Eisman Says 65% of the S&P 500 'Is Basically AI' as Ramsey Calls Veteran's $500,000 Safe

Three days before Dave Ramsey told a disabled veteran his $500,000 S&P 500 investment was safe, Steve Eisman told a podcast audience that "65% of the S&P is basically AI. So you own an index, it's basically AI." The clash matters well beyond one caller's portfolio: it cuts to the question of whether the most popular investment product in America still counts as diversified.

The caller at the center of the story is Jamal, 41, a 100% disabled veteran from New Orleans. He put $500,000 of a roughly $1 million oil field injury settlement into the S&P 500 and kept another $100,000 in a money market account. On Thursday's Ramsey Show, Dave Ramsey told him: "you were very conservative and very smart to get started." Ramsey projected the money would double roughly every six or seven years and reach $8 million by the time Jamal turns 69.

Eisman, the investor made famous by The Big Short, described the same index three days earlier in far less reassuring terms. In his view, investors who believe the index protects them "are not okay" if something bad happens to the AI trade.

What does each man actually mean?

The two statements aren't as contradictory as they first sound. Ramsey's advice works for this specific caller, but "conservative" is the wrong word for it, as the source analysis itself concludes. An S&P 500 fund is a reasonable long-term holding for a debt-free man with lifetime government income. It is also a pure stock position with a heavy tilt toward a single theme.

The arithmetic behind Ramsey's projection holds up. Turning $500,000 into $8 million over 28 years requires about 10% a year, roughly the index's long-run pace. His on-air market figure was nearly right too: the SPDR S&P 500 ETF Trust (NYSEARCA:SPY) was up 13% year to date, against the 12.5% he cited live.

The word "conservative" falls apart when you look at diversification. The Vanguard 500 Index Fund ETF (NYSEARCA:VOO) held 38% of net assets in information technology at midyear. Eisman's 65% figure goes further because it adds the utilities supplying data centers, the industrials building them and the banks financing them. Only staples and healthcare, "which add up to only 14% of the S&P," sit outside the AI trade in his accounting.

Apply that to Jamal's money and the stakes get concrete. Roughly $325,000 of his index investment rides on continued AI spending.

NVIDIA (NASDAQ:NVDA) shows what that exposure feels like in practice. The company's market value stands near $4.53 trillion, and the stock fell 22% over the past month, during what Fortune called tech's biggest selloff in years. Every S&P 500 owner absorbed part of that drop, whether they knew it or not.

The takeaway: an index fund is diversified across companies, but increasingly concentrated in one bet.

Why one question decides whether a crash hurts or just stings

The variable that matters is whether the portfolio ever has to pay the bills during a downturn. Consider a hypothetical 40% decline. Jamal's $500,000 becomes $300,000. Getting back to even then takes a 67% gain, or about 5.4 years at 10% a year.

Scenario one: he never sells. Ramsey's projection assumes Jamal lives on his VA disability income and leaves the money alone. In that case the 40% loss exists only on paper. He keeps every share, and they all recover together.

Scenario two: he withdraws $30,000 a year, an example figure. Before a crash, that's a 6% withdrawal rate. After a 40% drop, the same check consumes 10% of the portfolio. Shares sold near the bottom never come back for the recovery. This is sequence-of-returns risk: two investors with the same average return can end up far apart depending on when the bad years arrive.

Jamal's $100,000 money market account is his shock buffer. At the current one-year Treasury yield of 4.3%, that much cash generates about $4,280 a year with no AI exposure. Eisman's advice for money he wants protected: "I would just buy short-term Treasuries. The yield is decent and they are safe."

Ramsey's framing makes more sense once you hear his own setup. In the same episode, he mentioned holding no bonds and no gold, only paid-for real estate and growth stock mutual funds. For him, stocks are the default and everything else needs a reason.

The takeaway: a crash only becomes permanent damage if you're forced to sell into it.

What can an ordinary investor do before the next AI selloff?

The source lays out three specific moves:

• Measure your cash runway. Add up the yearly expenses your guaranteed income doesn't cover. Divide your cash and Treasury holdings by that number. If the answer is under three years, a long bear market could force you to sell stocks at a loss.
• Check your fund's real AI share. Pull your index fund's latest shareholder report. Add up the technology, communication services and utilities weights, then compare the total to Eisman's 65%. That number is your actual exposure to one theme.
• Run the 40% test. Multiply your stock balance by 0.6. If the result is a number that would push you to sell, review whether your cash and short-term Treasury holdings could cover your withdrawals until a recovery.

The bottom line is blunt: an S&P 500 fund is safe only for money you'll never need to sell during a crash.

Why this argument matters now

The Ramsey-Eisman split isn't an academic debate. The S&P 500's concentration in AI-linked stocks has grown with the market values of NVIDIA and its peers, and millions of retirement accounts hold the index through funds like SPY and VOO. When one investor famous for calling a bubble says nearly two-thirds of the benchmark is a single trade, and another famous for personal finance advice calls it conservative, ordinary savers get two contradictory messages about the same product.

For Jamal, the question resolves cleanly as long as his VA disability covers his living costs and the $100,000 buffer stays untouched. For everyone else holding the index, Eisman's math suggests the honest label isn't "safe" or "risky" but "leveraged to one story" and the job is deciding whether you can wait out that story going bad.