Lesetja Kganyago: South Africa's robust macroeconomic framework
Address by Mr Lesetja Kganyago, Governor of the South African Reserve Bank, at the New York Investment Conference, New York City, 5 October 2015.
South Africa has sailed a considerable distance over the past two decades. Perhaps the most critical macroeconomic markers over that distance were the efforts made in the late 1990s to establish fiscal credibility, the adoption of an inflation-targeting monetary framework, and the adjustment to a market-determined exchange rate. My good friend and colleague, the Minister of Finance, Nhlanhla Nene will address the first.
All three of these fundamental macroeconomic policy achievements are essential to understanding where we are now. And they are critical to thinking through why our policy stance and institutions will continue to support the economy through the adjustment needed in the face of major global forces beyond our control.
I would like to discuss the latter two achievements - monetary policy and currency flexibility - for these are central to the remit of the South African Reserve Bank but also because they are essential to keeping our particular ship afloat in present times.
For South Africa, the policy adjustment in response to the Asian, Argentinian and Russian crises of 1998-2001 marked a clear break from the past. We effectively gave up on the idea that a good way to support the economy was to closely manage the currency in an effort to hit multiple goals. Perhaps the most difficult part of this shift in thinking was the recognition that efforts to manage the currency really did impose major costs on the domestic economy. It was the acceptance of the Mundell trinity in the context of an ever more volatile global financial environment and awareness of massive built-up liabilities.
The outcome of this new thinking was the construction of a system which is sensitive to the dangers while welcoming the opportunities of global capital mobility. I believe they make us more resilient in the face of shocks, dramatically lowering the chances of a crisis. Moreover, they complement the diversity of the South African economy; currency adjustment in response to bad weather facilitates the moderation of imbalances in a way that encourages economic growth. It is not the growth we saw prior to the crisis - pumped up by high commodity prices, too rapid credit extension and excessive consumption. It is more patient growth, harder won, built on productivity growth and a greater focus on non-traditional (commodity) exports. And while a range of other emerging markets has adopted similar approaches to these common problems, the quality and credibility of our approach sets us apart.
In the wake of the Asian crisis
In developing robust macroeconomic structures, the crucial adaptation for South Africa was the shift to floating exchange rates. When we embarked on the era of policy reform and economic renewal in the 1990s, you could argue that the choices were difficult. You couldn't fix exchange rates, but unfixing them could be disastrous. As Paul Krugman wrote at the time, and I quote:
Nobody who looks at the terrible experiences of Mexico in 1995 or Thailand in 1997 can remain a cheerful advocate of exchange-rate flexibility. It seems that there is a double standard on these things: when a Western country lets its currency drop, the market in effect says "Good, that's over" and money flows in. But when a Mexico or a Thailand does the same, the market in effect says "Oh my God, they have no credibility" and launches a massive speculative attack.1
One of the accomplishments of emerging-market macroeconomic policy after 1998 has been burying that double standard. Emerging markets can build their own credibility, and this was best seen by the steady rise in capital flows into these countries through the 2000s and the expansion of investment analyst capacity to understand them better.
For South Africa, the period of integration with the world economy in the 1990s coincided with a steady fall in the price of major commodity exports. The rand was overvalued and needed to adjust downwards. It trended steadily weaker over the following decade, with periods of overshooting in 1996 and 1998. The South African Reserve Bank attempted to support the currency and failed. In the process, a large net open forward position in US dollars was created. This liability carried the potential for severe economic outcomes in the event of ongoing depreciation and meant that it had to go, but preferably as part of a new monetary framework. Inflation targeting, with its clearer policy anchor, implied dismantling the system of currency support and giving up on managing the exchange rate at all.
The initial test of the new framework was the sharp fall in the rand in 2001, in which intervention was avoided. The currency reached what remains an all-time low, in real terms, but then rebounded. Critical to this was the recognition that the rand's fall was part of a Dornbusch overshooting dynamic and that forecasts for inflation showed it moderating in the wake of the depreciation. This built credibility of the new monetary policy framework.
Over the following decade, the rand shifted its trajectory away from trend depreciation. Capital flows into South Africa, which had been quite weak and uneven through the 1990s, became much stronger and steadier. Large capital flows meant large current-account deficits, of course, but these were appropriate for a country with investment needs in excess of domestic savings.
