Nvidia trades at less than 17 times the profit analysts expect over the next twelve months, close to its cheapest level in more than a decade. As recently as May it was above 25 times. The multiple is roughly half what the stock commanded in 2025, a year when revenue and profit were growing more slowly than they are now.
The company is forecasting revenue growth of about 90 percent and net income growth of 99 percent in fiscal 2027, up from 65 percent the year before, and puts fiscal 2028 sales growth at 70 percent against a previous consensus of 45.
So the fastest growth in its history, at half the price the market paid for slower growth.
Eli Horton of TCW gave the cleanest read: the stock has de-rated significantly, which suggests a healthy dose of scepticism that current earnings power is sustainable, and while the performance is surprising given the fundamentals, it tells you the market expects less than consensus is estimating.
The specific worry is margin. Gross margin was 75 percent in the second quarter and is projected to slip below 72 percent in the fourth before recovering, squeezed largely by rising memory chip costs. David Russell of TradeStation put the structural version of it: customers want to reduce their reliance on Nvidia, so its market position may weaken over time, which means margins are more likely to go south than improve, and multiples expand for companies with room to get better.
Nvidia is up 22 percent this year, second best among the Magnificent Seven behind Apple. Within the semiconductor index, up almost 76 percent, it is the fifth-worst performer, while Intel and AMD have each gained more than 180.
Here is what a multiple actually is, and it is the most misunderstood number in business.
A multiple is not a measure of how good a company is. It is a measure of how confident the market is that the current earnings will still be there in five years. A roastery doing £2 million of profit on one enormous supermarket contract and a roastery doing £2 million across four hundred independent cafes are worth wildly different sums, and every buyer knows it. The second one sells for more, on the same profit, because the earnings are harder to take away.
Which is why a cheap-looking multiple on a company posting extraordinary numbers is rarely a gift. It is a statement that somebody has done the arithmetic on durability and not liked the answer. The cafe on the corner that is suddenly available at two times earnings usually has a lease expiring in fourteen months.
That does not make the market right. GuruFocus’s own value metric puts Nvidia about 42 percent undervalued against an intrinsic estimate of $393.71 against a $229.15 price, with a quality score of 95 out of 100. Fund managers are openly divided on whether the discount is a warning or an opening, and Jensen Huang has responded publicly.
Ninety percent revenue growth is a fact. Seventeen times is an opinion.
Source: Bloomberg, “Nvidia’s Stock Is Flashing a Warning Sign as Valuation Falls”
