Tuesday, October 27, 2009

Beyond The Consensus On European Bank Credit

Well, I never thought I would have to wait very long to get some confirmation of my last post on things that could go bump in the night in France, but even I wasn't expecting confirmation of what I was trying to get at so quickly. Now, according to Frank Atkins in The Financial Times this morning:
The eurozone has reported the first year-on-year fall in bank lending to the private sector, strengthening the case for the European Central Bank to maintain its ultra-loose interest rate policy. The latest eurozone credit statistics indicated lending had been scaled back at an unprecedented pace, even though signs have become stronger that the 16-country region’s economy has stabilised.

What are we talking about here?

Basically bank lending to the euro area private sector shrank by an annualised 0.3 percent in September, according to the European Central Bank's monthly report, making for the first contraction in lending since the series began in 1992. In fact, as Frank Atkins points out, there is some positive gleem in the data, since month-on-month there was €14bn pick-up in lending to households in September. Nevertheless lending to households was still 0.3 per cent lower than a year before. That compared with a year-on-year contraction of 0.2 per cent in August. However, before we start talking about whether to put a positive spin on the tealeaves we should make ourselves awar that this entire way of reading things is deeply problematic, since it ignores two vital points (which is why I head this post "beyond the consensus", since from time to time you can read things here on this blog that you normally won't even find in the analyst surveys):

i) when you get near turning points inter-annual data becomes increasingly inadequate, and hence we now need to follow quarterly and even monthly data, or we will miss the turn.

ii) aggregate data masque the big differences we have between the different euro area economies, and this is how Spain and Ireland got into the mess they are in. The big news of the moment, I would argue, is that the credit cycle has clearly TURNED in France, as I will show in the accompanying charts below indicating quarterly annualised movements. In other countries (and particular Spain) the downward drift continues. So basically relying on the average number hides a multitude of sins, as it did last time round when Spain got into the mess it is now in, and this is one of the things I think we should be learning this time round, since if not,..................

The French Credit Cycle Turns

The chart below (which comes from the Bank of France, based on data to September) shows total credit to the private non financial sector. As we can see, on a year on year basis, the rate of credit increase continues to fall (thick blue line). But if we look at the three month annualised rate, we will see that this rebounded after June (narrow black line). What I interpret this to mean is that the credit cycle in France has now turned, and looking at the interannual data you miss the bottom. This finding is pretty important I would say.

Corporate borrowing (SNF) has also bottomed, although even on a quarterly annualised basis it is still negative. Even corportate borrowing should turn positive in the next quarter, and it will be this that should allow the government to take the hand of the "G" button and start to rein-in the fiscal deficit, as win-win growth and inflation dynamics start to set in. But what this also will mean is that the ECB, at least in the case of France, now need to start take off the ultra-loose monetary policy. What a dilemma!

Household credit growth never even reached negative in France, and is now clearly on the rebound too, and with it the French housing market. (Menages in French is households).

For fuller explanation of the deep significance of having the credit cycle turning in France significantly ahead of the rest of the euro area see The French Rebound Continues In October While Germany Moves Sideways.

No comments: