This post could actually be summarized in one sentence: If you want Non-Performing Loans (NPLs) to fall, then you have to increase lending. Still, I don't really hope that I will be able to convince many just by stating this. Thus, what follows is an exposition of why the amount of loans (and more so of new lending) matters when it comes to NPLs and, in addition, when bank regulatory capital requirements are concerned.
NPLs are loans for which "payments of interest and principal are past due by 90 days or more". There is nothing more rational than to think of the rise in NPLs as an outcome of the crisis. This, nevertheless, is where most stop their arguments; the problem is that asking "why?" matters the most when it comes to policy. The answer is again so simple everybody has thought about it: it's because people lose their jobs and cannot repay their mortgages or other consumer loans, because there's no investment and no consumption forcing businesses to default and putting more people on the dole.
The correlation is obvious as can be seen in the case of Greece:
|Source: CEIC Network|
|NPLs as a percentage of total gross loans. Source: Index Mundi|
The additional problem here is that crises usually come around when banks are already contracting their balance sheets, so the hit in consumption and investment is even harder: less consumption, less demand and less money to go around as well. As a consequence, firms face serious trouble meeting up with their obligations. If they cannot make ends meet, then their loans enter the NPL category. When more NPLs are created, then the bank is more constrained by its regulatory capital needs as bad loans are assigned a higher risk weight than before. Thus, the higher the NPLs the lower the banks' ability to lend out more money.
The problem resembles the one of austerity: we need a government budget surplus but we cannot do it since by cutting expenses and transfers (e.g. pensions) we are reducing consumption and thus government income. Similarly, we want smaller banks, but we cannot do it without retracting money from the economy. As money is reduced, consumption and investment become more scarce and business struggle for survival; many go bankrupt. Driven by this lack of funds, unemployment rises resulting in even more NPLs.
How can we get out of this mess? The solution (ironically) is not that banks have to decrease their exposure. It's that they have to increase their lending in order to get the economy going again. If the economy does not have enough funds to pull itself out then countries experiencing these issues will face Greece-like situations: prolonged measures to make things better, but only making them worse (here's looking at you austerity!). When lending is increased then more investment is created; subsequently, more jobs and more consumption, leading to an increase in income and a decrease in the loans which cannot be repaid. When people have more money, loan payments which could not be paid before are met now. Nobody wants to lose their house, and no bank want to be stuck with one. In addition, less NPL's actually mean less capital needs, thus more funds to lend, thus more profit for the firm. Still, instead of lending and preventing this from happening, banks are forced by regulators (and themselves as well) not to lend out funds.
As said before, there is a time and place for everything. Just like it wouldn't make sense to continue expanding fiscal policy in a boom (see the UK experience), or performing QE operations when times are good, it does not make sense to use austerity measures when times are bad (Greece, Spain, Italy, Portugal, Ireland). Similarly, banks should not be pressured to reduce their credit exposures during downturns. Yet, unfortunately, policymaker decisions do not appear to be counter-cyclical.