When financial panics begin, they play out in similar ways. First, one asset class has a surprise drop. The leveraged investors sell the sinking asset, but soon the asset is unwanted by anyone. Margin calls roll in. Investors then sell good assets to raise cash to meet the margin calls. This spreads the panic to banks and dealers who were not originally involved with the weak asset.

Soon the contagion spreads to all banks and assets, as everyone wants her money back all at once. Banks begin to fail, panic spreads and finally central banks step in to separate winners and losers and re-liquefy the system for the benefit of the winners. Typically, small investors (and some bankrupt banks) get hurt the worst while the big banks get bailed out and live to fight another day.

What varies in financial panics is not how they end but how they begin. The 1987 crash started with computerized trading. The 1994 panic began in Mexico. The 1997–98 panic started in Asian emerging markets but soon spread to Russia and the big banks. The 2000 crash began with dot-coms. The 2008 panic was triggered by defaults in subprime mortgages. What will trigger the next panic?

This article by prominent economist Carmen Reinhart says the place to watch is U.S. high-yield debt, aka “junk bonds.” Right now, spreads between yields on emerging-market debt and U.S. high-yield debt are diverging. This suggests that either emerging markets are underpriced or U.S. junk bonds are overpriced. Reinhart suggests the latter.

Watch these spreads. If they start to converge with U.S. junk bonds falling in price, that may be a signal that a new financial panic is not far behind.

Institutional investors can schedule a proof of concept with the world’s first predictive data analytics firm combining human and artificial intelligence with complexity science. Check out Jim Rickard’s company at Meraglim Holdings to learn more.