4    Macroeconomic policy within cycles of international capital flows

The Indian experience

Sukanya Bose

Cycles in international capital flows

Emerging market economies have witnessed sharp swings in capital flows in recent years. From a high of nearly $1.3 trillion in 2007, net private capital flows to these countries dropped to roughly $600 billion in 2008. The recovery thereafter has been short-lived, with another downturn marked by financial turmoil, debt deleveraging and economic slowdown in the eurozone. Taking a slightly longer view, international capital flows to emerging-market economies (EMEs) seem to have entered their fourth major cycle in the past fifteen years (1996–2011), with the attendant features of boom, bust and then recovery. The first dip corresponded to the Asian crisis, when there was a mass exodus of private capital from Thailand and then other East Asian economies, bringing in its wake financial contagion, collapse of stock prices and currency values and huge increases in private indebtedness across economies. This was soon followed by many more EMEs – Brazil, Russia, Turkey, Argentina and other Latin American countries – falling into financial crisis and recession. This marked the second cycle of international capital flows, which hit its worst point in 2002. International capital’s ‘flight to safety’ from emerging markets in 2008, due to a sharp decline in global investors’ risk appetite in the aftermath of the collapse of financial institutions in the United States, marked the third major dip. With the latest forecast of $700 billion of net private capital flows for 2012, we are yet to reach the end of the fourth cycle. The centre of trouble for the last two cycles in capital flows is, notably, the core metropolitan centres rather than the periphery.

While the exact details of each of the cycles of capital flows vary, the endogenous unstable dynamics of the financial markets provide powerful clues to make sense of the boom–bust cycles driven by international capital flows. Palma (1998), in his analysis of the 1982 debt crisis, the 1994 Mexican crisis, the 1997 East Asian crisis and the Brazilian crisis in the late 1990s, found that they share certain common characteristics. ‘Over-lending’ and ‘over-borrowing’ are basically endogenous market failures of over-liquid and under-regulated financial markets. From the Keynes–Minsky perspective of financial instability, periods of deep recession associated with financial fragility are an outcome of financial excesses in the preceding booms. With rapid capital-account liberalization, procyclical international capital flows have increasingly dominated economic cycles in emerging markets. Given the herd behaviour intrinsic in modern financial markets, these tend to produce a cumulative process of credit expansion, asset-price bubbles and over-indebtedness which, in turn, adds to spending and growth. However, as balance sheets adopt smaller margins of safety, the system develops endogenous fragility, and financing positions are increasingly translated from hedge to speculative and, eventually, to Ponzi finance (Minsky 1984). With a cyclical downturn in economic activity and/or an increased cost of borrowing, incomes on assets acquired can no longer service the debt incurred, setting off a process of debt deflation and deepening the contraction in economic activity.¹ Because of extensive dollarization and in balance sheets and the substantial presence of foreigners in domestic asset markets, domestic financial distress combines with exchange-rate volatility in EMEs. Thus, boom–bust cycles in international capital flows determine boom–bust cycles in assets, credits and the foreign exchange market for emerging markets. Real economic activity is increasingly shaped by developments in the sphere of finance (Sen 2003).

This essential nature of international capital flows is important to bear in mind in a discussion of macroeconomic policy issues in the present world context. A country integrated through movements of international capital, as India now is, has to face and manage situations of incessant flows and then sudden disruptions driven by factors beyond the control of the recipient countries. It has to contend with herd behaviour and a substantial presence of foreigners in domestic asset markets. What limits does it set for macroeconomic policy? This question is most commonly answered using the Mundell–Fleming model. This paper holds up to the light the meaningfulness of standard economic theory in informing the debate on macroeconomic policy in an EME. It uses the canvas of the present policy debates in India to discuss first, what is happening in macroeconomic policy and the open economy dimensions; and second, the appropriate analytical framework within which to raise questions on macroeconomic policy.

The paper is organized as follows. In the next section, the present external situation of the Indian economy is flagged by reviewing a few important trends. Next, we pick up the ongoing policy debate in India regarding managed versus flexible exchange rates and the theoretical arguments in support of the position, which derive essentially from the Mundell–Fleming model. The Mundell–Fleming model is then examined for the reality of its assumptions and the mechanisms it implies. The review is limited to showing that many of the model’s important results do not hold in more realistic contexts of imperfect capital mobility, elastic exchange rate expectations and exogenous domestic interest rates (endogenous money). The main part of the paper is devoted to an enquiry of how the macro-relationships implied in standard economic theory hold up to the empirical reality of the Indian economy. We conclude by arguing that the manner in which the capital account regime bears on the scope and effectiveness of macro policy departs substantially from its understanding in standard economic theory; rather, the heterodox perspective on financial instability can add important insights into the way the economy functions in today’s context of boom–bust cycles in international capital flows.

Capital flows and external balances of India

Capital-account liberalization (CAL) in India began soon after 1991, when market reforms of the Indian economy were initiated following the balance-of-payments crisis and India acquiesced to the IMF–World Bank structural adjustment conditionalities. Almost immediately, the regulatory framework for foreign investment inflows was significantly liberalized. Over the years, the freedom to convert local financial assets into foreign financial assets and vice-versa increased, with first inflows and then outflows being liberalized. Tarapore committees on capital-account convertibility in 1997 and 2006 set preconditions on fiscal deficit and inflation targets, current-account deficit and debt-servicing targets and financial soundness indicators, such as NPAs of banks, to move into fuller capital-account convertibility. The underlying logic was that macro health (along with the development of a suitable institutional framework) can ensure financial stability with an open capital account. Endogenous instability of financial markets did not receive much consideration in these important policy proposals.

