Thread by @Brad_Setser

The transformation of the U.S. into a commodity exporter is (nearly) complete.

Looked at the real net exports data for q2 more closely, and the entire q2 exports surprise came from ag ('beans) and oil

Q2 is obviously an extreme case, and the jump in ag exports is magnified by the BEA's annualization process.

But it reflects a broader story, since the end of 2012, ag and petrol exports have contributed more to US growth than manufactured exports

If you exclude ag, the non-oil goods deficit has continued to deteriorate under Trump, tho the deterioration is now from rising imports and modest export growth, while the deterioration in late 14/15 was from a fall in non-petrol goods exports

And I looked at the available data for q3 -- e.g. real trade numbers for July and August. July ag exports were still strong, so barring a big fall in September, they will be flat. Petrol exports will be down slightly. The big swing will be in the non ag, non petrol deficit

So right now I expect a large drag from trade in q3, primarily on the manufacturing side (exports below their q2 average so far in q3, imports up).

the big fall off in ag exports may come in q4, as lots of farmers are storing their 'beans

dtnpf.com/agriculture/we…

This thread has a bit of traction, so adding one more, slightly complicated, chart. Shows the net contribution (e.g. fall in imports + rise in exports) from oil v the cumulative contribution of non ag, non petrol goods imports and exports (basically a manufacturing proxy)

Gap between the two lines is net manufacturing exports. idea is to capture the increase in oil production that that has gone into reducing imports along side the rise in exports (energy has a lot of 2 way trade, strangely).

Fairly clear, I think, that the dollar's 2014/15 appreciation has led to a shift in the composition of US traded goods output toward commodities

(the change in the services balance in this period isn't significant)