> how is million people independently trading a thing for $10 once meaningfully different than a single buyer/seller pair trading a thing for $10 a million times?

> Generating fake trades like this? It's injecting bad data into the market. It's poisoning the mechanism of price determination - which, to produce reasonably fair determinations, needs aggregate trades to average great many distinct buyer/seller negotiations. Amplifying a single datum with fake trades ultimately makes the market worse at efficient resource allocation...

I'm not a legal or markets scholar, so please excuse this possibly dumb question.

Why are wash trades specifically harmful, as opposed to book orders that don't self-trade?

Like, I think what Coinbase did not get in trouble for here was having a large order book that they did not intend to trade (spoofing), but they instead got in trouble for trading with themselves at prices they would have traded with any other market participant. (Seemingly, as a fluke of the way they structured their market-making apparatus as two separate bots, instead of a single bot.) Since this was on a public market, the trades inherently happened somewhere between the public bid and ask prices -- right? I don't understand how the self-trades manipulate the price information. The fakeness that comes out of this seems to be purely the metric of trading volume.

If I'm wrong, please let me know! I'd love to learn more. Thanks.