Figure 4.1 shows that the net private capital inflows to the Indian economy grew from USD 13 billion in 1996 to USD 102 billion in 2007 – an eightfold increase – before falling steeply as part of boom–bust cycles in international capital flows. As a proportion of the net inflows to EMEs, India’s share rose from 3.7 per cent to 8.7 per cent in the corresponding period. There is greater co-movement of the trends visible across the years. When the period 1996–2012 is divided into two halves, the correlation of net capital flows for India and total capital flows to the EMEs increases from +0.52 (1996–2003) to +0.81 (2004–2012). As India approaches fuller capital-account convertibility and integration through international capital flow rises, the economy is subject to the instabilities inherent in these flows.

Table 4.1 presents the summary BOP statement for India during three phases: first, the decade following the 1991 BOP crisis; second, the period of high capital inflows, 2003–4 to 2007–8; and third, the period of global financial crisis and beyond. The capital-account balance shows volatility typical of international c apital flows. Between the first and second phases, there was a substantial increase in capital-account balance as a proportion of GDP. This more than twofold increase, despite a high GDP growth rate, implies a very high rate of growth of capital inflows during the second phase. This was the period in which the Central Bank had to contend with excessive capital flows, and a significant part of the capital inflows were absorbed through increases in reserve assets with the Central Bank, which rose from 1.6 to 4.6 per cent of GDP. In the last phase, as the capital inflows thinned, surplus on the capital account fell substantially. On the other hand, the current-account deficit – a key concern for developing countries – jumped to levels reminiscent of the crisis in the early 1990s. For 2008–9 to 2010–11, the average CAB/GDP was –2.6 per cent, while the surplus on the capital account was 2.7 per cent of GDP. The overall balance therefore stood at the tipping point, 0.1 per cent. Thus, as noted by Chandrashekar and Ghosh (2012), India’s greater global integration has created patterns of dependence on international markets and international capital. This makes the economy significantly more vulnerable, especially because the growth has been reliant on capital inflows to generate domestic credit-driven bubbles, rather than trade surpluses.

The net international investment position (net IIP) substantiates the growing vulnerability in recent years. The difference between an economy’s external financial assets and liabilities is the economy’s net IIP.

In March 2011, reserve assets of around USD 300 billion constituted the bulk of total external financial assets (USD 440 billion in all), the other important component being foreign direct investment (USD 100 billion). External financial liabilities exceeded external financial assets by USD 211 billion, the net IIP (see Figure 4.2). External financial liabilities comprised direct investment (USD 211 billion), portfolio investment (USD 174 billion) and loans (USD 145 billion). Not only has the gap between external liabilities and assets been increasing, but also, a very substantial part of the liabilities comprises portfolio investment and loans that are recognized as the most volatile and unstable of the foreign capital flows. In recent months, with meagre export growth, weak portfolio capital inflows and increasing pressure on the rupee to depreciate, corporations have been urged to raise loans in foreign currency to cover the current-account deficit.² Debt-creating liabilities have dominated non-debt creating liabilities for the first time in many years (52:48 in December 2011).³ Also, an unprecedented portion of the external debt is short-term in nature. The present external-sector position, thus, is far from comfortable.

External balances and the macroeconomic policy

The exchange-rate regime in India since the early 1990s has exhibited features of managed float, with the Reserve Bank of India leaning with the wind to correct currency overvaluation, while leaning against it to prevent misalignment of the exchange rate from an equilibrium rate that was generally measured vis-à-vis the real effective exchange rate (REER), and hence expected to track the current-account balance. An intermediate exchange-rate regime of this kind, one felt, would contain the problems of currency misalignment and costly adjustment typical of hard pegs, and minimize the excessive volatility in exchange rates typical of flexible exchange rate systems. Thus, Patnaik and Shah (2010) note that although the rupee began its life as a ‘market-determined exchange rate’ in March 1993, this date is not identified as a structural break by the analysis of the data. A single sub-period of the exchange-rate regime is found to run from 1976 till 1998. The flexibility is even more limited in the period up to 2004, i.e. the period following the Asian financial crisis.

The intermediate exchange regime has come under consistent attack from mainstream economists who have argued that given free capital flows, intervention in the exchange-rate market is compromising monetary-policy independence. Asia as a whole has been typified as having excessive resistance to exchange-rate appreciation and running the risk of ‘exchange-rate dominance’, i.e. the danger that exchange-rate policy will come to dominate or pre-empt monetary policy (Hannoun 2007). In the context of rising capital flows since 2003, an increasing concern has thus been the risk of inflation consequent upon the abundant liquidity created due to exchange-market intervention. Whether one looks at the policy interest rates or the monetary aggregates as the benchmark of monetary policy, these economists argue that the management of the exchange rate has forced monetary accommodation on the Central Bank (see Patnaik and Shah 2010; Rosenberg 2010). Even when the intervention is sterilized, this sterilization remains imperfect and is conducted in ways that might not raise interest rates, i.e. through increased reliance on administered liquidity absorption tools such as reserve requirements. Large and prolonged interventions and imperfect sterilization have meant an accommodative monetary policy. A floating exchange rate is the only means to restore monetary-policy independence, according to this view.

