Surprise: The 2s10s Spread Might Bottom Early

Bloomberg: A Top Recession Indicator Makes a Lousy Sell Signal for Stocks

Investors ought to get bearish on U.S. stocks not when the curve is bear flattening or the 2s10s spread turns negative, but rather when the opposite trend is in effect, writes Mayank Seksaria, head of macro strategy and research at Macro Risk Advisors. That’s bull steepening, in which short-term yields fall faster than long-term yields as the market prices in monetary stimulus to compensate for a downturn in growth.

"Conventional wisdom focuses on the flattening toward zero or inversion in the yield curve as a precursor to the end of the cycle," he wrote in a note to clients on Wednesday. "While this is nominally true -- the curve does historically flatten prior to recessions -- the utility of this signal to risk management is unclear at best and questionable at worst."

If you're a short-term trader that is absolutely true. There were also false signals in 1989 and 2006. A 10-year Treasury yield below the 2-year isn't a fool-proof signal.

The above story got my attention because it didn't refute convention wisdom. Traders are smart and know how these signals work. Today, it took on more importance because I think it signals a common view in the market. The 2s10s going negative is an early signal. As the analyst above says, it's not the drop that matters but the rise. When the short end of the yield curve is falling and the Fed is cutting rates, that is when the stock market is falling. If you're looking at the 2s10s today, you're probably wondering (like me) when it will go negative and then when the rise in the spread might kick off. In other words, most traders and investors are looking to avoid moving too early. Investors worried about being early would find themselves late to the bear market. The real surprise move would be for the 2s10s to suddenly widen after bottoming between 0.3 and 0.4 percentage points. A rally in the stock market to a new high and a final 2s10s bottom near or below zero would also fit the historical pattern and the conventional wisdom.

The case for a surprise? A rising trend line. History shows the reversal in spread is starting earlier and earlier over time.

Another surprise would be a stabilization or continued drop in the spread because both yields rise in unison, similar to how the 2s10s stabilized after the 1994 bond market panic and didn't make a final low until the year 2000. The really big surprise would be a bond market panic that affects the 10-year more than 2-year, steepening the curve from the long-end. It's hard to see the stock market doing well in any of these other scenarios though.