In commenting on tariffs and the trade balance, Paul Krugman states that, in the "simple story", tariffs result in an exchange rate appreciation which leaves the trade balance unchanged.
A critical point here (which Krugman goes on to explore) is the idea that neither tariffs nor any associated exchange rate appreciation has any impact on capital account flows (or for that matter on the net flow of investment income). However, the simple story also contains an implicit assumption that the level of output is unchanged. If instead tariffs are accompanied by an increase in domestic activity, demand for imports may be unchanged and there may be no change in the exchange rate. It all depends on what else we think might be going on.
What interested me more here though was Krugman's assertion that trade deficits must eventually be replaced by surpluses (and thanks to Matias Vernengo for drawing my attention to this comment). The idea he is appealing to is that a country cannot have a negative net international investment position (NIIP) that grows indefinitely.
In fact this comes down to an r versus g question. g here is the rate of growth of the economy. r is the average rate of return on the NIIP. If r is less than g, then even with balanced trade, the NIIP will be declining as a percentage of GDP. No trade surplus is ever needed to correct a negative NIIP.
The thing about r here is that it is an average return and the NIIP is the difference between two larger numbers. For example the US NIIP of around negative $8 trillion is the difference between assets of $25 trillion and liabilities of $33 trillion. This means that the average return can vary greatly and can look nothing like a normal asset return. For example, in the US case, if the return on assets is sufficiently greater than the return on
assets liabilities, the average
return on the NIIP can easily be negative.
This means that, in cases where the NIIP is small relative to the gross positions, we cannot generally say anything meaningful about r and it is quite possible that a country can run a trade deficit for ever without any need for eventual surpluses. This is not, of course, the same as saying that a country can run any level of trade deficit for ever.