Jensen Huang Just Resigned From Nvidia. A Study of 187 Forced CEO Exits Shows What Could Happen Next

Jensen Huang resigned as Nvidia’s chief executive on Thursday morning, and a study of 187 forced CEO exits at S&P 1500 companies has a reading on what tends to come next for the stock. The resignation, as president, CEO and director, took effect immediately. Colette Kress, the chief financial officer since 2013, is interim CEO and keeps the CFO job while the board searches inside and outside the company for a permanent successor, and Harvey C. Jones, a former Synopsys CEO, became the board’s first-ever chair. Nvidia closed Thursday at $166.19, down 4.3%, and closed Friday at $162.85, 13% below its Nov. 6 close, the last before Huang admitted an affair with a woman U.S. officials suspect of spying for China, and 32% below its record close of $239.24 on Oct. 6.

The study is by Paul Farah of Notre Dame University-Louaize and Hui Li of La Trobe University, published in 2021 in the International Journal of Financial Studies. They examined 735 CEO changes at S&P 1500 companies from 2003 to 2012 and classified 187 as forced: the CEO was ousted, resigned under pressure, or left young and fast with no comparable job waiting. Then they split the forced exits by what the stock had done in the three years before. At 125 companies it had trailed the market. At 53 it had beaten the market, and those are the cases that matter for Nvidia.

Firing a CEO whose stock had lagged was good news. The stock gained 1.0% in the five days around the announcement, and over the next three years the median company beat the market by 18 points. Firing a CEO whose stock had beaten the market was the opposite. The stock fell 3.4% in the five days around the news, and over the next three years the median company trailed the market by 13 points. The study reads as a warning for anyone holding the stock: when a board pushes out a CEO whose stock had beaten the market, the stock has tended to trail the market for the next three years, by 13 points at the median; the limit is that the group is 53 companies from 2003 to 2012, none of them the size of Nvidia, the study does not separate founders from hired managers, and a median is a tendency, not a forecast.

Where Nvidia sits in the sample

Huang is on the outperforming side by a wide margin. From Nov. 13, 2023 to Wednesday’s close, the last before the resignation, Nvidia rose 258% and the S&P 500 rose 79%. The exit was forced by the study’s definition: The Wall Street Journal reported Wednesday that the board was preparing to oust him, and he left the company and the board the next morning with no other job. Two more things line up. Nvidia named no permanent successor, and 49% of the forced exits in the study had none either, which the authors read as a sign of how unplanned these departures are. And the reason is public, through Huang’s own Sunday statement and the reports since. That matters: when a board gave no clear reason for an exit, the stock fell 3.2% over five days, and when the reason was known it gained 1.5%.

What does not line up is the man. The average fired CEO in the study was 54 and had held the job for six years. Huang is 63 and ran Nvidia for 33 years from its founding; Kress’s own statement called it "the extraordinary foundation Jensen created over the last 33 years." That is the limit to carry through everything below.

A second, broader set points the same way on timing. PwC’s Strategy& looked at CEO changes at the world’s 2,500 largest public companies from 2011 to 2013 and found that forced exits cost shareholders mostly before the announcement: a median 13% of total shareholder return in the year before the change, and 0.6% in the year after. Nvidia fits the first half. Over the past 12 months it is down 13% while the S&P 500 is up 13%.

The retiree holding it

Take a couple, 68 and 65, a scenario built from the numbers above. He retired from an engineering job in 2021. In January 2016 he bought 200 Nvidia shares at $32.37, $6,474 in a taxable brokerage account, because he had used its chips at work. Two splits later, four-for-one in 2021 and ten-for-one in 2024, that is 8,000 shares. The rest of their money, $1.2 million, sits in index funds in an IRA.

The stake is about $1.3 million, 201 times what he paid, and 52% of everything they own outside the house. It is down $611,120 from the October peak and $190,320 since Nov. 6, a week in which the S&P 500 rose 2.6%. The resignation day alone cost $60,320. The study’s five-day reaction to firing an outperformer, 3.4%, would have been $44,295 on this stake; Nvidia has lost four times that since the admission, because the fall came in stages across a week of reports before the board acted.

The three-year number is the one that matters to them. Thirteen points behind the market on $1,302,800 is $169,364 less than the same money would earn in an index fund by late 2029. The industry-adjusted version, 15 points, is $195,420. If Nvidia instead behaved like the companies that fired an underperformer, 18 points ahead, the gap would run $234,504 the other way, but nothing in Nvidia’s record puts it in that group. These are medians: half the companies did better and half did worse, and the spread was wide.

The tax decision concentration forces

If you own Nvidia this way, the study does not tell you to sell. It says the base case after this kind of exit is the market’s return minus something, with a wide range, from a stock that is more than half your money. The question is how much of that risk to keep, and the tax bill sets the price.

Selling all 8,000 shares at $162.85 would realize a $1,296,326 gain. Stacked on their $54,000 of Social Security and $48,000 of IRA withdrawals, after the 2026 standard deduction, the first $40,500 of gain falls in the 0% capital gains bracket, the next $514,800 is taxed at 15% and the remaining $741,026 at 20%, and the 3.8% net investment income tax applies to most of it. The federal bill is about $268,754, and the sale would also set their Medicare premiums two years from now from this year’s income. They would keep about $1,034,046.

Selling a quarter, 2,000 shares, realizes $324,082, costs $48,921, about 15% of the gain, and leaves $276,779 to move into the index funds. Selling a tenth, 800 shares, realizes $129,633 and costs $13,370, with no net investment income tax because their income stays under $250,000. A tenth a year for several years is the path most planners draw for a position this size. The other way out is time: shares held until death pass to heirs at that day’s market value and the whole gain goes untaxed, which is why a 68-year-old with a $6,474 basis is usually told to trim rather than sell.

What to watch

Nvidia reports its fiscal third quarter on Tuesday, Nov. 17, after the close, with Kress presenting the results as both CEO and CFO. The permanent CEO is the other date. The authors’ explanation for the 13-point lag is that boards which pushed out a winning CEO brought in someone who could not sustain the run; whether Nvidia’s board finds the exception is the whole bet for a holder who stays.

The bottom line

A study of 187 forced CEO exits says the market’s first reaction to losing a winning CEO is a drop of about 3.4%, and the three years after run 13 points behind the market at the median. Nvidia has already fallen 13% in a week and 32% from its October peak. For a retiree with 8,000 shares worth $1,302,800, the median three-year lag is $169,364, a full sale costs $268,754 in federal tax, and a tenth a year costs $13,370 at a time. The study is small and old, and it does not separate founders. The concentration is not small.

Contact [email protected] for any questions or corrections.