Circular debt and the economic eco-system

Reading a recent leader story in this paper about the government’s intention to issue $5 billion worth of T-Bills/PIB’s in lieu of the circular debt has prompted me to write this article. The following article will deal with the dynamics of how this debt should be issued, what are the ramifications, general economic outcomes of such an issuance and what can be done to alleviate the arising liquidity problems after issuance.

First and foremost the policy makers have to understand that the economy is like a very finely balanced eco-system and if you unbalance one component, it has wider implications for the whole system. Secondly they need to understand that a massive T-Bill/PIB issuance in itself will only buy them time so that their reforms can be implemented, take root and hopefully coupled with an economy that has started to grow, better recovery rates, lower line losses will give them the ability to pay back the amount. In no way shape or form will this issuance make the problems disappear, it will merely push them forward, allowing for some breathing room.

The circular debt T-bill/PIB issuance will have to be framed keeping the following in mind. A major challenge awaiting the new PML-N government at the very outset is to make fiscal adjustment of an estimated Rs 350 billion by reducing expenditure and increasing revenue in its first year of rule. Low interest rate and single-digit inflation provide a good platform to initiate economic momentum and release investment potential that had been blocked in the energy sector. The side-effects of the desired economic impetus are palatable as long as it is a high growth/high inflation combination, rather than the present low growth/low inflation one.

In the first year fiscal adjustment will have to comprise a reduction in energy sector expenditure by about one per cent of GDP (about Rs 240 billion), and another 0.5%-0.7% of GDP (about Rs 120-150 billion) in the form of additional tax revenue. The existing difference between the applicable electricity tariff and the tariff approved by the National Electric Power Regulatory Authority (NEPRA) stands at about approximately Rs 3.0 per unit, entailing a tariff differential subsidy from the budget of over Rs 300 billion. However, the recent tariff determinations given by the NEPRA for electricity has increased the difference by almost Rs 6 per unit, this works out to a financing gap to Rs 550 billion. Therefore, the average electricity tariff would have to be substantially increased to reduce the gap by about Rs 200-250 billion at the start of the next fiscal year. Another 2.5 per cent of GDP (almost Rs 550 billion) adjustment would have to come from the revenue side, including Rs 350 billion (1.5 per cent of GDP) worth of new taxes upfront in the coming fiscal year.

In this back-drop an outstanding amount of Rs 500 billion of circular debt (accumulated on the balance sheet of generation, refining, supply and auxiliary concerns and not banks this time) has to be managed, plus the Rs 200 billion that is pending in delayed fuel adjustment costs with NEPRA since the last 2-years, and lastly the Rs 120 billion previously issued circular debt bond that is maturing in the middle of the PML-(N) governments tenure. This is all quite a task for the incoming Finance Minister and his team.

WHY YOU CANNOT TURN TO OUTRIGHT PRINTING OF CURRENCY The initial idea that was being floated (including by previous SBP Governor’s and prominent businessmen like Mansha) was to conveniently print the outstanding amount, take the one-time deficit hit and move on. If it had been done earlier it would not have been a problem but now the problem is so big that printing money will have serious systemic ramifications. To give the readers an idea of the enormity of the problem let us look at some numbers. The GDP of Pakistan is approximately $230 billion, the outstanding circular debt is approximately $5.0 billion (Rs 500 billion), which works out to about 2% of the GDP, this number does not include the $1.2 (Rs 120 billion) already issued 5-yr bonds and the approximately $2 billion (Rs 200 billion) in delayed fuel adjustment costs that are pending with NEPRA and will have to figure in any settlement as they will contribute towards any electricity price adjustments. Thus a stitch in time would have saved nine but that is not the case now.

Keeping the above deficit number in mind we have to remember that Pakistan naturally runs about a 3%-4% structural budget deficit in the economy. If we print this money and disburse it without the matching revenue source, this will only add to the existing structural deficit. Pakistan is projected to run a deficit of approximately 7%-8% and adding another 2%-3% to it will only add to the economic woes not solve them. Plus if we enter into any new IMF program in the coming fiscal year we will be expected to keep our budget deficit under 5%, which will not be possible in the above case.

