[pixl97]

Most chunks of the computing market should be thought of as a text book example of a 'free' market that operates with collusion with the few well monied "competitors" ensuring they don't put each other out of business.

Lets say you wanted to jump into the hard drive producing market. It's going to take you a few years to get there and a lot of billions of dollars. By the time you're close to producing units the existing players will suddenly drop the prices to the point where you cannot produce profits for as long as they need to. Aka, your competitors can collude longer than you can remain solvent. And yes, in two decades you will win the court case against them. And other than a fine nothing will happen because they're are so few manufactures that they are too big to fail.

[zozbot234]

The better way of putting it is that collusion is ultimately limited by the ease of entry into and exit from the market. The existing players don't need to "suddenly" drop the prices when a new entrant appears, because they've been doing that already.

It's only when supply bumps into short-term capacity constraints and price must rise enough to fully ration demand that this assumption fails. But then even if a new entrant appears that's no reason to drop prices unsustainably, because why would you? They're not taking any share of the market away from you, they're serving new demand.