I like to think that independent miners would keep running to secure the network but simple economics might keep that from happening. I see a possibility that large mining farms have so concentrated the mining power that if they were to all shut down one could switch on just long enough to do some damage. The concept of logical self interest isn't all encompassing and can't for tell every possible motivation for an attack.
I doubt that any hashrate drop will be so disruptive as to create any attack risk. Even if it drops by half you would still need a massive 150 PH/s to do a 51% and where would it come from?
I'm sure the network will be fine, diff drops, some miners switch off, perhaps exchange rate goes up a bit if coins are mined 10-20% slower until next adjustment, in other words it should be able to balance itself out.
The current selloff and any potential future selloff will only exaggerate this.
This will result in someone potentially being able to purchase a large amount of the network hashrate very cheaply with the possibility of being able to potentially attack the network.
Yes, that scenario is exactly what I'm talking about. I worry less about an attack when the price is stable and profitable for miners. When it becomes unprofitable and dropping or stabilizes at an unprofitable price point that creates an opportunity to exploit the weakness of the network.
At some point in the future I expect businesses like BitPay to take over mining in a big way. When this happens the network will be more centralized between a few businesses but safer because their profit will really exist in their primary business model. Mining, for them, will only be a sidebar necessary to secure their business investment. They will have no incentive to do anything but keep the network safe. Today the network is unstable because it's controlled by large pools/farms that have no other method of funding the mining operation except the mining. When they switch-off then anyone with enough power can switch-on and ultimately control the network. It would only take a small window of opportunity to do this.
I would however make a counter point to your concern, that is similar to a recent argument as to why PoW is superior to PoS. If someone were to buy up a lot of mining equipment cheaply and were to attack the network then they would lose out on their investment on the mining equipment (this may not matter if it was cheap enough) but also on future mining revenue. As it stands now miners will only be able to make roughly 1% (it is probably closer to 0.75%) of the total revenue they would mine over the course of their useful life. This percentage is probably lower when the difficulty has recently taken a huge drop. I think it is fair to say that any major attack on the network would likely be discovered within 48 hours of it starting, so an attacker would potentially lose out on ~98% of their potential mining revenue by attacking the network (this assumes that an attack were to cause confidence in bitcoin to be wiped out and have it's value go to essentially zero).
What if they want the value to go to zero? The current thinking is one dimensional. There are more reasons to attack the network than just to double spend and profit is only one reason for action. A government, a competitor (Visa/MasterCard) or several of them working together could kill Bitcoin easly and spend almost nothing to do it. They wouldn't even need the resources they threw at Silk Road. If I were going to attempt to destroy Bitcoin I would use your mining equipment to do it. I would host mining equipment somewhere electricty is cheap then get people to ship me their equipment. I would run it honestly until the farm was a large percentage of the network and execute my plan. Miners switching off just speeds up that process.