If the government decides to print the entire Rs 500 billion than the knock-on inflationary effects will be very hard to control if we take the quantum of the problem into consideration. Examining the problem from another number’s angle; our total M2 (as of the last SBP numbers published on, May 17th, 2013) stood at Rs 7.6 trillion. If we decide to print Rs 500 billion more that is tantamount to increasing this number by approximately 7%, keeping in mind that average nominal growth in money supply is around 13% an addition of 7% would take this number to 20% or above, which by any economic standard cannot be good. With CPI at the low level of around 5% right now the argument might be that the above is a tempting prospect, especially since the inflation levels are below the historic averages. The gapping hole though in this argument is that the core (non-food, non-energy) has remained sticky at above this number (the last reading was 8.7%, a new one is expected next week but it will only be marginally lower) and the core number is the one that is most important when making economic decisions.

If the government does decide to go down this route then their thinking is probably that inflation is a lag indicator and that on average it will take 6 to 8 months for the effects to become apparent. In the meantime what they are banking on is that the new government can initiate a sustained level of economic activity and as a consequence a sustained growth in revenue receipts, giving the government some space for fiscal manoeuvring. Although possible, this is a stretch especially keeping in mind the structural economic problems that the economy faces, which will not be easily overcome in the first 100 days.

WHAT HAPPENED LAST TIME AND WHY THIS TIME IS DIFFERENT? The idea of issuance of T-Bills/PIB’s arose from the fact that the government engaged in such a transaction once before. Last time the government issued Rs 391 billion (circular debt and commodity operation combined), split equally between T-Bills and PIB’s. Out of this amount approximately half of the outstanding that was issued as T-bills has matured and entered the banking system, with no great alleviation in the overall banking system liquidity drought. The last time was materially different because the build-up of circular debt was on the banking systems balance-sheets. This is because the government had borrowed money from the system in the shape of TFC’s that were issued to raise funds to help clear the systemic financial back-log in the power sector. When the government started to miss coupon payments on a regular basis the banking system was up in arms and the government was forced to swap the TFC’s for the more trusted instruments such as T-bills and PIB’s. Essentially last time there was a swap with no significant ramifications for market liquidity as the liquidity was already outside the system. The only major change needed were those made to the SBP rules that would allow for the issuance/re-tap of T-Bills/PIB’s with out an auction or tender. Most of the issued 5 year PIB’s were parked in the HTM portfolio so as to avoid IAS-39 related mark-to-market provisions. On the balance-sheet end the advances column was reduced and the investments column increased, this had a positive impact on the banking systems Advance to Deposit Ratio (ADR) which decreased and made it possible to do more lending, but which was not done due to systemic avoidance of lending by the banking sector as a whole, a dearth of customers and a preference for investing in government securities.

This time round the situation is different because the build-up of debt is not on the balance-sheet of the banks, rather the debt is on the balance-sheet of power producers against Central Power Purchase Agreement (CPPA), DISCOS, oil companies and others, the banking system is largely in the clear this time. What this means is that new funds will be raised through borrowing from the banking sector. This in turn will mean an outflow of approximately Rs 500 billion over time which will only add to the existing liquidity crunch in the system and in turn the outstanding OMO (presently at Rs 350 billion).

THE ACTUAL ISSUANCE AND STRUCTURES Now that it is pretty clear that there will be a T-Bill/PIB issuance. The question is that what will be the structure of any such issuance that is going to be done. As a precursor to any T-Bill/PIB issuance the government should cut the discount rate by at least 1.50%. This is as best a time as possible for a discount rate cut keeping in mind that Pakistan’s May inflation numbers measured by Consumer Price Index (CPI) which has come in at 5.13%. The slowdown in inflation in April to May 2013 echoes an easing of inflationary pressures and thus, gives SBP more room to stimulate the economy with another policy rate cut. This is as good a time as any to cut the rate although the grounds even by my standards and as earlier stated in this article are dubious at best. Keeping the government’s logic in mind and the fact that further down the road the pressures on the country’s balance of payments will increase as well as the ensuing impacts of increased government borrowings from the central bank will filtering down. Followed by fiscal measures of reducing subsidies through increasing energy prices, will only increase inflationary pressures in the months ahead especially around September. Thus, if the government wants to decrease interest rates further before entering an IMF program than this is as good a time as any.

