June Issue 2019

Asif Ali Qureshi, CFA, is the Executive Chairman of Optimus Capital Management.

One of the downsides of having a perennially weak economy is that it is open to almost every predicament. Just when a data point starts to look ominous, a simple tracing back of its historical series will reveal a time period when the situation was more or less the same, if not worse. This is true of Pakistan’s current level of indebtedness – though there is more to it than meets the eye.

As per the latest available data, Pakistan’s gross government debt (comprising domestic public debt of PKR 18.2 trillion and external government debt of USD 68.4 billion) had reached PKR 27. 8 trillion by the end of March 2019, equivalent to more than 74 per cent of the GDP. Depending on which GDP series you choose as the denominator (the current base GDP data is from the International Monetary Fund), the present level of the government’s indebtedness (debt/GDP) is the highest since June 2002 (reported GDP base), or since June-2001 (current GDP base).

As evident from the debt/GDP trajectory based on the reported GDP values for the respective years, Pakistan’s indebtedness was even higher in the 1990s than it is today. However, this should not offer any consolation, as the country later defaulted on its external debt and had to undergo multiple rounds of debt restructuring.

The current discourse on Pakistan’s debt challenge is focused more on form than substance. For instance, the two most debated issues seem to be the government’s borrowing from the State Bank of Pakistan (SBP) and the maturity profile of domestic debt, whereas the real issues are the ‘size’ of the government debt and its ‘sustainability.’ The IMF’s prescribed remedy to these professed twin ills (i.e borrowing from the SBP and the maturity profile of debt) which is, sadly, gaining traction with the local authorities, is to push the interest rates higher and higher. Before discussing why accentuating the interest rate tightening cycle would cause more harm than benefit, it is important to review the factors that led to the current situation.

Firstly, the combination of muted GDP growth and elevated budget deficit over the past 10 years lifted the government debt/GDP ratio to 74 per cent in March 2019, from 57 per cent in June 2008. GDP growth has averaged only 3.7 per cent, while fiscal deficit has been in excess of six per cent during the FY2009-19. The reported fiscal deficit and the public debt figures exclude the off-balance sheet expenditures/debt related to the power sector and commodity financing. The aggregate amount of commodity financing and public sector enterprise (PSE) domestic debt was over PKR 2.0 trillion in March 2019 – about 5.5 per cent of the GDP – while the external PSE debt amounted to another USD 3.5 billion, or 1.3 per cent of the GDP.

Government Debt/GDP (left) and Total Government Debt (PKR Trillion) (right).

Secondly, the declining share of external borrowing in the total government debt, from an average of 50 per cent till June 2008, to around 35 per cent in March 2019, has increased the burden of budgetary financing on the domestic markets. The size of the public sector debt dwarfs the domestic banking deposits. To put it in perspective, the total domestic public sector debt (public debt + commodity financing + PSE debt) amounted to PKR 20.2 trillion in March 2019, which was equivalent to 1.5 times the size of domestic banking deposits and 1.2 times the size of domestic banking deposits, plus the National Savings Schemes (NSS). In June 2008, the domestic public sector borrowing was only 0.7 times the size of banking deposits plus NSS. With such a massive build-up of domestic public sector borrowing, it is unsurprising that the government’s borrowing from the SBP has been increasing.

Thirdly, the net foreign assets (NFA) of the banking system have undergone massive contraction over the past three years, owing to the snowballing external account deficit. The NFA of the banking system shrank from a little over PKR 1.0 trillion, or eight per cent of M2 in June 2016, to (negative) PKR 822 billion, or (minus) 5 per cent of M2, in March 2019 (M2 is a measure of the money supply – including cash, checking and savings deposits, money market securities, mutual funds and time deposits). As NFA contraction reduces domestic liquidity, it has put further pressure of government debt financing on the domestic market.

Share of External Borrowing in Govt. Debt (left) and Government Borrowing from SBP (right).

Fourthly, the levy of tax on banking transactions of non-filers since July-2015, and the recent crackdown against fake accounts, has resulted in a massive surge in the currency in circulation (CiC). The CiC/M2 ratio, which was relatively steady at around 22 per cent till June 2015, jumped to over 28 per cent by March 2019. This translates into more than PKR 1.0 trillion in potential banking deposits/liquidity that would have otherwise helped reduce the government’s borrowing from the SBP.

Last but not least, the sharp rise in interest rates since mid-2018 (a 625 basis point increase in policy rate within 12 months, to 12.25 per cent in May 2019), coupled with the uncertainty about the levels where they would peak, explains the banks’ behaviour of limiting their participation in government securities to the shorter-end of the yield curve and even hoarding liquidity closer to bimonthly monetary policy announcements. This behaviour has been reinforced by the SBP itself, which has extended a virtual put option to the banks by religiously mopping up all liquidity at a repo (repurchase agreement) rate close to the policy rate.

