The contribution of capital flows to sustainable growth in emerging markets
Thank you for the honour of inviting me to give this lecture. Considering the people who have previously spoken at this forum, considering the stature of the audience here today, and also considering the legacy of Michel Camdessus, who led the International Monetary Fund (IMF) for so long, it is a great privilege for me to be here.
My subject today will be the role of capital flows in emerging market growth.
I want to start with a problem that has often bothered me. As the South African central bank Governor, I regularly meet with global investors to discuss economic conditions and policy settings in my country. The fundamental goal of these engagements is to encourage investment. Then I return from these meetings, and we have policy sessions where staff want to talk about the dangers of capital flows. But the investors I just met are the people who are responsible for the capital flowing. So, I wonder – which part of my time am I wasting? Do we want these capital flows or not?
This is a global discussion, and one that has evolved significantly over my time working in macroeconomic policy.
Twenty or 30 years ago, the mainstream view was that financial globalisation was good. Global markets could provide more financing, at lower rates, than countries could achieve by relying on their own resources. They would allow for better risk sharing, and they would create better incentives to get policy right. The standard policy recommendation was that controls on capital flows should therefore be liberalised, and where they were being applied, this was probably to cover for some other policy error.
Nowadays, the mainstream view has shifted. The IMF encourages policymakers to keep capital flow measures in their toolkits, both for pre-emptive purposes and to address capital flow surges. The guidance is nuanced, and there is still appreciation for the benefits of capital flows – but as Christine Lagarde put it a few years back, “[This] is not your grandmother’s IMF, and there has clearly been a big shift in the policy advice.
Outside of the IMF, attitudes to capital flows have been more bluntly critical. One part of this is unhappiness with spillovers from United States monetary policy, sometimes when the stance is loose, as in the ‘currency wars’ era after the global financial crisis, and sometimes when financial conditions tighten, as they did in the 2013 taper tantrum and as they have been doing during the current period. Another is the geopolitical tensions that have made ‘deglobalisation’ a buzzword. There have also been shifts in the realm of ideas, with even some mainstream economists condemning most forms of capital flows. For instance, in October 2022, Arvind Subramanian published an op-ed arguing that, “capitalism must be saved from its financial rentiers, and financial deglobalisation is a good place to start”.
Altogether, the reputation of capital flows is at a low ebb.
We should, nonetheless, respect the enormous opportunity presented by access to a global financial system. Indeed, taking a blue-sky approach, capital flows look much too small.
There are studies of optimal current account deficits for small economies, and they yield extraordinary estimates, for instance that it would be optimal to run annual current account deficits up to 60% of gross domestic product (GDP).
Relatedly, if you think about it, it is strange that interest rates in developing countries are not orders of magnitude above those in rich countries. Even in middle-income countries, capital stocks are typically less than a third of those in the United States, on a per capita basis. It would make sense to pay radically higher rates to attract more investment, which would then raise the productivity of labour. Yet real rates are not so far removed from advanced country levels: over the past two decades, real policy rates in rich countries have averaged about -1%, compared to just under +1%, on average, for middle-income countries.
Obviously, these observations are not policy recommendations. They do not pass reality checks. But they can help us approach the question of capital flows and financial integration with a more open mind.
Considering the empirical cases, critics of capital flows often point to the successes of Asian countries, most recently China. These fast-growing economies were typically capital exporters, despite starting off poorer and with smaller capital stocks than the economies in which they invested their surplus savings. These countries also suffered crises when they opened to financial flows in the 1990s, disrupting their remarkable development trajectories.
Where are the poster children for proponents of capital flows?
There are countries that have enviable growth records, and which have relied for many years on capital inflows. These include the United States, Canada, New Zealand, and perhaps the clearest case of all, Australia. As one study from the Reserve Bank of Australia pointed out, “Sizeable current account deficits have been recorded in Australia in almost every decade for at least 150 years”. These large and sustained capital inflows have allowed for a higher level of investment than could have been achieved with only local savings. And the success of the economic model is hard to dispute. Australia’s living standards have ranked among the highest in the world since the middle of the 19th century, and towards the start of the 20th century they were probably the highest of any country. It is highly unlikely that Australia would have performed better in the absence of capital flows.
Of course, we cannot read Australian economic history as one long vindication of free capital movements. Both the depression of the 1890s and the economic crisis of the 1920s had symptoms familiar from modern emerging market crises, including balance of payments pressures and foreign debt stress.