The boom of the 2000s era came to an end with the global crash of 2009. The rand depreciated dramatically but then, over the next two years, recovered strongly. It was helped by stimulus in China, which boosted commodity prices, as well as by quantitative easing in the United States, which encouraged capital flows into emerging markets. These forces, however, could not be permanent. Our terms of trade peaked in 2011 and have since fallen by 8 per cent. Meanwhile, world monetary conditions began to tighten, starting with the "taper tantrum" of May 2013. In line with these developments, the rand followed a downward trend for about four years. As before, our policy was not to intervene.
Maintaining a flexible exchange rate provides the economy with a number of benefits. One is the insulation from price shocks. In South Africa, as with many commodity exporters, terms of trade have deteriorated. But the impact is cushioned because production costs are denominated in rands whereas export prices are denominated in dollars. Our dollar-price index for South Africa's commodities is down 41,0 per cent since 2011. The rand-price index has actually increased slightly, by 4,8 per cent.
And the advantages are not limited to commodities. Rand depreciation makes the tradeables sector more competitive, creating the right import and export incentives - which are especially important in the context of substantial current-account deficits. In the context of a struggling global economy, it is not very surprising that our own current-account adjustment has been slow, but we do expect it to be better than previously elevated levels.
A flexible exchange rate also insulates the economy from some shocks. If you don't need to target a fast-moving variable like an exchange rate, then policy doesn't need to react and transmit the shocks that hit the currency on to the real, domestic economy. As an illustration, let me remind you that the rand has depreciated by 16 per cent against the dollar so far this year. We have raised our policy rate by 100 basis points over an 18-month period. In 1998, before the adoption of inflation targeting, the rand also depreciated by 16 per cent and we hiked rates by 690 basis points in a period of six months. Lower pass-through from the depreciation to inflation has also helped to moderate potential policy responses.
Of course, floating exchange rates are dangerous if domestic borrowers have run up large, unhedged foreign-currency debts. Where this is the case, the monetary authority may end up facing only bad choices. One is to use foreign currency to prop up the domestic currency to protect the economy, but this has clear limits and typically fails. The alternative is to step back and let sectors of the economy adjust on their own to the new currency level, implying much higher financing costs. The clear differentiator for South Africa against many other emerging markets is that this particular policy dilemma doesn't feature in our decision-making process. Renouncing the "forward book" and allowing the "fear of floating" to wear off over time has limited the extent of foreign-currency borrowing, encouraging the development of our own capital markets and reducing a major risk for non-resident investors in rand assets.2
When a country adopts a floating exchange rate, the reasons for holding foreign reserves also change. The goal is no longer managing the value of the currency. Instead, reserves serve to bolster the national balance sheet, reassuring investors that the country has access to safe, liquid resources even in extreme situations. The use of reserves is therefore very different. The reserves provide insurance, a guarantee that markets will be able to price exchange rates. The fact that emerging markets have accumulated many more substantial reserves than they held, for instance, in the 1990s or early 2000s, suggests that markets will be able to function in more orderly ways, and that large, long-term investors will feel comfortable trading in those markets.
The exchange-rate policy fits within the broader policy of inflation targeting. Inflation targeting provided us with a policy goal more realistic than the previous eclectic approach to policy, because we could hit it. That goal was also more meaningful for citizens, because we could focus on protecting the buying power of regular consumers in domestic markets - as opposed to, say, trying to hold down the costs of holidays in New York. These approaches to policy have served us and others well for more than a decade. But just because we replaced a bad policy with another that returned better results doesn't mean that monetary policymaking has become easy or obvious. This is a challenging period. It is worth discussing our policy response.
The most straightforward decisions for an inflation-targeting central bank are those when inflationary pressures are coming from the demand side. In these circumstances, monetary policy acts to hold growth closer to its potential level. The primary mechanism is clear: by raising the costs of borrowing and rewarding saving (which is deferred consumption), central banks moderate the level of economic activity, returning it to a non-inflationary level.
However, South Africa has rarely enjoyed such straightforward conditions. Instead, inflationary pressures from the supply side have dominated. The standard accusation levelled at inflation-targeting monetary policy which responds to these shocks is that it is procyclical and therefore wrong.
But I think that this charge misses the subtlety of the inflation-targeting response, with its emphasis on flexibility and the primacy of second-round effects. Take the current period as an illustration.
