In 1995, the earthquake in Kobe cost approximately 2% of GDP. The area directly aff ected by this year’s earthquake and tsunami refl ects approximately 6-7% of total Japanese private capital stock, number of households and output. If we assume the cost of the current earthquake to be around 5% of GDP, the nominal amount would be around USD 250 bn (0.05 * USD 5 trn USD). If it were all Government debt fi nanced, the debt level would increase by approximately 2.5% of GDP. Not insubstantial, but not enough to have a long-term signifi cant impact when taking a closer look at the debt position: approx. 95% of the STOCK of Japanese government debt is held by domestic investors. 55% of the assets of Households are held in cash which, via the banks, fi nances the government debt. Japan can fi nance any amount of FLOW of new debt (and the debt / GDP ratio is just over 200%) so long as NET Domestic savings EXCEED the budget defi cit. They can roll the STOCK of existing debt so long as, very largely, the level of household assets EXCEEDS the stock of debt. For the last 20 years, both conditions have been met (see chart 1), enabling the Japanese bond market to rally and grow. From a domestic Japanese investor perspective, buying 1-2% yielding JGBs in defl ation gave a REAL yield of 2-3%; few assets globally have been better in the last decade or two. From here, household assets will be LESS than the stock of debt within 3 years on the longer term trend and net domestic savings will fall short of the defi cit: Japan is now at the turning point. The earthquake will probably not break it, but it prevents any reversal of the current trend to fi nancial crisis. They already spend 21% of revenue on debt service and each 100 bps rise in the cost of capital causes a further 20-25% of revenue to be diverted to debt service. A rise in yields of just a couple of hundred bps will bankrupt the government. The fact that the debt STOCK is largely domestically held is irrelevant if they cannot roll it and they won’t be able, AT THE MARGIN, to fi nance indefi nitely a FLOW of NEW debt domestically.

As regards to economic growth, Japan is roughly 6% of Global GDP, so a 10% decline in GDP would reduce global GDP growth directly by 0.6% points; that would not be suffi cient to severely damage global expansion directly. However, the impact on global confi dence and risk premia suggests that the eff ects will be larger. Also, there could be a signifi cant impact on China, less expected by the market. Indeed, 23% of Japanese imports come from China and 21% of Japanese exports go to China. At the end of 2010, the Japanese defi cit with China was almost closed, and at the lowest level since 1993. This represents a negative exogenous shock to an already slowing Chinese economy.

Another obvious issue is the shutdown of the nuclear plants. It represents today a loss of about 6% of the power generating capacity for Japan. There is suffi cient spare capacity in other energy sectors to accommodate the loss. Moreover, an exogenous shock of this nature when the economy is below trend (IP is 17% below trend) should not have severe and lasting impact. However, although it appears unlikely and premature at this stage, the risk of all out meltdown and catastrophic emission of radiation needs to be considered.

Concerning assets, just after the Kobe earthquake, the Yen appreciated against the US Dollar from 100 to 80 as capital repatriation (by insurance companies and private agents to cover losses) was enough to drive the Yen higher. The BoJ reacted very slowly then and fi scal stimulus was also slow in coming. This time, liquidity has been injected very rapidly and in large size. The prospects for the Yen will be a product of the NET capital fl ow (repatriation by Japanese investors minus repatriation by foreigners plus liquidity injections by the BoJ). The BoJ has added almost USD 330 bn in liquidity to the system since the earthquake hit (as of Wednesday 16th), or 6% of GDP. A loss, on that scale, entirely bourn by the private sector, is beyond even the most extreme estimates and so it appears that the NET impact of liquidity injection plus the net private capital fl ows is unlikely to be Yen positive as it was after Kobe.

As for equities, we know that implied equity returns is equal to the dividend yield plus the EPS growth plus the multiple change. Does the shock STRUCTURALLY reduce expected trend EPS growth? We believe not. It simply depresses earnings below trend now and boosts them above trend later. Permanent destruction of the capital stock and/or a substantial permanent reduction in the labour stock and/or a permanent/structural reduction in multifactor productivity are necessary to structurally reduce EPS trend growth: none of those is evident. On that basis, we continue to value the market on a trend earnings basis. The market trades at 20* trailing 10 year reported earnings, over half of Japanese companies trade below their book value (liquidation value) and cash on corporate balance sheets amounts to almost one-third of market capitalization. The market is fundamentally cheap in a way that hasn’t been the case since 1989 (see chart 2). The Price to Book ratio trades at a discount of over 40% to the rest of the world, the price to 10 year reported earnings ratio is at levels unseen since the early 1970s and the risk premium is positive under all scenarios including a decline to record low valuation multiples.
A deterioration of the nuclear situation via a signifi cant increase in radiation levels would lead us to review our positioning; but if authorities manage to contain the situation, the Japanese equity market would appear to be extremely cheap by historical standards in a particularly loose monetary environment.