We also cannot say Australia has always been happy with large current account deficits. In the late 1970s and early 1980s, these deficits were a major concern for policymakers, especially because they were being driven by fiscal policy and were draining foreign exchange reserves.
However, significant current account deficits persisted even after the floating of the Australian dollar and fiscal consolidation. This gave rise to the so-called ‘consenting adults’ thesis, that current account deficits produced by private sector decisions could be optimal and sustainable, and policymakers would not have to worry about them.
Reflecting on other country experiences, the simple fact that deficits were privately contracted seems insufficient. We know private sector flows can be dangerous: clear recent examples are Spain and Ireland’s experiences during the euro area crisis. It therefore seems relevant that foreign investors in Australia were willing to accumulate claims denominated in Australian dollars, and that the Australian dollar floated.
Another crucial fact appears to be a bedrock of investor confidence, based on credible macroeconomic policies – including a reasonable degree of price stability – and a resilient financial system.
One also senses some deeper mechanism here, which ensured capital was channelled into productive assets that generated good returns. Of course, this reflects more than the resource endowment, as we could name many countries with ample natural resources which have not absorbed capital flows productively. There has also been something else going on, since at least the 19th century, that has made capital in Australia productive, whereas that same capital deployed elsewhere would have produced boom-bust cycles and default.
Part of this is a story about the quality of institutions as well as the human capital available and empowered to run them. Another theme is the development of local capital and financial markets, and their capacity to turn capital to productive purposes. As I will discuss below, these capacities intersect with policy choices, with capital flows supporting or weakening the productive potential of an economy.
On the whole, the Australian case teaches us that a country can absorb large capital flows over very long periods of time, and use these to support high levels of prosperity. This contrasts starkly with the Asian examples, where a range of countries likewise achieved impressive gains in living standards, but mostly did so without foreign capital.
Most of the world’s countries would be happy if they could be Asian tigers, and ecstatic if they could be Australia. But many of us have a long way to go. What attitude to capital flows would help us on our way?
In South Africa, we have long favoured the Australian option.
Given ample investment opportunities and limited domestic savings, growth and capital inflows have typically been correlated. In 1985, when the apartheid government was hit with sanctions, access to capital flows was largely cut off. Of course, this was not a developmental policy for South Africa; it was a punishment. When sanctions were lifted at the end of apartheid, we looked forward to restoring access to global financial markets. We also appreciated that the end of sanctions did not mean the taps were open. Foreign investors all loved South Africa, but they would not invest based on warm feelings. And there was going to be a limit on how much investment we could attract, even with good policies.
With a low domestic savings rate, if the public and private sectors were both borrowing heavily, this meant we were going to hit a balance of payments constraint. Specifically, we anticipated an unsustainable current account deficit, which would weaken the rand and drive up inflation. Interest rates would then have to rise to rebalance savings and investment, slowing growth. This was the core problem statement of the macroeconomic strategy adopted by the Mandela government in 1996. The goal was to attract more foreign savings, apply some fiscal discipline to improve the country’s investment profile and reduce government’s demands for savings, thereby permitting lower interest rates to allow more private sector investment.
In hindsight, I would say the strategy was mostly successful.
We experienced some of the downsides of openness to capital flows. These included a huge depreciation of the rand in 2001, which only loosely reflected fundamentals, as well as a phase of currency strength during the mid-2000s, which may have affected export competitiveness. The stronger currency and low interest rates also fed dramatic house price appreciation and risky mortgage growth ‒ a dynamic arrested by the 2008 crisis.
Despite those blemishes, it was still a success. Indeed, as time has gone by, it looks more and more like a golden age of South African macroeconomic policy. Living standards were rising and growth was outpacing the global average. Our investment rate rose from around 15% of GDP at the end of apartheid, to around 20% of GDP, even as the domestic savings rate remained low at about 15%. This naturally entailed significant net capital inflows, much of it through portfolio flows rather than foreign direct investment. These we de-risked, in large part, by committing to a free-floating currency and minimising foreign currency borrowing across the economy.
It seems extremely unlikely that we could have had better results by closing ourselves to global capital. What causes me great concern, by contrast, is what happened next.
