The Global Capex Debate - Odds and Ends

Claus Vistesen: Copenhagen

The global economy today is surrounded by many vexing questions which are subject to much debate amongst economists. These questions include interesting topics such as global macroeconomic imbalances, the existence of a saving glut, and excess liquidity. Of course all these topics are intimately related and my impetus for this entry is a recent note over at MS GEF entitled the global capex debate (link is down right now, but it is the 19th february issue), where a number of MS analysts engage in a discussion on the topic. So in this entry I am going to address yet another derivative of the global economy, namely that of global capex (investment and capital accumulation), what it is, where it is and is there too little, too much, or just about enough of it, and lastly, but by no means leastly, what drives it?

What is Global Capex?

Generally, capex can be thought of as the investment and/or capital accumulation which accrues as a response to increased capacity needs. As such, there is both a supply and a demand dimension here since global capex can be operationalized as the supply side to capacity which then translates itself further downstream into demand. This conceptualization might seem false since it by-passes the tradtional definition of AD which in itself includes investment as a by-product of savings. However, as I move along here I will make it clear why we need to look at it this way.

But why is global capex important then? To answer this question we need to include the concept of the global capital to labour ratio (K/L ratio) and crucially to ask ourselves what has happened to this ratio in the last decade or so. One of the most notable changes in the global economy over the last 10-20 years has been the massive supply shock to the global economy in terms of how big emerging economies such as China, India, and Brazil have come onstream effectively skewing the K/L ratio in favor of the latter. This supply shock, as it were, has most widely been conceptualized as the so-called China effect which is normally thought to have had two overall effects on the global economy. Firstly, this labour supply shock has often been seen as a positive growth shock to developed economies or perhaps more accurately to their domestic labour and product markets. The point, in its most simple form, is that the emergence of cheap labour has enabled central banks to keep interest rates low without fuelling inflation in many domestic economies. The underlying point is that the abundance of cheap labour in, for example, China has lead to the import of deflation in consumer goods. A further consequence of this is also said to be one of excess global liquidity (a saving glut) as a result of low interest rates. Secondly, we have the global labour arbitrage argument which is used to explain why wages or perhaps more specifically labour compensation has been subdued in developed economies especially in certain sectors. In the US this discussion is for example packaged as a discourse about inequality and how to explain the seeming correlation between productivity growth and income inequality.

So what was global capex again? Well, if we follow the definition advanced in the introduction to the MS' debate we can say as the global K/L ratio has become effectively skewed towards labour, and that the global economy is suffering from a shortage of capital investment or more specifically capex. With this in mind, we should also take note of the point made by Stephen Jen in the above-linked debate (as well as elsewhere) that the whole concept of 'excess' global liquidity is itself a result of the low level of investment relative to the volume of capital available. This effectively means that Jen himself turns the traditional argument (as cited above) upside down when he advances the claim that excess liquidity will continue until global capex has sucked up the 'saving glut.' More specifically, Jen refers to the fact that excess liquidity is 'real' and not nominal which in Jen's own words means that;

Many, such as some ‘old paradigmers’, paint the image that we are standing knee-deep in liquidity flooded by the G10 central banks, with emerging economies as the accomplices through their currency interventions. But I believe it is the low level of global investment that has led to excess savings, which, in turn, have artificially depressed the real long bond yields in the world. This investment (I) and savings (S) framework is a real concept, not a monetary or nominal notion.

What we have above then is more or less what you initially need to juggle around in your head as we venture along. Moreover, it serves to brings us a step further as we can now ask where global capex takes place, what drives it and perhaps also my personal favourite question: where is the capacity (i.e. demand) for this capex?

