As a Quantitative Analyst at Pinnacle Advisory Group, part of my job consists in applying quantitative methods to financial and economic variables in order to try to distinguish between perceived relationships and relationships that are actually quantifiable in a statistically significant way. In my research for variables that could have a quantifiable and statistically significant relationship with the future growth rate in S&P 500 earnings, I found consumer comfort to be one of the most relevant ones, and by far more relevant than crude oil prices. This makes intuitive sense, as consumer spending represents about 70% of the U.S. economy.

The chart below contains two ways of looking at consumer comfort, as measured by the Bloomberg U.S. Weekly Consumer Comfort Index. The blue line, based on the left scale, represents the level of the index, using a four-week moving average to smooth out some of the short-term gyrations. The red line, based on the right scale, represents the year-over-year change in the index. Over the last year or so, while the overall level of consumer comfort (blue line) has been moving more or less sideways at a level that is fairly positive, the yearly change in consumer comfort (red line) has been on a nosedive, falling from over +30% down into negative territory.

Related: 3 High Octane Value Trades for a Bear Market

Our research indicates that it is the latter way of looking at consumer comfort (i.e. its rate of change) that matters for the future growth of S&P 500 earnings. In other words, it doesn’t matter how the consumer is feeling currently, what matters is whether the consumer is feeling better or worse than it was feeling prior. More precisely, we estimate that each 1% decline in the Bloomberg U.S. Weekly Consumer Comfort Index could be detracting 0.48% from S&P 500 earnings growth over the following year, which means that the recent dive in the rate of change of consumer comfort is taking away a +15% tailwind from earnings.

I have one more piece of good news: Consumer comfort is only one of the variables that make up our quantitative model that we use to forecast S&P 500 earnings growth, and one should never make a forecast based on just one variable. While consumer comfort is clearly heading in the wrong direction, other variables included in the model (crude oil prices and credit spreads, just to name two) have been improving, which is providing the model with some support. At present, the overall model is forecasting that earnings growth will remain moderately negative over the next six to twelve months, and that the speed of the decline should ease and stabilize. This is one of the reasons why we fear that current earnings estimates may be too optimistic. As always, only time will tell.

Sauro Locatelli is the Quantitative Analyst at Pinnacle Advisory Group, a participant in the ETF Strategist Channel.