Anything that ''cannot go on forever,'' in the economist Herbert Stein's very wise epigram, will eventually stop. This is worth keeping in mind as attention focuses again on imbalances in the world's accounts, especially between Japan and the U.S., and the related problem of currency instability.

Public debate on the issue has largely been a contest between the complacency of the supply-siders and the alarmism of the protectionists. To supply-siders, both the U.S.'s trade deficit and its foreign borrowing can go on more or less forever, and no action is necessary to correct them. To protectionists, these matters are so immediately pressing they must be stopped by any means at hand.

It is true, as supply-siders maintain, that a trade deficit in itself is no cause for concern, any more than a trade surplus is cause for rejoicing: the prosperity of a nation is not measured by the size of its foreign reserves, but by the wealth of its people.

Even if the trade deficit were something to worry about, protectionism wouldn't make it go away. It does not matter whether this involves raising barriers to Japanese imports, or, as every American protectionist insists, one's only aim is to pry open Japan's markets: trade balances are not a function of trade policy. The U.S. could be much more protectionist, and Japan much less, and the American trade deficit would remain.

Trade policy determines how open the economy is to trade, without saying whether that trade will balance. Exchange rates influence the balance of trade, without regard to how open or closed the economy may be. In the long run, a country can only export as much as it imports: the exchange rate, other things being equal, sees to that. So if a tariff closes off imports, it must ultimately close off the same amount of exports as well.

MOVEMENTS IN VOLUMES

I say ''in the long run'' and ''other things being equal'' because exchange rates are not the only factors influencing the volume of exports or imports. This is why a given decline in the American dollar does not guarantee an improvement in its trade balance, any more than a fall in price guarantees a company will increase its revenues. It depends also on movements in volumes.

But if the trade deficit is not worth worrying about, foreign borrowing can be, even though in accounting terms one is simply a restatement of the other. Supply-siders are fond of pointing out that no one worries if California is a net debtor vis-a-vis the rest of the U.S. Why, they ask, should it be any different for capital flows between nations? But people do worry if West Virginia has a chronic trade deficit with the rest of the U.S. That's called ''regional disparities.''

The reason for both phenomenons is the same: the exchange rate, or rather, the lack of one. California, West Virginia and the other 48 states all belong to a common currency area.

The good side of this is that investment can flow from one part of the U.S. to another without worrying about local currencies. The bad side is that if the terms of trade are fundamentally altered to one region's disadvantage, the exchange rate can't move to take that into account. The burden of adjustment must be borne by wages and prices; if not by them, then by employment and output.

For trade between two countries with different currencies, the problem is reversed. Trade imbalances - other things being equal - will even out over time. But if a large stock of debt is accumulated in the interval, and if foreign investors suspect the debtor country intends to cut the value of their holdings by depreciating the currency, then the adjustment in the exchange rate may not be gradual, but a sudden, cascading decline, as foreign investors sell their holdings and their fears become self-fulfilling.

But if the nation is thought too much in debt to others, the answer is not to cut off the flow of foreign capital at the border, whether through controls on foreign investment or barriers to the exported goods and services that finance it. It is to reduce the need for it in the first place: to cut back on the level of total borrowing, at least as it involves not the voluntary actions of private investors, but the political imperatives of governments.

This is what George Grant called ''tight-money nationalism,'' and is to be distinguished from recent Canadian practice, which has been to raise the level of public borrowing exponentially, then to discourage anyone from lending us the money to cover it. But even that cannot go on forever.