Gunjan Banerji

Two financial advisors I spoke with said, hey, this does get a passing grade.

You know, this isn't exactly what I would say to investors, but this is not a bad starting framework in terms of a broad portfolio of diversified index funds.

So Gunjan, first of all, let's get the disclaimer out of the way.

We don't advise letting AI make your money decisions and wealth managers generally say it's a bad idea.

That said, tell us about this experiment you set up.

So a few months ago, I decided to test what it would be like to ask ChatGPT to serve as a fiduciary financial advisor.

Basically, I crafted a prompt giving it my age, how much money I had to put to work.

I had around a million dollars to invest.

I stated what my risk tolerance was, what my goal was, and this would be in a taxable investment account.

And then over the course of many months, I kind of peppered it with questions on major market events, everything from the government shutdown to Trump's trade war and then recently the war with Iran.

And how did it perform picking stocks?

When I asked ChatGPT, hey, there's a trade war underway.

How should I position?

It did suggest making several different shifts to my portfolio and then asked it to select some stocks for me that might outperform.

Through the trade war, when I checked in on how that basket was doing, interestingly, it had outperformed the S&P 500 over a period of time.

Recently, I had a publication.

I checked in on it again.

The basket has risen around five, six percent since mid-October when I first asked it that question.

And that underperforms the S&P 500's roughly eight or nine percent gain since then.

And, you know, what financial advisors told me was like, hey, this is not a situation in which you should be making too many shifts to your portfolio.