With 30-year Treasury yields hovering around 4%, a central question for investors is if and when inflation will strike.

The near-term risk looks negligible, but central banks' extraordinary laxity during the crisis and $1,200-an-ounce gold warn of trouble down the road.

One defense of the Federal Reserve's current policy is its observation that "substantial resource slack" means inflation will remain subdued. Despite some improvement since early 2009, industrial-capacity utilization stood at a miserable 73% in the first quarter. The average since 1967 has been 81%. With that much idle capacity, why worry about inflation?

Like any single measure, it is wrong to put too much faith in this one in isolation. Capacity utilization has a spotty track record in predicting inflation. Big increases in utilization in the 1970s prefigured spikes in core inflation, but oil-price shocks are the likelier culprits. Meanwhile, inflation slumped in the 1990s even as utilization rose.

In a 2005 paper, Philadelphia Fed officials Mike Dotsey and Tom Stark found capacity utilization proved moderately useful in explaining inflation trends until the mid-1980s, but its predictive capacity waned sharply after that.

The timing suggests a range of possible rationales for this change, from globalization shifting the source of inflationary pressure from U.S. factories to Chinese ones, to rapid technological progress, which might distort the data themselves.

When gauging inflation risks, therefore, it is important not to focus on one indicator like capacity utilization. Price expectations can change rapidly, especially under such an extraordinary monetary policy. In particular, it is hard to count on the Federal Reserve scaling back reflationary efforts at the right time, given its record of blowing bubbles.