Category: Economics and Banking

The Bank of England recently published a working paper “Bank leverage, credit traps and credit policies” which looks at why, following a financial crisis growth tends to stagnate for a long period and how macroprudential policy tools should be used both before and after a crisis.  They look at “credit traps” which arise when shocks to bank equity capital tighten banks’ borrowing constraints, causing them to allocate credit to easily collateralisable but low productivity projects. Low productivity weakens bank capital generation, reinforcing tight borrowing constraints, sustaining the credit trap steady state.

Financial crises tend to have severe negative effects on real activity, and recoveries following crises tend to be weak and slow. In Japan, for example, real GDP remained some 30 per cent below its pre-crisis trend 10 years after the onset of its financial sector distress in 1991. In the UK, the gap between realised real GDP and the level implied by the pre-crisis trend was around 20 per cent five years after the onset of the crisis in 2007. And in the USA, Japan, the UK and the euro-area, the rate of credit growth collapsed around the onset of the crises. In Japan, anaemic credit growth continued for at least a decade.

These consequences have triggered various policy responses. On the one hand, reform of financial regulation continues apace. Across jurisdictions, macroprudential policy authorities have been established and tasked with conducting system-wide prudential policy, including the use of countercyclical bank capital requirements. At the same time, central banks and governments have introduced a range of ‘unconventional’ monetary and credit policies, including asset purchases, policies to support bank funding, and recapitalisation of financial institutions. In light of these sweeping changes to the policy landscape, there is a real need to understand the mechanisms, costs and benefits of these interventions, and the conditions under which they can be effective. This paper enhances the understanding of such ‘credit policies’ – both ex-ante (to avoid credit crises), and ex-post (to escape their consequences) – by presenting a novel, tractable macroeconomic model to understand their effects.

We do this by constructing a simple overlapping generations model featuring financial intermediation and credit frictions, and use it to study the credit policies mentioned above. The key feature of our model that makes it particularly useful for studying these policies is its ability to generate a ‘credit trap’ steady state – that is, a steady state of the economy that features low real activity, low productivity, low bank capital, and weak bank profitability. In our model, the borrowing constraints facing banks depend on the health of the banking system: when the net worth of the banking system is low, banks’ ability to finance productive investment through borrowing is severely constrained. The economy enters a ‘credit trap’ when a large unanticipated shock to bank assets reduces bank capital below a critical threshold, causing banks’ funding conditions to tighten, inducing them to invest in less productive assets that, nonetheless, have higher pledgeability to creditors. Thus, even a temporary shock can have extremely persistent effects if it causes a large reduction in bank capital. And it is the possibility of entering a credit trap that has profound implications for policy that have not been examined by existing work in this area.

Concluding remarks
The recent financial crisis has raised the question of whether there is something fundamentally different about economic recovery following a severe financial crisis and, if so, how macroprudential policy tools should be used both before and after a crisis. Most modern macroeconomic models are unsuitable for addressing this question, with their economies quickly returning to health once a negative shock is unwound. In this context macroprudential policy tools play the role of reducing volatility, rather than avoiding a catastrophe or supporting the recovery from a crisis. By contrast, in this paper we explicitly consider a model in which the economy can become trapped in a steady state featuring permanently lower output, bank credit and productivity following a sufficiently severe financial shock, a confluence of characteristics we call a credit trap.

In this paper we have developed a simple, tractable OLG model for analysing credit traps. We have examined the effectiveness of policy both at preventing a credit trap occurring, and helping the economy to escape (which becomes necessary as it will not recover without intervention). Our analysis shows that a leverage ratio cap is effective in increasing the resilience of the economy against shocks and reducing the probability of a credit trap. However, this comes at the cost of lowering the level of output in the ‘good’ steady state, and hence the policymaker needs to set the cap to trade off these costs and benefits. Relaxing the leverage ratio cap is effective in encouraging faster recovery after a negative productivity shock, provided that the shock is sufficiently small. But if the shock is large enough to tip the economy into a credit trap, then relaxing the leverage ratio cap will not help the economy get out of it. To escape a credit trap other policies are needed, and we consider the efficacy of a set of ‘unconventional’ credit policies: direct lending; bank recapitalisation; and discount window lending. These policies present rich, realistic trade-offs which vary with their relative efficiency costs. Their effectiveness depends on the state of the economy, with all more effective when the economy is weaker.

In future work, it would be interesting to analyse more thoroughly the optimal leverage cap that would be set by a policymaker in advance of a trap. We have shown that the level of the leverage cap that maximises resilience is countercyclical: it would be interesting to analyse numerically if the optimal level of the leverage cap is too, and whether this would vary with the state of the economy in a non-linear way. This would be particularly interesting when the economy is just above the trap threshold, and the policymaker has to trade-off rebuilding the health of the banking system with the possibility of further negative shocks. .

Note: Staff Working Papers describe research in progress by the author(s) and are published to elicit comments and to further debate. Any views expressed are solely those of the author(s) and so cannot be taken to represent those of the Bank of England or to state Bank of England policy.