The Spanish economy embarked on its current phase of recovery in 2013 Q3 after five almost unbroken years of shrinking GDP. As usually happens with economic recessions preceded by a credit boom and banking crisis, the real economy started to recover while the outstanding credit balance was still contracting, giving rise to a creditless recovery. Calvo et al. (2006),1 in a pioneering study on creditless recoveries in emerging economies, called this kind of phenomenon «phoenix miracles»: like the mythical bird, production seems to rise again from its «ashes» without the help of credit. The «miracle» part of this term refers to the somewhat paradoxical observation that although a lack of credit is an essential component in explaining the depth of the recession after a credit boom, the recovery nevertheless occurs without its support. This Dossier reviews the empirical evidence on creditless recoveries at an international level to establish their main features and understand why they are exceptional in order to analyse the role of bank credit in Spain's current economic recovery and attempt to determine whether we are witnessing yet another phoenix miracle.
Creditless recoveries are not very common and occur relatively infrequently in advanced economies given their high degree of financial development and wider range of sources of financing. Specifically, according to a study by the IMF based on 388 recoveries in 172 countries between 1960 and 2009,2 20% of recoveries are creditless, a percentage that halves (10%) in developed countries. However, creditless recoveries are much more likely after a banking crisis and a credit boom, occurring 80% of the time under such circumstances. The depth of the preceding recession also plays an important role in the evolution of credit during a recovery. The IMF study also states that average annual growth during the first three years is one third less in a creditless recovery than in a recovery with credit (3.8% compared with 5.7%, see the table). Nonetheless, it is important to point out that GDP growth is still considerable. Hence the «miracle».
In Spain, therefore, it comes as no surprise that seven quarters after GDP started growing, the outstanding balance of bank credit to the private sector, which started to shrink in 2009 Q3, should still be posting negative rates of change (–4.9% in 2015 Q1) although its rate of contraction has decreased notably since the recovery began. However, although it is true that the outstanding credit balance is still shrinking, it is important to note that the credit variable affecting economic growth is actually the increase in credit growth.3 Note that GDP is a flow and, as such, its variation is due to variations in other flows such as new loans granted. In this respect the data are highly revealing. New credit operations4 posted their first positive year-on-year rate of change in June 2014 and are showing a clearly upward trend. This variable, moreover, is very closely related to GDP. By credit type, of note is the good performance by new loans to households for consumption which started growing in September 2013 and posted a 17.7% increase in March 2015, the most recent figure available. These data show that, in effect, growth in GDP and especially in private consumption has been accompanied by growth in credit (see the graph).
The «miracle» aspect has therefore been somewhat debunked as all the evidence points to credit actually playing an important role in the trend in GDP. Specifically, in the first six quarters of the recovery starting in 2013 Q3, credit grew by 1.6% annualised on average. To be able to compare this figure with the existing studies, which cover three years,5 we have used our forecasts for up to 2016 Q2 and estimate that average annual growth will be around 2.2%. This figure is slightly lower than the IMF's 3.8% but it should be remembered that their study includes recoveries in emerging countries whose growth rates have little in common with those of advanced economies. Darvas (2013) presents a breakdown by income level of the country in question and finds that economies with high incomes grow by 3.2% in creditless recoveries (by 4.1% with credit), a figure more in line with Spain's forecast. At this rate of growth, the Spanish economy will return its pre-crisis level of real GDP by next year, following the usual pattern in this kind of recovery.
The pattern of growth in a recovery is also affected by the availability of credit. Investment, highly dependent on external financing, is more affected than consumption. For the Spanish economy's current phase of recovery, we expect consumption to contribute 1.6 pps to the annual growth in GDP, double the contribution made by investment (0.8 pps). This relationship between the contribution of consumption and investment is very similar to the one documented by the IMF's study. In contrasty, the growth pattern is more balanced in normal recoveries (see the table). With regard to net exports, whose contribution to growth is usually zero, we predict a slightly negative figure in Spain's case.
With a view to the coming quarters, several factors related to both supply and demand will strengthen Spain's positive trend in credit. So far, the demand for credit by households and companies has been conditioned by the adjustment in private balance sheets. After a credit boom, especially when accompanied by a bubble in the real estate market, the value of assets falls and existing debt becomes excessive and must be reduced, limiting the capacity to take on more debt. In this respect, the good news is that private deleveraging is now at quite an advanced stage, especially in the case of non-financial firms and its contractive effect should therefore be fairly limited from now on.
On the supply side, there are increasing signs that the factors limiting the granting of private sector loans are also being resolved: the restructuring of the banking sector is now almost complete, banks passed their stress tests with good marks, European banking union is being implemented as planned and regulatory uncertainty is receding. Moreover, in an environment of low interest rates, banks need to grant more loans to increase their profits. In this respect the high demand for TLTROs points to a greater willingness to grant credit on the part of the banking sector.
In short, the current phase of Spain's economic recovery resembles other episodes that have been classified as creditless recoveries, characterised by slower growth and a pattern biased towards private consumption. Luckily, the process of adjustment, both in the supply and in the demand for credit, is almost complete, as can be seen in the rise in new loans granted. We therefore expect the Spanish economy to benefit from this boost to credit over the coming quarters, helping the recovery to gain steam.
Macroeconomics Unit, Strategic Planning and Research Department, CaixaBank
1. Calvo, Izquierdo and Talvi (2006), «Phoenix miracles in emerging markets: Recovering without credit from systemic financial crisis», NBER WP 12101.
2. Abiad, Dell'Ariccia and Li (2011), «Credit-less recoveries», IMF Working Paper 11/58. These percentages are very similar to the ones estimated by Darvas (2013), «Can Europe recover without credit?», Bruegel policy contributory 2013/03. It should be noted that these studies do not identify a causal relationship between the lack of credit and lower growth but restrict themselves to noting the relationship between both variables.
3. Biggs, Mayer and Pick (2009), «Credit and economic recovery», DNB Working Paper, no. 218.
4. We have excluded loans in excess of one million euros as these are usually requested by large firms which are resorting to the bond market.
5. Economic literature defines a creditless recovery as one in which there is negative growth in credit during the three years following a recession.