Gorilla in the closet Problem of NSS

Reducing the budget deficit has traditionally been hard for Pakistan, given the country’s problem with stagnant tax collection and unyielding public spending; it is only exacerbated by endemic corruption. Countries like Pakistan are unable to reduce current expenditure due to fears of unemployment or protest against wage reductions, thus they often reduce long-term public capital expenditure, ie, PSDP in our case.

This reduction has, in turn, the effect of reducing the productivity of the public and private sector and hence national income. The alternative policy, namely, to increase private savings, seems not to really be a policy, since saving is a behavioural variable. That is, savings is the endogenous outcome of optimising behaviour by individual agents and hence is an endogenous outcome for the macro-economy. Nonetheless, to increase private saving is the most attractive for Pakistan, a country that suffers from large budget deficits, a lack of foreign capital inflows, and a low domestic savings rate.

Pakistan has a rather unique financial institution, the National Saving Scheme (NSS), for this purpose with the aim of increasing private savings by broadening the pool of savers. A number of different schemes are offered under NSS in the investment horizon of 3 years to 10 years. All the money raised in this scheme is essentially unfunded debt. The total share of unfunded debt in the government’s domestic debt stood at Rs 2064 billion out of which Rs 1,939 billion was NSS related.

Approximately 23% of all GoP debt outstanding is unfunded as of the last SBP report dated May 2nd, 2013 on the SBP website, consisting of data until the end-March 2013. The stock of unfunded debt increased by Rs 338 billion or approximately 21% percent compared with the same time last year; mainly because recently institutions were again allowed direct participation into the NSS, as earlier they were not. Net receipts in the various schemes have also registered impressive double digit increases ranging from 12% to over 20%, depending on the scheme, even though rates on the schemes were revised downward in response to the decrease in the benchmark discount rate.

In particular, a key goal of the NSS is to bring the county’s rural populace into the country’s financial system. By allowing institutions back into the scheme it might be the case that the NSS has not actually increased the volume of saving. Rather, it has only brought in existing saving from the informal, non-financial sector or other government institutions. There is no reliable data on either savings in the informal sector, for example in goods, or on the share of savings in the urban and rural sectors. There has been a constant significant increase in the total deposits in NSS and it is speculated that most of this increase has been contributed by governmental institutions and departments such as the EOBI, WWF, MoIT and so on. Since there is no reliable data on the individual contributors to the NSS, or the quantum of rural/urban saving, hence speculating about the success of NSS in generating truly new savings is a little doubtful.

The question arises that what are the problems of unfunded debt? The NSS has a number of unusual attributes. Given Pakistan’s chronic difficulties with public sector deficits, the NSS, which is a key instrument for financing this deficit, is of considerable budgetary importance. Little is known, however, about the impact of NSS upon the macro economy. In addition, the determinants of flows into the NSS are essentially unknown. Is there a positive interest elasticity of savings into NSS? Do flows into NSS reduce flows into other financial instruments? Is NSS, which pays subsidised interest rates to depositors, an effective means of financing the public sector? What is the tiered maturity profile? What is the average duration of the liabilities? What is the actual quantum of savings in the NSS system? And lastly but not least how are we going to meet the unfunded liabilities of the occurring exposure in the long run?

All of the above are important questions that need to be answered and we essentially have very opaque answers and in most cases, none. The answers that we have are also based on intuition and observations rather then empirical, hard, factual data. For example it is clear to anyone who has been in the finance industry that the NSS clearly inhibits the fixed income market at all levels. Be it the recent introduction of short-term papers or the longer term DSC’s, SSC’s they all directly compete with T-Bills and PIB’s; and can clearly be labelled substitutes of each other. Ever since institutions have been allowed to reinvest in the NSS schemes the trend and flows of institutions towards market-based fixed income products has decreased. Some clients like the Workers Welfare Fund have stopped investing in the fixed income market altogether and prefer the NSS outright.

Duration of Domestic Debt is estimated to stand at approximately 2 years excluding SBP Market Related Treasury Bills (MRTBs) as of the last economic survey. Duration including MRTBs was probably under 2 years. This estimate of duration may be misleading, but why? This is owing to the non-availability of actual maturity profile of NSS and manual operations of Central Directorate of National Savings (CDNS); although I have recently seen an advertisement a few months ago soliciting proposals for the computerisation of the directorate. In here, the unknown lays the crux of the problem. The behavioural study undertaken to estimate the maturity of NSS instruments mentioned in the most recent economic survey was not academically citied or available for review. So essentially what is happening is that the government is incurring lower annual debt servicing cost while completely ignoring portfolio risks and without truly knowing the quantum of it liabilities, or what the pattern of the asset liability mismatch is. The very basics of asset liability management (ALM) merit that you know what you owe and when. Prudent ALM management dictate that it is important for the government to take necessary measures to lengthen the maturity profile of domestic debt so that it is not servicing the current maturities with current inflows. Although increasing the duration may result in additional debt servicing cost in the short term, it would certainly help in reducing the associated liquidity and refinancing risks in the domestic debt portfolio that is prevalent from an analysis of the available data. Another problem of the manual operations of the CDNS is that there is probably comprehensive duplication in savings, especially in capped retirement and income products. Meaning that the same person has invested above the prescribed limit allowed in a specific scheme by just simply using a different saving center. Once these anomalies are properly accounted for the true quantum of unfunded debt outstanding will become clearer.

