A Note on Foreign Currency Reserves

In today’s context of exchange rate instability and the currency wars there is an urgent need to address the problem of “reserves” that many countries have piled up over the last decade or two.

This note presents the simple economics of foreign currency reserves.  I begin with what I call the expanded balance of payments equation:

X – M = + R – (F + H + B)   ………………. (1)

Where X = exports; M = imports; R = reserves; F = FDI inflow; H = portfolio investment inflow and B = net borrowings (inflow).  Note that outflows of F, H and B would mean a negative quantity.  Suppose the country has a surplus in current account of $100 (X – M = 100), then if R = 0, we must have F + H + B as -100 so that X – M = 100 = 0 – (-100) = 100.

Now consider the situation where F = H = B = 0, then X – M = R.  In this case if R = 0, then X – M = 0 or X = M.  Let $ (USD) be the foreign currency and Rs. (Rupee) the domestic currency.  What this means is that if a country does not hold foreign currency reserves, then we must have current account balance, ceteris paribus.  Figure 1 clearly shows what happens when a country is at equilibrium at E₀ followed by a shift in its supply of foreign currency curve (S$) curve.  In this case, the excess supply of foreign currency will mean that the domestic currency (Rs. = rupees) appreciates till we reach point E1 where S$’ = D$.

 

To prevent the appreciation in currency and the subsequent loss in exports as we move from E00 to E1, the central bank could intervene in the forex market by buying dollars at the initial exchange rate, er₀.  The dollars purchased by the central bank are what we call as “reserves” or R in equation (1) above.  Whatever the purpose of a country holding reserves of dollars may be, it is clear that reserves would lead to a depreciation of the domestic currency (Rs.) vis-à-vis the dollar ($).  There is extensive literature on why countries hold reserves and what is the optimum amount of reserves that a country should hold; whether reserves are held based as part of a mercantilist strategy or to protect themselves against forex market turbulence or as an insurance against financial risks. In addition to the why it is nonetheless important to understand the implications of holding reserves.  It is also clear from Figure 1 above, that when the domestic country holds reserves, from equation (1) it is clear that X > M or X – M > 0.  If so, in a two-country model, we must have M > X or a current account deficit for the USA.  Simple as this may seem, many people begin with the USA running large current account deficits as the basis of attack against its economic policy.  However, failure to consider the implications of its trading partners holding high volumes of reserves, render such arguments incomplete.

Now consider a situation where there are capital inflows into the domestic economy (assume B = 0).  Let A = autonomous transactions in the balance of payments where:

A = (X – M) + F + H ……………….. (2)

These would then give us the S$ curve.  If we consider R as the only accommodating transaction (on the balance of payments) which gives us the D$ curve, then in Figure 2 we can see how the central bank can manipulate the exchange rate by “choosing” an appropriate level of R.

In Figure 2, with an increase in hot money inflows, A shifts to A’.  To maintain the exchange rate at er0, the central bank must increase its reserves from R to R’.  Reserves then become the instrument to control the exchange rate.

 

But there are consequences for the domestic economy on account of increasing reserves. When the central bank buys $s, it injects rupees into the economy.  This, as we know, shifts the LM curve outwards, reduces interest rates, increases output.  However, from the AS-AD framework we also know that a shift in LM curve implies a shift in the AD curve.  The higher output comes at the cost of higher inflation.

We have so far considered a two country model.  If we extend the model to include more than country, say, the EU, we could have more than one reserve currency, say, euros (E) and dollars ($), i.e. R_E and R_$.  We could also include gold (G) as a component of reserves along with SDRs (S).  The balance of payments equation will then be:

X – M + F + H + B = R_$ + R_E + S + G ……………. (3)

It is clear from equation (3) above that an increase in G may not have an impact on the Rs./$ exchange rate.  If, for example, the gold is bought from Australia, then an increase in gold reserves may mean a strengthening of the Australian dollar but not of the US dollar.  If the US is a country’s major trading partner then the outcome in terms of the Rs./$ rate may not be the intended one.

Consequently, we propose the following hypotheses for further empirical study:  countries, especially those which intend to stimulate growth through exports and foreign investment (like China or India) will hold reserves of foreign currencies in proportion to their trade and investment shares with those countries.  For a currency to become a reserve currency for these export-oriented countries, trade and investment share with these countries must grow.  In other words, for the US$ to lose its position of a reserve currency its trade and investment share with south and south east Asian countries will have to decline.

This brief note also throws light in the question we raised earlier.  Why does Europe hold a large percentage of its reserves in gold and countries like China and India hold dollar denominated reserves?  Obviously Europe does not engage in the same strategy of increasing its exports to the US through currency depreciation as many South and Southeast Asian economies have pursued.