If you're worried about predicting the next economic recession, there's a key indicator in the financial markets to watch: the yield curve. While it typically slopes upward, the curve has flattened out recently and even turned downward — a phenomenon many observers warn is our best predictor of a recession.

"The market’s most reliable recession indicator is finally flashing red," said James Mackintosh in the Wall Street Journal. "With the Treasury yield curve inverting on Friday—the 10-year yield fell sharply to be lower than the three-month for the first time since 2007—is it finally time to prepare for an economic downturn?"

So what does this actually mean?

Basically, the yield curve reflects the relationship between the yield of Treasury bonds over different lengths of time. Usually, the yield on a bond — the interest that will be paid after the bond's term — is larger the longer the term is. But in an inverted yield environment, as is beginning to emerge, the three-month yield can actually be higher than the 10-year.

Why should we care about this?

Well, many analysts note that this phenomenon typically precedes a recession, as economist Paul Krugman showed in a recent Tweet:

It's not that the inverted yield curve triggers a recession. The curve is just determined by the bond markets. What it does is give us a glimpse into the what investors are expecting for the future — and when the curve is inverted, it means investors are nervous.

Bond purchasers "see economic weakness ahead, and believe that the Fed will be cutting rates," Krugman explained.

"That isn’t quite the same thing as predicting recession, since the Fed can cut rates without recession," explained Mackintosh. "Indeed, the two times that yield curves inverted on most measures without recession were in 1998 and 1965-66, both times when the Fed slashed rates and the economy continued to grow."

But the Fed often does cut rates because of a recession, and as Krugman explained, the inverted yield curve is frequently followed by a downturn.

So should we be worried?

For most people, the best financial advice is broadly the same — you're not going to beat the market by watching different economic indicators, and you're best off with a diversified and stable set of investments. But recessions are bad above and beyond what happens to your own investments or whether you even have any, so it's important to know what might be coming down the pike.

While the yield curve inversion is making some analysts nervous, others are less worried.

Mackintosh, for instance, suggests we should be more skeptical about the historical evidence:

The yield curve might be less reliable than its recent U.S. history suggests. It has a terrible record internationally, for instance. It flat-out hasn’t worked in Japan, also has a poor record in the U.K. and in Germany provided no advance warning of the 2008 recession, the worst since reunification. At the moment the curve isn’t inverted in any of them thanks to superlow or negative interest rates, even though all are struggling with greater economic troubles than the U.S.

He also noted that the inversion doesn't necessarily reflect immediate recession; it could arrive in two months or two years. His conclusion is that all the development tells us right now is that the economy is weakening — but the Fed had already sent that warning sign loud and clear.

Krugman, on the other hand, thinks people might be underestimating the risks.

"There is, however, a twist right now which makes the inversion even scarier than it looks, at least a bit," he explained. "Previous inversions took place at much higher short rates -- 5 or 6 percent, versus 2.5 percent now."

He continued: "So this inversion actually reflects a worse outlook for the economy than the number itself suggests. Again, it's not a direct read on the economy; it's a read on what the average bond investor thinks is happening to the economy. But still not encouraging."

When Vox's Matt Yglesias wrote about the phenomenon on Monday, he noted that in developed countries "yield curves seem to have grown structurally flatter over time" — which would make an inversion a much less significant indicator than we think.

But on Wednesday, he tweeted: "When I started working on a yield curve inversion explainer, I was prepared to conclude that worrying about this is overblown but then it totally dropped out of mainstream coverage and I actually think people aren’t panicking enough."