State and federal unfunded liabilities now total more than Rs 2064 billion and this amount is rising rapidly. Why should we care about it? Before we get into the detail of how unfunded liabilities are paid for and managed by our governments, we need to first understand what unfunded liabilities are. An unfunded liability is a debt that is not covered by the value of assets, savings or investments that have been allocated to pay the debt. What this means is that the Pakistan federal governments have a debt that is now more than Rs 2437 billion in overall size (inclusive of Rs 375 billion prize bonds that are shown separately in the SBP Statements), yet is not directly related to the budget debt that the governments are trying to overcome.

To understand where this debt hides, we need to take into account that governments compartmentalise the business of running the state or country into departments. Each department is responsible for managing a portion of the annual budget. Departments manage two types of debts that we’re interested in. The first is a cost that occurs during the year and is paid for out of the annual budget. There is no future impact of this cost. The second cost, on the other hand, does have a future impact. It is either identified or created during the year and will, eventually, become due. An example of this is when a department buys a fleet of cars and pays for the cars out of the annual budget. The department now has a fleet of cars that has been paid for, but what about future running costs? There are two options: either set aside funds now to pay for the future running costs, or pay for the future running costs out of future income. The first option appears to be prudent financial planning. By setting aside funds to pay for the future running costs there is no detrimental effect on future budgets and we can be sure that the funds needed to pay for the running costs will be available. If a government decides to pay using future income, then this cost will be factored into future budgets until such time as the cost ends – essentially, when the fleet of cars is retired and sold off. So how does this relate to the budget deficit we hear about and the unfunded debt/NSS? What the government does not want is for the public to become concerned about future liabilities, so they simply do not mention them when talking about the next budget. They get around these discussions by focusing more on current liabilities (what is due to be paid in the next budget only).This sleight of hand trick causes the unfunded liabilities to become lost in a haze of smoke and mirrors. By now, you might be wondering why you need to be worried about unfunded liabilities. If you aren’t, you should be. Every year, some portion of the unfunded liabilities becomes due for payment. And in the case of NSS we have no way to plan for this maturing of the liability as we don’t know that it is due till it is physically presented for encashment. If the government cannot pay, then the only option is to borrow to pay, which we are essentially doing by soliciting new NSS deposits. This adds to the overall size of the unfunded liabilities and means the cost of servicing this debt is constantly growing, as is the case with NSS. You might think that if the government returns to a surplus budget, this will reduce the size of the unfunded liabilities. This is possible, but not always the case and especially not the case in Pakistan, which has not seen a sufficiently surplus budget in a very long time.

Hidden costs

NSS related superannuation and saving benefits are by far the largest contributors to the current unfunded liability debt. Successive governments have maintained the system, albeit with some minor tweaking. What this means is that as the number of current savers grows, the size of the debt increases. If you add to this the years that the NSS schemes were not fully funded due to a deficit budget (and were therefore funded by borrowings), it becomes obvious why the size of the unfunded liabilities now exceeds the above stated figure by much more than Rs 2064 billion, and we don’t even have a proper idea because no one was keeping very accurate records.

Government by credit card

NSS related savings are of critical social benefits that savers contribute to throughout their working lives. Though you would think that the government would do everything possible to reduce this cost to the public purse, yet Pakistan now has more than Rs 2064 billion in unfunded liabilities. And this amount is growing rapidly. The annual cost of servicing this debt is now in billions. An unfunded liability is equivalent to a credit card in many ways and every Pakistani gets warned of the danger of over using a credit card. We’re often told that we are one of the poorest and indebted countries on the planet, so why are we using this credit card without impunity? I often wonder why our government has such a deep unfunded liabilities debt, and constantly do their best to hide this from the public eye.

A partial solution

The solution requires some innovative thinking, some sort of “Future Fund” should be set up and a fixed contributions should be made by the government into it. The fund should be given constitutional cover and the government should be unable to withdraw from it before a specific date in the future as determined by actuarial calculations, or till the fund grows large enough to assist the government in reducing the future detrimental effect of unfunded liabilities; to prevent future governments from accessing the Future Fund too early. Also once the funds are withdrawn they should be allowed to be used for anything but the specified purpose of paying the unfunded liabilities. The idea is to grow the Future Fund until it is larger than the unfunded liabilities. Then, theoretically, the Future Fund will be able to pay down the liabilities as they fall due. The one major problem is that until the Future Fund grows to be larger than the unfunded liabilities debt, there is the cost of servicing the debt itself, inflation and debt increase due to increases in the size of the current and future liabilities. Calculations on how much or unfunded liabilities are growing at and at what rate yearly and the contributions in the fund will have to be adjusted for these growth rates. What cannot happen is that the contributions into the fund are outpaced by the growth in liabilities. If we keep doing what we’re doing we could easily find ourselves in a bottomless pit of debt. We have to realise that Pakistan not a highly productive economy, our net foreign liabilities as a percentage of GDP are very high in the world, we have this persistent current account deficit. In short, all the entitlements (what little we have) are being funded at the pleasure of multilateral agencies and precarious finance. And there comes a time if we keep going this way, there will come a time when we will crash and burn under the burden of our own local currency debt, funded and unfunded. It has happened before in the world.