Financial Crises, Reserve Accumulation and Capital Flows

EAST ASIA: A DECADE AFTEREconomic & Political Weekly december 15, 200745Financial Crises, Reserve Accumulation and Capital FlowsPrabhat PatnaikThe article constructs a theory to understand a financial crisis in an open third world economy in the context of a sequence of stock equilibria, ensured by inelastic expectations. It then explores the predicament of such an economy when it “opens up” to global financial flows. In the absence of central bank intervention it has to face financial crises. But central bank intervention aimed at avoiding financial crises by stabilising the exchange rate and holding foreign exchange reserves pushes the economy to a perennial stock disequilibrium.The end of the dirigiste era has brought in its wake a proc-ess of liberalisation of financial flows into and out of most third world economies. This process began early in some parts and later in others; and it has proceeded to different extents in different countries. Nonetheless it is a fact of life over much of the third world. Any such liberalisation brings third world assets into the ambit of portfolio decisions of first world wealth-holders. This has certain implications, one of which is the unleashing of financial crises in the third world. Let us first look at a simplified picture of the financial crisis.1 Understanding a Financial CrisisThe introduction into the portfolios of first world wealth-holders of some of these assets that were hitherto inaccessible to them, is likely to entail an improvement in these portfolios. Hence their demand for these assets will go up, through an ex ante reduction partly in their demand for money, and partly in their demand for other non-money assets. This will have an impact on the prices of the third world assets, and a new stock equilibrium will be estab-lished, where, under standard assumptions, the returns at the margin, or what Kaldor (1964) had called the “own rates of money interest”, from each of the assets belonging to this aug-mented universe and entering into the wealth-holder’s portfolio, is equalised for each wealth-holder. A sufficient condition for such a new equilibrium to come about, if we follow Keynes, is that the elasticity of price expecta-tion must be less than unity for all assets for all wealth-holders, even though the expectations themselves may be divergent among them. This divergence of expectations played a crucial role in Keynes’ analysis. It meant a division of the wealth-holders into “bulls” and “bears”, which explained both how asset transac-tions actually took place, and also how equilibrium was estab-lished, through a shifting of the line demarcating the two. But it precluded any notion of a “representative wealth-holder”, such as was invoked in the analyses of Kalecki (1954), Kaldor (1964) and Hicks (1946). In what follows however, since our concern will be with examining equilibrium conditions, and not with the process through which equilibrium is arrived at, we shall invoke the notion of a “representative wealth-holder”, which, as Kahn (1972) had pointed out, assumes a “dense concentration at the margin”. But we shall distinguish between the “representative wealth-holder” from the first world and the “representative wealth-holder” from the third world.While inelasticity of price expectations in the asset market is a sufficient condition for equilibrium, it is obviously not a necessary Prabhat Patnaik (prabhatptnk@yahoo.co.in)is at the Centre for Economic Studies and Planning, Jawaharlal Nehru University, New Delhi.

