GDP growth forecasts made to within a few 10s of basis points are, quite obviously, not meant to be taken entirely seriously. But getting the trend right is important, even if difficult. Today the ECB cut its forecasts for Eurozone GDP for 2012 and 2013, giving a range of minus 0.9% to +0.3% for 2013), whilst assuring us that a recovery should start 'later in 2013' with sufficient strength to produce growth of 0.2% to 2.2% in 2014. Are they likely to have the trend right?
You will have to scroll down for my conclusions. But it's fair to say they are not cheerful.
Here's how I set about narrowing the odds on getting the trends right. The unknown unknowns will always surprise us, and plenty of the known unknowns will do too, but the starting points I use in determining the outlook for 2013 are these:
- What are the trends in the structural underpinnings of the investment cycle
- Are they likely to change, or be changed?
Returns on Capital & Labour
And the starting point for the investment cycle (and thus also for employment) is shifts in returns on capital and labour. To estimate whether returns on capital are likely to be rising or falling I use a crude asset-turns model, first estimating the stock of capital by assuming all gross fixed capital formation is depreciated over 10 years, and then expressing nominal GDP as an income from that stock of capital.
For the Eurozone, it seems return on capital has been broadly stable for the last 2 years, managing a very modest rise during 3Q12. But there are three things to notice. First, that stability has been won only by allowing growth of capital stock to slow: on my calculations capital stock growth slowed to 0.5% yoy by 3Q, but it took a fall of gross fixed capital formation of 3.2% yoy in 3Q to achieve this slowdown. Second, by pre-crisis standard, ROC remains at historically low levels. Even if one knew nothing else about the Eurozone economy, one would expect that the current pattern of falling capital investment necessary to maintain currently sluggish rises in ROC would be maintained.
Germany's Cycle is Not the Eurozone's Cycle
The third point is that Germany's situation is not the same as in the rest of the Eurozone. In Germany, the ROC may be flat, but it is also at historically quite high levels – higher, indeed, than before the financial crisis. As a result, capital stock is still growing, by around 2.3% yoy. Meanwhile, outside Germany, ROCs are hardly rising from historically very low levels, and there is no longer any growth at all in capital stock. Given these differences, one would expect German investment and growth to continue even as recession continues and deepens in Eurozone-ex-Germany. To be clear: on this basis we can and should expect Germany's investment cycle to be maintained even as the rest of the Eurozone's fails to recover. Germany's economic preponderance over the rest Eurozone will intensify during this stage of the cycle.
And where investment leads, employment follows. During the early stages of a recession, companies lay off staff faster than they can reduce their stock of capital, so in the early stages of a recession labour productivity tends to rise. Subsequently, changes in labour productivity tend to normalize, with changes in output per worker usually simply reflecting changes in capital per worker. When one adjusts to take account of changes in capital per worker, you also find that changes in employment tend to follow changes in labour productivity – as one would expect.
In the Eurozone's case, the news isn't entirely bad. Yes, labour productivity (adjusted for capital per worker) is declining, but on a 12m basis the fall is rather modest compared with pre-crisis normality. And that surprisingly good productivity performance does seem to be providing a backstop for employment so far: the fall in employment during 2012 has been far less pronounced than during the initial phase of the crisis. If these trends can be maintained, the relatively modest rise in unemployment seen so far this year may not accelerate too much next year. (Considering the background, the rise in unemployment rate is modest – from 10.6% at the end of 2011 to 11.7% by October).
(NB. Believe it or not, Europe's unemployment trends are one of the less-dark parts of this picture.) The flip side is that this is also where there is room for disappointment / shocks).
Structural Issues – Money and Financial Institutions
In the absence of other structural issues, these relationships would tend to set the cycle direction. But of course, the Eurozone is a nest of intractable structural issues which concern the relationship between the economy and the Eurozone's financial institutions, and the relationship between the Eurozone's citizens and money. These issues are absolutely fundamental to decisions to save and invest, and so cannot be ducked.
The first issue can be stated quite simply: the Eurozone's financial institutions apparently have insufficient capital (or believe themselves to have insufficient capital) to buy or create risk assets. In addition, their deposit customers are currently reluctant to take the credit/duration risk of making longer-term deposits. Taken together there are two consequences: first, when the private sector secures a positive cashflow, any money they put into the bank is not re-lent to the private sector, so private sector credit contracts. Secondly, partly because these savings are not re-allocated to the private sector (but rather are merely lent to government), the total efficiency of money allocation falls, which we can see expressed in a fall in monetary velocity (GDP/M2). The lack of bank capital not only means a smaller proportion of savings are allocated to investment, it also means that the proportion which is on-lent is also allocated with decreasing efficiency (ie, with less marginal positive impact on GDP).
Structural Issue: Money, Fear & Savings
The second, and related, issue is the change in the relationship between the private sector and money, and in particular the impact of financial risk aversion. When people (or companies) are in the grip of financial fear, they will try to save more, and, having saved, will be cautious about how those savings are stored. We can track the rise of the savings impulse via movements in the private sector savings surplus; the second can be inferred from changes in the types of money people keep. In particular, we can track changes in the proportion of their money people/companies choose to keep in cash (ie, liquidity preference).
The trends of 2012 are very clear: both the rise in the private sector savings surplus (from 4.1% of GDP in 2011 to 4.8% in the 12m to Sept 2012) and the record rise in liquidity preference (with 56.5% of all money currently being held in cash or checking accounts) are both the product of financial fear. The relationships people have with money, and their attitude towards financial risk rarely change quickly – although when they do, the economic effects are profound. There seems little reason to expect any significant retreat in financial fear in the short term. Rather, it seems more likely that 2013 will see private sector savings surpluses climb back towards 2010 levels, whilst liquidity preference at best remains within the levels seen since 2010.
In normal circumstances, monetary policy has usually seemed to manage to influence levels of savings surpluses: depressing bond yields far below 'fair value' (a concept chipped out of blocks of short-term rates, GDP and inflation forecast) is usually followed by a fall in rates of surplus savings; pushing up bond yields far above 'fair value' has in the past been regularly followed by a rise in net savings behaviour. But as the ubiquity of 'quantitative easing' among the world's major central banks betrays, it is precisely those powers which have failed and are still failing.
- Germany's investment cycle is not the Eurozone's cycle – they are operating from fundamentally different starting points and are more likely to diverge than converge over the next couple of years.
- The current trends in return on capital and in investment spending by themselves point to continued contraction – the numbers give no reason to expect the cycle to turn 'later in 2013'.
- With no upturn in the investment cycle, there is no reason to expect labour markets to improve either. For the last two years, the performance of labour productivity has underpinned the relatively mild rise in unemployment. If this can be maintained, we may have no further dramatic acceleration in unemployment levels.
- The structural trends give no reason to expect an upturn in the investment cycle in the short or medium term. But the structural problems of the financial sector, and in particular its inability to buy/create risk assets will tend to exaggerate the negative bias.
- Quite reasonably in the circumstances, the financial behaviour of Europe's private sector suggests fear. Unless this can be changed, we should expect private sector savings surpluses to continue to rise, at the expense of consumption and investment demand. One result of this should be a continuing flow of savings into government bonds and foreign assets.
As I said at the beginning, the unknown unknowns will always surprise us. So let's hope the ECB knows something we don't.