"Generational wealth" is a lucrative term for many investors. While it's nice to set yourself up for retirement, it's even better to leave your family and other loved ones a substantial amount of money that would significantly improve their lives as well.
Many investors may be searching for the next stock that can achieve returns by a multiple of 100 to set up their families, but often it's the stocks that consistently deliver market-beating returns that can truly produce generational wealth.
The long-term return for the market is around 10% per year. Let's say you can beat that return by a few percentage points -- call it 13% per year. If you start with $1 million, a 10% return transforms that into $17.5 million over 30 years. However, the 13% return translates to $39.1 million.
That's a huge difference and underscores why beating the market by just a few percentage points each year is more important than finding the next 100-bagger.
I have two stocks that I believe can deliver this level of outperformance over the next few years: Nvidia (NVDA 0.48%) and Taiwan Semiconductor Manufacturing (TSM +0.42%). Both of these are huge beneficiaries of the buildout of AI computing capacity and have tremendous tailwinds in their favor.
Image source: Getty Images.
Nvidia
Both Nvidia and Taiwan Semiconductor (TSMC for short) have benefited from the massive amount of AI spending going on.
Nvidia's graphics processing units (GPUs) have become the industry standard for training and running AI models, despite numerous alternatives claiming they can do it better and cheaper. The reality is that its products are extremely flexible and can be used in multiple ways, whereas some of the alternatives that haven't caught on yet are focused on a single workload type.
While the company experienced a growth setback in China due to the Trump Administration's change in export restrictions (which differed from the restrictions by the Biden Administration), Europe could fill the gap left in Nvidia's results. And domestically, it still has a ton of growth left, as AI hyperscalers are spending a record amount on data centers this year.
NASDAQ: NVDA
Key Data Points
That trend is likely to continue, according to a third-party estimate quoted by Nvidia that 2024's data center capital expenditures (capex) were $400 billion. However, they project that figure to rise to $1 trillion by 2028.
Considering that the company generated $115 billion from data center sales in fiscal 2025 (encompassing most of 2024), this indicates that it captures around 29% of all such spending. Even if its market share fell to 20% and capex still hit $1 trillion, Nvidia would generate $200 billion in data center revenue by 2028.
That indicates a compound annual growth rate (CAGR) of about 15%, far exceeding the mark necessary to surpass the market.
Taiwan Semiconductor Manufacturing
With all of the various competitors in the AI race, few of them have internal capabilities to produce chips. That work is farmed out to a handful of semiconductor foundries, with TSMC being the leader by far. Nvidia and many other notable names are its clients, and it has won this work through its constant innovation and by delivering unmatched production yields.
And with TSMC's position within the industry, management has an excellent view into what the future holds. Chip orders are often placed years in advance, giving management clarity about demand.
Over the next five years, it projects AI-related revenue will have a 45% CAGR. And since TSMC supports a wide range of industries beyond AI, its overall revenue CAGR is expected to reach nearly 20% over the next five years.
NYSE: TSM
Key Data Points
That's still far greater than the stock market's 10% average annual return, making the company an excellent candidate to produce generational wealth.
With Taiwan Semiconductor Manufacturing and Nvidia, investors should keep a close eye on their growth as we approach the end of the AI buildout cycle. This could cause them to lose their ability to reliably beat the market. However, we have barely scratched the surface of what's necessary for an AI-first world, and that could cause these two to soar for many years to come.
