One of last year’s Albie winners was Ken Rogoff’s Boston Globe op-ed arguing that a Trump administration could stimulate a serious economic recovery by unleashing the “animal spirits” of investment. Even he acknowledged, however, that, “all of this is an optimistic spin on a Trump economy. If the new administration proves erratic and incompetent (a real possibility), dejection will quickly overwhelm confidence.”

In the months after Rogoff wrote that op-ed, there was a curious divergence between the “soft data” and “hard data” about the American economy. It is undeniable that Trump’s election had a salutary effect on the soft data, which measures perceptions about the state of the economy through surveys. Whether one looked at consumer confidence, small business, or manufacturing surveys, the results were the same: Trump seemed to be good for business.  The president tweeted about this earlier in the month:

Credit where it’s due: president Trump has made a lot of Americans feel better about the economy.

As I was typing those words, however, I was struck by a powerful sense of déjà vu. Because that data bears a strong resemblance to another country that was feeling pretty great about it’s economy a few years ago: Russia. As I noted at the time, Russia’s objective economic conditions had worsened, but Putin’s annexation of the Crimea had caused Russians to feel better about their economy anyway.

Sure enough, if you look at the “hard data” — you know, things like economic output, home sales, or commercial bank loans — the “soft data” seems wildly, absurdly optimistic. It’s not that the hard data is truly bad, it’s that it shows no real difference or surge in real economic growth.

As a result, the divergence between the soft data and the hard data seems pretty big:

So which set of figures is right, the hard data or the soft data?

One could spin ways for one set of figures to converge to the other. If enough businesses and consumers are confident in the economy, they could start spending and investing and generate real economic growth. If, however, investors and consumers start to lose faith in the Trump administration, the soft data could come back to earth.

The past week suggests that its the soft data that’s about to give. After Trump’s health care fail, it has begun to dawn on markets that maybe all the expected tax and regulatory benefits won’t be forthcoming:

The Dow Jones Industrial Average just notched its longest losing streak in six years.  Bloomberg reported on other data suggesting that markets are starting to lose the contact high that comes from listening to Donald Trump:

The dollar, sometimes a bellwether for U.S. economic confidence, has slumped almost 2 percent in March and is close to erasing a more than 6 percent rally since Trump’s election in November. The Bloomberg Dollar Spot Index, a gauge of the American currency against a basket of global peers, is down 4.9 percent since peaking at a record Jan. 3….
After an extended leave of absence, the so-called fear gauge is back. The Chicago Board Options Exchange Volatility Index jumped 15 percent last week, the biggest such gain this year, as it became apparent the health-care bill was not going to survive a vote in the House.

This has international ramifications as well. Global markets had been engaged in the “Trump trade,” betting that U.S. interest rates would rise, the dollar would appreciate, certain sectors would benefit from a Trump presidency. As the Financial Times’ Robin Wigglesworth reports, however, ” Many popular bets — such as on a stronger dollar, rising bank and small company stocks and more volatility — have fizzled.”

None of this means that the U.S. economy is headed for a recession. And the Trump administration has been bound and determined to make big business feel welcome, whether through access to Trump himself or through Jared Kushner’s new “SWAT team of strategic consultants.”

Still, each day the hard data doesn’t match the soft data, the latter will be more likely to converge towards the former. What Mohamed El-Erian calls “The Confidence Economy” will come to an end.

If this happens, count me among those who will not be surprised. As I wrote in January, when the stock market was surging:

Political commentators should actually interpret the markets’ reaction to Trump … by acknowledging that markets are not the great predictor that some scholars have claimed. Social scientists have long assumed that investors have a powerful interest in understanding how politics works. Certainly, since 2008 more and more financial actors have invested in political analysis.
Actual political behavior can confuse them, though. Market participants are adept at identifying the economic pressures that could force politicians to act. Predicting how politicians will react to those pressures is where traders may fall down … Politicians have different incentives than market participants — a fact that sometimes escapes for-profit actors when they think about the world.

The hard-working staff here at Spoiler Alerts was pretty confident that the current president would not be good at his job. The political soft data backs up this perception, and the AHCA fail is the first piece of hard data to do the same (it is possible that the political soft data has its own problems, of course).

Financial markets are late to this party. It appears, however, that they are starting to recognize reality.