How Costly Are Contractions In Bank Credit?

Summary

• A new IMF working paper examines the importance of bank lending in periods of economic expansion and assesses how a lack of credit impacts economic activity.
• It finds that, during economic boom phases, supply-side credit constraints happen every five years on average, cutting GDP growth by up to 9pp throughout the credit cycle.
• The results emphasise the importance of macroprudential policies beyond economic downturns, thereby ensuring bank credit expands simultaneously with economic activity.

Introduction

A raft of prudential regulations sprung up to protect financial stability in the aftermath of the 2008 Global Financial Crisis. Yet at the onset of the 2020 COVID pandemic, these policies were relaxed to avert a contraction in bank credit. This is because supply-driven credit contractions are costly in terms of economic activity – a new IMF working paper examines just how costly.

The paper considers the relative dynamics of bank credit and economic activity throughout credit cycles, using event study techniques. In particular, they quantify the impact of episodes in which credit growth contracted and GDP growth expanded (referred to as credit reversals) on economic activity at different points of the credit cycle. They find:

• Credit growth and GDP growth generally co-move, although the relationship has loosened since 2009.
• Credit reversals occur more frequently than banking crises and are more likely to happen during negative credit-to-GDP gaps, which the authors refer to as negative phases of the credit cycle.
• These negative credit cycle phases usually last five years but become more protracted, deeper and wider during banking crises and credit reversals, lasting 1.7 years longer on average.
• Credit reversals reduce cumulative GDP growth throughout the credit cycle in which they occur by 9pp.

Credit Reversals and Banking Crises

Aggregate credit conditions reflect both banks’ willingness and capacity to supply loans and individuals’ ability to borrow (demand). During a credit crunch, for example, both supply and demand forces work together to worsen credit conditions as banks consolidate balance sheets and falling economic activity (and incomes) reduces individuals’ ability to obtain loans.

To discern only the impacts of supply-induced credit contractions, the authors look at two sets of events as proxies of supply-driven, negative dynamics in credit markets. The first is where a contraction in bank credit happened alongside expanding economic activity. They refer to these events as credit reversals, which filter out the demand effects on the availability of credit.

Another proxy for when the banking system is unable to provide credit and normal financial intermediation is during banking crises. The authors define these as events in which there are significant signs of distress in the banking system (bank runs, losses in the banking system, and/or bank liquidations) and significant banking policy intervention measures in response.

Credit Cycles

In the analysis, the authors use credit cycles as the reference frame to study the relationship between credit reversals and banking crises and their effects on credit cycles and economic activity. There is no widely accepted definition of a credit cycle, and so the authors use the evolution of the credit-to-GDP gap (CYGAP) as their credit cycle. A good definition of the credit-to-GDP is provided by the BIS, but in short it refers to the difference between the ratio between total credit relative to GDP and its long-run statistical trend.

In total, the paper exploits yearly data for 179 countries from 1960 to 2017. The authors measure credit by domestic bank credit to the private sector, excluding cross-border lending. And they convert the series to real terms by using the GDP deflator to capture the underlying structure of the credit portfolios. Overall, they capture 371 cycles, with each cycle lasting around 10 years, equally split between negative and positive gaps.

It is worth noting that a positive (negative) gap or phase of the credit cycle refers to a set of years with an uninterrupted positive (negative) CYGAP. Furthermore, credit reversals can occur in either positive or negative gaps. Gaps refer to the difference between the credit-to-GDP ratio and its long-run trend, while reversal episodes refer simply to periods where credit growth is negative and GDP growth is positive.

Analysing the short-term dynamics of credit cycles, the authors find that credit-to-GDP ratios increase by about 9pp during positive gaps and drop by about 2pp during the negative gaps. Also, average credit growth tends to be positive in both phases, which means negative gaps reflect periods where GDP growth is more positive than credit growth.

Co-Movement Between GDP and Credit Growth

The paper starts by analysing how GDP and credit move throughout the credit cycle. First, the authors determine whether the two co-move. On balance, the two appear to move together, with credit growth and GDP growth increasing simultaneously (Chart 1). However, on one in four occasions, credit growth shrinks as activity expands (credit reversal events).

A closer analysis using the concordance index (which measures the proportion of the time where the two variables move in the same direction) shows that the average concordance between credit growth and GDP growth is moderate (0.68 for developing countries and 0.73 for industrial countries). Since 2009, there has been a sharp drop in concordance, which the authors believe reflects excess credit in bank balance sheets and the rapid development of shadow banking.

Credit and GDP Growth During the Credit Cycle

In terms of credit cycle dynamics, credit growth is positive on average in both the up and down phases of the cycle. It is also more volatile than GDP growth – during positive CYGAP phases, credit growth is three to five times greater than GDP growth, while credit contractions can be dramatic during negative episodes.

On the characteristics of the cycle itself, more sizeable and longer-lasting positive phases precede deeper and longer negative episodes. They also lead to lower credit and GDP growth during the subsequent negative phase. Also, sharper increases in the credit-to-GDP ratio during the positive gap episodes lead to larger subsequent declines. Lastly, greater credit growth during the positive gap episodes tend to dampen their subsequent dynamics during the negative phase.

Credit Reversals and Banking Crises

Next, the authors study the relationship between credit reversals and banking crises. Credit reversals occur, on average, once every five years and are more likely after banking crises – in industrial countries, the probability of a credit reversal increases from 8% to 35% following a banking crisis. Credit reversals are also more frequent in developing countries and more likely to happen in the credit cycle’s negative phase.

Meanwhile, banking crises occur, on average, every 7.6 years and are equally likely across industrial and developing countries (roughly once every eight years). However, consistent with the view that crises are credit booms gone bad, banking crises tend to occur in the positive phase of the credit cycle. Specifically, they are most likely to occur four years before the CYGAP turns negative.

Credit cycles feature different characteristics during banking crises and credit reversals. Cycles with banking crises display a sharper collapse in credit growth during the negative CYGAP phase and are also more protracted, deeper and wider – the negative phase of the credit cycles with banking crises is 1.7 years longer on average. Cycles with credit reversals (i.e., ones in which bank credit is contracting and growth increasing) are also associated with deeper, faster and wider drops in credit-to-GDP ratios.

The Economic Costs

There has been extensive research into the costs of banking crises, with findings suggesting a cost of 4-15% of GDP. So instead, the authors of the IMF paper focus on the cost of credit reversals. They find that for credit cycles with credit reversals (solid line in Chart 2), GDP growth is lower versus a cycle without a credit reversal (dotted line in Chart 2) in both positive (right of red line in Chart 2) and negative (left of red line in Chart 2) phases and is more persistent in developing countries.

Quantitatively, the authors find credit reversals have a substantial negative impact on economic activity, particularly in industrial countries. Specifically, for a credit cycle that lasts 10 years, cumulative GDP will be 9pp lower if the cycle included a credit reversal episode. This is only 3pp lower than the cumulative GDP effect of banking crises, providing strong support to macroprudential policies that try to prevent supply-driven contractions in bank credit.

Bottom Line

Central banks have looked towards loosening credit conditions at the onset of economic downturns to avoid any potential credit crunch. However, the authors of this paper show that a lack of credit growth in the subsequent economic recovery phase can have significant negative long-term implications on growth. And so, expect fiscal and monetary authorities to continue to provide favourable credit conditions as economies reopen after Covid to avoid longer-lasting drags on economic activity.

Citation

Vazquez, F. (2021), ‘Credit Reversals‘, IMF Working Paper No. 2021/103