If you ever get reincarnated, make sure, as James Carville said, to come back as the bond market. That way, you'll be able to predict who'll win civil wars.

That, at least, is what researchers found when they looked at bond prices for governments facing armed and dangerous internal oppositions. The idea, of course, is that bond investors don't care about whatever -ism people are fighting for. Bond investors only care about getting paid back what they're owed. So they don't let wishful thinking, ideological or otherwise, cloud their collective judgment about who's likely to win a civil war. That's because you won't get paid back if you lend money to the losing side.

From the victors, in other words, go the spoils of bond payments.

You can see just how prescient the bond market has been in the chart.  Bond prices showed the Confederacy didn't have much of a shot during the Civil War. Even before Gettysburg, European markets only put the odds of a Southern victory at 42 percent. That, as you can see below, promptly declined to around 28 percent after Pickett's Charge failed and Vicksburg fell the following day.

The fighting, of course, went on for 18 months, but financial markets could see what the South perhaps couldn't: it didn't have the industrial capacity or the manpower to stop the North (at least not with Ulysses S. Grant in command).

The same phenomenon could be seen in China's civil war. The following chart uses the Chinese Nationalist government's bond prices to show what chance markets thought it had of beating the Communists. The answer: not much of one. As soon as it became apparent that there would, in fact, be a civil war, and not a negotiated settlement, the Nationalist government's bonds tanked. And that didn't even change when they captured the Communist capital at Yenan. Instead, markets kept marking down the odds of a Nationalist victory, a full three years before Americans got around to debating who had "lost" China.

A house divided against itself may not be able to stand, but once it's together again, one side won't be paying back its debts—and markets are pretty good at figuring out which one that will be.