The first discernable impact that a decreased discount rate will have is that the interest servicing cost for any new circular debt T-Bills/PIB’s issue, which will decrease substantially. Since these T-Bills/PIBs will be picked up in regular auctions such a mammoth increase will follow the rules of supply and demand and yields will be pushed upwards, so the discount rate cut will hence help nullify or limit any market exuberance for a steep move upwards in yields. Once the desired amounts are picked up the rates can be adjusted according to economic fundamentals and or prevailing realties ie the IMF programme.

Any issuance of Rs 500 billion circular debt related in either T-Bill or PIBs exclusively is probably not the best of ideas. If we were to just issue Rs 500 billion in PIBs this would increase the government’s permanent stock of debt by 25% and if we were just to issue T-Bills this would increase the stock of floating debt by approximately 11%. The total quarterly issue size of T-bills is approximately Rs 1.0 trillion, maybe slightly more and since the T-Bills for circular have to be issued in one go, issuing such a large amount in one quarter will not be feasible. The downside of only issuing T-bills is that although the overall amount of debt does not increase substantially, since the maximum horizon for T-bills is one year this might not be enough time for the changes in the power sector reforms to take hold and thus we would open ourselves to the possibility that the maturity and a bit more might be upon us, without the reforms kicking in, this would leave us at the square one all over again.

My recommendation would be for a tiered issuance spread between 6-months to 10 years, with the load of the issuance being between the 7-year or shorter tenors. My recommended break-up would be 35% 1-year T-Bills (Rs 175 Billion), 35% in 5-year PIBs (Rs 175 Billion) and 30% (Rs 150 billion) in the 7-year tenor. In the first tranche Rs 175-250 billion should be issued that will add approximately 0.75%-1.0% to the deficit, a manageable number and enough to alleviate the immediate power problem related to outstanding amounts. In other words this should be more than sufficient to get the wheel rolling in the right direction. Why I say that 0.75%-1.0% increase in the deficit will be manageable is because according to the latest State Bank numbers reported on in the second half of May, the borrowing for budgetary support rose to Rs 1.042tr during ten months and 10 days of the current fiscal. However, this is still less than the borrowing of Rs 1.076tr made last year during the same period. The expected fiscal deficit will be around 8pc at the end of the current fiscal year not much different from last years. Surprisingly, most of the important figures regarding the government borrowing are almost same of the previous fiscal year. The size of the budgetary borrowing is almost same with a fractional difference. The borrowing from State Bank rose to Rs 454bn; it was Rs 424bn during the same period last year. The growth rate of monetary expansion for this period was 9.86 percent while the growth rate was 9.06 percent last year. Even the government borrowing from the banking system was almost identical. The government has so far borrowed Rs 631bn from commercial banks; the amount was Rs 624bn last year. So since most of our monetary aggregates are the same as last year an addition of 0.75%-1.0% increase could be managed with out causing many knock-on domino effects.

The biggest difference was noted in the credit off take by the private sector which shrunk to record low this year in the 10 months. The private sector credit off take during this period was just Rs 92bn compared to Rs 235bn last year. This fresh borrowing for the circular debt T-Bill/PIB issuance will add to the liquidity constraint in the banking system and will contribute to the already prevalent “crowding out” effect. Thus managing the fallout from this increased borrowing in order to get the economic engine moving forward will be of paramount importance.

Any bond issued should have some sort of optionality built into them ie some make whole clause at a certain price or a floor that gives the government the right to buy the bonds back at any time at Par plus some pre-designated premium. This is important because hopefully when the economy picks up and interest rates are lower due to the humming economic activity the government will have the ability buy back the high yielding debt and maybe reissue at lower yields instead of permanently locking themselves in at a higher yield.

Once the initial issuance has been done and the government has made the necessary adjustments to prices then the remaining outstanding amount of circular debt bonds can be issued.

HOW TO MITIGATE THE ENSUING LIQUIDITY CRUNCH AND SPUR LENDING This will need more creative thinking than the SBP has displayed until now. Firstly the SBP will need to decrease the CRR reserve requirements for the PkR and USD deposits to immediately give the system a liquidity injection. The CRR on Rs has to be decreased to 1% from 5%; this will add Rs 24 billion to the system. The USD CRR has to be cut from 20% (15% Special CRR and 5% CRR) to 5%; this will add $750 million to the system. The SBP would then enter into Buy-Sell transaction’s to convert into equivalent rupees, Rs 75 billion approximately, making the total liquidity injection around Rs 100 billion. Since we are looking to issue approximately Rs 250 billion in the first round of issuance this new liquidity will be good enough to cover approximately 40% of the issuance.