The theoretical paradigm underlying this policy advice is the Mundell–Fleming (M–F) model (Kohli 2010). For an open economy with perfect capital mobility, monetary policy is a powerful stabilization tool when the exchange rate floats but ineffective when the exchange rate is fixed; the Central Bank cannot indefinitely control both the nominal exchange rate and monetary policy. The ‘impossible trinity’, as this result is known, has become the workhorse of open-economy macroeconomic policy. A consistent macroeconomic policy with a freely floating exchange rate and free capital flows is all that is required to ensure an automatic mechanism for external and internal balance. This automatic mechanism works as follows. An expansionary monetary policy, by causing the domestic interest rate to fall below the international rate, results in capital outflows and depreciation of the nominal exchange rate (at a fixed price level). Depreciation helps export demand. Because of the expansion in exports, aggregate demand expands more than in the closed-economy case. Consequently, the higher demand for money increases domestic interest rates when the money supply is exogenous. The domestic interest rate then rises towards the international level. This is also the mechanism behind the idea that fiscal policy is impotent, as a fiscal expansion would cause exchange-rate appreciation and reduction of net exports until output, the demand for money and the interest rate return to their initial levels. The following discussion raises some of the important problems with the model – a model that has dominated academic thinking and influenced policy across decades.

We have seen that capital flows are subject to sharp swings and international capital flows dominate the economic cycles in EMEs. On the other hand, the basic assumption of the M–F model is that short-run mobility of international capital is both free and perfect and the economy faces an infinitely elastic supply of foreign exchange at any interest rate slightly higher than the international rate plus some country risk spread. Serrano and Suma (2011: 4) argue that it is not realistic to assume that the international interest rate itself is ‘independent of the size of its external financing operations’, or that any current-account deficit of any magnitude can always be financed at a constant interest rate. Imperfection in international capital markets means that the assets of one country are not perfect substitutes for assets of others, and agents require a growing interest premium to retain them in larger quantities in their portfolios. Market imperfections and asymmetric information will lead to increased loan risks when a country increases its foreign currency liabilities. Further, beyond a certain limit foreign credit rationing is bound to occur, and the country will not be able to obtain more foreign currency, regardless of how high the domestic interest rate is. In other words, the BP curve, in addition to being positively sloped, becomes vertical at a certain point (Serrano and Suma 2011: 9).

Another unrealistic assumption of the M–F model relates to the absence of any role for expectations. If expectations are minimally elastic with respect to the exchange rate actually observed, the short-run Mundell–Fleming model can become quite unstable. Each time the domestic interest rate falls and the current exchange rate begins to depreciate, the spot exchange rate expected for a subsequent period will also change in the same direction, increasing the rate of change of the expected exchange rate and therefore shifting the BP curve leftwards. Thus, unlike in the stylized model – where the BP curve shifts rightwards and depreciation of the exchange rate causes exports to increase and output expand – the impact of depreciation can be contractionary through the process of expectation build-up. This is particularly true for the developing countries, and has been experienced in every crisis or near-crisis situation when depreciation in currency fuelled expectations of further depreciation and triggered speculative attacks that were self-fulfilling.⁴

The link from balance of payments to monetary base, and hence broad money, has been questioned effectively by the endogenous money school (Lavoie 2006). Post-Keynesians argue that the fixed exchange rate does not prevent central banks from setting interest rates while money creation is still demand-led. When foreign currencies are exchanged for domestic currency at the fixed exchange rate, commercial banks’ reserves actually increase, but there is no reason to suppose that the banks are forced to lend these additional resources if nothing occurs to increase the number of customers considered creditworthy by banks. The only direct impact of the initial monetary-base increase is a drop of the money multiplier. Eventually, banks do not want to retain these extra, idle and unpaid reserves, and invest these excess bank reserves in government securities. As the Central Bank, when setting the interest rate, has agreed to sell bonds at the exact amount that is demanded, this leads to an increase in public debt and sterilization of the increase in foreign-exchange reserves. Thus, the sterilized intervention is almost always the rule and not the exception to the rules of the game. In the context of a flexible exchange rate, if we assume that the domestic rate of interest is exogenous and the money supply is endogenous, it is evident that the automatic adjustment mechanism stops working. When the Central Bank lowers the interest rate, continuous devaluations occur and even if the economy expands, the interest rate will not increase, as the amount of money expands in line with the increasing activity levels (and even prices) of the economy.

This brief review of the literature clearly demonstrates that none of the basic results of the M–F model survive in more realistic contexts of imperfect capital mobility, elastic exchange-rate expectations or exogenous domestic interest rates. The automatic adjustment route attributed to flexible exchange rate is based on certain very limited assumptions. Boom–bust cycles, herd behaviour, procyclicality and uncertainty of these flows are outside the scope of the model. Instead, the finance–investment link is reduced to the assumption of perfect capital mobility. The empirical evidence reviewed below only reinforces the gap between the existing theoretical framework and the reality of macroeconomic policy in the context of capital flows. We argue that the heterodox perspective on financial instability can provide important insights to make sense of the processes at work and guide policy.

Empirical evidence on India

We analyse below the relationships across major macro variables in the light of the theoretical discussion.

Interest rate and capital flows

An important relationship underlying the M–F model is the interest sensitivity of capital flows. A number of empirical studies have obtained evidence that places in question this basic underlying relation, and the simplistic understanding surrounding it. In an estimation of the determinants of foreign capital flow for the period from 1991–2 to 2009–10, Mundle, Bhanumurthy and Das (2011) find that the capital flows to India are not interest-sensitive, i.e. the differential in domestic and foreign interest rate does not explain the variation in capital inflow. Among the statistically significant determinants of capital flows are domestic GDP, US GDP and Chinese GDP. While the first two have positive coefficients, the latter is negatively significant in the relation. Another recent study published by the RBI confirms this result. Using causality and cointegration analyses, Verma and Prakash (2011) find that FDI and FII equity flows, which together on a net basis accounted for around three quarters of the total net capital inflows during the ten-year period from 2000–01 to 2009–10, are not sensitive to interest-rate differentials. Only debt-creating flows, such as ECBs and NRI deposits, are found to be interest-sensitive. Gross capital inflows appear to increase by 0.05 percentage points in response to an interest- rate differential increase of one percentage point – a very small degree of elasticity. Both studies find that the growth performance of advanced countries positively affects the quantum of capital flows to EMEs.

