Economies have memories even if markets aren’t supposed to. By looking at the capital stock to labour ratio I try to explain why the global excess savings post the equity crash prompted a residential and commercial real estate binge. The problem is that the surplus economies are likely to keep on generating surpluses while the investment needs in the west remain meagre on an ongoing basis given the excess capital stock outstanding.

Thus, I’m worried that the global interest rate will have to take the burden of adjustment again, implying ever lower real interest rates further out the curve, mediated only by western governments appetite to absorb the excess flow through deficit spending on more structures! I can’t believe that’s a stable equilibrium.

—-

Markets may not (allegedly) have memory, but economies do.

The crisis was the result of many factors but I think we can lay part of the blame on rather strange sectoral behaviour in the US and persistent “excess” savings in countries such as Japan, Germany, OPEC and the important new kid on the block, China.

The chart above (click to enlarge) shows the net saving balances of the household, corporate and government sectors in the US as a percentage of GDP. In the run up to the equity crash some rather unusual things were going on in the corporate and household sector balances.

Re-drawing the chart with a shorter time frame and by showing the balances as a difference to the whole period median level makes this clearer. (Again, click to enlarge.)

That the corporate sector went on something of a balance sheet binge from the mid-1990s onwards is well known. What isn’t is that they went on an unprecedented period of surplus afterwards for a total of 18 quarters in a row. The previous record of consecutive corporate quarterly surpluses wsa in 1991 at 8 quarters and you have to go back to 1964 to find anything on a similar scale. Something unusual happened to American corporations during this period which turns out to have been critical to our story, and I’ll come back to it in a moment.

The other strange outcome was in the household sector, not just the size of the deficit being ran but the persistence. The household balance entered deficit in Q2 2002 and remained there until Q1 2008, which at 6 years is a record post war by a huge margin. The only other times the household sector has ran a deficit was between Q2 1999 and Q2 2001, one quarter in 1951 and two quarters in 1955. And that’s it, not one other quarter of deficit since 1950.

Compared to their median levels, the picture is more obvious. The corporate sector went into an unusual surplus from late 2000 and stayed there pretty much throughout the 2000s. They were either generating super normal retained earnings or adding to their capital stock at a very slow pace compared to post-war history. On the other hand the household sector ran a balance that was slightly below the median from the early 1993 onwards, paused during the equity crash and then went for it with a vengeance as rates plummeted. It is only in the past year that the net-household balance has gone above its median.

Lets consider the corporate side of this strange behaviour first, since I said I’d come back to it and I believe it’s a key element to the story. The charts below shows measures of the net capital stock to labour force ratio for the US since 1950. These are a little unusual for a Street broker economist to be showing but they can prove immensely useful, not just at telling us what we already know, but predicting future behaviour (the ratio’s increase over time reflects the normal capital deepening process – in other words the capital stock grows faster than the working population).

The first chart on the left below  looks at the business sector’s total non-residential capital stock against available labour versus its trend line. A few things stand out. First of all the 1981 and 1974 recessions occurred when this ratio was above its long-run trend line. And the 2008, which is the last number available, ratio was substantially above the trend level. Indeed in percentage terms it is a larger gap than 1983. Interestingly the evidence of cumulative over investment in capital from the early 1970s through to the early 1980s (ie above the line) fits well with the historical record and the fact that the capital ratio went sideways from early 1980 to the early 1990s.

Consider the same chart but this time only looking at the stock of equipment and software capital (a sub-component of the capital stock in the first chart) against the labour force. Now the reason for the 1970s and 1980s over-investment is clear – it was in equipment and software not non-residential structures. This time around it appears that this area is bang in line with its long-run trend and as such there doesn’t appear to be much need for anything beyond a cyclical adjustment. Note the slight bump around 2000 – that was the ICT bubble and its subsequent relatively mild burst. It appears to me that this part of the capex cycle didn’t react to the excessively cheap capital on offer because there wasn’t any need and it was so unfashionable and balance sheet constrained from the 1995-2000 episode that another boom was simply unconscionable.

Not so for the other components of business capex clearly from the first chart. Structures are clearly the area to blame. Indeed if we break out the residential, non-residential structures and equipment and software capital stocks against the labour force (see chart above) it seems rather clear where the issues are. Residential is well on its way through the adjustment process, corporate structures isn’t, and equipment and software appears to be reasonably close to its long-run sustainable level as of end 2008.

From my economic cycles have memories point of view this is interesting. When the Fed cut rates aggressively during the 2003 crisis the sector that was responsible for the crisis couldn’t respond because it had plenty of capital and debt. The sector that could was structures and residential, setting us off on an unbalanced expansion path. This is because the excess capital stock today tells us a lot about the relative growth of net and gross investment versus GDP in the future.

The chart above tries to show this by comparing the corporate capital stock to GDP ratio against the growth differential between corporate investment and GDP over the following 3 years. This is trying to answer the question as to whether capex outgrows or lags GDP depending on the capital stock level in the past. The scatter diagram suggests that this sensible view is in fact that case; high levels of capital stock versus GDP today tells me a lot about the relative growth of both in the future.

Indeed we can use our scatter diagram relationship in the chart above to predict what did and what should happen based on the capital stock today. The chart below shows this. In essence the black line in the chart is telling us that given the current capital stock, corporate investment should grow slower than GDP over the 3 years from 2008, ie up until 2011. Using the sectoral insight from above, this is of course a reference to corporate structures rather than equipment and software. Sadly, the former has been a substantially higher component than the latter. It’s also worth noting that the rate of depreciation on equipment and software is generally longer than on structures which will only serve to elongate the issue over time.

Now here’s the rub. If the previous economic crisis set up this one via producing an undesirably distorted and bubbly increase in the residential and structures capital stock, and both are now in slower than GDP growth mode for several years then are we left with any lessons for this recovery? Well it seems to me that the equipment and software side can and visibly is responding to easy money.

But more worryingly is the other side of the equation – the source of savings. I and many others would argue that the other key component of the last crisis was the excess of savings to investment outside the US in other currencies. That excess of domestic savings to investment I think reflected a change in behaviour from previous major crisis but one – the Asian currency crisis. That money had to find a home. Policy ensured its home was in the USD, and the mechanism for clearing the market was to drive down the “World” real interest rate – in this instance the US government and agency bond curve. That, it seems to me is what drove the capital binge.

Let me put that another way. Several large key economies are persistently exporting capital – in rank order its Germany, Japan, oil producers and of course China that matter. That behaviour either reflects a lack of domestic investment opportunities worth undertaking or to some extent BOP policy and sterilised intervention. By needing to find a home when the US corporate sector’s appetite for investing in capital with actual future returns to pay of the loans was low the money found two new homes; financial sector accumulation of financial assets and residential and commercial real estate construction.

What now? I’ve hopefully demonstrated that the likelihood of another capex boom in places like the US and UK etc is low, yet it doesn’t seem clear to me that the surplus economies will or will want to generate domestic investment of a countervailing nature fast enough. Thus, I’m worried that the global interest rate will have to take the burden of adjustment again, implying ever lower real interest rates further out the curve. I can’t believe that it’s a stable equilibrium, a point I’ll return to when I get back from Australia next week.

RBS small print.

This entry was posted by Kevin Gaynor on Wednesday, March 3rd, 2010 at 10:15 and is filed under Capital markets, People. Tagged with Capital stock, gdp, Kevin Gaynor. Edit this entry.