You know the story. A bank lends you money to buy a property in a bubble-inflated market, securing the loan with a mortgage over the property. The bubble bursts, so that the market value of the collateral is now less than the outstanding debt. For whatever reason, you stop paying the mortgage. The bank repossesses the property and sells it, receiving a price that is insufficient to cover the debt.
Now why should you be liable for the shortfall? You shouldn't. Protecting the lender against default is the purpose of the collateral. When the collateral has been seized, any further claim against the borrower is double dipping. As the lender has more power and more information than the borrower, it is the lender's responsibility to make sure the collateral will cover the debt. If the lender has failed to discern a bubble or to leave sufficient margin for the possibility of a bubble, that failure is the lender's fault and its consequences should be the lender's problem.
And why should your credit rating take a hit if you walk away? It shouldn't. The return of the collateral in undamaged condition should constitute an alternative fulfilment of the contract. Again it is up to the lender to ensure that the collateral is sufficient for the purpose.
The ability to walk away without penalty is equivalent to a put option whereby the borrower can sell the property to the lender for the amount owing. That put option will have its market price, which will be expressed in the interest rate or the set-up fee or some combination of the two. Obviously that market price will reduce the amount that the buyer can borrow to spend on the property. But, as property prices are determined by borrowing capacity, prices will adjust to compensate, and affordability will be unaffected.
Would non-recourse loans make a crash more likely, by encouraging defaults? Only if they're introduced at a time when the market is already overvalued, in which case, by forcing lenders to be more conservative, they might precipitate the crash, which in turn would encourage defaults. If they're introduced while the market is rising but not yet overvalued, that same conservatism will tend to prevent the formation of a bubble. If there's no bubble, there's no crash, hence no encouragement to default.
Yes, the great U.S. housing bubble inflated in spite of non-recourse loans in many states. But that was because the mortgages were resold, so that the original lenders didn't care if collateral values fell, whether the loans were non-recourse or full-recourse. That made the lenders reckless. The problem wasn't the non-recourse loans, but the circumvention of their essential feature: exposure of the lender to any fall in collateral values.
Even more iniquitous than full-recourse mortgages is the practice of evicting tenants when their landlords default, then selling the vacant properties -- or worse, holding them off the market in order to prop up prices and rents, and generating income as necessary by drip-feeding repossessed properties to buyers instead of letting them to tenants.
If you are a lender repossessing a rental property, you should have to honour the rental contract. If you sell the property, you should have to sell it subject to the rental contract. If that requirement makes you more cautious about lending for investment than for owner-occupation, so be it. In Australia, where negative gearing helps investors to outbid owner-occupants, that caution would tend to level the playing field.
In the mean time, if you are a tenant being evicted because your landlord has defaulted on the mortgage, you might try telling the lender that you are considering a lawsuit and/or private criminal prosecution against the lender for being an accessory to the landlord's insolvent trading. You might be able to extract a very favourable rental deal in return for dropping that threat.
If you pay off a loan early, why should the lender be allowed to charge you an exit fee? Does the fee compensate the lender for loss of interest? No, because interest is itself a compensation for alternative uses of the principal; and by returning the principal early, you allow it to be used for those alternative purposes. Does the fee compensate the lender for the cost of re-lending the principal? No, that's the purpose of set-up fees, not exit fees.
Do exit fees mean lower set-up fees? Perhaps. The ability to pay out the loan early with no penalty is equivalent to a call option whereby the borrower can buy the mortgage for the amount owing. That call option will have its market price, which can be expressed in the interest rate or the set-up fee or some combination of the two. Converting it into an exit fee makes it less visible to the borrower entering into a mortgage -- in other words, less honest. The Federal Government is therefore to be congratulated, howsoever grudgingly, for deciding to ban exit fees.
What about variable-rate mortgages, in which the stronger party to the contract can unilaterally vary the most important term in the contract after it is signed? If the "variable" interest rate were tied to some agreed measure of the lender's borrowing costs, that would at least retain an element of mutuality (not to be confused with safety). But the lender gets carte blanche. Why should the lender be allowed to vary the rate at will, or at all?