The Anatomy of Global Capex

Let us begin with the first question in terms of who does the global capex? This is of course a tricky question since everybody obviously does some of it, but as we read along in the discussion over at MS two interesting points emerge. Firstly, we have the point made above that capex is conducted generally to rebalance the global K/L ratio which would then mean that for example China and India with their large share in the positive labour supply shock to the global economy should suck up a substantial and growing component of global capex. China and India are of course a little different in this respect in terms of their respective growth paths (and where they are in the development process) but for the purposes of this note I will simply leave that to one side.

The second important point to be found in their debate is the one forwarded by Robert Alan Feldman, who argues that ageing societies also will demand more capex as the only way they can grow is through increasing their investment shares and their underlying productivity. With this idea in mind, we can now begin to approach a framework from which we may be able to explain the current drivers of global capex. There seems to be two; 1), the skewed K/L ratio translates itself into a demand for capex (too much money chasing too little investment?) to meet the situation of excess liquidity and the saving glut and 2) the ageing process in some key developed economies, which in the most affected economies then acts as a driver for increasing capex as these economies enter a growth path where rising investment shares and an increase in TFP (total factor productivity) become the only viable way to sustain economic growth and living standards in the face of increasingly flat domestic consumption.

However, if we accept this explanation, I still think that we are missing a crucial part of the picture and this is also why, although I do sincerely believe that the Morgan Stanley Team and Brad Setser are right on cue with many of the points the make about the current conjuncture in the global economy, I am going to argue that we still need to adequately factor in one last aspect here, and I do not think it should comes as any sort of surprise to many readers, that I am once again going to invoke the importance of demographics.

In fact, as I have also argued before here at GEM, Robert Alan Feldman over at MS has already contributed markedly to the discourse by suggesting that the ageing process, via the consequent structural decline in consumption, will tend to transit economies onto a growth path driven by investment and productivity gains. However, what seems to be missing here is an account of the flipside to all this talk about the need for more capex, or more specifically an account of where the capacity is going to come from? In order to grind down to the core here it might serve us well to re-call the classic identity which makes up the GDP in an economy.

Y = C+I+G-(X-M)

Now, if we lock-in (growth in) Y at a constant nominal value and imagine a given country X with a median age of around 35 years we could imagine a composition of GDP where C (consumption) equals 70% of GDP. Then, and this follows from Feldman's point as well as arguments advanced a number of times by Edward and I, we should expect this figure to gradually decline as the populations steadily age, coming down around 60% of GDP as media age approaches the 45 mark due to the life cycle component of consumption and saving patterns (empirical evidence for this estimate can be currently found in places like Germany, Japan and Switzerland).

Now, if we hold the share of G (government speding) constant, we could begin to think about this transition by arguing that what gradual population ageing means is a relative shift from a consumption-driven growth path to one driven by saving/investment dynamics (i.e. capital accumulation). However, since we are dealing with open economies we can also expect that this transition also is a transition from one where the economy (perhaps) runs a trade deficit (i.e. domestic excess demand) to one where the economy runs a trade surplus (i.e. domestic excess supply). In fact, it is difficult to argue against this argument I think since the decline in consumption is a proxy for a shrinking demand side in the economy and as such the extent to which investment (capex) can propel domestic growth relies on the extent to which this investment is a countpart to foreign demand or capacity.

We need to remember here the argument that capital accumulation or growth through capital deepening runs into decreasing returns over time and this is exactly what would happen in a closed economy where the population ages and consumption declines accordingly; in such a case investment will enjoy ever decreasing returns relative to the decline in domestic capacity to absorb the capex. So all this is simply to say although I agree with Feldman we need to consider that this also means that ageing societies will be structurally prone to running trade surpluses (Germany and Japan are important test cases here).

This, I would argue, is the only way that these countries can, in Feldman's words, 'fight off' the structural effects of ageing'. In general, of course, consumption still constitutes the lion's share of GDP in all countries, but I would argue that there is structural limit as to how far this number can decline without having consequences for absolute real GDP growth and in the meantime this structural push will only increase the economy's reliance on exports. This is also the link to my initial comment on the supply and demand made in the introduction since whereas the demand for capex can be operationalized as a function of the skewed K/L ratio we also need to think about the demand for capex in terms of what the capex is used for, that is to say that an ageing population will (in real terms) demand fewer goods and thus require less domestically-oriented capex. Of course a theoretical end limit point here might entail what some have coined as 'rapid' dissaving process further on down the line, but we need to think about what happens between now and then even if we firmly believe this may well finally happen.