in the third world economy which has a perpetual life, so that its rate of return is simply its nominal earning e (assumed to be constant over its lifetime) divided by

e

EAST ASIA: A DECADE AFTEREconomic & Political Weekly december 15, 200747turn elastic over a whole range of prices, when the prices are declining). But this model can illustrate the argument I wish to make.2 Regulating Demand for Third World Country’s AssetThe rise in the dollar prices of the third world assets consists of two parts: a rise in the dollar value of the local currency, and a rise in the local currency value of the assets. Before the economy opened itself up to financial flows, the first of these considera-tions did not figure in anyone’s calculations. There was a stock equilibrium in the economy with the “own rates of money inter-est” being equal (calculated in local currency) across all assets for the representative domestic wealth-holder. If one of these, say the deposit rate, is fixed, then all the others must adjust to it in equilibrium. Now, suppose after the economy has got “opened up”, banks continue to accept any amount of deposits at the same (fixed) interest rate. Then the same own rates of money interest in local currency must prevail in a stock equilibrium after the “opening up” as the ones that prevailed before. But, for this new stock equilibrium to prevail, it is necessary not only that there should be inelastic price expectations in the markets for local currency assets, but also that there should be inelastic price ex-pectation in the foreign exchange market. If price expectations are elastic with regard to the local currency price of the asset, then there would be no equilibrium to start with, even prior to the opening of the economy to financial flows. But in addition, after the economy is “opened up”, a stock-equilibrium cannot exist in the absence of inelastic price expectations in the foreign exchange market. Thus, inelastic expectations with regard to the dollar value of the third world currency is a condition for the ex-istence of a stock equilibrium in a third world economy “opened up” to financial flows. Stabilising the Dollar Value of Local CurrencyNow, suppose the government wants to avoid financial crises. The most obvious way in which it can intervene is by stabilising the dollar value of the local currency. This is because, unlike other assets, the supply of local currency is within the powers of the central bank, and hence (whether directly or indirectly) of the government, and can be augmented at will. The central bank can simply buy dollars at the going exchange rate and hold them as foreign exchange reserves when the demand for local assets by first world wealth-holders increases. Let us look at the implications of this. If we denote the “own rates of money interest” (in local currency) on local currency assets before and after “opening up” byε(which does not change for reasons just discussed), then, with the exchange rate remain-ing unchanged through central bank intervention, the condition for a new stock equilibrium will ber* =ε– ρ… (G)But, ex hypothesi the reason why first world wealth-holders wanted to move into the third world economy’s asset in the first place was that the rhs in(G) exceeded the lhs. An equilibrium can be reached only if this inequality is converted into an equality by the very process of the increase in the ex ante demand for the third world economy’s asset. But the only equilibrating factor in (G) isρ, the marginal risk-premium, which represents compensa-tion for foreigners against two kinds of risk: the foreign exchange risk and the risk associated with the asset market (compared to the risk of holding the first world asset). The very stabilisation of the exchange rate however actually reduces the first kind of risk, while the second kind of risk remains unchanged. (As for the principle of increasing risk, it becomes relevant in the present context only if there are substantial increases in the actual holding of local currency assets by foreigners. Withε remaining unchanged, there is no reason why such substantial transfers of local currency assets should occur at all from domestic to foreign owners.) Hence central bank inter-vention reduces the value of ρ, if anything, and allows little scope for ρ to increase as the foreigners’ demand for local assets increases. The attempt to eliminate financial crises therefore prevents the existence of a stock-equilibrium altogether. Looking at it differently, inelastic price expectations are a condition for stock-equilibrium only whenprices are allowed to move around. But if there is a restriction on price movements, then even though thismay itself ensure inelastic expectations, it cannot ensure theexistence of a stock-equilibrium.The prevention of crises through government intervention in short entails that no mechanism to arrest the rising demand for the local asset remains. True, we have assumed so far that the as-set that “rules the roost” among local currency assets, and to whose rate all other rates are tethered, sees no shift in its rate of return.2 If this asset happens to be bank deposits, then, it may ap-pear, that this assumption loses its rationale, and that variations in the deposit rate can be a policy instrument for achieving a stock equilibrium. But, even assuming that the correct deposit rate for