As these will be my last words in these pages, it’s time to look back and take stock. It has been nearly four and a half years since I began writing this column, which works out at roughly 240,000 words altogether. As these will be my last words in these pages, it’s time to look back and take stock. If part of your job is to be a pundit then, as the University of Pennsylvania political scientist Philip Tetlock argues in “Superforecasting: The Art and Science of Prediction,” you need to keep score.

As Tetlock had a dig at me in that book — which was published in 2015, before I began writing for the Sunday Times of London and then The Globe — this is also a good opportunity to settle a score.

Tetlock was of course quite right about most public intellectuals (a term that always makes me think of public convenience). They seldom hold themselves to account. Nor do they get fired if their predictions are consistently wrong, so long as they are entertaining enough to engage readers’ eyeballs.

However, since I set up an advisory firm nine years ago, my approach has been different — out of necessity, as fund managers differ from newspaper editors in their attitude to predictions. Not only do they notice when you’re wrong, because one or more financial indicators make that clear; they also let you know about it (with grim relish, usually). If you’re wrong too often, it’s goodbye.

So at the beginning of each year we at Greenmantle make predictions about the year ahead, and at the end of the year we see (and tell our clients) how we did. Each December we also rate every predictive statement we have made in the previous twelve months, rating them either True, False or Not Proven. In recent years, we have also forced ourselves to attach probabilities to our predictions (not easy when so much lies in the realm of uncertainty rather than calculable risk). We have, in short, tried to be superforecasters. And with some success.

Now it’s time to apply the same retrospective scoring to this column.

By far the biggest issues I wrote about in my first 12 months were Britain’s vote to leave the European Union and the election of Donald Trump. How did I do?

On Brexit, I was wrong. From the outset, I was a Remainer. “The idea that we can … separate ourselves from Europe is an illusion,” I wrote on February 21. “For the future of Europe without us would be one of escalating instability.” Impolitely, I called Brexiteers “Angloonies” and “happy morons.” When Remain lost, I predicted a “stairway to hell” — or at least a recession (June 26). Wrong.

At the end of the year, I made a confession. I had been motivated to back Remain more because of “my personal friendship with David Cameron and George Osborne” than out of any deep allegiance to the EU. I regretted — and still regret — not urging Cameron to reject “the risible terms that the European leaders offered him back in February on EU migrants’ eligibility for benefits.” That was the moment he should have called their bluff by backing Brexit.

Yet the humiliation of Brexit gave me an advantage over American commentators on the 2016 presidential race. I had moments of doubt, admittedly. I compared Trump to unsuccessful Republican candidates Wendell W illkie and Barry Goldwater. I also predicted the bursting of the Trump bubble in the Wisconsin primary. Ted Cruz won that, but it didn’t burst the bubble. Much more often, however, I went against the conventional wisdom that Trump was doomed to lose.

“Trump has the face that fits the ugly mood in America,” was my headline on November 1, 2015. “Trump has both the resources and the incentives to press on. In the current national mood of disaffection with professional politicians, he could seem an attractive alternative to Hillary Clinton …

The point about Trump is that his appeal is overwhelmingly a matter of style over substance. It is not what he says that a great many white Americans like — it is the way that he says it.”

I was against Trump. I was a signatory of a “Never Trump” letter. I repeatedly condemned his “open expressions of racial prejudice and xenophobia,” his isolationism and his fishy bromance with the Russian President Vladimir Putin. I regretted that Mike Bloomberg chose not to run.

But I also saw clearly the strength of Trump’s appeal. “He is winning,” I wrote “because no other candidate has a more convincing explanation of why so many Republican voters genuinely are worse off today than in 2000 … No one can rule out Democratic defections to Trump when it comes to the crunch on November 8.” I later imagined him winning and running for an unconstitutional third term in 2024. “Trump can beat Hillary Clinton,” I wrote in May.

“Can Trump succeed where Mitt Romney failed?” I asked in July. “Yes. … Many young voters will fail to show up for Clinton. Meanwhile, the white lower class, especially the older cohorts, will turn out for Trump in droves, just as their English counterparts turned out for Brexit.” The choice between Clinton and Trump was a choice between snafu and fubar, I wrote in September, “but wouldn’t you risk being fubar … if it was your only shot at avoiding four more years of snafu?”

“The rage against the global,” I wrote a week later, “is why Trump could win this election. It is why Brexit happened. It is why populists are gaining ground wherever free elections are held.”

I marked my first anniversary of column-writing with a piece that compared Trump with the Chicago Cubs, the outsiders who had just won the baseball World Series. “If there is a differential in turnout between his supporters and hers in the battleground states comparable to the age- and ethnicity-based differentials in the UK referendum,” I wrote, “he can win."

Now, dear reader, you are burning to know what I think will happen this November. Bad luck. You will have to seek my superforecast in Bloomberg Opinion.

Niall Ferguson is the Milbank Family Senior Fellow at the Hoover Institution at Stanford University, and managing director of Greenmantle.