The 15 years from 2009 onwards are almost a mirror image of the first 14 years of democracy. From 1994 onwards, we achieved steadily high investment and growth, interrupted by temporary setbacks; from 2009 onwards, we have had steadily lower investment and growth, punctuated by incomplete recoveries. The IMF’s projections have 2023 investment at 16% of GDP ‒ far below the ratio needed for adequate growth. And yet even this level of investment is posing a serious funding challenge because the domestic savings rate is just 13% of GDP ‒ the lowest level since at least 1980 ‒ and the investment case for external investors has weakened significantly. Growth is projected at a mere 0.1%.
This is where the capital flow story comes in.
For much of the decade after the global financial crisis to date, South Africa had ample access to foreign capital, helped by ultra-low interest rates in major economies. The average current account deficit, until the onset of COVID-19, was over 3% of GDP. The financing for this deficit mainly came from portfolio flows, as it did before the global financial crisis. However, the composition of investment for this period shifted markedly towards government debt, and away from private sector assets such as equities. During the boom of the 2000s, government and public corporations absorbed just 16% of portfolio flows. In the next decade, this rose to 78%.
Recalling the Australian case discussed earlier, what we see is that South Africa moved away from a ‘consenting adults’ arrangement, where a stable fiscal position and a current account deficit were driven by private sector decisions, to a classic twin-deficit situation.
This had three destabilising implications.
First, the volumes of money available to South Africa after the global financial crisis undermined good policymaking. In the 1990s and 2000s financial markets helped with fiscal discipline; in the 2010s they enabled excess. The underlying problems were homegrown, but the ready availability of foreign savings after the global financial crisis made it harder to win policy battles. Investor scrutiny is good for policy: it obliges everyone to double-check the figures and cut back on things you do not really need. But when you know the money is coming anyway, it becomes much harder to insist on policy rigour.
Second, these flows permitted the build-up of a large sovereign debt position. Debt, famously, is a troublesome form of financing because the lender shares relatively little risk with the borrower – unlike, say equity investments, where unsuccessful projects directly affect share prices and dividends. Sovereign debt is particularly problematic, because declining government creditworthiness also spills over to the credit profiles of firms and households. With time, it leads to higher taxes and lower public sector investment to accommodate higher interest payments. An unsustainable fiscal position can therefore become a drag on the whole economy.
Third, capital flows eroded potential growth. We often talk about the importance of institutions to growth, but debt can be used to weaken institutions by funding systems of patronage and corruption, driving out skilled and diligent public servants. Many private sector firms will also follow the money, redirecting their efforts from productive enterprise. Through these two channels, capital flows helped subvert the market incentives and competent bureaucracies that power modern economies. Our macro framework delivered resilience through a floating exchange rate, low foreign currency debt exposures, and careful regulation of the financial sector. But resilience is not enough; if you are going to absorb capital flows, you also need to get allocation right. Huge non-resident flows into public sector debt can actually make this more difficult.
Today, we face the consequences. With too much borrowing, not much domestic saving and limited non-resident appetite for our assets, interest rates must rise to restore balance. The alternative is an inflationary balance of payments problem, which is plainly against the South African Reserve Bank’s mandate. This does not make the Reserve Bank popular, but facts are facts. We have gotten ourselves back in the trap we escaped in the mid-1990s.
Reflecting on this whole experience, it is easy to sympathise with Daron Acemoglu’s argument in a recent Project Syndicate piece, that South Africa shows how capital flows, instead of promoting good government and development, can ‘facilitate’ a “hollowing out [of a] country’s economy and institutions…”.
Nonetheless, I do not think we should jump from diagnosing bad consequences to urging a prohibitionist approach to capital flows, and giving up on the benefits. I would much prefer a risk management approach. A source of inspiration here is the airline industry, where regulators and companies work together to fly as many planes as possible, as safely as possible, rather than responding to the occasional accident by adopting a zero-tolerance attitude to risk, which would sharply reduce the number of flights and blow up costs. I think this is the right way to approach capital flows.
A great strength of the IMF’s ‘institutional view’ is that it acknowledges the benefits of capital flows upfront and then moves on to risk control, using a toolkit of capital flow and macroprudential measures.