bringing about an equilibrium can be accurately estimat-ed, monetary policy is not free to fix the deposit rate anywhere it likes, since it must inter aliacover the minimum lenders’ risk.3 In other words, the deposit rate may be ostensibly what “rules the roost” (if it does), but underlying it will be something more solid. The conclusion that government efforts to overcome financial crises negate the possibility of a stock equilibrium itself, there-fore, remains valid.Crises, Marx had said, are a way of resolving, forcibly and tem-porarily, the contradictions of capitalism. Of course Marx often tended to look at crises in purely cyclical terms, i e, assumed an automaticity about recovery from the downturn, much the way we have assumed above. There is in fact no such automaticity. Even so, a crisis is a way of rectifying, no doubt in a most explo-sive and painful manner, the unrestrained movements of the economy, or of particular markets, in particular directions. In the simple model sketched above, a financial crisis was the mecha-nism for preventing an unrestricted explosion of the demand for the third world economy’s asset. In the absence of crises, there is nothing to prevent such an unrestricted explosion of demand.Financial Crises as Equilibrating MechanismThe fact that the above model holds the occurrence of financial crises as the factor responsible for arresting the unrestricted explosion of demand for financial assets, may not be obvious at first sight. It may appear as if the restraint on demand for the

EAST ASIA: A DECADE AFTEREconomic & Political Weekly december 15, 200749worse over time. Since ε in (H) is given, and since the stock equi-librium before “opening up” must have ensured that the “own rate of money interest” on all assets equalledε, the fact that foreigners wish to hold local currency assets after “opening up” implies that ε exceeds (r* +ρ). Under these conditions however exchange rate stabilisation makes it impossible to attain a stock equilibrium after “opening up”, since no equilibrating mecha-nism now exists. It follows then that exchange rate stabilisation makes it impossible to achieve a stock equilibrium.We have not till now discussed either the disequilibrium be-haviour of the system or the process through which equilibri-um is reached. We have talked of foreigners’ demand for local assets raising their prices, but we have said nothing about the actual financial flows. Strictly speaking, from an examination of equilibrium conditions we cannot jump to any theoretical conclusions about the magnitude of actual financial flows. But it is plausible to assume that the magnitude of financial flows into an economy will be linked to the magnitude of the foreigners’ demand for its assets. It follows that since the excess demand for the economy’s assets is not eliminated in this case, foreign exchange reserves keep piling up. And since with the piling up of reserves the marginal risk premium associated with holding the local currency declines further from the al-ready low level to which the stabilisation of the exchange rate had brought it down, the excess demand for the local currency asset will keep increasing and reserves will keep piling up over time even faster.Inability to Prevent Occurence of Financial Crises Let us look at this disequilibrium behaviour briefly. The fact that the economy does not “explode” in a situation of perennial excessdemand for domestic assets is because in practice only a certain finite amount of financial inflows occurs in any period notwithstanding this excess demand, which affects domestic asset prices but keeps them within bounds. Even within this overall disequilibrium however, portfolio adjustments made by domesticwealth-holders will ensure that all “own rates of money interest” in local currency terms equal ε. This will happen through an increase in their prices relative to their expected prices. The sequence of such states of rest across periods (they cannotstrictlybe called “stock equilibria”) can still generate a domestic financial crisis, in the form of a sharp drop in the price of some local currency assets, for exactly the same reasons as were discussed earlier. When such a drop occurs, foreigners will not be unaffected by it. In other words it will also entail a reduction in demand for the local currency itself. But even if there is no foreign exchange cri-sis, because the reduced demand for the local currency is han-dled through the decumulation of foreign exchange reserves, the collapse of the price of some local currency assets will have seri-ous adverse consequences.Central Bank intervention, in the form of holding foreign ex-change reserves, to prevent financial crises, in a world with free financial flows, therefore is ineffective in two senses. It prevents first of all the achievement of a stock equilibrium altogether. Secondly, even within this overall disequlibrium, it still cannot prevent the occurrence of financial crises. The most it can do is to prevent domestic financial crises from spilling over into foreign exchange crises, but even that is achieved at the cost of com-pounding the problem of instability over time. Of course, this does not mean asking for a removal of central bank intervention. On the contrary, such intervention is essential in a regime of free financial flows for preventing increased unemployment and