In recent years, emerging markets have been buffeted by two kinds of shocks. The first was falling commodity prices, caused in particular, if not exclusively, by slowing and shifting growth patterns in the Chinese economy. The second was portfolio shifts, occasioned mainly by changes in US monetary policy. Both of these shocks have caused emerging-market currencies to depreciate. In South Africa, that depreciation has been marked. Since 2011, when our terms of trade peaked, the rand has fallen against the dollar and on a trade-weighted basis. This has been a persistent source of inflationary pressure in the South African economy.
Literature emphasises the importance of the origin of a shock. A terms-of-trade shock is likely to reduce net exports and thus aggregate demand. This lowers the output gap - offsetting the depreciation with some disinflation. By contrast, a sustained portfolio shock has little effect on aggregate demand and provides no offset to the inflationary consequences of the currency's fall.3
Central banks are supposed to respond to currency depreciations caused by portfolio shocks but not necessarily terms-of-trade shocks. However, as the Governor of the Central Bank of Chile, Rodrigo Vergara, recently reminded us, output gaps do not appear to explain much about the currently observed inflation, particularly in emerging markets.4 Currency depreciation, by contrast, is a clearer source of inflationary pressure.
So what are emerging-market central banks to do?
Our answer is to go after the so-called second-round effects, ensuring that inflation corresponds to the target over the medium term. First-round effects, either directly through the higher cost of imported goods or indirectly through those costs feeding into other goods and services, are relatively hard to control. Under flexible inflation targeting, we have some leeway to look through these shocks. But when wages and inflation expectations shift away from the inflation target, we are required to act. We are also inclined to do so pre-emptively, before expectations stray from the target, potentially requiring greater efforts to be shifted back downwards.
Our tools for addressing these forces are our communications strategy and the policy rate. These instruments affect inflation in several ways. By convincing price and wage setters of our commitment to the target, we prevent them from assuming higher levels of inflation in the future and from creating a self-fulfilling prophecy. Higher rates which lower the output gap put downward pressure on inflation. It is also possible that higher rates will appreciate the currency, although, as I have said, this is not a policy goal.
The extent of any policy response is proportional to the challenge. In South Africa, we have adjusted the repo rate by a cumulative 100 basis points since January 2014. In real terms, it remains close to zero, as it has since the crisis - a very low level from a historical perspective, and also quite accommodative relative to other countries. The adjustment has been smooth, without large increases or abrupt changes of direction. During this period, headline inflation has made temporary departures from the target range, but core has remained within the target range. Longer-term inflation expectations have stayed close to the top of the target, and we watch them very closely for signs of breaking away. If longer-term expectations were clustered at the midpoint of the target range, we could look beyond near-term volatility. But a small margin of safety has made this more difficult.
We expect our gradual tightening cycle to accomplish our primary mandate: controlling inflation as defined by our inflation target while showing due concern for growth.
Minister Nene has already spoken about the fiscal aspects of macroeconomic policy. I now turn to aggregate debt dynamics and in particular private-sector debt.
In the years before the Great Recession, South Africa's debt-to-GDP ratio declined markedly due to robust growth and booming revenue collection. Yet, despite the space created for fiscal stimulus and accommodative macroeconomic policy settings - and very high commodity prices - growth persistently disappointed, and has slowed steadily since 2011.
As the Bank for International Settlements and other observers have argued, emerging-market debt is a worrying fault line in the world economy.5 This is partly the old foreign-currency story, in which a stronger dollar and tighter monetary policy in the US render debt burdens unsustainable. It is also a debt-accumulation story, in which growth is accomplished through greater leverage. The concern is that, as the boom ends, it will become obvious that productivity growth has been stagnant and the returns on investment will be too low to service the debt incurred - widespread attempts to deleverage will then further weaken aggregate demand, setting off a vicious cycle. This dynamic can be exacerbated by the coincident slump in commodity prices. In the press, this scenario - of financial and commodity cycles moving in opposite directions simultaneously - is typically introduced with Warren Buffett's axiom that you only discover who's been swimming naked when the tide goes out.
I am happy to tell you that South Africa is wearing a swimming suit. In fact, the peak of our financial cycle came before the Great Recession. This is something we have in common with advanced economies - although our overall debt levels are closer to emerging-market averages. Since the crisis, private-sector borrowing appetites have been broadly muted.
