Global Macro Monitor

QOTD – Quote of the Day

h/t:  Dr. W., the best!

Weekly Summary

Global markets correlated as theory would predict:  interest rates higher, the dollar stronger, and stocks and commodities lower.

Ten-year yields were up across the globe as the U.S. Treasury note yield approaches 4 percent on sticky inflation data. Consequently, world stocks fell across the board, with the S&P 500 down 2.7 percent on the week and the Dow now down 1 percent, YTD.   Biotech, real estate, and retail were hammered. 

The dollar index strengthened about 1 percent with big moves lower in the Chilean Peso (copper related) and the Russian Ruble.

Copper fell below $4.00 though natural gas put in an impressive rebound but the vapor is still down yuuge on the year.  

We are baffled that the markets are baffled that disinflation may have bottomed.  We see the issue as the myopic memory of the massive stock of central bank injections into the global economy coupled with too much focus on the price rather than the quantity of money, which is still at flood stage levels.  

Levels do matter folks. 

Good piece in the Washington Post today on the “cost of thriving index  (COTI),” which provides some granular insight into how America’s middle class is faring.  

In our opinion, the economy’s unexpected mojo and sticky inflation can be explained in two simple macro charts.  M2 as a percent of GDP remains freakishly high, though we don’t put much stock in the old monetary aggregates, and Household Wealth as a percent of GDP remains off the historical charts.  

Two of the deadly inflationary sins:

1) Money for nothing
2) Wealth without production 

Levels matter.  

Wow!  The wing-nuts are coming for AI and ChatGPT…err…WokeGPT.

Prepare for an invasion of the Luddites. 

Senator Ted Cruz of Texas, for instance, has called attention to conservative complaints that the AI refused to write a song praising him, but it did so for Fidel Castro. – Bloomberg BW Daily

Not that the new generative AI chatbots are without sin….

Mistakes are inevitable, because large language models don’t have any inherent sense of truth. ChatGPT has a tendency, when confronted with a question it can’t answer, to simply make up something. These “hallucinations,” as AI experts call them, can be extremely convincing. Last year, Meta Platforms Inc. pulled down a large language model known as Galactica, which was supposed to be “a curated corpus of humanity’s scientific knowledge,” after critics pointed out it had a tendency to invent science behind queries such as “the benefits of eating crushed glass.”  –  Bloomberg BW Daily

DALL-E 2 – An Andy Warhol painting of Luddites smashing a computer