im-miserisation. But it is this regime itself that needs transcending. One can go further. Since this regime represents the latest phase of existing capitalism, transcending this regime means tran-scending existing capitalism.Putting it differently, crises, though painful, play a certain role under capitalism. The prevention of crises, by foreclosing this role, creates instability in a different way, by undermining the modus operandi of the system. To say this, echoing Marx, is not to plead for an acceptance of crises; it is to plead against the ac-ceptance of capitalism as we have known it, since this knowledge has shown that “crisis-free capitalism” is a chimera. This does not ofcourse mean that we should not demand intervention against crises, but we should do so without any illusions, and only as part of a process of going beyond existing capitalism.4 Concern with the Accumulation of ReservesThe accumulation of reserves by several third world economies is often seen as a factor reducing global demand.6 This is not true. A distinction must be drawn here between reserve accumulation that is accompanied by a current account surplus and reserve accumulation that is unaccompanied by such a sur-plus. The cases of China and India belong respectively to these two categories. In the case of the latter, reserve accumulation, as already discussed, has an expansionary consequence on the level of aggregate demand of the reserve accumulating economy,for in its absence there would be a currency apprecia-tion causing domestic unemployment. But since such an appre-ciation would be accompanied by a corresponding increase in the aggregate demand of those countries whose currencies have depreciated relatively, the level of world aggregate demand would remain unchanged. In short, the level of world aggregate demand is unaffected whether or not there is reserve accumulation of this sort.Even in the case of the former, reserve accumulation per se cannot be said to have caused a decline in world aggregate de-mand. If China for instance decided not to accumulate reserves but to let her currency appreciate, then there will be a decline in the level of aggregate demand in China which will be accompa-nied by an increase in the aggregate demand of countries cur-rently out-competed by China, notably the US. In principle, there is no reason to believe that the level of world aggregate demand will change on account of China’s reserve decumulation. It is only if reserve decumulation in China is accompanied by a corre-sponding increase in some other item of aggregate demand that we can expect an increase in the level of world aggregate de-mand. If for instance, the Chinese economy instead of running current account surpluses decides to enlarge domestic absorption through enlarged government expenditure while keeping the ex-change rate unchanged, then there will be an increase in world

EAST ASIA: A DECADE AFTERdecember 15, 2007 Economic & Political Weekly50aggregate demand. But then this has nothing to do with reserve decumulation per se: all that it says is that the level of world aggregate demand goes up if the domestic absorption of one par-ticular country goes up without that of any other country reduc-ing, which is but a truism. The conclusion in both cases that the level of world aggregate demand remains unchanged in the event of a decumulation of reserves through an appreciation of the exchange rate, assumes that the economy, whose currency has depreciated relatively and which experiences an increase in aggregate demand as a conse-quence, is unconcerned with the associated increase in its domes-tic prices. If the price effects of an expansion of its aggregate de-mand entail high social resistance, as might be expected in the case of theUS perhaps, which has a restraining effect on the mag-nitude of its demand expansion, then reserve decumulation on the part of the “emerging market economies” through an appre-ciation of the exchange rate has a contractionary effect on world aggregate demand. It follows that the accumulation of reserves through central bank intervention in keeping the exchange rate fixed,makes the level of world aggregate demand no lower thanitotherwise would have been; but it invariably has a net expansionary effect on the level of aggregate demand of the re-serveaccumulating economy itself, compared to what it would beifreserves did not accumulate and the exchange rate was al-lowed to appreciate.The concern with the accumulation of reserves represents however a refracted perception of something altogether differ-ent. While such accumulation has no effect per seon the level of world aggregate demand (except for the fact that its substitu-tion by larger domestic absorption would be beneficial for all, which is a very different proposition), it does undermine the position of the leading capitalist economy, by accumulating claims against it. Throughout the history of capitalism, the diffusion of industrial capitalism from the core to “newly industrialising countries” has been accompanied by the leading capitalist power of the time running a current account deficit vis-a-vis the “newly industrialising economies”, thus offering them space in its domestic market. But this never caused any net accumulation of claims against the leading capitalist power because it always had access tuo a drain of surplus from