The warning said three to four years. The sale came in days

The inverse Cramer trade is back in circulation, and this round of it has a clean paper trail. On 31 July the CNBC host asked IBM's chief executive on air whether quantum computers could break the cryptography behind bitcoin. Days later he said he was selling his position. Less than a month after that he told a caller on the same programme to buy it.

The part almost nobody quotes is the answer he got:

“I think that you should give yourself three or four years, and at that point, I would get rather paranoid about it.”

— Arvind Krishna, IBM, on Mad Money, 31 July 2026

Arvind Krishna, chief executive of IBM, asked whether quantum computing threatens bitcoin

Three or four years is a horizon, not an alarm. Read against it, the numbers line up like this:

• The warning's own horizon: three to four years, or 1,095 to 1,461 days.
• The delay before acting on it: days.
• Bitcoin since the sell call: from about $63,700 to $84,194, up 32.1% in 52 days.

Selling roughly 1,100 to 1,500 days before the risk window described by the person issuing the warning is not following the advice. It is reacting to the headline of the advice. Whether quantum computing eventually matters to bitcoin is a serious question with a real literature behind it; nothing about that question was answered in the days between the interview and the sale.

The fund that tried it professionally

Now the other half, because the meme deserves the same scrutiny as the man. Tuttle Capital actually built this trade into products in 2023: an Inverse Cramer Tracker ETF and a Long Cramer fund running alongside it. The long fund closed first. The inverse fund followed in February 2024 with $2m in assets. A strategy that everyone agrees works could not hold enough money to keep a fund open, which is the single most useful fact in the whole genre and the one that never makes the headline.

Both things are true at once. Someone who faded the sell call in early August is up 32.1% in 52 days, and the fund built to do that professionally is closed. The difference is that the meme is a story about one man, while a fund has to be right often enough, across enough names, to cover its costs. We ran into the same gap last week when three AI models produced one identical price target: a signal that looks sharp in a headline gets thin the moment you have to size a position with it.

Why the 32.1% is not a verdict

The market context is worth a line so the 32.1% is not read as a personal verdict. Bitcoin is 43.4% above its July low, and our four tests on whether the winter ended still disagree. Anyone who sold anything in early August has underperformed since, whatever they said on television. The lesson is not that he is reliably wrong. It is that a horizon of three to four years was in the room, and nobody used it.

Informational material, not investment advice. Figures are the bitcoin price at the time of the reported sell call and on 24 September 2026; a 52-day return is not a strategy and is not annualised here for that reason.