Households have deleveraged gradually, reducing debt relative to incomes. When it comes to firms, as the International Monetary Fund has recently noted, South Africa is one of a few emerging markets where corporate debt stocks have declined since 2007.6 One lesson from this is that South Africa does not have a burgeoning private-sector debt problem which could soon be revealed as an important vulnerability. Another lesson is that episodes of rapid debt accumulation, even when they do not end in crisis, commonly give way to extended periods of weak growth. From a private-sector perspective, this is not a useful depiction of South African conditions, even if it is accurate for other emerging markets.
In short, there is little impeding corporate borrowing from a balance-sheet view in the South African economy, and households have made decent progress in reducing their own debt levels. This is likely to continue in a moderate way over the next year.
In the short run, there is a strong case for countries with macroeconomic space to use it in a countercyclical way, restoring growth. Many emerging markets have done this. Unfortunately, growth in emerging markets has also been slowing for five years, ever since the post-crisis rebound in 2011.
Permanent stimulus is not countercyclical. One of my greatest concerns is that we let the fear of slow growth scare us into a deeply uncomfortable macroeconomic position in which monetary policy accepts stagflation.
In South Africa, our inflation target is 3-6 per cent. We do not want to let that become a policy of 6 per cent plus whatever shocks appear. Inflation higher than that of our trading partners affects our competitiveness. It erodes incomes, especially those of the more vulnerable members of society. Over the medium term, extra inflation does nothing for growth.7 These are important lessons, too easily overlooked when the focus is on the latest data and the short run.
In the past, emerging-market crises tended to centre on currency crises. Prominent examples include the Tequila Crisis of 1994, when Mexico devalued the peso, and the Asian crisis of 1997/98. Given these experiences, it is not difficult to imagine the Fed tightening, generating further volatility and creating expectations of more emerging-market crises.8
But emerging markets are in better shape than they used to be, in particular because they have flexible exchange rates. South Africa, perhaps more than other emerging markets, has developed a robust track record of enabling the currency to carry the brunt of the adjustment to global volatility. Importantly, the low level of foreign-currency liabilities helps to reduce the risk of a sudden sharp crisis and cannot act as a complicating factor in policy decisions.
Currency adjustment helps the market to find its balance again but also, importantly, works in the direction of the unwinding of macroeconomic imbalances. In other words, in the current conditions of weakening commodity prices, currency adjustment works in line with fundamentals. This helps the real economy to find its own balance and rebound. Relative to other emerging economies, let alone the developed markets, our private and public debt levels remain low, a clear advantage as the global economic recovery strengthens further in coming years.
In South Africa, our task is to maintain our robust macroeconomic architecture, work through the structural reforms necessary for raising potential growth, and ensure that we remain at the forefront of emerging markets. Our macroeconomic policy framework remains sensitive to the challenges of global policy normalisation and flexible enough to take advantage of the few opportunities underlying the difficult global conditions of our times.
1 Paul Krugman, Latin America's swan song (http://web.mit.edu/krugman/www/swansong.html).
2 Guillermo Calvo and Carmen Reinhart, Fear of floating (http://www.nber.org/papers/w7993.pdf), November 2000.
3 Christopher Ragan, "The exchange rate and Canadian inflation targeting", Bank of Canada Working Paper 2005-34 (http://people.mcgill.ca/files/christopher.ragan/ER-FinalBankofCanadaWP.pdf), November 2005.
4 Rodrigo Vergara, Inflation dynamics in LATAM: a comparison with global trends and implications for monetary policy (https://www.kansascityfed.org/~/media/files/publicat/sympos/2015/econsymposium-vergara-remarks.pdf?la=en), August 2015.
5 See, for instance, the Bank for International Settlements quarterly review (http://www.bis.org/publ/qtrpdf/r_qt1412.pdf), December 2014, pp. 20-21.
6 International Monetary Fund, World Economic Outlook, "Chapter 3: corporate leverage in emerging markets: a concern?" (http://www.imf.org/external/pubs/ft/gfsr/2015/02/pdf/c3.pdf), September 2015.
7 As discussed in Andrew Haldane, How low can you go? (http://www.bankofengland.co.uk/publications/Pages/speeches/2015/840.aspx), 18 September 2015; see also Robert Shiller, Why do people dislike inflation? (http://www.nber.org/chapters/c8881.pdf), 1997.
8 See, for instance, Nelson D. Schwartz, "Watching the Fed, and remembering the Tequila Crisis", New York Times (http://www.nytimes.com/2015/09/19/business/economy/if-rates-rise-global-markets-fear-effects-of-any-aftershocks.html?_r=0), 18 September 2015.