The World's Liquidity Trap
Guest contribution by Andrew Smithers on Japan and the World's economic problem
Andrew Smithers has been my mentor and friend since I arrived in Japan in the mid-1980s. It is he who convinced me to leave academia and engage with the real existing world of finance — in mid-May 1989 I joined him at his SG Warburg in Tokyo, literally a couple of days before the Bank of Japan started its attack on the bubble economy. Timing is everything - there is no question in my mind that the collapse of the bubble and consequent asset deflation super cycle was a much more exciting and stimulating professional training ground than the bubble up-cycle could ever have been. Policy makers were as confused as the best investors; and it is because of Andrew that I got to engage with both groups as someone with ‘skin in the game’. More importantly, Andrew has always been focused on “real macro” and I trust you will enjoy his latest thoughts on global imbalances, the misguided messages sent by new ways to measure fixed capital formation, and what the world could learn from Japan’s experience.
The World’s Liquidity Trap
by Andrew Smithers
The World’s private sector is seeking to save more than it wishes to invest - the same problem caused the slump in the 1930s. Then the whole world suffered, though Japan less than most because Finance Minister Korekiyo Takahashi increased fiscal spending.[1] He was right and, as Keynes showed, in times he called ‘liquidity traps’, large fiscal deficits are needed, despite ultra-low interest rates, to avoid large scale unemployment.[2] The 1930s were one such period and the 21st Century is another.
We now have a new problem. The national debts of several important countries, including France, Japan, the UK and the US, appear to be on track to grow faster than their economies’ ability to pay the interest on them. Conventional economic theory has no policy to answer this, so a new policy based on a different theory is needed.[3] The Stock Market Model (SMM) recommends tax restructuring to prevent the current crisis being followed by a 1930s’ style disaster.
The SMM shows that corporation tax (CT) is not a tax on profits but on investment. Shifting government revenue from CT to income and value added taxes, boosts demand even when the fiscal deficit is unchanged, because the rise in investment that results is greater than the decline in consumption. This policy of tax restructuring solves the current problem in two ways. It lowers the interest governments must pay by reducing the deficit needs to prevent unemployment and accelerates countries’ ability to pay by speeding the growth rate of national income.
Japan is a bit different.
Japan is different from France, the UK and the US as it has a large current account surplus, which makes its problem less acute;, its interest rates are much lower, which makes it likely to get worse and to which its huge national debt makes it sensitive. So long as China runs a large current account surplus, the rest of the world must run a deficit, and this shows no sign of ending. President Trump is trying to end the US deficit, but the problem won’t go away even if he succeeds, it will simply shift to other countries. Countries borrow from foreigners when they run current account deficits. This boosts their economies’ growth if they spend the money on investment and ends in disaster, as the Greek euro crisis showed, when spent on consumption. The solution to the world’s problem is not to cut fiscal deficits, it is to spend the inflow of foreign savings on tangible investment rather than consumption.
Some years ago, I received an unofficial question from friends in the Bank of Japan. “Why” they asked, “had Japan’s growth slowed so much when investment had remained so strong?” My answer was simple. “Real investment has fallen sharply, but the data don’t show this because the way national accounts, including GDP, is calculated has been changed.”[4] The result is dramatic, as shown in Figure 1.
What Went Wrong.
The alteration in data involved a new definition and a new way of measuring investment. Those who supported the change claimed that the value of intangible investment could be measured and had not previously been included in the national accounts data. Both assumptions were, I think, wrong.[5] The value of capital equipment depends on the technology which it uses; even if this requires no new physical investment – the type termed “putty-putty” - and we cannot tell how much of the value of capital stock depends on its intangible (IP) element and how much on its tangible one. It’s not a matter of cost, as equipment which has not benefitted from a putty-putty advance in technology does not cost less than stuff that does, it’s just worth more.
In all countries included in the G5 group (France, Germany, Japan, the UK and the US) living standards have slowed sharply this century, in line with a marked deceleration in the growth rate of the value of capital stock. As Figure 2 illustrates, had the trend rates of growth continued this century, G5 countries would have been between 25% and 50% better off by 2022. This slowdown could be due to less money being invested or to the technology used advancing less rapidly than before. It would be helpful to know, but we can’t tell and, as others have also remarked, it is dangerous to pretend to know things we don’t.[6]
The Technology Bias.
Confused by bad data, the consensus view is that growth depends on the rate at which technology advances. In fact, it depends much more on how much we invest. There have been huge advances in technology not yet exploited because they need new tangible investment. The big stock market successes this century have been the companies with new technology, and this has misled public opinions and policymakers into thinking that technology is the sole key to growth.