Anecdotal evidence points in a similar direction. Faced with a surge of capital inflows in 2003, Bimal Jalan, then-governor of the Reserve Bank of India, had noted that arbitrage was unlikely to have been a primary factor in influencing remittances or investment decisions by NRIs or foreign entities in the recent period:

Jalan (2003) concluded that external flows into India have been motivated by factors other than pure arbitrage.

Patnaik and Shah (2010) juxtaposed short-term (real) rates of interest against business cycle conditions to find that during high growth phases in the business cycles, monetary policy in India was forced to yield negative real rates. While their main point was to prove how monetary policy has been constrained by exchange rate interventions, a la the impossible trinity, the negative real interest rate, high growth and procyclicality of capital flows together prove that capital flows were determined by a host of factors, and negative real interest rates did not prevent capital from flowing in.

Figures 4.4a to 4.4c plot the trends in capital flows, the real sector and financial-sector variables over the years, showing that the response of capital flows to domestic economic conditions is procyclical (Figures 4.4a and b). The correlation between growth of capital inflows and domestic GDP growth is +0.47. Similarly, the correlation between growth of capital inflows and year-on-year change in BSE SENSEX is +0.27. Theoretically, the causality is expected to run both ways. Economic expansion and booms in asset markets attract foreign investment and lending. Also, the effect of capital flows on domestic spending tends to be procyclical, with booms in capital flows and currency appreciation encouraging spending, which is reinforced by the wealth effect and real balance effect.

Figure 4.4c plots the trend in interest-rate differential, i–i*, where interest rates for India and the US are measured as short-term policy rates, domestic repo rate minus Federal Funds effective rate.⁵ The capital-flow variable is not correlated with the interest-rate differential. The correlation coefficient between capital flows and the interest-rate differential of –0.13 is not only small, but has a negative sign. This is in line what others have found on the lack of interest sensitivity of capital flows.

However, when the (actual) exchange rate changes are included and the net interest-rate differential (i–i*–Δe) is considered, the correlation improves substantially and is in the expected direction (+0.44; Figure 4.4c). The differential in earnings between an overseas investment and home-country investment after taking into account the exchange-rate movements is represented by i–i*–Δe. What this indicates is that while there is little relation between i–i* and net capital flows, there is a strong relationship between the capital movements and changes in exchange rate, even though the system is a managed float. A surge in capital inflow by causing the exchange rate to appreciate adds to the potential gains from investment for the foreign investor. Elastic expectations play an important role. Exchange-rate appreciation creates expectations of further appreciation, which in turn would bring in larger capital inflows. This is unlike the prediction in the interest-rate parity condition, that foreign capital inflows would bid down the returns in the capital-recipient economies and bid up the interest rate in the home economies. Rather, capital flows by causing the exchange rate to appreciate, generating expectations of further appreciation, which attracts more capital, and so on. This is the classic mechanism by which the carry trade is sustained.⁶ There is no self-correcting mechanism.

There are some far-reaching implications of these observations. First, capital inflows are by and large exogenous, so a recipient country can neither know a priori nor influence through macro policy the quantity of foreign capital it might receive over a certain period. The only handle that the domestic economy then has to limit capital flows is capital control measures. Second, capital flows are fair-weather friends and do not help to tide over adverse domestic conditions. Economic expansion and booms in asset markets attract foreign capital. Whereas the standard economic theory does not take cognizance of the procyclicality of capital flows, the essence of the financial instability hypothesis is the procyclical response of financial market to impulses emanating from the real economy. Capital inflows, Akyuz (2008) notes, cause appreciation of the currency, add to asset-price inflation and raise aggregate demand and growth in a cumulative process. The process also generates vulnerability by causing unsustainable trade-deficit and maturity mismatches on balance sheets. When capital flows stop as a result of rapid accumulation of risks, a negative shock to growth, deterioration in global financial conditions with respect to liquidity and risk appetite or contagion from another developing country, this process is rapidly reversed, resulting in sharp depreciations, credit crunch, debt deflation and economic contraction. Third, the empirics show that the link between capital flows and interest rate is much more uncertain than is often assumed. At times of favourable risk assessment a much smaller interest-rate margin can attract large inflows of capital, whereas during episodes of capital flight, steep hikes in interest rates are unable to retain capital. Fourth, the herd behaviour characteristic of international capital movements, combined with flexibility in exchange rates, would ascertain that foreign investors reaped additional gains through appreciation of the exchange rate. The demand for flexibility in the exchange rate will obviously gain huge support from certain quarters.

Central Bank intervention, reserve accumulation and monetary-policy independence

During surges in capital inflows, the central banks in developing countries intervene in the currency markets for two ostensible reasons: first, they wish to avoid sharp appreciation of the exchange rate; second, a policy of accumulating reserves at times of strong capital inflows and using them during sudden stops and reversals appears to be a sensible counter-cyclical response to instabilities in capital flows. In the M–F model, attempts by the Central Bank to prevent currency appreciation through direct intervention, when faced with capital inflows, leads to an expansion of the monetary base and broad money, and causes the interest rate to fall via a shift of the LM curve. In turn, an expansionary monetary policy – when economic growth is already high, and with procyclical capital flows – contributes to inflation (Patnaik 2007).

