Once again, the trade figures have confounded the "experts": the June trade deficit rose sharply to $15.7 billion from $14 billion in May.

And although the old homily about not paying too much attention to one month's data is still true, the bigger deficit is a disappointment -- especially when the smart money expected a decline.

As a result, we've had a weaker dollar -- a sign that financial markets believe that the worst is not over, despite the Reagan administration's earlier assurance, endorsed by many private economists, that the trade deficit would begin to come down.

"We underestimated the extent to which rising foreign prices offset the real improvement in the trade picture," said C. Fred Bergsten of the Institute for International Economics -- in effect, speaking for all economists.

The fact of the matter is that America's trade deficit is shrinking, even though the commonly used statistics don't seem to show it. Import prices have skyrocketed 14.5 percent from June 1986 to June 1987 -- and more than two-thirds of the rise has occurred since Jan. 1 of this year.

Thus, even though the volume of imports is down, their value is up -- more than offsetting a modest increase in exports that has stirred some enthusiasm in the country's manufacturing centers. Robert Ortner, chief economist for the Commerce Department, estimates that the "real" trade deficit, using constant 1982 prices, has declined about $30 billion (at an annual rate) since last September.

Ortner's figures are based on a deficit rate of $161.6 billion in trade and services in the third quarter of 1986, which was $33.8 billion higher than the $127.8 billion shown in the preliminary estimate for the second quarter. Allowing for the bulge in June just announced, Ortner says, that would leave the $30 billion improvement.

The question thus arises -- if the current data on the value of exports and imports is disguising an improving trend, how can that message better be conveyed to the public? Can figures be developed, similar to the "real" gross national product, that will give a more valid impression of what is going on?

Sen. John C. Danforth (R-Mo.) has sponsored legislation -- part of the omnibus trade bill -- that would require the Commerce Department to release each month, coincident with its traditional publication of monthly trade numbers in dollar-value terms, the equivalent deficit or surplus in volume terms.

"If we had it, it would be very useful," Ortner said in an interview, "because we are doing a lot better than the {present set of} figures show." But he says that the government does not have available, on a sufficiently timely basis, the monthly price data with which to determine the "real" or deflated trade volume.

There may be ways around that dilemma, if it proves politically useful to have the volume number. Most important, if the Danforth proposal passes, there would need to be an appropriation that would enable the Bureau of Labor Statistics to put the necessary manpower into collecting import price data.

Right now, the Commerce Department makes a private, curbstone guess of trade volumes for its preliminary quarterly estimate of the gross national product. Ortner agrees that more effort might be put into that, with the goal -- perhaps -- of coming up with a published monthly volume figure after a 30-day delay.

The economic and political potential of this new indicator could be significant. For example, the probability is that the raw announcement of the June figures -- the increase in the deficit to $15.7 billion in June from $14 billion in May -- not only disturbed the foreign exchange markets but also bolstered the chances for tough trade legislation.

But suppose that -- at the same time -- the Commerce Department had been able to announce that in volume terms, the deficit had not increased? Goldman Sachs International vice chairman Robert Hormats believes that the availability of such a number might have a calming effect on exchange markets, because a declining trade deficit volume in effect would be predicting a decline in value terms.

It isn't possible, Ortner says, to develop a precise volume figure for the June deficit. But we do know this much: from the third quarter last year to the second quarter of this year, the deficit didn't improve much -- it declined only from an average of $14.9 billion to $14.2 billion monthly.

But in constant 1982 prices, the deficit actually dropped from a monthly rate of $16 billion to $13 billion. Ortner cautions that for technical reasons, the two series of numbers are not exactly comparable. But a decided trend for the better catches the eye.

The problem is that the deficits, on the traditional value basis, could get even worse before getting better, if the dollar starts to go down again. In part, it's a Catch-22 situation: the dollar weakens because the trade deficit fails to improve. And with a weaker dollar, which pushes up prices, the trade deficit tends to increase as an immediate response -- the famous "J-curve" effect.

Over the past 30 months, in which the dollar plunged so sharply against the yen and German mark, Japanese and German manufacturers at first successfully maintained market shares by cutting profit margins. More recently, they have been forced to boost their actual prices.

Now, once again, the dollar is weaker after a period, starting with the Louvre Accord of February, when the major countries had successfully established and defended "target zones" for their exchange rates -- for example, the yen around 153 to the dollar.

But after the June trade deficit figure was published, the dollar began to decline again. At mid-week, for example, it had dipped to 144 yen. A continuation of the dollar slide, as Bergsten and other economists think necessary and expect, will further boost import prices -- exacerbating the American trade deficit in value terms.

It isn't clear when the ultimate goal of exchange rate adjustments -- the discouragement of imports because of high prices -- may actually happen.

To the extent that domestic manufacturers raise their prices, as Chrysler has just done for 1988 model cars, the beneficial effects of a lower dollar are diminished, and inflation gets an upward thrust.

But in the interim, the big trading nations need to climb out of a statistical trap. The Danforth amendment and anything else that would strip emotions from the trade figures would be a big help