Spending on research (R&D) is subsidised heavily in both the UK and the US, and investment on it has risen as it has fallen on equipment. We can’t tell whether tangible investment would have been even lower if less had been spent on R&D, but we do know that the policy of subsidising it has not produced growth.
Persevering with a policy that has failed and expecting the results to differ in the future is a form of insanity.
If Japan’s current account surplus falls as the US becomes protectionist, it will need to boost domestic demand to avoid unemployment. Tax cuts will make its debt problem worse, so Japan should, as it wisely did in the 1930s, ignore consensus theory, which today means subsidising tangible investment. I hope that this will be followed by other countries, preferably before the next financial crisis.
Andrew Smithers, London January 2026.
Andrew Smithers :
Andrew received an MA in Economics from Cambridge University, having been at Clare College 1956-1959., where his supervisor and director of studies was Professor Brian Reddaway. Andrew was born in 1937 and attended Winchester College 1951-1955.
Andrew founded Smithers & Co in 1989. Before that he ran SG Warburg’s asset management business for many years (now part of Blackrock). A regular financial commentator and columnist, and author of many academic publications, he co-authored “Valuing Wall Street” with Stephen Wright (March 2000). He wrote “Wall Street Revalued: Imperfect Markets and Inept Central Bankers” (July 2009), and “Can we identify Bubbles and Stabilize the System?” in The Future of Finance (LSE September 2010), “The Road to Recovery: How and Why Economic Policy must Change” (September 2013), “Productivity and the Bonus Culture” (July 2019) and “The Economics of the Stock Market” (Oxford University Press 2022).
Andrew seeks urgent change to accepted economic theory as the policy on which it is based is the cause of regular financial crisis and bouts of rapid inflation. He is currently working on a book with the working title “Managing an Economy with Multiple Sources of Disequilibria” which he hopes will be published soon.
[1] Cambridge History of Japan Vol 6 p468 quoting Kinshuku seisaku.
[2] The General Theory of Employment Interest and Money by John Maynard Keynes (1964) Macmillan
[3] Coming to Terms with Fiscal and Trade Deficits by Andrew Smithers (2025) American Affairs (Vol. 9 No.3).
[4] The change was internationally agreed under the United Nations’ System of National Accounts (SNA) and implemented in most countries e.g. in 2012 by the US Bureau of Economic Analysis (BEA) in its 14th amendment.
[5] As I explain in Tangible and Intangible Capital by Andrew Smithers (2025) World Economics (Vol. 26 No.2).
[6] Macroeconomics after the Crisis: Time to Deal with the Pretense-of-Knowledge Syndrome by Ricardo Caballero (2010) Journal of Economic Perspectives Vol. 24 No. 4.





Here a question via Michael T., and Andrew’s reply :
Q: I wonder if Andrew Smithers has any knowledge of, or answer to, the Cambridge capital controversies? As far as I know, the problem raised - that you cannot calculate “real” capital stock - has never been answered or disproved. Mr Smithers argument hinges entirely on the assumption, or rather assertion, that it can. For me, though the argument is interesting, it means it is holed below the waterline. Sorry about that. Always interesting, though.
Andrew : I agree with Michael T’s excellent point that no one has, so far as I know, been able to calculate the size of the real capital stock. It does not, however, stop it existing and it is extensively used in growth equations. For example, the definitions set out in Neoclassical Growth with Fixed Factor Proportions by Solow, Tobin, Weizsacker and Yari (1966) The Review of Economic Studies Vol. 33 No. 2 is:- I (ν)dν the amount of capital stock (physical units) installed in the period (ν, ν+ dv).
In its Fixed Asset Tables, the BEA includes Table 1.9. Current-Cost Average Age at Year End of Fixed Assets and Consumer Durable Goods and I have discussed with Martin Weale and Yu Zheng whether these data cannot be used to construct, using a perpetual inventory calculation, a data series for the volume of the capital stock, but it seems too sensitive to assumptions about the unknown data on scrapping vintages.
Things don’t cease to exist simply because we cannot measure them – the absence of data for it does not imply the absence of something. I make no assumption or assertion that the volume of the capital stock can be measured so my argument is unaffected by the unavailability of the data.
I broadly agree with Smithers, particularly when considering Japan’s current output gap. With demand largely no longer deficient, broad-based stimulus or cash handouts risk fueling inflation or higher interest rates rather than real growth.
Fiscal policy still has a role, but the emphasis should clearly be on the supply side. Investment that expands productive capacity and removes bottlenecks — infrastructure, energy, logistics, and capital deepening — is far more appropriate at this stage than demand-boosting measures.
In short, if Japan uses fiscal policy today, it should aim to raise potential growth, not artificially stimulate consumption.