I can still remember how ragey the business community got over facebook's early valuations. facebook is not worth $33 billion, they wrote. There was an oddly righteous tone to it, like, I can't believe these guys have the arrogance to raise at that valuation. Who do they think they are? No revenues and they're asking for HOW MUCH!?

I can't say I blame them; it still seems absurd, six years later, that facebook paid billions for two social networks in 2012-2014, neither with significant revenue: first $1 billion for Instagram (2012), then an even more eye-popping $19 billion for WhatsApp, just two years later (2014). I remember standing in my kitchen in San Francisco thinking about it; facebook paid half the value of Ford (the car company) for a company with around 50 employees and minimal revenue. It was unprecedented; never had a company with so few tangible assets or cashflow been purchased for that much.

I took me a long time to see this, but I realized it's all just supply and demand. Business school teaches that securities (e.g. stock, how ownership is transferred) should be valued on more objective criteria, things like discounted cashflow: project the cash flows, and calculate based on those. Perhaps that's a good for calculating what we ought to pay—a finance academic would probably argue that's correct—but what people do pay is based on a market process, just like houses or public companies.

And that market process runs on supply and demand. Specifically, when raising funding for early-stage companies, it seems to work like this:

• At any point, a certain amount of equity (ownership) is on offer, "for sale" to investors, and
• A given level of money—"risk capital"—is on offer, which purchases the ownership.

As I said, no different from houses or public companies.

The key insight of this model: supply of risk capital drives valuations (prices).

So how do these factors (supply and demand) vary over time? Interesting question. One thing I've noticed: company formation happens much more slowly, and at a much more constant rate, than investment, which is why capital availability predicts valuations. Company formation is affected by things like:

• Confidence—how many people are leaving their jobs to found companies?
• Perception of opportunity: are investment bankers and lawyers spilling over into long-term careers in tech? (I've seen a lot of this over the past decade)
• Do people think there's specific opportunity in a space? (Social networking, robots, AR/VR)

Whereas risk capital is more of a "Wall Street" thing, driven by:

• Is early-stage tech perceived to be a good investment? (2008: Nobody investing in anything. 2012: INVEST IN ALL THE THINGS. 2018: Most people say "yes", smart money beginning to wobble)
• What else is available? Are other asset classes yielding enough? (2008-2016: No. 2016-now: interest rates going up)

In the past year, I've noticed a bit of tapering. Still lots of money flowing, but more to later-stage companies. Apartment rents are softening; I see more office vacancies. The overall tone is still optimistic, but more cautious than a few years ago. Fewer new things seem to be getting traction.