A shortcoming in using these tools, however, is their weakness where the problematic flows spill over into the public sector. And these cases are hardly outliers. Capital flows into sovereign debt have been a major source of crises since at least the 19th century. The ongoing African ‘funding squeeze’ is fundamentally about government borrowing. Nonetheless, the IMF’s 2022 review paper on capital flow measures says relatively little about fiscal policy. The word ‘fiscal’ appears only nine times in that document, compared to 119 instances of the word ‘bank’. The policy advice is simply that if fiscal policy is the problem, it should be adjusted.
Fair enough, but what if that does not happen? Do we have additional policy tools to manage these risks?
One option is to adjust the regulatory treatment for government bond holdings, for instance by obliging banks to hold capital against them instead of treating them as riskless. But this would not directly affect non-resident investment decisions.
A second tool is developing a proper sovereign bankruptcy procedure. This would give lenders stronger incentives to scrutinise borrowers. Where debts become unsustainable, countries would also have a better option than prolonged debt distress which delivers restructuring only after years of misery. Still, I have limited faith in the ability of lenders to exercise adequate caution in the boom phase of the cycle. And where government debt is an asset held throughout society, default is probably a cure worse than the disease.
A more benign tool is foreign exchange reserves. Standard accounts traditionally emphasise the role of reserves in meeting balance of payment needs, especially in the context of inflexible exchange rate arrangements. But foreign exchange reserves are arguably more important for risk management, especially now that floating exchange rates are normal practice. A new wave of research is now also making direct connections between foreign exchange reserves and sovereign debt vulnerability.
With floating exchange rates and reserves financed in local currency, negative shocks to the country generate positive valuation effects on foreign exchange reserves. Central banks can therefore accumulate reserves to hedge the public sector balance sheet against adverse outcomes, driven by factors that include unsustainable fiscal policies. Furthermore, central bank independence provides a technology for protecting these assets from spending demands.[33]
This reserve accumulation approach may well work better than trying to restrain surges with capital flow measures, in a general sense, with reserve growth during inflow phases and the option to release reserves during outflows.[34] And it is a particularly useful option where flows are going to government debt and regular capital flow measures are not viable.
In addition to these tools, we should consider our macro policy narratives. For a start, we need to rediscover the dangers of government borrowing. Responsible policymakers never forget that fiscal debt is risky. But the nature of policy discussions is that while many claims are valid, some points get more emphasis than others. In the past decade, one such point was that fiscal consolidation hurts growth and is therefore self-defeating. Another was that higher government debt levels were safer than previously thought. I have personally observed these claims justify sustained fiscal slippage in South Africa. If we had felt the urgency of debt sustainability more keenly, we would have had a wiser conversation. We need a more responsible set of narratives around fiscal risks.[35]
We also need to think more clearly about allocative efficiency. One of the strongest lessons I have learnt as a policymaker is that poor countries are poor not simply because they do not have money, but because they do not use money effectively. Too often, there is a tendency to look at a problem, cost out a solution and focus on raising the cash. Implementation is just a black box. But good policymaking starts with implementation and the financing need should reflect what can be used efficiently. Indeed, one might cast the volatile and often damaging history of capital flows as a conflict between budget constraints and capacity constraints. Capital flows provide spending power and can radically shift the budget envelope, but implementation capacity is stickier, and budgets can easily overshoot capacity.
This point is relevant, once again, in the global dialogue about climate change justice and the financing that should be directed from rich countries to poor ones. There is a strong focus on costing the climate change impact for poor countries and using those estimates to lobby for massive inflows. But we have seen many times that the sum of money is secondary to the quality of policies, the incentives they create and the capacity of the institutions available to invest funds. The capital flow sceptics and the climate justice activists should exchange notes.
Ladies and gentlemen, to conclude, I remain impressed by the power of global capital flows to support investment, reduce financing costs and accelerate convergence in developing economies – especially where domestic savings are below investment needs. Nonetheless, this is a force that is dangerous as well as useful and powerful. The South African case shows both sides of the coin: intelligent use of capital flows in one period, and abuse in the second.
For countries where investment opportunities exceed local savings rates, doing without capital flows means giving up on significant growth. It is not an attractive strategy. A better one is to welcome capital flows, control risks and nurture institutions that can deliver productive investment choices. That applies to climate finance, too.
We need to remain optimistic about capital flows and vigilant about the risks, rather than pessimistic about the flows and allergic to the risks, or naïve about the flows and blind to the risks. My hope is that when the next boom comes, we will have learnt lessons that make that boom as safe, as long and as large as possible.
Thank you.
