By Jac Laubscher, 19 February 2014
The economic background to the national budget that will be delivered later this month is no exception and future budgets will, among other things, be severely affected by the recent decision of the US Federal Reserve to gradually taper its quantitative easing (QE) policy.
The risk emanating from an uncomfortably high current account deficit that needs to be financed repeatedly by capital inflows that are now at risk requires a tighter fiscal policy. Government dissaving, amounting to 4% of GDP in 2012 and declining slightly to 3,8% in the first three quarters of 2013, should be eliminated as a matter of urgency. The economically sound way to do this is by reducing current expenditure, specifically the public sector wage bill.
In addressing the feasibility of reducing public sector wage bills, Pierre Cahuc and Stephane Carcillo point out that periods of fiscal drift, defined by simultaneous increases in public sector wage bills and in budget deficits as a percentage of GDP, usually occur during booms and election years. Fiscal tightening, with public wage bills and budget deficits simultaneously declining, occurs more often during recessions. The South African experience is of course the exact opposite, with fiscal drift occurring during recessionary times and between elections, indicating how difficult it will be to correct the situation.
However, if the government does not want to be saddled indefinitely with this albatross around its neck, it has no choice but to act. To demonstrate its commitment to sound finance, instead of implementing a freeze on further expansion of the civil service as proposed in the 2013 Medium Term Budget Policy Statement (MTBPS), a freeze on filling all vacancies, existing and future, should be implemented. A similar freeze on all promotions should be announced.
Infrastructure projects should be completed expeditiously because of their high imported capital goods content, and because of their role in removing important logistics bottlenecks that are holding back exports. The longer it takes to complete an infrastructure project, the longer the time lag between incurring the expense of the investment and reaping the benefits thereof.
Government finances should be managed in a disciplined manner to protect South Africa’s sovereign credit rating to safeguard its attractiveness as a destination for international capital and to minimise its cost. More generally, the national budget should support the removal of structural impediments to higher growth as these are ultimately the decisive factor in attracting the right kind of capital, i.e. foreign direct investment.
The changed financial market conditions that will be brought about by the Fed’s decision will furthermore have important implications for the debt management dynamics faced by the National Treasury.
The end of QE is likely to affect financial markets via three channels, viz. reducing global liquidity and capital flows to relatively risky asset classes, including emerging markets, the normalisation of market volatility at a higher level, and an increase in long-term interest rates globally. The first two channels will require a one-off adjustment to the change in conditions, while the last factor (viz. higher interest rates) will have a more lasting effect.
As intended, the Fed’s security purchases caused long-term interest rates to decline sharply, with the US 10-year government bond yield falling from 4% at the start of QE to a nadir of 1,6% in Q3 2012 (currently 2,7%).
The question is to what level interest rates globally are likely to rise in the next few years as a result of the reversal of central bank accommodation and what that will mean for the cost of capital and in particular the cost of government debt. What is to be expected for US bond yields is of crucial importance as bonds globally tend to be priced off their US equivalent according to market conventions.
Although tapering has already caused an increase of approximately 100 basis points in interest rates at the long end of the yield curve, the short end remains well anchored and will only start moving when an increase in the Fed Funds rate becomes a certainty, which will probably not be before 2015. This will be the real game changer.
A useful rule of thumb is that the long-term interest rate should be roughly equal to the growth rate of the economy, whether in real or nominal terms. History shows that in practice the long-term interest rate has been somewhat higher than the growth rate, reflecting the desire of the current generation to improve their living standard by borrowing from the future. If we assume a potential growth rate of 2,5% for the US economy and that the Fed will be successful in stabilising inflation at its target level of 2%, the US 10-year yield should be expected to be slightly higher than 4,5%, say 5%.
This conclusion is supported by forward rates putting the US 10-year yield at 3,5% by the end of 2014 and 5% by end 2015, compared to its current level of 2,7%. In its 2014 Global Economic Prospects the World Bank estimates the average10-year yield for the G4 at 3,6% by mid-2016.
