( ... ) When Mr. Keynes comes back to the problem of adjustment of disturbances in the balance of payments in his general treatise on money, he is contributing to the general theory and no longer confines himself to the particular case of German Reparation payments. There he takes his position on the classical doctrine.
Before going into his theory it is desirable for us to explain some of the symbols employed by him: In his book [,]
- B＝the Foreign Balance＝“the balance of trade on income account, resulting from the excess of the value of home-owned output of goods and services (other than gold), whether produced at home or abroad, placed at the use and disposal of foreigners, over the value of the corresponding foreign-owned output placed at our use and disposal,”
- L＝the values of Foreign Lending＝“the unfavourable balance of transactions on capital account, i.e. the excess of the amount of our own money put at the disposal of foreigners through the net purchase by our nationals of investments situated abroad, over the corresponding amount expended by foreigners on the purchase of our investments situated at home” ;
- G＝the exports of gold.
By definition we have L＝B＋G. There cannot be, however, equilibrium in international trade so long as there is a continuous movement of gold into or out of the country. “The condition of external equilibrium is, therefore, that G＝0, i.e. L＝B.” 
- Thus the existence of external equilibrium depends upon the factors governing L and those governing B.
- He finds that the former are the relative interest rates at home and abroad, while the latter are the relative price-levels at home and abroad of the goods and services which enter into international trade.
- But “there is no direct or automatic connection between these two things; nor has a Central Bank any direct means of altering relative price-levels, The weapon of a Central Bank consists in the power to alter interest rates and the terms of lending generally.” 
Thus, he still denies the doctrine that B is directly a function of L.
The direct weapon that the Central Bank can use in regulating external disequilibrium is the bank rate. But to employ the bank rate as a weapon to meet external disequilibrium would, for a time, mean a divergence of the terms of credit from their domestic equilibrium level, if they are originally in equilibrium. Hence the first effect of using bank rate to preserve external equilibrium will be to produce internal disequilibrium. That, however, is true only of a transitional period. When the change in bank rate has worked out its effect via prices on B so as to bring about a stable equilibrium between L and B, the internal equilibrium will once again be restored. Thus he says:ㅡ
Bank-rate is both an expedient and a solution. It supplies both the temporary pick-me-up and the permanent cureㅡprovided we ignore the ^malaise^ which may intervene between the pick-me-up and the cure. ... The essence of the matter can be set out briefly. Raising the bank-rate obviously has the effect of diminishing L, and the net amount of lending to foreigners. But it has no direct influence in the direction of increasing B. On the other hand, just as the dearer money discourages foreign borrowers, so also it discourages borrowers for the purposes of home investmentㅡwith the result that the higher bank-rate diminishes ... the volume of home investment. Consequently total investment falls below current savings (assuming that there was previous equilibrium), so that prices and profits, and ultimately earnings, fall, which has the effect of increasing B, because it reduces the costs of production in terms of money relatively to the corresponding costs abroad. On both accounts, therefore, B and L are brought nearer together, until in the new position of equilibrium they are again equal. ... At the new level of equilibrium we shall once again have ... I＝S. But we shall also have B＝L. For since B moves in the opposite direction to P and P in the opposite direction to LㅡB [? L－B], whilst L moves in the opposite direction to the bank rate, for every value of bank-rate there is a value of P at which B＝L; and since S moves in the same direction as bank-rate and I moves in the opposite direction, there is always a value of bank-rate for which I＝S. Consequently there is always a pair of values of bank-rates and of P at which both I＝S and B＝L. 
 ^Treatise^, vol. i, pp. 214-5. I＝the value of increment of new investment goods. S＝the amount of saving.
Mr. Keynes calls attention to the growing importance of the role of the international movement of liquid capital:ㅡ
In modern times, when large reserves are held by capitalists in a liquid form, comparatively small changes in the rate of interest in one centre relatively to the rate of others may swing a large volume of lending from one to the other. That is to say, the amount of foreign lending is highly sensitive to small changes. The amount of the foreign balance, on the other hand, is by no means so sensitive. ... This high degree of short-period mobility of international lending, combined with a low degree of short-period mobility of international trade, meansㅡfailing steps to deal with the formerㅡthat even a small and temporary divergence in the local rate of interest from the international rate may be dangerous.
When the bank rate is put up against an unfavourable balance of payments, the existence of capital in liquid form in the international market would cause the gap between L and B to be filled by the flow of short-term capital. That will quickly restore apparent equilibrium at a stage quite insufficient to work out its effect on B and to re-establish underlying equilibrium. The result is to delay the credit measures, while the progressive accumulation of short-term indebtedness becomes itself an independent threat to equilibrium.
As indicated above, the fulfilment of the condition of external equilibrium depends on two things: first, relative price-levels which affect B; and, second, relative interest rates which affect L. External disequilibrium may thus arise either from disequilibrium in relative price-levels or from disequilibrium in relative interest rates. He finds that there is a radical difference between those two cases. "In the first case, the disequilibrium can be cured bu a change in price-levels (or, rather, of income-levels) without any permanent change in interest rates, though a temporary change in interest-rates will be necessary as a means of bringing about the change in income-levels. In the second case, on the other hand, the restoration of equilibrium may require not only a change in interest-rates, but also a lasing change in income-levels (and probably in price-levels). That is to say, a country's price level and income-level are affected not only by changes in the price-level abroad, but also by changes in the interest rate, due to a change in the demand for investment abroad relatively to the demand at home" 
( .. .... )