In reality, interest rates are announced as a matter of policy, and these rates are accepted by the financial markets. Even the new consensus view on monetary policy accepts that interest rates are not set via an interest-elastic demand for money. When base money rises, so that liquidity is excessive, banks and other participants can usually place these excess reserves with the Central Bank at the going interest rate (reverse repo rate in India). Thus exchange-market interventions do not prevent the Central Bank from setting interest rates wherever they want to. Grenville (2011) shows that for a cross-section of Asian countries, nominal short-term interest rates have been quite varied, hence demonstrating independent monetary policy, whereas exchange rates in these countries have shown a good deal of stability in real terms, the latter being an important objective for export-led growth.

China is a classic case in which intervention has not only been successful in stabilizing the exchange rate but has also been less costly to the government because of the control over the banking system. Thus, the link from exchange-market intervention to increases in base money is broken by effective sterilization.

In the case of India, it may be said that the exchange-rate stability has been more limited and varied across time. Figure 4.4 shows the movement of the exchange rate (monthly average) and RBI intervention measured as net purchase of foreign currency by the Central Bank. The nominal exchange rate fell between April 1998 and May 2002 from 39.7 to 49 per USD, before recovering to 47.3 by March 2003 – phase I. In phase II, which is marked by continuous increases in capital inflows, there was a gradual appreciation of the exchange rate up to Oct 2007 (phase II). Of the three phases, this was the only period without a major financial crisis in the world economy. Exchange-rate volatility was low. Ramachandran and Sambandhan (2007: 1324) note that ‘the excessive volatility in the nominal exchange rate was minimized by not allowing capital flows to exert their full weight on the exchange rate through judicious interventions’. Beyond October 2007, the movements in exchange rate were sharper and there have been two sharp peaks – phase III. Volatility of exchange rate in this phase is higher than in the previous phases. The RBI’s intervention (even in gross terms) has been parsimonious. It has intervened only to defend the currency depreciation of the exchange rate through sale of foreign currency. Intervention through purchase of foreign currency is negligible.

To what extent has the RBI’s intervention to defend the currency endangered monetary-policy independence? A scrutiny of the trends in monetary and credit variables suggests the following (see Table 4.2):

1. Intervention by the RBI brought about a change in the composition of reserve money. The RBI’s net foreign-exchange assets have emerged as the dominant source of reserve-money expansion. This source of base-money creation was offset by the reduction in net RBI credit to the government through sterilization (Table 4.2, columns 1 and 2). Thus, reserve-money growth remained well within limits except in 2007–8, when it grew by 27.5 per cent. Even in this year, the broad money growth was not as high.
2. Despite the monetary expansion, policy interest rates were successively raised from 2004–5 onwards, when the capital inflows were the highest. It does not appear that the setting of interest rates by the Central Bank was compromised by the RBI’s intervention. This directly challenges the claims made by Patnaik and Shah (2010).
3. Investments in government securities by the commercial banks (sterilization) do not seem to have happened at the cost of traditional lending activities. Between 2003–4 and 2007–8, the ratio of commercial banks’ investments in government securities fell as a proportion of commercial banks’ credit to the commercial sector, from 0.69 to 0.39 (column 7). Banks were not forced to invest in government securities and lending and investment activities were still largely propelled by the demand for loans, which in turn is a function of real economic conditions.
4. The highest credit growth to the commercial sector did not occur in the year 2007–8, the year of highest-ever intervention, but in the preceding years. Ramachandran and Sambandhan (2007) draw attention to this. In many years when base-money growth was high, credit expansion remained low, and vice-versa. For inflation, what matters is the credit growth, which they show to be determined by the growth rate of manufacturing output. The authors conclude that exchange-rate intervention does not have one-to-one correspondence with credit expansion.
5. However, there may be a more direct relationship between capital flows and credit growth that is recognized in the Keynes/Minsky tradition. Instead of central banks’ intervention leading to procyclical monetary policy, Serrano and Summa (2011) show that the correct chain of causality arises from more favourable external interest rates and credit terms in relation to domestic credit and interest-rate conditions, inducing local banks and businesses to take more credit to finance their decisions to increase spending on productive and unproductive investments. Empirical evidence on capital inflows and expansion of credit would thus be least surprising.

In sum, the evidence shows that exchange-market intervention did not dissolve the central banks’ ability to formulate monetary policy in the traditional sense.

Having said that, it is important to recognize that with open capital accounts, monetary policy is circumscribed in ways that are not recognized by mainstream theory. At times of increasing capital inflows, if the monetary authorities intervene through sterilized intervention, there is a cost to the government in terms of the difference in yield on the government securities and the return on the foreign currency assets it receives. The difference between i–i* in Figure 4.4c provides a rough approximation. The greater the difference in returns and the higher the capital inflows, the greater would be the quasi-fiscal costs.

During episodes of sudden stops and reversal of flows, hikes in interest rates are often unable to retain/attract capital and check sharp currency declines. Still, the central banks typically deploy monetary measures, along with direct intervention in the foreign-exchange market, to relieve buying pressures. Excess demand pressures in the foreign-exchange market in India have prompted the RBI to increase cash-reserve ratios, repo rates, bank rates and other liquidity management measures, in what the literature has described as the interest-rate defence (Bose 2002). Domestic conditions require the monetary authorities to cut interest rates to prevent financial meltdown and stimulate economic activity, whereas the need to prevent capital outflow forces monetary authorities to pursue procyclical policies. A higher interest rate aiming to retain capital might be perceived to signal an increased risk of default, however, and may precipitate a currency depreciation (crisis) and economic contraction (Figure 4.1). The continuous turbulence in financial and currency markets and the fire-fighting by the RBI since the last quarter of 2011 reflect this phenomenon.

