Heading out the door for the Labor Day weekend, with blue skies and 85° temperatures, I had to share a chart. It’s from the San Francisco Federal Reserve (via Torsten Slok of Apollo) and shows Wall Street expectations for Federal Reserve rate action.
It’s the perfect explainer for why forecasts tend to be so inaccurate.
As you can see, as recently as February 2026, market participants expected a series of ongoing rate hikes—a simple extrapolation from the prior trend.
Then the war began sending food and energy prices higher.
But the Middle Eastern adventure was promised to be short and indeed already over, so the next set of expectations were flat. A few months later, the “short military operation” turned into a war; inflation remained sticky, and expectations were for modestly higher rates.
Now we are six months into a war that shows no signs of ending, voters are angry, and a bad actor is in control of the Strait of Hormuz, with the likelihood of an ongoing tax on Middle Eastern oil. Once again, expectations were adjusted upwards, and now we see “higher for longer” as the consensus.
Unless, of course, something else unanticipated occurs…
Forecasts are for the most part simple extrapolations of the status quo or the current trend; they also fail to include random or unanticipated events – the kind that happens all the time in the economy, markets, and geopolitics.
When you stumble across a forecast that turned out to be more or less correct, it usually means nothing happened, and the extrapolation proved to be randomly correct.1
But most of the time, $h*t happens: wars break out, Pandemics occur, terror attacks happen, new technology comes along and fails or succeeds, and governments fail to fund their annual budgets or wildly overspend their fiscal limits.
The parade of endless random events derails even the most thoughtful of predictions. A year is simply too short a time to guarantee that the dominant secular trend asserts itself, and too long a period to avoid random events.
Previously:
The Folly of Forecasting (June 7, 2005)
Nobody Knows Anything (Archive)
Source:
Productivity -Driven Growth Confronts Elevated Inflation
Huiyu Li
Federal Reserve Bank of San Francisco, September 3, 2026
__________
1. Please note that I said randomly correct — that is not he same as being prescient.