Currency in Circulation/Money Supply (left) and Public Sector Debt/Banking Deposits & NSS (right).

As with all IMF programmes, reducing the current stock of the government’s borrowing from the SBP seems to be a key prior action/quantitative target under the recently negotiated USD six billion Extended Fund Facility (EFF). The IMF’s prescription for reducing the government’s borrowing from the SBP, is to hike interest rates to such levels where the ‘markets’ meet the government’s borrowing needs. The SBP has been quick to sign up to this philosophy. The Monetary Policy Committee (MPC) noted in the MPS of May 2019, that the increased government borrowing from the SBP “reflects a shift away from commercial banks which were reluctant to lend to the government at prevailing rates.” This was a surprising statement, considering that the increased government borrowing from the SBP was the result of a combination of factors discussed earlier.

In the face of high government debt/GDP, elevated fiscal deficit and anaemic GDP growth, pushing the interest rates – especially the long-term bond yields – to historic peaks works against the objectives of fiscal consolidation and the revival of growth, both of which are needed to reduce the government’s indebtedness. The debt servicing/GDP is set to cross five per cent in FY2019 – the highest in the last 17 years, and will climb further in FY2020. Moreover, high borrowing costs will deter private investment, thereby keeping the country trapped in a low-growth cycle for much longer.

The authorities must realise that without ‘expansionary fiscal consolidation,’ public indebtedness cannot be reduced. This requires changing the engine of growth from the public sector to the private sector. Keeping interest rates higher for longer only works against this objective.

Govt. borrowing from SBP (2013-16) (left) and Govt. borrowing from SBP (2001-04) (right).

The idea of letting the markets determine the interest rates on government securities, is flawed at multiple levels. Firstly, the domestic fixed income sector is a far cry from an efficient market. It is dominated by a handful of large banks. There are, also, serious awareness and access issues with the general public’s investment in treasury bills (T-bills)/Pakistan Investment Bonds (PIBs). Moreover, the NSS itself is the biggest distortion in market-based pricing. Secondly, with one-way (inward) capital account convertibility, the local banking system surplus liquidity remains invested in government securities in some shape or form. For instance, even in an open market operation (OMO) mop-up, banks enter into repo transactions with the SBP, against T-bills. Thirdly, it is cheaper for the government to borrow in USD from the international markets than to attract money into rupee-fixed income securities. The global liquidity pool for USD securities is much bigger than that available for investment in local currency securities of a developing country. Moreover, taxation and repatriation issues further increase the costs for foreign investors.

There are four main ways in which the government’s borrowing from SBP could be reduced:

A. The banks shift any surplus liquidity that they have parked with the SBP (OMO mop-up), into the government securities.

B. The SBP provides additional liquidity to banks via OMO injections and the banks, in turn, invest it into the government securities.

C. A part of the currency in circulation (CiC) flows back into bank deposits, providing additional liquidity to the banks for investment into the government securities.

D. The NFA expansion, resulting from either the government’s direct external borrowing, or from other net FX inflows, can help lower the government’s borrowing from the SBP. The government’s direct external borrowing reduces its recourse to the SBP, while other FX inflows generate additional liquidity, which is, in turn, invested in the government securities.

Of the four ways of reducing the government’s SBP borrowing listed above, the first two (A and B) do not require the jacking up of interest rates. In a stable or declining interest rate environment, the banks not only invest their entire surplus liquidity in government securities, but even borrow from the SBP (OMO injection) to invest more. This happened during the IMF programme periods of 2013-2016 and 2001-2004. However, the amount of the government’s SBP borrowing that can be reduced by shifting the surplus banking liquidity parked in OMO mop-ups and additional liquidity injections into government securities, is limited, due to the SBP’s net domestic assets (NDA) targets and banks’ internal limits.

The increase in CiC is a structural issue and cannot be reduced without increased documentation of the economy. The newly unveiled amnesty scheme for assets declaration may help in bringing some of the CiC into the banking system.

NFA expansion is the best approach to reducing the government’s borrowing from the SBP. This could be achieved by a combination of long-term external borrowings by the government, incentivising the private sector to raise external financing by offering forward cover, and encouraging FDI/FPI inflows.

FY2002-07 was the only period during which Pakistan achieved significant reduction in government indebtedness, from over 81 per cent in June 2001, to 52 per cent in June 2007. This was the period of high GDP growth –  running at six per cent –  a low fiscal deficit of around 3.6 per cent and low interest rates, with a six-month T-bill rate averaging 6.8 per cent. The authorities must realise that without ‘expansionary fiscal consolidation,’ public indebtedness cannot be reduced. This requires changing the engine of growth from the public sector to the private sector. Keeping interest rates higher for longer only works against this objective.