Bonds that are priced off the US yield curve will be expected to show similar increases, plus any change in risk premium. If volatility normalises as expected, with the VIX returning to an average level of approximately 20 – 25 index points compared with fluctuating between 10 and 20 in the past three years, it will have a knock-on effect on emerging markets via increased risk premiums.
The yield on the generic 10-year government bond yield has already increased by approximately 200 basis points to 8,5% since the markets took fright at the Fed’s suggestion in May 2013 that tapering of its bond purchase programme was drawing closer. This increase is not unexpected in view of the dramatic increase in the correlation between movements in the SA 10-year yield and the US 10-year yield since the start of the Fed’s quantitative easing programme in November 2008. Whereas 40% of movements in the SA yield could be explained by movements in the US yield in the five years up to October 2008, the percentage has increased to 86% in the period since then.
The IMF estimates that the 10-year yield for South Africa will increase by more than 150 basis points based on a combination of assumptions, including a 30% reduction in foreign holdings of local currency government debt and a 100 basis points increase in the US 10-year yield, which will take it close to 10% if not higher. This conclusion is supported by the increased correlation between SA and US bond yields, based on data for the past five years as referred to above.
Regardless of the level at which government bond yields settle in future, the message is clear – it will become substantially more expensive to finance the public sector borrowing requirement and the average cost of government debt will rise at a time when the budget is already under stress.
Government debt-service costs are estimated to amount to approximately R107 billion in the current fiscal year, i.e. 9,5% of main budget expenditure or 2,8% of GDP, compared with 8,5% of main budget expenditure and 2,3% of GDP in 2008/09. Debt-service costs are lamentably the fastest rising expenditure item, set to increase to 10,7% of main budget expenditure or 3,1% of GDP based on the assumptions in the 2013 MTBPS. The revisions required by events since the release of the MTBPS will push up these numbers further.
The National Treasury has wisely extended the average maturity of government’s debt portfolio from 97 months at the end of 2007 to 151 months in September 2013 (the last available number) for domestic debt and from 84 months to 101 months for foreign debt by issuing longer-term paper and conducting switching auctions. Although this will ease the pain of higher interest rates, it will not prevent higher debt costs from eventually becoming a reality. The projected decline in debt-service costs as a percentage of gross government debt from 6,8% in 2013/14 to 6,5% in 2016/17 may therefore not materialise.
Rand weakness will add to government’s woes as it increases the outstanding value of foreign denominated debt. South Africa has followed an extremely prudent policy over the years in keeping the foreign currency component of its debt low by issuing the minimum new debt required for establishing a benchmark for offshore borrowing by local borrowers. (National government foreign debt amounted to R140 billion or 4,1% of GDP at the end of September 2013.) The negative impact of rand weakness on South Africa’s debt dynamics could nevertheless hardly have come at a worst time.
The bottom line is that South Africa finds itself back in the situation it faced in the mid-1990s, viz. that it has to drastically reduce its debt and debt-servicing burden to create fiscal space for other policy initiatives to be affordable. Government’s commitment to align the budget with the National Development Plan will otherwise not be executable.
It therefore once again puts the primary fiscal balance in the spotlight. For the debt burden to decline, government needs to quickly move to a meaningful primary surplus and sustain it for long enough to return its debt burden to a manageable level. The path set out for the primary balance in the 2013 MTBPS is just not ambitious enough – it projects the primary balance to move into surplus only in 2016/17, and then only just (0,1% of GDP).
The easy option would be to increase taxes, but with a tax burden that has steadily increased in the past 20 years to 25% of GDP currently, an economy that is struggling to grow, and business confidence at a low, the options are limited. An increase of 1% to 2% of GDP in the tax burden will nevertheless not come as a surprise in view of the primary deficit being of a structural nature rather than cyclical, as indicated in the 2013 Budget Review.
References Cahuc, Pierre and Stephane Carcillo: “Can Public Sector wage Bills be Reduced?”, NBER Working Paper 17881. March 2012.