Now, these numbers are all forecasts, of course, and are subject to the usual problems with forecasts. For instance, the White House budget assumes that the U.S. economy will grow by 3.6% in fiscal 2004 (remember that's the 12 months starting in October 2003) and by an average of 3.3% a year for the fiscal years 2004 to 2008. Inflation will stay at a low 2.1% in fiscal 2004 and average 2.2% for the period. Interest rates for the 10-year Treasury note will hit 5% in fiscal 2004 and average 5.2% for the period.

Shift any of these assumptions and the end result bears little resemblance to the current projections.

Change the time period and the budget forecast looks totally different as well. The budget produced by the White House this year has a serious case of back-loading costs. For example, the Bush administration has proposed creating vast new categories of savings and retirement accounts that would have the effect of exempting gains from taxation when they're withdrawn at some point in the future.

According to the budget, these proposals have no net cost because taxes paid in the early years by investors converting from current IRAs and the like to the new plans would offset any revenue losses from forgone taxes on distributions. And because the projections in this year's budget stop after five years with fiscal 2008, the heavy losses of tax revenue produced after 2008 don't count.

This kind of back-loading of costs produces especially large distortions, because the Bush budget team released budget estimates for just the next five fiscal years instead of the 10-year projections in use beginning in 1996. The White House is perfectly correct in saying that the country got along with five-year budget projections from 1971 to 1995, but the change comes at an awkward moment in the fiscal life of the federal budget, because it is the period after 2010 that will see the big explosion in costs for health care and retirement as the "boomer" cohort shows its age.

For example, the Treasury Department, which still produces 10-year projections, estimates that the cost of the tax reductions that President Bush has proposed as part of his economic growth package will be $1.5 trillion over 10 years. That's more than double the $700 million price tag for five years. The Center on Budget and Policy Priorities figures that the 10-year figure is closer to $1.9 trillion.

How Long Does the Deficit Last?

Most of the argument about the effect of the deficit concentrates on its size. One side claims that the deficit is so big that it will force borrowers to pay higher interest rates as they compete with the federal government for cash. The other side's retort basically boils down to: "Will not." And the experts on either side are off and yelling.

The truth is the U.S. deficit as a percentage of gross domestic product, or GDP, is still very modest. There's a lot of leeway in a $10.5 trillion economy, and the Bush administration is absolutely right to point out that the proposed 2004 deficit is smaller, as a percentage of GDP, than in 12 of the last 20 years. The projected fiscal 2004 deficit of $307 billion is only 2.8% of projected U.S. GDP. Japan would kill for that low a deficit, and France, which recently became the third country in the European Union to face fines for running a deficit above 3% of GDP, can only look on with envy.

But the U.S. role in the global monetary system isn't like that of France or Japan. The U.S. depends on the rest of the world to buy U.S. dollar-denominated financial assets to fund our massive consumption of imports. And the world's economy depends on the continued ability of U.S. consumers to buy those imports. That makes foreign central banks very, very reluctant to dump dollars and dollar-denominated assets onto world financial markets.

For example, you'd expect that Asian central banks would have been selling dollar assets recently. After all, the euro has picked up better than 20% against the dollar since the start of 2002, and investors who switched currencies have seen a sizable profit. But instead, Asian central banks, which have about 75% of their reserves invested in dollar-denominated assets, mostly U.S. Treasury notes, have actually been buying more dollar-denominated assets.

According to CrossBorder Capital, an investment research company, foreign central banks have increased their holdings of U.S. Treasuries by about 10% in the last six months of 2002. Much of that buying has come from Japan, China and South Korea.

It's all part of a global economic deal -- these countries buy our Treasury notes, which keeps the dollar relatively strong vs. their own currencies, and U.S. consumers buy their products.

But this isn't a deal that can withstand all pressures. Foreign holders of dollar-denominated assets do see the value of those assets erode if the value of a dollar sinks vs. the value of their local currency. At some point, the losses from a falling dollar become large enough that U.S. Treasuries become a harder and harder sell, unless they pay a higher rate of interest.