colo-nies, in addition to having the colonial markets “on tap”, which more than paid for its current account deficits vis-a-vis the “newly industrialising countries”. Its currency therefore was never under pressure even as it ran current account deficits against the NICs of the time. The fact that the leading capitalist power of today finds its currency under pressure and itself be-coming a heavily indebted nation, is because the possibilities of colonial drain no longer exist, though misadventures for resur-recting the system of colonial drain, especially from oil-rich third world countries, have not been given up.7M.S. SWAMINATHAN RESEARCH FOUNDATION, CHENNAI Invites Application for the following positions: 1. Ford Foundation Chair in Women and Food SecurityJob Description: Develop and carry out a programme of research and education, guide research, conduct training and orientation for policymakers and trainers, and undertake advocacy in the areas of women and food security.Qualification: Ph.D. in any social science (preferably Economics) and at least ten years experience, with evidence of strong interest in women’s/gender studies, and good record of work in research and teaching, and/or field programme related to women’s/gender issues and food security. Qualifications in women’s/gender studies will be an advantage.Salary: Appropriate level in 22,000-1,000-30,000 grade, with permissible allowances. 2. Uttara Devi Resource Centre in Gender and Development – Project CoordinatorJob Description: Develop and carry out projects within the broad mandate of the Centre in the areas of: research and documentation; training and capacity building, and advocacy and communication related to gender in development, preferably in one of the priority areas of the Foundation, such as, biodiversity, rural livelihoods, agriculture, NRM and food security. Interact with and provide support related to gender dimension in other programmes/activities in the Foundation.Qualification: Ph.D. in any relevant Social Science with five years (or M.Sc. with 7-10 years) experience, along with evidence of work experience and strong commitment to women’s/gender issues. Grassroots experience and/or qualifications in women’s/gender studies will be an advantage. Good communication skills both in English and Tamil/ or any other Indian language.Salary scale: Appropriate level in 17,000-1,000-25,000 grade, with permissible allowances. Application should include complete biodata showing academic record, area of specialization, field and duration of experience and address of past employers, if any. Applicant may also mention names and addresses of two referees, who during the last five years have had close familiarity with the candidate’s professional strength. Application may be sent by post before 31.1.2008 to: Manager (Pers. & Admn), M. S. Swaminathan Research Foundation, III Cross Street, Institutional Area, Taramani, Chennai 600 113. Visit our website www.mssrf.org for details.

EAST ASIA: A DECADE AFTEREconomic & Political Weekly december 15, 200751Notes 1 For a discussion of “own rates of own interest” see Kaldor (1964). 2 Kaldor (1964) had raised the whole question of why money “ruled the roost” among all the assets, in the sense that the yield on money determined the vector of own rates of money interest on all other assets. 3 The question may be asked: if the opening up of third world assets to first world wealth-holders makes them demand these assets, and if among assets in each place the one that “rules the roost” is tethered to the minimum lending risk, then this must presuppose that the lend-ing risks are higher in the third than in the first world. This is an entirely plausible supposition, since the higher level of development of the institutions of capitalism will normally be associated with lower lenders’ risk. 4 One is reminded here of Dennis Robertson’s (1940) re-mark about “the dying embers of liquidity preference” having to be periodically stoked through falls in bond prices. Likewise the runaway demand for third world financial assets has to be periodically curbed through falls in the prices of these assets. The difference between the two cases however consists in this: the memory of these falls produces actual falls,which in turn perpetuate this memory. 5 The concept of a Non-Accelerating Inflation Rate of Ex-change is discussed in Patnaik and Rawal (April 2, 2005). 6 Ocampo, Kregel and Griffith-Jones (2007, 156) write: “Irrespective of the exchange rate regime adopted, to insure themselves against sudden shifts in market senti-ment, most emerging economies have kept increasingly high stocks of international reserves. This ‘self-insur-ance’ option entails significant costs and could constrain global growth as it reduces global aggregate demand.” 