Why we bought more in Japan
We moved about 2-4% of our discretionary portfolios into Japanese equities (using ETFs for speed and ease of execution) over two days following the tragic events of last Friday. These moves were a response to very large changes in absolute and relative levels of the markets we use as portfolio building blocks. For clients with short horizons or low risk tolerance, about half the move was funded from risk free assets but there was also a switch from other equity markets. For clients with long horizons and typical risk tolerance, fully-invested in equities, it was funded by sales of US and European equities.
Our Defined Outcome portfolios are managed to individually-customised applications of a quantitative model, designed at the asset allocation level to deliver goal-specific outcomes within explicit tolerance ranges at defined future dates. The allocation decisions follow from monthly computer runs of a unique model for every goal-based portfolio for every client. For the first time since the collapse of Lehman Brothers, we reran models mid month, using Tuesday’s Tokyo close and Monday’s close for all other markets.
Several recent posts have focused on the attraction of Japanese equities. In this one, we focus on changes in exposure, rather than the allocations themselves, in the belief any investor, whatever their views, should change their exposures in response to such large price changes. As an explanation, we show below in full what we sent to clients on Monday, in advance of the changes. Though specific to our process, we think it is of interest to investors with different approaches.
Why we should follow the model
Our model is by design ‘contrarian’ but it is also blind to emotional influences on how it allocates capital (other than those emotions captured in the risk preferences for each portfolio goal). It is hard headed and also hard hearted.
The assumptions the model has been given about extreme short-term volatility in absolute and relative returns, from whatever source, allow for the kind of shock experienced in Japanese stocks. These assumptions are what keeps us out of equities for short-term liabilities.
The assumptions the model has been given about uncertainty in the long-term trend of real returns from any equity markets are also extremely prudent. In Japan’s case, the shortage of history (we only use data since the initial post-war reconstruction phase which we think ended in the mid-1950s) means that our trend projections allow for greater uncertainty than either the UK or US, which have over a century of data. But in fact there is widespread agreement amongst expert professionals that we ought not to expect any impact on the long-term trend of real returns from investing in Japan as a result of these tragic events.
The near-term impact on markets is likely to be explained by three factors:
The effects on measured activity – output, GDP, exports etc
Perceptions about the underlying valuation of Japanese equities
The effects on the mood of Japanese investors.
We should comment briefly on all three influences on the impact.
counting the cost
While the destruction of national resources is always a loss, in economics the replacement of scrapped assets is counted as a gain. We can also anticipate that economic output will be temporarily lost due to power shortages. We do not yet know how public capital will be applied to making good personal loss, so that the neutral effects of reinvestment we should expect in the national accounts are not prevented by financial incapacity at the individual household level. Overall, however, the profile of activity will definitely change but not , over a decent interval, its eventual level. Temporary effects should not greatly alter the price investors are willing to pay for long-term profit and dividend streams for Japan’s businesses.
The market’s capacity to absorb such a shock ought therefore to be partly dependent on whether the valuation of share prices is thought to be particularly high or low at the time of the shock. Valuations are generally much less demanding than at the time of the Kobe earthquake in 1995, thanks to the combination of modest returns and gradual improvement in both profitability and profits in absolute. If we examine all incidents of major earthquake activity in industrial Japan since the start of the 20th century it is arguably the health of the banking system and the state of the property market (its main collateral) that has explained best the market impact. Both are infinitely healthier than in 1995.
You will read or hear that the Japanese public sector could not be in a worse shape to deal with a disaster. This misses the point that governments never have money; only the private sector and foreign sector have money. Governments merely attach some of it, through taxation or borrowing. Japan is fortunate as an immensely wealthy and highly successful trading nation that it is not dependent on foreign creditors. In an era of low personal incomes growth and high risk aversion, the government’s financial deficit necessarily follows from the personal sector’s preference for high, liquid and low-risk holdings of financial assets and so has not made borrowing difficult.
If the economic impact of the tsunami is temporary and the valuation context is supportive, the market impact may mainly reflect its unpredictable effect on the mood of Japanese households and their instincts for how to manage their own balance sheets better in response to this sort of risk. We will see some national characteristics come to the fore in this crisis involving fortitude and discipline but we cannot assume either will extend to their risk tolerance. They are likely to be glad they have been so risk averse to date but we cannot be sure they will not decide to be even more risk averse. If they are, the job of markets is to lower prices to tempt them back to risk taking.
Unlike 1995, markets look much more attractive to foreign investors and so their reaction could prove an important factor in determining the subsequent price action in Tokyo. Several major investment houses had said before the earthquake that they were bullish of Japan at current valuations and had been increasing exposure. Amongst conventional valuation measures, yields are now higher than in the US and price to book values lower than either the US or UK. Earnings and cash flow multiples will be less useful for comparative purposes because of the short-term dislocations to sales.