For the monetary authorities, the most important challenge around capital inflows is to maintain financial stability and prevent the build-up of financial fragility. Capital flows are associated with asset booms, credit expansion and overindebtedness. As argued by Kindleberger (1995: 35), monetary-policy authorities need to use judgement and discretion, rather than ‘cookbook rules of the game’, when speculation threatens substantial rises in asset prices and exchange rates, with the possible subsequent harm to the economy. However, they often refrain from doing that in the belief that their task is to keep inflation under control: a monetary-policy stance that maintains price stability would also promote financial stability, and financial markets do not need intervention as they regulate themselves (cited in Akyuz 2008). Further, it is common to shift the burden of adjustment on fiscal policy. Kohli (2010: 44) advises that running a counter-cyclical fiscal policy weakens the link with the domestic cycle – reduction in domestic demand mutes the response of consumption growth to equity, debt and other investment flows. ‘Fiscal responses are more effective than capital controls…’ It is expected that capital regulation will not be the preferred option under the present policy dispensation.

Capital flows, currency appreciation and trade balance

The appreciation of the exchange rate following foreign inflows is a classic question of open-economy macro-economics. Foreign inflows cause the price of non-traded goods to rise relative to that of traded goods, bringing about an appreciation of the real exchange rate. The result is a decline in competitiveness, a decrease in export growth and worsening of the current-account balance. In the M–F model, the erosion of competitiveness through a decline in exports lowers aggregate demand and therefore money demand, which should cause the interest rate to fall. With lower arbitrage margins, capital inflows are expected to moderate in a self-correcting mechanism, and exchange-rate appreciation cease.

Figure 4.5 traces the real effective exchange rate trade-based (REER), trade balance and capital-account balance as a proportion of GDP from 1993–4 to 2010–11. Higher capital-account flows have been unambiguously correlated with the appreciation of REER (correlation coefficient: +0.56). Before 2002–3, the relationship was weak (+0.19). The Central Bank, through intervention in the forex market, exercised much greater control on REER, such that the influence of capital flows on REER was less dominant. Chakraborty (2003), using a VAR model for the period 1993Q2 to 2001Q4, confirms that the Central Bank’s action was effective in avoiding any serious distortion in the real exchange rate, unlike the case of Latin America. Beyond 2002–3, the intervention continued, but with huge increases in capital inflows there was a constant upward pressure on both the nominal and real exchange rate to appreciate (+0.68). Unlike what was postulated in the M–F model, active exchange-rate management was necessary to keep the REER near desirable levels. Despite intervention, the pressure of capital flows was such that the REER appreciated steeply in 2007–8, before falling with the boom–bust cycle of international capital flows.

Figure 4.6 shows the deteriorating trade balance, a trend that began after 2003–4 and has continued since. Trade balance and REER are negatively correlated, higher REER being associated with lower trade balance. The correlation coefficient is –0.22 for the period 1993–4 to 2007–8.

While the impact of capital flows on REER is irrefutable, the link from REER to trade balance has been denied openly in recent years by proponents of the flexible exchange rate. Under pressure, the RBI has observed that REER appreciation may not have implications for external-sector competitiveness (see RBI 2011). This is surprising since a number of authoritative studies had shown that India’s supply and demand for exports are both responsive in the long run to changes in the REER (Joshi and Little 1994; Srinivasan 2001). In the exchange-rate management policy of the 1990s, when the surge of capital flows was still limited and the memory of the balance-of-payments crisis of the early 1990s still fresh, stable and non-appreciating REER was repeatedly stressed as the guiding consideration for the conduct of exchange-rate management. The important question raised for the Indian monetary authorities in the 1990s was ‘whether the nominal exchange rate in India was managed in such a way as to produce the appropriate real exchange rate … defined as that value of the REER which would produce a sustainable current account deficit’ (Joshi and Little 1994: 80).

It is possible that the relationship between REER and export growth or trade balance has changed, since capital flows enhance the volatility and uncertainty in exchange-rate movements. Trade theory has explained how a volatile exchange rate reinforces the wait-and-see attitude of firms that must pay a price to enter or exit markets. Paul Krugman (1989), in his collection of lectures entitled Exchange Rate Instability, suggested that this response to uncertainty leads to a multiplier process of real exchange-rate volatility: the more volatile the exchange rate, the less responsive trade is; the less responsive trade is, the more volatile the exchange rate. The policy implication is that exchange-rate volatility must be consciously contained to encourage trade.

On the contrary, in the past few years (phase III in Figure 4.4), both the quantum and the frequency of intervention by the RBI in the currency market has reduced (EPW Research Foundation 2010; Rajwade 2012). While there has been no explicit change in exchange-rate policy, the unofficial word is that the G-20 is putting pressure on the EMEs to prevent distortion in the determination of the exchange rate. That official intervention is a cause of distortion in the exchange rate and exchange rate determined by the vagaries of capital flows is the undistorted rate is obviously a fallacious argument. It has long been known that in an era of free capital mobility, left to itself, the exchange rate would be found at levels far from anything that makes sense, given the fundamentals. Even in conservative circles it is maintained that monetary authorities must take a long-term view on where equilibrium lies and push it back towards the proper equilibrium, or maintain a band around the equilibrium.

