The Fed is going to have a hard time cutting rates

Outlook

Securities and metals markets are overshadowing economic data and perhaps even central bank narratives. If the stock market is fainting and sentiment turning from greed to fear, why is gold not a beneficiary of renewed risk-off? As for crypto, Bloomberg reports the industry has lost $460 billion in the last week. Added to the vast decline in the tech equities, we are easily looking at a trillion or more.

Where did that money go?

It didn’t go to any other asset (like bonds) as far as we can tell from prices, so it must be sitting in cash.

Panic selling feeds on itself. Whatever the specifics, full-bore faith in AI is now shaken and in doubt. It’s a little funny that a push over the cliff came from Google announcing fresh investment in AI double the expected number. A bridge too far.

At some point, the bottom feeders will emerge and then the regular guys in the gang, and we will get a recovery, if with the pieces on the board re-arranged.

The dollar seems unfazed by all this. Currency moves look normal and not reflective of any lurch into its supposed safe-haven status. But that may be in the cards. Yesterday Fed Gov Cook (yes, her) said she is more worried about inflation than about the labor market, one of the first hawkish remarks. Richmond Fed Barkin has never changed his tune about inflation on the rise.

Reuters’ Dolan points out that PCE inflation, expected around 3% when it comes out late this month, is not only well over the 2% target but headed in the wrong direction. See the chart. Also, “According to UBS, online retail prices captured by Adobe's Digital Price Index rose in January by the most in the series' 12-year history.

“Even though January typically sees higher monthly inflation than other months, price rises this year were the fastest ever seen in that survey.

“On top of this, annual online price inflation of 2.9% is the fastest rate since the post-pandemic period between 2020 and 2022 - the point at which the Fed began its steep credit tightening. This metric remained negative for most of the next three years up until last summer.”

There’s more evidence, like input prices. One important point—we expected tariffs to deliver a jump, but instead the exporters and importers ate most of the higher costs. Smaller companies just went out of business. The point is that tariffs and super-charged growth are indeed lifting inflation, just not in a one-time bump. Fed chief Powell had thought supply chain problems during Covid would be a one-time thing. After tariffs, he thought inflation would be a one-time thing. He was wrong. It’s a creeper, like a garden weed.

The Fed is going to have a hard time cutting rates.

Forecast

We don’t yet know if the euro collapse was a dead-cat bounce but it is starting to look like it. Sterling and the euro are likely responding to their central bank decisions and comments, while the dollar/yen is back on track for that line in the sand at 160. We need to expect jawboning, again. In other words, the dollar is not on an upswing because of better yields or much of anything else except exceptional growth. We don’t see safe-haven conduct. And creeping in may also be acknowledgement that the Fed really shouldn’t be cutting rates but instead thinking about raising them.

Tidbit: Faith in AI is now in doubt. It’s more than a disruptor of conventional methods of doing things and replacing humans in jobs. It’s up-ending the whole tech world and thus all the rest of us, too. However, we just did a many months-long investigation into using AI for technical analysis and auto-trading. The AI’s for that are multiplying like rabbits.

No doubt the top coder-types at big funds have tackled AI to the ground and can use it. The rest of us can go pound sand. You need to understand advanced coding and perform like a developer to use it. The AI’s invent words, use old words wrong with new meanings, have devised a payment system on “tokens,” and do not deliver what a sensible tech analyst wants—a chart with indicators and buy/sell signals clearly marked as the optimum over some past period. Some of the AI’s are good at pattern recognition but have no moving average, momentum or bands/channels.

You have to translate your data feed to the AI (Open AI will have the London data sometime). One method of doing that is to translate your data using CSV. Who wants to do that? And it costs the earth, a minimum of $200/month for the AI plus a bunch of other critical things, for a total of s much as $1600/month. Bah.