Your third link “simply don’t correlate” has a column headed simply “trade”. My original statement was that a weak dollar made US goods cheaper to other nations, and their goods more expensive to US. And, further, that a weak dollar led to increased US exports.
Some government stats add imports and exports together and call that total “trade”. Not sure what is represented in your column “trade”, but we’d have to look at either US exports, or imports alone to test what I (and many others) said: that a weak US dollar typically increases our exports.
And you are also looking at weighted average exchange rates, and the total Balance of Payments, or the trade deficit since that’s all we’ve experienced for years. Again, there can be so many ups and downs with various nations’ currencies that you end up with a big average where all the high and low flucuations in specific dollar exchange rates have been smoothed out, and those average exchange rates and total balance of payments just don’t really test the statement we’re discussing.
All those weighted averages and totals arrived at by taking data for 100 to 200 nations are just not mathematically or statistically valid for testing the statement.
A test would be: find a year where the dollar became significantly stronger or weaker against the yen. Look at our imports and exports with Japan for that year and following years (or months or quarters). Determine if the exchange rate ups or downs are reflected in our imports and exports with Japan. - It should really show in our imports from Japan if the yen is weaker. If the dollar is weaker, it should show in higher exports to Japan. (Japan would often eat the negative exchange rate impacts if the dollar were weaker and their goods more expensive to the US, to keep their exports to the US going).
Or you could whine some more?
Right, and it's got a bunch if 'up's and 'down's like the dollar one after it. It has to do with the correlation, that whenever we get a year of both the balance of trade and the dollar going up at the same time then it's called a direct correlation, and whenever one goes up while the other goes down it's and inverse correlation.
You don't need the table (also shown in post 24) because the graphs and the numbers all show the same thing, that half the time the correlation is direct and the other half of the time it's inverse. This could be done month by month or by quarters and it wouldn't make any difference. Go ahead and run through the numbers yourself --let me know if you need any help downloading.
Let's get clear on what's going on first, and only after that's settled can we think about why it works the way it does.