In the absence of direct intervention by the Central Bank to manage the exchange rate, REER cannot be targeted for competitiveness. With exports rendered exogenous and the willingness to use tariff and non-tariff barriers limited, expenditure compression is then posed as the only possible alternative for improving trade balance. In the absence of expenditure switching, expenditure reduction – in particular, fiscal response – has to restore the external balance.⁷ An overall contractionary bias to macro policy in a regime of free capital flows and flexible exchange rates appears inevitable.

Conclusion

An increasing trade deficit, volatility in portfolio flows and exchange rate, slowdown in other investment inflows and a growing dependence on external private debt to balance the BOP define the present predicament of the external-account position of India’s economy. The proximate cause of the problem is the eurozone crisis and the flight to safety of international capital which has affected the worldwide economy, including that of India. Taking a slightly broader view, it is clear that the cycles in international capital flows are fairly frequent and tend to dominate the economic cycles in the recipient countries during both the boom and the bust phases. The M–F model is the standard framework to make sense of the consequences of capital flows and the associated policy constraints. Citing the relevant theoretical literature and using evidence from the Indian economy, this chapter argues that the M–F framework is lacking in realism and fails to capture the essential nature of the flows. Uncertainty, procyclicality, herd behaviour, elastic rather than inelastic expectations and sudden stops are integral features of present-day international capital flows. The intersections of capital flows with other financial markets – stock markets, foreign-exchange markets, the credit market – and the real economy, and consequently the scope of policy, need to be discussed within this overall understanding.

Where the M–F model fails, the financial instability hypothesis provides useful insights to explain the facts surrounding the Indian situation. The essence of the financial instability hypothesis is the procyclical response of the financial market to impulses emanating from the real economy. Capital inflows to India are found to be procyclical with regard to both the domestic economy and advanced countries’ GDP growth. These fair-weather friends have caused overexpansion during the expansionary phase, including higher spending, asset-price booms, exchange-rate appreciation and worsening of the trade balance. The interventions by the Central Bank to minimize exchange-rate appreciation and excessive volatility in emerging Asian economies, including India, can be seen in this light. However, there has been continuous pressure to withdraw intervention and allow the exchange rate to appreciate, so as to restore independence of monetary policy, a la the impossible trinity. The examination of the link between RBI intervention, monetary aggregates, interest rates and bank credit growth does not support the argument that intervention caused the monetary authorities to lose control in setting the policy rates or caused an explosion of monetary aggregates/liquidity during the phase of rapid capital inflows. Sterilization effectively dampened the excess base-money creation and credit growth increased depending on the demand conditions in the economy. The debate on forex market intervention and inflation has in fact sidelined the real challenge for monetary policy when faced with surges in capital flows – that is, how to prevent the build-up of financial fragility.

While the markets support appreciation of the domestic currency, they are less tolerant to depreciation.⁸ The central banks thus respond to flight of international capital and the resulting turbulence in the financial and currency markets through a variety of steps such as direct intervention, hikes in interest rates and other measures to attract/retain capital. The empirics reveal that the capital inflows are not sensitive to interest-rate differentials in any predictable manner. At times of favourable risk assessment a much smaller interest-rate margin can attract large inflows of capital, whereas during episodes of capital flight, steep hikes in interest rates are unable to retain capital, which makes the relationship between interest rates and capital flows uncertain. In other words, capital inflows are by and large exogenous, so that a recipient country can neither know a priori nor influence through macro policy the quantity of foreign capital it might receive over a certain period. Regulation and control of capital flows, rather than macro policy, provides the principal tool for financial and economic stability.

This chapter also argues that the supposed lack of relationship between REER and trade balance is a dangerous logic of convenience that is being used to legitimize the Central Bank’s hands-off policy on the exchange rate on the one hand, and to impose a contractionary macro policy on the other. Capital inflows push up the value of the emerging market’s currency, make the exchange rate highly volatile, erode export competitiveness and may even rupture the response of exports to relative costs. Without a handle on REER, expenditure contraction remains the only means to adjust the trade deficit. In the interests of the real economy, regulation of capital flows is a must.

Notes

1  Though capital flowed towards EMEs after the second cycle in very high magnitudes, the net resource transfer was in the reverse direction from EMEs to advanced economies (Boratav 2009). The coupling of capital flows with over-liquid and under-regulated financial markets in these economies was a sure recipe for crisis in these economies.

2  The Union budget for 2012–13 included a series of measures facilitating External Commercial Borrowings (ECBs) by the corporates in select sectors, and in a few sectors the withholding tax on ECBs was slashed. The Economic Survey 2011–12 (Government of India 2012) has a long list of measures undertaken by the RBI to incentivize ECBs, FIIs, NRI deposits, etc. Though intended as short-term expedients, these are steps towards further liberalization, and none are likely to be reversed.

3  See India’s Quarterly International Investment Position (IIP) for December 2011.

4  See the case of Brazil discussed by Paul Krugman (1998) in ‘Latin America’s Swan Song’, ####

5  The difference i–i* is positive and varies between 1.5 per cent and 9 per cent, with an overall negative trend. Until 1999–2000 the interest-rate differential declined steadily, after which the gap widened.

6  See FRBSF Economic letter Nov 2006–31, Nov. 17, 2006, ‘Interest rates, carry trades and exchange rate movements’.

7  If inflation is higher than targeted, which is usually the case in rapidly growing economies, policy rates will generally be raised. This is the case even when inflation is not, due to demand pull factors. See Bose (2012) for the debate on the monetary-policy response to inflation during the recent inflation episode. As for government expenditure, the twin deficit identity is extrapolated to argue that the fiscal deficit is the cause for the current account deficit.

