This tweet gets to the heart of far too much economic analysis.
Now Larry Summers is a brand unto himself and has held economic positions that most could only dream of; and he earned them. But I'm going to say my piece anyway.
Far too much of economic analysis is looking for historical parallels and overlaying them over the present. That's natural and it's the precise reason we study history but laying the charts over directly is non-sense. We see the same thing over and over again as a chart of one thing is laid over another or a historical chart matches up with a current one.
It never works.
A year ago, everywhere you looked there were charts highlighting things like how there had never avoided a recession in the following year with UMich consumer sentiment so low.
Well, a year later the US consumer continued to power the economy. As I often wrote at the time:
"The University of Michigan consumer sentiment survey is a joke of an economic data point...What the UMich survey measures is the political mood and gasoline prices, nothing more."
Whatever predictive power this survey had died long ago.
The same argument about a recession was made about the declines in US manufacturing and related indexes. All of that ignored that manufacturing was nowhere near as large a part of the US economy as it was 40 years ago. And even more importantly, it failed to understand the bullwhip effect from the pandemic -- which was obvious.
I've ranted about all this before and I refrained from bringing it up two weeks ago when the chart that Summers posted was first doing the rounds.
So much changes in an economy in 40 years. In the 1970s, the US wasn't globalized and millions of women were piling into the workforce, doubling the earnings of many households. The geopolitical backdrop is also different, as it always is.
It also irks me how US-centric so much economic analysis is. There are dozens of countries with good economic data to draw on yet the economic elite only seem to pull from the US data set. A recent example is the San Francisco Fed's look at rental inflation using only an extremely limited set of US data. I could go on but I'll save that for a different post.
As for chart overlays in general, the best thing to do is completely ignore them but in the click-baity world of 2023, they're an irresistible impulse, even to someone like Summers who should know better. He also stole it. And if that's not bad enough, note the different axis' on the left and right of the chart, which is always a clue that something dishonest is going on.
Now, if an overlay is something like gold and gold stocks then it's probably valid but when it comes to the broader economy, it's never this simple. The bottom line is that there is far more creativity and critical thinking needed in economic analysis; and in trading too.
