Syammondsandrew, that response is a tautology.
It's like saying "ninjas have swords because they're ninjas" in response to someone asking about the possibility of a ninja having a gun. It doesn't do anything to address the point that there would be obvious advantages for a ninja who gets a gun.
Now one could play games with taxonomy and say "No one with a gun is a ninja." But it's a pointless argument. The important thing is "Are people going to get shot?"
It's not really important to answer "Some assassins in black robes just shot someone while acting in all other ways like ninjas. Should we call them ninjas?"
The guy who got shot doesn't care. It's missing the point.
In this circumstance, the original question was a hypothetical "What would happen if a market maker sells more shares than exist and fails to deliver?"
To simply say that it wouldn't happen, is pointless. WHY wouldn't it happen? Would they immediately get arrested? Would a bomb strapped to the bottom of toilet blow up?
Would someone give them a wet willy? Banana in the tail pipe?
When evaluating the possibility that someone might do something that they're not supposed to, it's useful to ask:
"What is the mechanism that would cause the nefarious behaviour to be discovered."
and
"What would happen once it is discovered?"
Here's some more direct, on-topic info:
"That point is made abundantly clear by tapes obtained by Rolling Stone of recent meetings held by the compliance officers for big Prime Brokers like Goldman Sachs, Morgan Stanley and Deutsche Bank. Compliance officers are supposed to make sure that traders at their firms follow the rules — but in the tapes, they talk about how they routinely greenlight transactions they know are dicey."
https://www.rollingstone.com/feature/wall-streets-naked-swindle-194908/
"The attack spiked on September 9th, when there were over 1 million undelivered shares in Lehman. On September 10th, there were 5,877,649 failed trades. The day after, there were an astonishing 22,625,385 fails. The next day: 32,877,794. Then, on September 15th, the price of Lehman Brothers stock fell to 21 cents, and the company declared bankruptcy."
Personally I had never heard that there were 32 million fails on Lehman stock. It would certainly be interesting to know who was failing to deliver. With a transaction of that size, I don't think you can rule out a "market maker" without information to the contrary. I would think some random "non-connected"/market maker firm would have been blocked from trading prior to racking up that many fails. Think of the losses their broker could have been on the hook for if the stock had recovered.
Given that the source of those failures hasn't been identified, let alone been given a wet willy or banana in the tailpipe as a punishment, I don't see why it can't happen again.
One other interesting tidbit from this article:
"If you own stock that pays a dividend, you can even look at your dividend check to see if your shares are real. If you see a line that says “PIL” — meaning “Payment in Lieu” of dividends — your shares were never actually delivered to you when you bought the stock. The mere fact that you’re even getting this money is evidence of the crime: This counterfeiting scheme is so profitable for the hedge funds, banks and brokers involved that they are willing to pay “dividends” for shares that do not exist. “They’re making the payments without complaint,” says SusanneTrimbath, an economist who worked at the Depository Trust Company. “So they’re making the money somewhere else.”"
So it looks like my idea that dividend paying companies offer some protection must be taken with a grain of salt. It seems there is a mechanism to pay dividends to shareholders of counterfeit stock. I still think payment of a regular and significant dividend would be strong disincentive to counterfeiting, but I wanted to point that section out.
