scorecardresearch Skip to main content
Evan Horowiwtz | Quick Study

How can you tell there’s a recession coming?

By one metric, stock prices have reached the second-highest level ever, behind the dot-com spike of 1999-2000.Richard Drew/AP

It’s inevitable. Someday our long run of economic growth will end in a damaging recession. Jobs will disappear, paychecks will shrink, investors will panic, and (hopefully) Washington will mobilize to minimize the damage.

This prospect may seem rather remote, following recent news that the nation’s GDP grew at a healthy 4.1 percent annual rate in the second quarter, but economies sometimes sour quickly. Just look at the circumstances leading up to the last recession. A bare year before it began, the economy was expanding by 3.5 percent.

In fact, myopia has become one of the hallmarks of modern recessions, cousin to the “irrational exuberance” that takes hold at the peak when people start to feel that nothing can go wrong.


Which means there are really two different kinds of indicators to watch when trying to suss out the odds of a coming recession: evidence of worrisome economic trends but also signs of excessive psychological optimism.

It’s not hard to spot examples of both right now.

Consumer sentiment is one useful gauge of that over-confidence. Higher numbers are generally considered good news, a show of faith in the economic climate. But at some level it seems to signal collective blindness, often reaching a peak about a year ahead of recessions.

Right now, consumer sentiment registers at a relatively high 97.9, well above its long-term average of 86. That’s high enough to be approaching records that we may not want to break.

Before the 2007 recession, the peak reading was 96.9. In January 2000 — a year before the dot-com bubble burst — it was 112.

The stock market is another place to look for excess enthusiasm, though it won’t do simply to check the level of the Dow Jones industrial average or the Standard and Poor’s 500. What’s needed is some way to track how stock prices compare with fundamentals, like actual corporate profits.


One widely used approach is to compare today’s stock prices with corporate earnings over the last 10 years — rather than just the most recent quarter.

By this metric, stock prices have reached the second-highest level in history, behind the dot-com spike of 1999-2000.

Just as these overconfidence-metrics are starting to smoke, there are also more direct signs of economic trouble, including in the housing market.

Recent months have seen a slight uptick in housing inventory —

meaning a growing number of unsold homes on the market. That tends to be a good, leading indicator of recession, as does the number of new home sales, which has flatlined in recent months.

Housing indicators have this weakness, though: they missed the dot-com recession entirely. And it’s possible they won’t capture the full depth of the next one, if the demographic forces reshaping the housing landscape — growing demand from millennials, aging boomers — swamp the cyclical issues.

There is, however, one indicator that has predicted every major recession since the 1970s, and it, too, is approaching problem territory.

Ignore the technical-sounding name — the inverted yield curve — and the concept is straightforward. When short-term government bonds offer higher rates than long-term bonds, that’s a bad sign for our future.

The reason? Usually long-term bonds have to pay a premium, in order to persuade investors to lock their money up for an extended period, at the risk of missing good investing opportunities in the meanwhile.


This changes when investors see trouble ahead. Suddenly, they prefer to have their money tied up, so they can earn steady returns even as the economy crashes and rates tumble.

This inversion hasn’t happened yet. Currently, 10-year Treasuries pay a little bit more than 2-year notes, but not much. And the difference has shrunk considerably, from over 2 percentage points in early 2014 to just under 0.3 percentage points as of Thursday.

It’s possible the yield curve will linger at this low level for some time, but if it does invert that would be the clearest sign yet of a coming calamity.

None of this speaks to the question of what will actually cause the next recession. Candidates abound, everything from the escalating trade war to potentially overhasty interest rate hikes by the Federal Reserve.

But tracking causes is often a better topic for the post-mortem, when we can map out the full impact of subprime mortgages, and wonder at the unworkable business model of

For now, it may be better to bracket the search for causes and just watch for worrisome symptoms. Like when everyone becomes convinced that this time really is different.

Evan Horowitz digs through data to find information that illuminates the policy issues facing Massachusetts and the nation. He can be reached at Follow him on Twitter @GlobeHorowitz.