Secondly the SBP needs to remove the money market corridor. What it does is that it effectively takes away the 7% minimum that we are paying the banks. When the banks are not doing their respective jobs as financial intermediaries properly then why should they get a minimum guaranteed return. Plus removal in the corridor will introduce some volatility into the money markets, once there is no floor and when the markets overestimate or underestimate liquidity and are penalised for it, this will allow the markets to become more self regulating and slowly the rates will sink to an equilibrium level below what the existing minimum range of the corridor is at right now. Since most of the liquidity is controlled between the 5 largest banks pleasant market equilibrium should be found without any major turmoil. Another positive outcome of eliminating the corridor is that it will also decrease the pressure on the OMO. When Banks will know that their funds can also go unutilised then they will avoid overbidding in the OMO.

Thirdly, the SBP will need to rethink its conventional view of discount rate. The discount rate will have to be viewed as a more dynamic policy tool. A dual discount rate system should be introduced. Dual discount rates are used in places like Germany. It is called the Lombard rate there and is the penalty rate above the official policy rate that private financial institutions can temporarily borrow at from the Deutsche Bundesbank by pledging approved securities. Now this situation needs to be replicated in our country but in reverse, meaning that since the GoP is the one locking up the banking sectors liquidity in illiquid securities they should facilitate the banks in getting that liquidity released through the SBP. Meaning that the banks should have the ability to borrow against these circular debt issued securities at up to 2% below the prevailing discount rate. In order to avoid the banks abusing the system the amount that they can borrow at the subsidised interest rate should be capped at a quarterly limit or should be directly proportional by some formula to the amount of new credit disbursed by the bank in the given quarter, which should be easily trackable by the SBP. This is akin to providing the banks a subsidised interest rate so that they may lend in order to spur economic activity and be minimally hindered by the tying up of liquidity in the circular debt issues. The concept is similar to the subsidised interest rates on the Export Refinance Facility (ERF) available to exporters, but in this case the end beneficiaries of this subsidised interest rate would be local PkR credit consumers.

As a second phase to all the above once the governments start to get fiscal house in order it needs to introduce a Fixed Income desk at the SBP. The primary function of this desk would be to increase quantitative easing by providing more liquidity in the market by engaging in outright purchases from the market. All these expansionary tactics if done in a controlled, tightly monitored and targeted operation would put downward pressures on the interest rates. An excess availability of liquidity for banks to lend would eventually put exponential pressure on the KIBOR to decrease. All this is similar to what Central Banks all over the developed world are doing in order to support growth.

As a next logical step to this the SBP should also follow the above mentioned activity introducing SLR caps, similar to the ones we have on PIB holding’s (presently at 10% of Demand and Time Liabilities, commonly referred to as DTL). As of this moment on average banks have invested between 40%-50% of their balance sheets in SLR eligible securities, this is almost twice the required amount. The SBP should introduce an SLR cap and the new SBP desk would help in the orderly exit from the excess security position. A cap would also force the banks to re-evaluate their roles as financial intermediaries and since they would have excess liquidity that they would need to deploy, they would hopefully be forced to lend again.

CONCLUSION In the end it is my conclusion that whatever the solution pursued the SBP will need to expand its balance sheet temporarily in order to relieve the circular debt and stoke economic activity, the latter being part of its stated mandate. It is important to note that all these measures will be temporary and not a permanent solution to our lingering structural problems. Structural adjustments to electricity prices will have to be made, losses at public and state enterprises will have to be limited, aggressive belt tightening will have to be pursued and privatisation’s undertaken. People will be more willing to take swift pain in the first year of the PML (N) government, while the incompetency of the previous government are still fresh in their minds, rather than further down the road. If permanent changes are not made to the economic fundamentals of the county, permanent new sources of revenue not utilised and gapping balance sheet holes not plugged all of the above measures if undertaken will only worsen our position and we will be not at 0 but -1.