Too Much, Too Little or Just About Enough?

Leaving this theoretical example for a moment and going back to idea of the global K/L ratio it seems to me that although there are a lot indications which point to too little capex relative to an increase in global labour we might end up with too much! Why you ask? Well, let me explain ...

While we have indeed experienced an immense positive supply shock to the global economy in terms of the emergence of China, India, and Brazil (etc) there is another factor pushing in the other direction and that is quite simply ageing and we need I think to consider this a global phenomenon. We then get an inverse effect. As such, if capex needs to rise to accomodate a growing global labour force (or global supply shock) we also need to realize that the global labour supply might begin to fall at some point. This of course is not yet a reality but ageing and the demographic transition are not synchronously occuring on a global scale from the one and the same point of departure. In fact, as I have been arguing, some countries, such as Germany and Japan are now ageing rapidly, while the whole process of population ageing is nowhere near as advanced in many other develped economies. So, I am sketching a rather different scenario here than the one forwarded by the discussion over at MS. Consequently, we might end up having too much global capex and ageing can help us explain why. Consequently, if ageing pushes economies towards growth paths driven by investment and productivity gains we also need to take into account the likelihood that this growth path will be driven by the need to export. As such, if ageing is a global phenomenon, it also means that as countries join the club of 'demographic decliners' as Feldman has aptly put it, we are left with a long reaching situation where too much capex will need to be absorbed by too few importers of capital.

Investment is Bad then?

I would not want to walk away from this one leaving the impression that I am advocating saving/investment dynamics to be a bad and un-wanted thing for the the global economy since this would make me a pretty poor macroeconomist. Indeed, there have been many profound empirical economic studies which show how long term growth to a great extent is correlated with the savings rate. In this regard I also think it is important to note the point made by Stephen Jen (linked above) about saving rates and investments rates in Asia ...

'Exhibit 2 shows that in Asia, one of the fastest-growing regions in the world, while the savings rates for NE and SE Asia have not changed that much in the past 15 years, their investment rates have collapsed, even including the massive investment that has taken place in China in recent years.'

This clearly suggests that, at this point at least, there is indeed a shortage of investment relative to savings. Moreover we also have the US economy which is increasingly driven by consumption rather than capex. This of course conversely to my argument above indicates, at least in theory, that current growth is being traded for future growth. Indeed, some are hard at work predicting a recession in US which of course would have marked consequences for the rest of the world. So we also need to look at the imbalances here and crucially how demographics affect economies' growth paths and thus saving investment dynamics.

However, what I ultimately propose is that ageing and its impact on the economy might turn around the short term/long term view in the sense that we need to think about long term growth driven by for example technology and investment relative to a given level of consumption to GDP. In essence, we need to think about the idea of a balanced growth path and what happens when consumption as a share of GDP decreases to below 60%? Of course offloading your goods abroad might provide a brief asylum, especially if you are good at it, but in the end this is not structurally viable since many countries also have debt to pay and expensive welfare systems which after all depend on growth rates in the absolute level of GDP and not simply on changes in GDP per working member of the population (productivity).

Ending on the remarks made by MS I am not at all at odds with the general analysis that the skewed K/L ratio demands capex from a theoretical point of view. I am also much intrigued by the proposed relationship between this and the concept of excess liquidity and a savings glut. I like the idea that the world might, in fact, be in the 'early stages' of a whole capex cycle. I don't diasgree with much here but I think other structural forces are at play here in the long run which are important to consider when we discuss the global economy, macroeconomic imbalances and other related topics.