large capital gain.  

A proper structuring of the 50% windfall tax can ensure that the taxes fall on those making the windfall profits.  

A tax of the type proposed will clearly be borne by the wealthy, who would have been the purchasers of these Bonds, and go some way towards balancing the burden imposed on less wealthy parties as a result of the DDR.

To place this suggestion in perspective, between April 8, 2022 and March 13, 2023 all Bonds issued by CBSL had yield rates in excess of 20%. 

The total face value of such Bonds was Rs 1,233 billion. 

The weighted average yield on these Bonds was 27.89% and the weighted average coupons in these Bonds was 19.09%.  

The latest 4-year Bond yield rate is approximately 15% and this is likely to drop further.  

As a result, any seller today of an average high yield Bond would make a significant capital gain on the sale and would need to pay substantial taxes to the Government at the proposed 50% windfall tax rate.  

As per CBSL statistics in the original CBSL presentation to the Cabinet, approximately 63.5% of all Bonds at that time were held by the banks and the ‘Other’ category. 

If this percentage is also applicable to the high yield Bonds referred to above, the banks and the ‘Other’ category would be holding Rs. 777 billion of these Bonds.  

Advocata has recently analysed EPF’s Bond portfolio prior to the DDR Bond exchange and concluded that EPF’s share of high yield Bonds was proportionately much less than for other Bond holders. 

Therefore, banks and “Others” would probably be holding even more than the above mentioned Rs.777 billion in such high yield Bonds which, in turn, will mean that a 50% windfall tax on the profits from these Bonds will result in substantial tax revenues for the Government.

An aside at this point is that for some unclear reason there is no withholding tax (WHT) applied to the interest earnings on Bonds as in the case of interest earnings from deposits in banks and finance companies.

I have made enquiries as to why this is so from several senior Government officials and have not received any explanation for the practice.  

It may well be that some Bond holders do not even have income tax files and that they are evading all taxes payable on Bond interest. 

Since Bond holders are all likely to be at the highest tax bracket given the minimum sizes of Bond investments, I suggest that a WHT of 30% be applied to all Bond interest payments.

Holders are free to file tax returns and seek a refund if they have been taxed in excess of their liability due to such a WHT. 

I have been told there is some complication in applying WHT to Bond interest held by foreigners. 

If that is indeed the case, foreign holdings of Bonds could be excluded.

Conclusions

The economy of the country is in dire straits as a result primarily of bad/foolish policies of recent past Governments (including those by CBSL under its previous two Governors and Monetary Boards at that time) coupled with the endemic corruption inherent in the system. 

The present Governor of CBSL and his professional colleagues are facing very difficult conditions and are striving to get the country’s finances back on track. 

I can understand trade unions, workers and other similar entities objecting to the implementation of policies which directly impact them, irrespective of whether such policies are needed to solve the country’s issues. 

But, why is it that so called ‘experts’ and think tanks do not recognise the ground realities of what is practically achievable and support the efforts of CBSL rather than criticising these efforts in print and at discussion forums?