Banks can. They do it everyday of the week. It's called fractional reserve banking; by the time they have redeposited the deposits half a dozen times they can then lend out much more than the real amount of cash originally deposited.
But the issue is why would a bank want to take a big hit that would take it years to recover from when the customer is still paying the interest on time.
Discussions around the multiplier effect of FR invariably confuse the impact at a system level vs that of the individual bank. At the individual bank level your funding task comprises making sure you have sufficient liabilities to fund your assets.
People also tend to confuse capital requirements (how many $ you have to hold for a given asset size that you don't owe anybody in order to meet unexpected losses) vs. liquidity reserves which is what you need to set aside from your funding base to meet your obligations to depositors as and when they fall due.
They aint the same thing.
And there are several reasons why CBA would prefer to take a hit early.
First and foremost, if the deal is underwater and the borrower is in trouble, unless you're confident the borrower will right itself, waiting simply adds to your losses.
Also, if you're in negative LVR territory and there is a reasonable chance the borrower won't get back in good order, you have an impaired loan which is now a deduction from your Tier 1 capital. You don't want it sitting there any longer than is reasonable.
That said, no bank wants impaired assets on their balance sheet so you certainly don't go out and create them. My guess around the 4C examples would simply be that on the yearly review it became apparent that the borrowers were in breach of their LVR covenants or other bits, looking at current circumstances. there was little chance of things getting better, so CBA figured better to accept today's known loss than risk a far bigger one down the track.
Alternatively, maybe CBA is just mean.