The Chart That's Keeping Goldman Up At Night

The spread between "hard" and "soft", or survey and sentiment data, ever since the election has been extensively noted and discussed on this website in recent months (especially since over the past two months the soft data has rolled decisively over, while the Citi economic surprise index has crashed at the fastest pace on record). Which is why it will come as no surprise to readers that, as Goldman writes in a note looking at "Peak Sentiment", over the past six months, US “sentiment surveys” have outpaced both “activity surveys” and “hard data”. This rise in sentiment has accelerated in the post-election period, prompting many (Goldman included) to link the surge in sentiment with optimism about the new administration’s pro-growth policy. It has also spilled over into equity markets in general, and institutional and retail traders in particular.

That is a problem for one simple reason: as Goldman points out, based on empirical studies, "positive sentiment negatively correlates with future asset returns", or the more confident the market, the greater the subsequent drop. Additionally, Goldman notes that "the correlation between consumer sentiment and market sentiment suggests that the former might be useful as a contrarian market timing signal."

To shows this possibility, the following Exhibit plots year-ahead 12-month returns on the S&P 500 vs the Conference Board Consumer Confidence index. The plot suggests that for values of the CBCC below 50, median returns are significantly higher.  Conversely, when the CBCC moves much above its current level of 120, the expected median return begins to decline and eventually turns negative (the median return is 9.7% for the CBCC at 120, falling to 5.0% at 130 and to below zero above 136). In results not reported here, we have also explored this timing signal on other assets and found a similar pattern for high-yield credit spreads, but not for 10-year yields or the trade-weighted Dollar.

But forget broad consumer confidence. What is the result when focusing only on sentiment among market participants? Well, for one, a very nervous Goldman Sachs.

As the firm's credit strategist Charles Himmelberg explains, Goldman is worried because "more direct measures of market sentiment provide still more reason to worry that market sentiment may be over-extended."

To demonstrate this point, Goldman shows data from the Yale Stock Market Confidence Indices, plotted in the following exivit, which show that the “one-year-ahead confidence” of institutional investors has recently risen to its highest level on record, with a similar jump in the sentiment of retail investors. 

This index calculates the percentage of respondents reporting positive expected returns for the year ahead. In February of this year, this ratio for institutional investors had risen to an all-time high of 99%. In August of last year the fraction was at 73%. And for retail investors, the current  fraction is 88%, up from 67% in August.

Putting it all together, this is the one chart the keeps Goldman up at night.

And two bonus charts: this is who is most optimistic about the future - mostly Republicans and Trump supporters.

It's not just record investor confidence that is worrying Goldman.

In addition to “peak sentiment”, we see other reasons to think the risks from here are skewed to the downside. The most recent China PMIs weakened meaningfully, perhaps portending a slowing of growth as policy tightens. US reported growth in the first quarter was softer than expected, and the hard data remain only so-so. With the US labor force seemingly at full employment, the economy now runs the risk of overheating. Whether or not one is inclined to put much weight on the view that animal spirits have gotten too far ahead of reality, a lot now rides on the hard data.

And the best indicator for that will come tomorrow in the April payrolls report. As for whether's ongoing bearishness is correct, the big - and maybe only - wildcard is what the central planners at the Fed and other central banks will end up doing: they have an unpleasant habbit of crushing even the most confident negative forecasts.


No comments yet! Be the first to add yours.