(FORTUNE Magazine) – Most of the time our economic worries are of a comprehensible sort: Jobs are scarce, health-care costs are high, retirement funds are meager. Every month or two, though, we are wrested away from those quotidian concerns and told to worry about the dollar.

As in late September, when the finance ministers of the big rich countries (a.k.a. the G-7) seemed to say they wanted a weaker dollar: Suddenly the financial pages fill with jittery currency talk and the stock market catches a chill. Then, in part because there's no clear consensus as to just what is so worrisome about sickly green notes with pictures of dead Presidents on them, everybody moves on to freaking out about some more tangible potential disaster. (In September it was rising oil prices.)

The dollar scares will keep returning, because they serve as shorthand for what people at the IMF call "widening global imbalances." The main imbalance is that the U.S. has been doing most of the world's spending for the past decade, while Europe and Asia have been doing most of its saving. That presumably can't go on forever, and the fear is that the "adjustment"--to slip again into IMF lingo--will be painful.

Those who forecast the economic future follow one of two paths. The first and most popular is to assume that the immediate past will repeat itself: Trends will continue, this recession will look like the last one, etc. That technique works fine in quiet times but misses every economic shock and turning point (which is why only a handful of Wall Street economists saw the 2001 recession coming).

The second method of forecasting--its most prominent practitioner these days is Morgan Stanley economist Stephen Roach--is to identify the biggest economic and financial forces at work in the world and spin theoretical scenarios of how they might interact in the future. The global economy is far too complex for any such scenario to capture its essence, which means that the prognosticators are often spinning whole cloth. But sometimes they do succeed in predicting busts (or booms) unforeseen by the economist masses.

Right now most economic indicators are showing improvement, so the Wall Street consensus is that they will continue to improve. As for those "imbalances"--a $550 billion U.S. current-account deficit and a $560 billion federal budget deficit this year--the past-repeats-itself crowd isn't too worried. There was a lot of portentous talk in the mid-1980s about the dangers of the "twin deficits," yet both eventually went away (see chart) without causing great trauma.

But this is 2003, not 1985. At 5% of GDP, the current-account deficit (the sum of a huge trade deficit, a tiny surplus in investment income flows, and a substantial deficit in "unilateral current transfers" like foreign aid and money sent home by foreign workers) is bigger than it's ever been. The same is true by definition of the equally large net inflow of foreign capital to the U.S. But those foreign investors who put money into the U.S. are hoping eventually to take even more out. And for the past couple of years, instead of putting it into direct investments like factories or banks or cellphone companies that imply a long-term commitment, they've been buying securities that they can sell in an instant: Treasuries, mortgage securities, stocks.

The scariest scenario that can be spun from this is that if the current-account deficit keeps growing--and with it foreigners' U.S. investment holdings, already $9 trillion at the end of 2002--foreign investors will one day get scared and flee the U.S. The dollar will crash, interest rates in the U.S. will go way up, and a bad time will be had by all. No economist we came across is actually predicting that, but many do see a resumption of the dollar's slide (it rose during the summer but is down 10% overall since early 2002; 25% against the euro).

That would make exporting from the U.S. more attractive and importing to it less so, thus shrinking the current-account deficit. So it would seem to be good news. But unlike in the late 1980s, when Treasury Secretary James Baker engineered a nearly 50% fall in the dollar with few side effects, the world's other major economies may not be capable of picking up the slack in global demand that an extended dollar swoon would bring. What does that mean? No disaster, but yet another bit of unfinished business from the booming 1990s that's likely to weigh on our recovery.