Economic Policy

The ‘balance-sheet recession’ hypothesis and the UK

  • Published: Jun 30, 2011
  • Author: Teodor Todorov
  • Category: Economic Policy

Does the balance sheet recession experienced by Japan after its property bubble account for the rising UK corporate savings since the tech bubble burst in 2001?  At least until recently, the answer must be ‘no’.

Japan and the balance-sheet recession

The balance-sheet recession hypothesis is strongly associated with Nomura Chief Economist Richard Koo. Following the asset price collapse in 1990, Japan is said to have experienced such a recession as corporations moved to repair their balance sheets which had been severely damaged in the property price crash at the end of the 1980s.  The recession only ended in 2005. Koo’s 2008 book, The Holy Grail of Macroeconomics: Lessons from Japan’s Great Recession, tells the story.

Despite near-zero interest rates, firms switched to minimising debt instead of maximising profit. Firms continued to generate positive cash flow but their priority was to reduce leverage and, in the meantime, divert investors’ and creditors’ attention away from their balance sheets and towards the continuing fine performance of their core businesses.

Figure 1 reproduces Koo’s exhibit 2.5 which shows what was happening at the aggregate level.

Figure 1: Japanese non-financial corporations financial surplus and net increase/decrease in financial assets and financial liabilities


Source data (Bank of Japan). Non-financial corporations: Financial Assets FF'FOF_FFYF410A900; Financial Liabilities (FF'FOF_FFYF410L900 - FF'FOF_FFYF410L700); Financial Surplus/ Deficit FF'FOF_FFYF410L700 (Government of Japan, National Accounts of Japan) GDP (Nominal GDP Fiscal Year.  The figures do not include revaluations of financial assets and liabilities. The flow of financial assets and liabilities reflect their transaction amounts as these are a far better guide to the intentions of corporations than if revaluations were included. 

The key point is that after 1990, Japanese non-financial corporations (NFCs) first reduced the rate at which they took on new financial liabilities significantly (the rate of growth of liabilities declining almost every year). From 1997, liabilities shrank year by year, a period of net debt repayment which continued until 2005. 

This switchover from ever-falling rates of growth of liabilities to a period of net debt reduction corresponds to the point at which the corporate sector financial balance moved from deficit to surplus.  Between 1990 and 1997, as the rate of growth of liabilities was slowing so the size of the financial deficit was shrinking.

Up to 2003, NFCs also drew on their stocks of financial assets periodically (for example in 1990-1993 and 1997-1998) in order to meet obligations.  From 2003, some companies started to restore their stock of financial assets.

The ‘balance-sheet recession’ hypothesis and international financial surpluses since 2001

Does this way of looking at things provide a clue to explain the corporate surpluses of other advanced economies that began in the period around the tech bubble.

Koo and others have extended these insights from the Japanese experience to explain the financial surpluses in advanced OECD economies like Germany, the US and the UK in the period following the tech bubble as well as the 2008 financial crisis.

A country may experience certain characteristics of a ‘balance-sheet recession’ (i.e. corporations paying down debt) without going through a recession, or even experiencing poor macroeconomics performance if the reduction in aggregate demand from corporations is replaced with increases in aggregate demand elsewhere in the economy. For example, Koo argues that ‘by replacing the IT bubble with the housing bubble, (Greenspan) managed to keep the US economy going so that US companies had the revenues needed to clean up their balance sheets’ (Koo 2008: 222).

Even if a country manages to avoid a prolonged recession, Koo links the corporate sector financial surpluses since 2001 in advanced OECD countries with a ‘debt rejection syndrome’ following the bursting of the IT bubble (Koo 2008: 158).

Have UK NFC been repairing their balance-sheets since 2001?

However, as figure 2 shows, the behaviour of UK NFCs during the 2001-2010 period has been very different from that of their Japanese counterparts during their ‘balance-sheet recession’ period.

Figure 2: UK non-financial corporations financial surplus and net increase/decrease in financial assets and financial liabilities


Source data (all ONS UK Economic Accounts updated 28/06/2011). Non-financial corporations: net acquisition of financial assets NRGP; net acquisition of financial liabilities NRGR; capital account surplus EABO (all table X18); GDP at market prices YBHA (table A1).



The key point is that corporate saving of UK NFCs since 2001 has not been associated with debt repayment.  The financial surplus  emerged in 2002.  Yet in only one year (2009) did financial liabilities actually fall.  In only four years, did the rate of growth of liabilities decline year on year.  UK also NFCs increased their stock of financial assets every year except for 2009.  The increase was far from modest, averaging over 10% of GDP each year. 

In figure 1, 1990 marks a clear break-point for all three series – assets, liabilities and the surplus – shown there.  By contrast, in figure 2, although the corporate sector balance moves into surplus shortly after the 2000 spike, the assets and liabilities series really don’t change their behaviour in any obviously appreciable way.  If figure 1 is the characteristic signature of a balance-sheet recession, figure 2 shows that that is not what the UK experienced after 2001.

Has the UK been in a ‘balance-sheet recession’ since 2008?

It is possible that Koo’s hypothesis is more relevant in explaining the UK’s current situation following the 2008 financial crisis.  The corporate saving data since 2008 certainly does not provide unequivocal support. Financial assets and liabilities rose in 2008 before turning negative in 2009 (in line with the hypothesis) but then increased again in 2010.

Of course conclusions should not be drawn from just three data points, especially as they may be subject to substantial revision in the coming years.  A reasonable inference from figure 1 is that it can take several years for the aggregate series (1990 to 1997) to change their behaviour completely.  Koo’s theory of corporate saving behaviour can therefore only be properly analysed in the medium term.


If the evidence is against Koo’s hypothesis as an explanation for the corporate saving behaviour of UK NFCs, then it also poses a new puzzle.

If the growing corporate surplus from 2002 were a result of cutbacks or postponements in fixed capital formation, we might expect to see a growing accumulation of financial assets accompanied by a fairly steady rate of accumulation of financial liabilities.  As figure 2 shows clearly, however, this was not the case.  Why this might have been and what its implications might be will be the subject of a future post.


blog comments powered by Disqus