Bubbles last until they feel like fundamentals,” my old Credit Suisse colleague Neal Soss likes to say. A simple comparison of the S&P 500 with total corporate profits as reported in the US National Income and Product Accounts suggests that equity prices have run far ahead of underlying earnings, just as they did before earlier market crashes.

With 10-year US bond yields at just 1.5% and nearly $20 trillion of bonds globally offering negative yields, investors are starving for income. High-dividend “defensive” stocks like health care, consumer staples and utilities offer an alternative to bonds, but with a difference. As I warned in this column July 24, supposedly defensive stocks have levered their balance sheets to record levels of indebtedness in order to pay out money to shareholders.

As of the second quarter, several S&P sectors paid out more in dividends and buybacks than they earned (I use EBIT, or earnings before interest and taxes): industrials, health-care, real estate and consumer staples.

The sectors who return the most money to shareholders have increased leverage the most during the past 10 years. Utilities, traditionally a bond-like sector suitable for retirees, now have net debt equal to five times their earnings before interest and taxes.

Among the largest-capitalization constituents of the S&P, notionally defensive stocks are returning money to their shareholders faster than they can earn it.

The S&P may be flat over the past 365 days, but most of its sectors are deeply in negative territory.

Positive returns have come from the tech sector, which is up slightly, and from utilities and consumer staples, which are up substantially. This means simply that the S&P is suffering from the so-called Red Queen effect: It has to run faster (that is, put on more leverage) simply to stay in place. That can’t be good.

Consumer spending has been stronger than I expected it would be. Strong jobs growth explains part of the consumer story, although job growth hasn’t been as strong as first reported. The latest revisions from the Bureau of Labor Statistics reduced the jobs count during the year through April 2019 by 500,000. Hours worked have also been weak. What the data suggest is that the stock prices of the big consumer companies have performed well not because their revenues are stronger, but because they are borrowing to return money to shareholders.

That’s as close to a rigorous definition of a financial bubble as I can find. And the one thing we know about bubbles is that they eventually pop.