It Has Been a Near-Perfect Investing Environment. But It May End Soon.

The perfect investment is one that only goes up. Almost as good is an investment that does well when the rest of your portfolio hits a rough patch, but over time still makes money.

Such a perfect investment shouldn't exist. Yet, for the past two decades, government bonds have offered exactly this free insurance, moving in the opposite direction of shares in the short run but producing gains almost as good as equities in the long run.

The scale of the magic is stunning: From the start of 2000 to the end of last year, holding the latest 10-year Treasury and reinvesting coupons returned 155%, the S&P 500 with dividends 158%, while a 60-40 equity-bond portfolio beat both.

But the magic can't continue forever. If the link between equity and bond prices were to return to what once counted as normal, the magic disappears -- and there are good reasons to fear that could happen soon.

The danger is that bond yields rise without any corresponding strength in the real economy to protect profits and stock prices. The two most obvious causes would be the return of inflation or a shift of stance by the Federal Reserve to stop protecting investors from losses.

Both of those possibilities are worth worrying about.

Start with how shares and bonds behave. Prices of the two biggest asset classes have tended to move in opposite directions since the late 1990s, measured as a strong negative correlation.

This pattern is so well-established it seems like the natural order of things. But since the start of the 19th century, there has been only one other significant period where stocks and bonds behaved this way, according to Ian Harnett of Absolute Strategy Research. The late 1950s and early 1960s had a similar stock-bond relationship to the past few years, and were also the last time inflation was quiescent.

Caution on inflation

The stock-bond link is complex, but depends to a large extent on inflation, uncertainty about inflation and more recently the central bank.

When investors are confident that inflation is under control, they focus instead on the real economy, and economic news pushes bonds and equities in different directions. A strong economy generally means bond yields rise (and so bond prices fall) in anticipation of higher inflation and higher interest rates, while share prices rise in anticipation of higher profits. When there are fears of slowing growth, investors dump stocks and buy bonds.

Fear of inflation alone usually has the same upward effect on bond yields (and so downward effect on bond prices) as economic growth. But inflation doesn't help corporate profits much, while higher yields mean a higher discount rate applied to future profits, which -- in theory at least -- should push down stock prices.

It's too soon to be sure that inflation is awake again after lying dormant for a decade, but there are signs that the tight U.S. jobs market is leading to higher wages. Technological advances such as online shopping still weigh on prices, but with little spare capacity, inflation should pick up. If investors switch focus from the economy to inflation, the nightmare would be higher bond yields and lower share prices.

Inflation itself isn't the only concern. Alongside low inflation has come a belief that inflation has been conquered. The extra yield on Treasurys that investors demand to compensate them for inflation uncertainty, known as the term premium, is extremely low. Inflation options are pricing the lowest chance of inflation being badly behaved over the next five years -- that is, inflation being above 3% or below 1% -- since at least 2009, according to Minneapolis Fed calculations.

It's hard to see how investors could be much less concerned about inflation, so the risk is that anxiety returns, bringing with it higher bond yields and arriving with enough force to pummel share prices.

Fed risk

The final risk is the Fed. Almost everyone thinks that the Fed's multitrillion-dollar bond purchases succeeded in lowering yields and pushing up stock prices. Quantitative easing has only just been put into reverse, and the Fed's $4 trillion balance sheet ended last year only $3 billion smaller than it started.

As the balance sheet shrinks this year, the effects the Fed had on stocks and bonds should also go into reverse, creating upward pressure on bond yields and downward pressure on stock prices.

Worse would be if the Fed's new leadership decided that investors have had it too easy. The late-1990s switch in the stock-bond relationship came as investors realized the Fed would bail out the market with rate cuts in bad times, while letting the good times roll. This asymmetric "Greenspan put" has continued, and will probably become the "Powell put" when Jerome Powell takes over this year. However, if Mr. Powell wanted to take a hawkish tone, he could make clear that the Fed will no longer mollycoddle the markets.

None of these dangers is sure to materialize in 2018. Inflation can stay low for longer. The economy can improve even further. The Fed can keep feeding its friends on Wall Street. Or correlations might be overwhelmed by a new market mania; after all, the S&P 500 managed a near-20% gain in 2017 even as bond yields ended the year where they began. But high on the list of things to worry about is that higher bond yields will finally arrive in 2018, and bring with them not even more new stock-market highs but a correlation crisis.

Mr. Mackintosh writes The Wall Street Journal's Streetwise column. Email james.mackintosh@wsj.com.