8  The Economic Survey (2011–12) notes: ‘rupee volatility impairs investor confidence and has implications for corporate balance sheets and profitability in case of high exposure to ECBs when currency is depreciating. A more aggressive stance to check rupee volatility is therefore necessary’.

References

Akyüz, Yilmaz (2008) ‘Managing Financial Instability in Emerging Markets: A Keynesian Perspective’, February 29, 2008. http://www.networkideas.org/featart/feb2008/fa29_Markets.htm.

Boratav, Korkut (2009) ‘A Comparison of Two Cycles in the World Economy: 1989–2007’, IDEAs Working Paper Series, Paper no. 07/2009.

Bose, Sukanya (2002) Money and Finance in Deregulated Economies: The Indian Experience. PhD Thesis, Unpublished.

Bose, Sukanya (2012) ‘Inflation: Sources, Challenges and Policy Options’, Economic and Political Weekly, Volume 47 No. 03; pp. 27–30.

Chandrasekhar, C.P. and Jayati Ghosh (2012) ‘India’s External Sector’, http://www.networkideas.org/news/jan2012/news11_External_Sector.htm.

Chakraborty, Indrani (2003) ‘Liberalization of Capital Flows and the Real Exchange Rate in India: A VAR Analysis’, CDS Working Paper, No. 351.

EPW Research Foundation (2010) ‘Hands-off Policy on the Exchange Rate’, Economic and Political Weekly, Volume 45 No. 38, pp. 70–76.

Grenville, Stephen (2011) ‘The Impossible Trinity and Capital Flows in East Asia’, ADBI Working Paper Series, No. 319.

Government of India (2012) Economic Survey, 2011–12.

Hannoun, Herve (2007) Policy responses to the challenges posed by capital inflows in Asia, Speech, Bangkok, 28 July 2007, Bank for International Settlements, http://www.bis.org/speeches/sp070728.pdf.

Jalan, Bimal (2003) ‘Exchange rate management – an emerging consensus?’, BIS Review 36/2003.

Joshi, V. and I.M.D. Little (1994) India: Macroeconomics and Political Economy, Delhi, Oxford University Press.

Kindleberger, C. (1995) ‘Asset Price Inflation and Monetary Policy’, Banca Nazionale del Lavoro Quarterly Review 192, pp. 17–37.

Kohli, Renu (2010) ‘Managing Capital Flows, c.2010: Policy Options for India’ Economic and Political Weekly, Volume 45 No. 46, pp. 40–44.

Krugman, P.R. (1989) Exchange Rate Instability, MIT Press Cambridge, MA.

Lavoie, M. (2006) ‘Endogenous Money: Accommodationist’, in P. Arestis and M. Sawyer (eds), Handbook on Alternative Monetary Economics, Edward Elgar, Cheltenham.

Minsky, H.P. (1984) ‘Can “It” Happen Again?’, in M.E. Sharpe (ed.), Essays on Instability and Finance, Armonk, New York.

Mundle, Sudipto, N.R. Bhanumurthy and Surajit Das (2010) ‘Fiscal Consolidation with High Growth: A Policy Simulation Model for India’, NIPFP Working Paper No. 73.

Palma, Gabriel (1998) ‘Three and a Half Cycles of “Mania, Panic, and [Asymmetric] Crash”: East Asia and Latin America Compared,’ Cambridge Journal of Economics, Volume 22, No. 6, pp. 789–808.

Patnaik, Ila (2007) ‘Indian Currency Regime and Its Consequences’, Economic and Political Weekly, March 17, pp. 911–13.

Patnaik, Ila and Ajay Shah (2010), ‘Asia confronts the impossible trinity’, NIPFP Working Paper, 2010-64, January.

Rajwade, A.V. (2012) ‘The Fall of the Rupee: Background, Remedy and Policy’, Economic and Political Weekly, Volume 47, No. 2, pp. 10–14.

Ramachandran, M and D. Sambandhan (2007) ‘Resisting Rupee Appreciation?’ Economic and Political Weekly, Volume 42, No. 15, pp. 1322–1325.

RBI (1997) Report of the Committee on Capital Account Convertibility, Chairman: S.S. Tarapore.

RBI (2006) Report of the Committee on Fuller Capital Account Convertibility, Chairman: S.S. Tarapore.

Reserve Bank of India (2011) ‘The External Sector’, http://rbidocs.rbi.org.in/rdocs/Publications/PDFs/MMD300411C3.pdf.

Rosenberg, Irma (2010) ‘The Impossible Trinity and Constraints on Policy options’, Reserve Bank of India, First International Research Conference, Mumbai, India, February 12–13 2010.

Sen, Sunanda (2003) Global Finance at Risk: On Real Stagnation and Instability. Palgrave-Macmillan, Houndsmill, 2003.

Serrano, Franklin and Ricardo Summa (2011) ‘Mundell–Fleming without the LM curve: the exogenous interest rate in an open economy’, Conferenciainternacional del aniversario de la revista, Faculty of Economics, UNAM, September 7–9 2011. www.depfe.unam.mx/70aniversario-ie/.

Srinivasan, T.N. (2001) India’s Reform of External Sector Policies and Future Multilateral Trade Negotiations, Yale Economic Growth Center Discussion Paper No. 830. http://ssrn.com/abstract=275307.

Verma, Radheshyam and Anand Prakash (2011) ‘Sensitivity of Capital Flows to Interest Rate Differentials: An Empirical Assessment for India’, RBI Working Paper Series, WPS (DEPR): 7/2011.