Different groups of overseas investors and institutions reach that switching point at different moments. Individual investors, who aren't charged with protecting their country's trade surplus, already have started to re-evaluate their willingness to hold dollar-denominated assets.

Foreign central banks are still on board, because taking currency losses is less important -- as long as the losses are reasonable -- than making sure that a rising yen or yuan or won doesn't price Japanese, Chinese or Korean goods out of U.S. markets.

But even central banks don't have endless patience. The disquieting thing about the White House budget numbers, to an overseas investor or banker, is that the deficit doesn't come down very fast or very far over the next five years. The fiscal 2004 shortfall of $307 billion does indeed sink to just $208 billion in 2005. But then it sticks there, falling to $201 billion in 2006 and finishing the period at $190 billion. Five years is a long time if you're a banker watching the value of your reserves slowly decline because the U.S. government can't get its deficit down.

What Are the Causes of the Deficit?

In his budget message, President Bush said that "a recession and a war that we did not choose have led to a return of deficits." Call this the cyclical view. A major cause of the surplus that during the Clinton administration, according to this view, was the revenue bubble that went along with the stock market bubble.

Soaring stock prices produced soaring tax revenue. And when the bubble burst, tax revenue went into a steep decline. Tax revenue fell in 2001 and 2002, the first back-to-back decline in 40 years. The 7% revenue decline in 2002 was the steepest drop since 1946.

The good news in this bad news is that, if the deficit is a result of a cyclical decline in tax revenue, then an economic recovery that produces a cyclical increase in tax revenue should rapidly eat into the deficit.

But what if the deficit isn't cyclical but structural? According to this view, the Clinton-era surpluses were temporary and misleading. The surplus was an accounting fiction caused by baby boomer revenue that was arriving before the costs of benefits and retirement for that generation came due. If you looked at the surplus in a longer time frame than an annual budget, it disappeared -- eaten up by increases in health care costs, for example. If you agree with the structural viewpoint, then the temporary surplus should have been saved to meet the bills that were rapidly coming due.

The Bush administration and its allies among supply-side believers have an answer for the structural deficit camp. The Bush tax cuts will create enough extra growth and generate enough extra revenue to meet these future bills.

The structural deficit believers argue the tax cuts make the structural problem worse. They guarantee that when the economy does come out of its cyclical trough, the tax rates in effect won't generate enough revenue to get us back to even a Clinton-era surplus, let alone the kind of surpluses necessary to fund the bills coming due.

I don't think there's any way to settle this argument in the abstract now, and I'm not even sure that the data will be clear enough to declare a winner. After all, none of the parties to the great supply-side experiments of the Reagan administration is willing to admit defeat 20 years after the numbers are in.

What Investors Can Take Away

But using my three-part framework, investors should be able to calculate how much play there is in the system and how much (or little) time there is before truly negative consequences start to appear.

Interest rates go up under all of these scenarios. Even the Bush budget sees rates on the 10-year Treasury note climbing to 5.6% in 2008 from less than 4% now. If the deficit turns out to be worse than currently projected, or if it comes to be seen as structural rather than cyclical, investors can expect interest rates that are higher than that projection.

We're likely to know the answer to the cyclical/structural debate some time in calendar 2004 or so. If the economy does indeed start to pick up -- which in itself will put some upward pressure on interest rates -- but the actual and projected deficit stubbornly won't fall significantly, then investors here and abroad will start to rethink their bets on the U.S. dollar and the assets denominated in dollars.

And that, too, could easily push interest rates above the 5.6% projected by the White House for fiscal 2008.

Exactly how high rates could climb and how fast depends on far more than just the size of the federal budget deficit. But reading the numbers from Washington, investors should plan on an end to 20 years of falling interest rates and building portfolios with climbing rates, even if they are just gently climbing.

Picking the right asset classes for the new investing environment is the topic of my next column. In that piece, I'll take a look at asset classes that most investors don't own but that should now be part of every investor's portfolio planning.