7 The arguments of this paragraph have been devel-oped at length in Patnaik (2005). ReferencesHicks, J R (1946): Value and Capital, Clarendon Press, Oxford.Kahn, R F (1972): ‘Notes on Liquidity Preference’ in Selected Essays on Employment and Growth, Cambridge University Press, Cambridge.Kaldor, N (1964): ‘Own Rates of Interest’ inEssays on Economic Stability and Growth, Duckworth, London.Kalecki, M (1954): The Theory of Economic Dynamics, Allen and Unwin, London.Ocampo Jose Antonio, Kregel Jan, Griffith-Jones Stephanie (eds) (2007): International Finance and Development, Orient Longman, Hyderabad.Patnaik, P (2005): ‘Contemporary Capitalism and the Diffusion of Activity’ (Inaugural Lecture for the Sukhamoy Chakravarty Chair), Economic & Political Weekly.Patnaik, P and V Rawal (2005): ‘The Level of Activity in an Economy with Free Financial Flows’,Economic & Political Weekly.Robertson, D H (1940): ‘Mr Keynes and the Rate of Interest’ in D H Robertson,Essays in Monetary Theory, Staples, London, pp 1-38.5 The Indian CaseThe proposition advanced above that central bank intervention for stabilising the exchange rate, in order to prevent unemploy-ment through exchange rate appreciation, and future financial crises, has the effect of making things worse over time, is borne out by the Indian experience. It was argued earlier (in Section 3) that even within the overall stock disequilibrium that character-ises a situation of central bank intervention to stabilise the ex-change rate, the “own rates of money interest” in local currency (in Kaldor’s terminology), will be equal to ε and to each other. Since this comes about through movements of actual asset prices relative to their expected prices, and hence entails significant movements in actual asset prices, it does not prevent financial crises. (In India’s case it is the stock market crisis that is perti-nent.) Government intervention in India therefore has taken the form ofbothstabilising the exchange rate andkeeping the stock market boom going, through selective interventions by way of “liberalisation” and fiscal incentives.But the very logic of a combination of a stock market boom and stabilised exchange rate has meant a continuous shift of foreign demand towards the local currency assets and hence towards the local currency as well. This has saddled the Reserve Bank of India with burgeoning reserves which have in the course of three years climbed to $200 billion. And precisely because such large reserves make the rupee less vulnerable to collapse, the marginal risk premium associated with holding the rupee has gone down, thereby further stimulating financial inflows and hence adding to reserves.This situation of disequilibrium would not matter, except for the fact that there is a massive difference between the rate of return earned by those who are bringing in financial inflows and the rate of return earned on the reserves. If a minimum of 20 per cent is taken as the rate of return, inclusive of asset price appre-ciation, for those who bring in financial inflows, then, given the fact that the reserves earn no more than about 1.5 per cent on average, the annual loss owing to this difference comes to about 6 per cent of the GDP of the country. In other words, the country, as it were, is borrowing dear to lend cheap, to an extent where its annual loss at present amounts to 6 per cent of the GDP. This is not to say that 6 per cent of the GDP is being actually drained out of the country every year. But the fact that it is not being actually drained out now, only implies that the size of the drain will be even larger in future. The so-called “sterilisation” operations, meant ostensibly to control money supply, are really a means of preventing damage to the banking system under the weight of this loss. Precisely because we are in a disequilibrium situation where financial inflows continue to occur, and foreign exchange reserves con-tinue to get built up, the magnitude of reserve money increases sharply which finds its way into the banks and far outstrips the demand for credit from “worthwhile borrowers”. Banks’ profits are threatened by this discrepancy, and the Reserve Bank steps in to shore up these profits by putting income earning govern-ment securities into the banks’ portfolio. The Reserve Bank’s doing so however means a substitution of foreign exchange re-serves for government securities in its own asset portfolio, with the former earning much lower rates of return than the latter. This therefore lowers the Reserve Bank’s profitability, and hence reduces a whole range of development finance activities which the Reserve Bank uses its profits for promoting. In short the Reserve Bank is taking upon itself the losses that the com-mercial banks would otherwise have incurred owing to this disequilibrium situation.Many in India, in a throwback to mercantilism, have applaud-ed the accumulation of reserves, deeming it to be a great success for the neo-liberal policies. For reasons just discussed, nothing could be further from the truth. Very recently the Reserve Bank has allowed an increase in the dollar price of the rupee (partly perhaps to control incipient inflationary pressures), but the ef-fects of this move are already apparent in the form of lower ex-port prices for the already hard-pressed peasants.What the Indian case clearly demonstrates is that measures aimed at avoiding financial crises push the economy into a per-ennial disequilibrium. This choice between the devil and the deep sea may well be avoided in the first instance through the imposition of capital controls. Such controls will bring in their train a whole range of further developments, the cumulative impact of which will be a shift in the balance of class forces that underlies contemporary Indian capitalism. But this shift can occur only if there is no sudden panic-stricken or externally-imposed change of course in between, entailing an abandon-ment of controls.