Korva is a scroll of the internet’s best ideas.
Buying Nvidia at four times earnings
The market is focused on the next year and the next quarter, and very few people believe you can accurately predict two, three, or four years out. A very strong business model on the right part of an S-curve produces exponential unit growth, and its earnings compound exponentially instead of linearly.
Understanding the S-curve, knowing the moats, and knowing how to model lets an investor predict these results. Earnings growing from one dollar to ten happens way more than most people think. When Amazon was bought for AWS, AWS came for free.
Underappreciated long-term earnings power is the last piece of a framework that begins with the S-curve and moves to competitive advantage.
Why does the market focus on the next quarter?
The Securities and Exchange Commission began requiring US public companies to file quarterly reports in 1970. Analysts publish estimates every quarter, and investors judge companies on whether they beat or miss them. Fund managers are measured and paid on short stretches of performance. Those habits make near-term results the market's main focus.
How did Nvidia end up at the center of AI?
Nvidia released CUDA in 2007, a software platform that let programmers run general-purpose computations on its graphics chips. In 2012 the AlexNet image-recognition network, trained on two Nvidia GPUs, won the ImageNet competition by a wide margin. Researchers then built their tools around CUDA. That ecosystem of software made Nvidia's chips the default for training neural networks.
How fast did Tesla's deliveries grow?
Tesla reported delivering about 367,500 vehicles in 2019. In 2023 it reported about 1.8 million. That is roughly a fivefold rise in four years, driven by new factories in Shanghai, Berlin, and Texas. Its Model 3, launched in 2017, was its first car built for the mass market.