Over the weekend, the NYT ran The Stock Market Is Shrinking. That’s a Problem for Everyone, a piece detailing the halving of the US public company count—8000 to 4000 in the past 25 years. That is indeed a problem as the article says, as retail investors are missing a lot of tech-driven growth.

Almost parenthetically, the article added:

Without deep knowledge of a company’s critical research — which businesses may be reluctant to share, for competitive reasons — it’s difficult for outsiders to evaluate a start-up’s worth. That makes it harder to obtain funding, and it may be partly responsible for certain trends: why there are fewer initial public offerings these days, why smaller companies are being swallowed by the giants, and why so many companies remain private for longer.

That modern accounting doesn't tell the whole story in tech, is so obvious to anyone paying attention. Accounting has always been a lagging indicator, and not all firms have high-tech levels of R&D spend. But reading a typical tech income statement, I can't help but perceive an enormous gap between what's measured in the statement, and the true drivers of high-R&D company performance.

A few examples:

• The key to success in R&D-intensive companies, whether drugs, software, film, or semiconductors, is making something people want. But we have no standard metrics for how much people want what a company is shipping today, or what's under development.
• Tech products are differentiated; the good ones have no close substitutes, giving them a lot of pricing power. iPhone fanatics don't buy Androids. Unit profitability, a major focus of traditional income statements, just doesn't matter that much. Bad movies lose money. If you invent the cure to cancer, you can charge a lot and people will still pay it.
• We say companies spending more than revenue are "making a loss"; that's such sloppy language, as not all losses are created equal. A grocery store making a loss faces quite a different outlook than a tech company spending like crazy to acquire customers who want the product (provided they don't churn).
• In high tech, talent acquisition is important to an extent people from more traditional industries (retail, hospitality, manufacturing) can scarce comprehend. 100 of the wrong people often can't do what 10 of the right people can. It's not a matter of 10x productivitiy, it's a zero-to-one, can-vs-can't, difference.
• "Above the line" vs "below-the-line" doesn't matter as much in R&D-intensive industries. There are basically only three expenses: product development, production/manufacturing (zero for films and software, nonzero for internet services, drugs, and iPhones), and distribution. Below-the-line spending isn't "overhead" to minimize; it can make, or break, a company long-term.