Smithers Revisited: The Irrelevance of the Restructuring Debate

Date February 8, 2002
Speaker Andrew SMITHERS(Chairman, Smithers & Co.)


Japan has two key problems. First, it has a structural savings surplus. Second, there is excess debt in both the public and private sectors. The same medicine can deal with both problems, but it has nothing to do with structural reform. For Japan to recover, it needs a much weaker yen for two reasons. First, a weaker yen in real terms is needed to increase net exports. Second, a weaker yen in nominal terms is needed to trigger inflation, thereby removing excess debt.

Japan's structural savings surplus is associated with its demographics. There are relatively many people in the high savings ages of 30s and 50s, and relatively few over 65 (when people draw on their accumulated savings). The result is that the savings rate is high. Also, there are relatively few people under 20. This means that, without immigration, the workforce will be falling for the next 20 years at around 0.6% per annum. Accordingly, Japan can only absorb, on a profitable basis, a reduced level of investment.

The age structure causes the equilibrium national savings rate to be 4% of GDP above a steady state equilibrium. It also causes the equilibrium investment rate to be 3.5% below the steady state equilibrium. Thus for a sustained recovery, Japan needs a current account surplus of about 7.5%.

So either Japan allows for mass immigration, which is not popular in Tokyo, nor is it politically or economically practical or Japan runs a current account surplus, which is not popular in Beijing, Seoul, or Washington. This hard decision has led policymakers into denial and they therefore chose to talk about less relevant things like bank restructuring. Restructuring is important but less relevant over the next two years.

Japan needs lower investment, as the return on capital needs to be competitive in Japan with that in other mature economies. Labor productivity in Japan is unlikely to improve faster than that in other such economies. The capital income share in these mature economies is stable. Japan must therefore reduce the level of its domestic investment.

Cost cutting is a fallacy of composition: if all companies reduce costs, profits will fall (as was witnessed in the 1930s). Because the capital income share of mature economies is stable, it is not possible for companies in aggregate to cut costs.

Currently, Japan invests 3% more of its GDP in plants and equipment than does the US. With the same capital income share and improvement in productivity, its net investment needs to be about half the US level. It takes 45% more corporate capital to produce the same output in Japan compared with the US. If the capital income share is the same in both countries, returns can only be the same if this gap is closed.

Japan needs net export growth. Its budget deficit is around 7% of GDP and must fall to near zero. Investment in plants and equipment needs to fall by around 7% of GDP. The household savings rate should fall as the budget deficit improves. The demand gap requires a massive rise in net exports.

Reduced investment can raise the return on new spending, but it cannot increase the return on existing capital. Therefore, the current stock of capital must be written down. Writing down capital would be easy if it were equity financed. But 75% of corporate capital comes from debt and most of this is from banks. Write downs require massive write-offs of bank debt.

Low historic return on Japanese assets is not a reflection of bad Japanese management. Japan grew far more rapidly than other G7 countries before 1990. In such a catch-up economy, the real exchange rate appreciates. Japan's exchange rate has risen by nearly 3% per year over the last 40 years. A 7% return in the US has been equal to a 4% return in Japan. The low returns of the past have been due to Japan's rapid growth.

Investment returns must now be the same. To do this, 120 trillion yen of debt must be written off. There are three ways to write off debt: banks default on their deposits (which happened with disastrous consequences in the 1930s), taxpayers pay (courageous but not politically feasible), or inflation. Inflation is the least bad option.

Japan has a structural demand problem. Since it is structural, budget deficits are no answer. Since it is a demand problem, supply-side solutions, such as restructuring, are irrelevant. Devaluation is the solution. Intervention or non-funding provide the route.

Questions and Answers

Q: You say Japan has a demand problem, yet you recommend capital stock be slashed. This story does not seem consistent.

There are two issues here: the return on the flow of new capital (a real economy issue) and the return on the stock of existing capital (a financial economy issue). The return on new capital can only be raised by a reduction in the amount of investment. Japan's return on new capital must be competitive with the return on new capital around the world. There is a limit to economic growth because labor productivity miracles do not happen. Meanwhile, it is impossible to change the return on the existing stock of capital. Destruction of companies is useless, quite bad. You have to do a write down of the book value of those assets. This is needed because the stock of debt held by Japanese companies is greater than the real value of their stock of assets.

Q: Can you explain the transmission of inflation from bond issuance?

If the government issues a bond, people who buy the bonds send money to the government and get bonds in exchange. Money supply falls: bond supply rises. If instead of issuing bonds, the government borrows from banks, the effect would be to increase money supply. What is the transmission of rapidly growing money supply and a recovery of the economy? I would suggest that Japan depress the exchange rate. This is based on the idea of asset allocation. If the supply of money increases rapidly, the supply of money relative to other assets grows rapidly. People are therefore tempted to shift some of that money to other assets. I believe in the intelligence of the Japanese people, so I think that they would buy overseas assets. A fall in the yen would follow.

Q: If the government borrows from banks, the government could loose its independence because a bank's problem would become the government's problem.

My impression is that if this problem exists, it exists already. I doubt the problem would get worse.

Q: Your solution sounds more like accounting rather than economic. A write down is accounting and inflation is a monetary phenomenon. The proportion of consumption to GDP is smaller in Japan compared to other countries. There is scope to expand demand. You need a supply side change because resources are not being channeled to where demand is.

There are two issues: the debt problem and the structural demand problem. The demand problem is a real economy issue. With regard to the debt problem, the first round effect is a monetary problem but it does have a real economy second round effect. You cannot get real interest rates down when you have deflation. The value of Japanese debt is greater than the assets being used to finance it. The debt must be brought down to its underlying value. The best way to do this is inflation (as an accounting answer). It is not a demand change, but it has important consequences. There is a powerful demand effect: you get real interest rates down.

Q: You assume the return on capital should be at global standards and that the demographic situation will continue for a while. If Japan is a capital-rich country, why can't we say that Japan is different, that Japan has a lower return on capital. Non-performing loans could actually be semi-performing given these assumptions. If people invest overseas to find higher returns, the yen would naturally weaken.

If the yen fell to, say, 250 yen to the dollar, you would expect for the next 30 years a rising real exchange rate. You would have a mature economy with a rising real exchange rate. In those circumstances, you would achieve exactly what you are talking about-an equilibrium model for Japan, in which Japan has poor returns on capital. (This seems to be Professor Paul Krugman's model for Japan's economic recovery.) The savings rate would then have to be boosted from the figures I am using because when people are saving for retirement, the lower the return, the higher the savings needs to be. So the situation would look worse. You could therefore treat Japan as an artificially closed economy, by a massive one-off change in the exchange rate. I feel that this model is lacking in realism. I presented you with what I believe to be a more realistic model, where the return on capital has to be on par with other mature economies.

Q: There is a psychological effect of reform on consumers, if Prime Minster Koizumi succeeds.

My model is for the medium term. If economic hopes are raised by rhetoric, I worry that when these expectations are not met, consumers could become more depressed.

Q: Is there a risk of stagflation?

It does not concern me. Demand can go up quite a bit before inflation starts.

*This summary was compiled by RIETI Editorial staff.