April 2006

The Importance of Dividends in Japan


Contrast the front page summary of a brokers' company research report in Japan and the UK: the salient difference is that for most Japanese companies there is no mention of dividends.

This is because for the last three decades average dividend yields in Japan have been around 1% and investors therefore considered them immaterial to future returns. Actually, as is often the case with dividends over longer periods of time, their contribution to overall returns is much larger than some realise. From 1975-2005 Japanese dividends contributed 31% to overall returns and over longer and shorter periods the contributions were significantly higher, at 69% from both 1955-2005 and 1995-20051 . These dominant contributions are explained in the case of the last 10 years by the poor capital performance of the market, which hardly changed and in the case of the last 50 years by high dividend yields in the early years. Dividend yields in the 1960's averaged 7.6% and in the 1970's 5.1%, which made the compounding effects of these high dividend returns over the first 25 years responsible for the large influence that dividends had in returns over the whole period.

In the second half of the 20th Century dividend yields in all major stock markets fell, but in Japan the fall was much more pronounced. For instance UK dividend yields were 4.6% in 1950, falling to 2.4% in 2000 but yields in Japan fell from 9.7% to 0.7% over the same period2. Clearly, the pace of dividend growth failed to keep up with the capital performance of markets. In the case of Japan, where dividend growth was worse than most, dividends grew in real terms by 0.4%2 per annum which compared to the total real return from the market of 9.1% and real GDP growth of 6.0%.

One of the major reasons for this paltry level of dividend growth in Japan must relate to the ownership structure of the market. For most of the post-war period the ownership of the Japanese market was dominated by cross-shareholdings. Cross-shareholdings, generally between financial institutions and corporations, were designed to cement business relationships in order to provide protection against unwanted stake building and to boost mutual revenues and profits. As an example, a Japanese life insurance company might take a stake in a corporation which would than encourage its employees to buy life insurance policies from the life insurance company. The cross-shareholders were relatively insensitive to conventional investment returns like dividends with the effect that, as the concentration of ownership amongst major corporations and financial institutions increased, shareholder returns were stifled. As an illustration, in the heyday of the 1980's boom from 1987 to 1991 operating profits increased by an annualised rate of 14% whereas dividends only advanced by 9%1. With the benefit of hindsight, it was clear that cross-shareholdings were instrumental in creating many of the distortions that led to the explosive growth in asset prices over that period. Banks in their role as lenders allocated most of their abundant capital to those companies with which they had the strongest relationships or deepest cross-shareholdings even though these companies had no obvious need for it, encouraging rampant speculation in real estate and other such assets. Only when asset prices fell did the banks and other cross-shareholders realise the folly of their ways. As a result from the early 1990's cross-shareholdings began to fall as companies sold shares in each other to shore up balance sheets. Initially this was a slow process as companies first milked other hidden gains, but as the asset price decline intensified so corporate treasurers were forced to consider the investment consequences of tying up so much capital in cross-shareholdings. Thus cross-shareholdings, which had exceeded 50% of the market in 1989, approximately halved by 2005, continue to fall today and are most likely to carry on falling in the future as companies recognise that some of the benefits of cross-shareholdings could just as easily be substituted with a transparent commercial relationship. We believe that such a decline is already affecting the Japanese equity market and will have even bigger implications for it over the long term.

If cross-shareholders have been the major sellers of the market since 1989 the major buyers have been foreign investors and Japanese pension funds whose ownership has risen from 5% to 24% and from 4% to 12% respectively3. This is important because both these groups of investors can be loosely described as portfolio investors (as distinct from cross-shareholders) whose remit is to generate tangible medium-term financial returns for their customers, whether they are pension plans or individuals. As investors, they will likely soon dominate the ownership of the market, if they do not already, and should expect their priorities as major shareholders to prevail over others, where they differ. Crucially, they recognise the vital contribution that dividends make to overall returns and will likely emphasise dividends as an important cornerstone contribution in the future especially because capital gains in Japan have been an unreliable source of return for so long. We know from our own experience that this is already happening. As an illustration we have learned of a consultancy established to represent the views of major foreign institutional investors to Japanese companies. As the consultancy may be representing a number of foreign institutional investors, whose aggregate ownership could amount to a significant minority shareholding of say 20-30%, companies are bound to take note of such representations. In some instances capital allocation policies have been modified. Furthermore, in a culture that shies away from direct confrontation, the involvement of an intermediary allows the company to present any decisions to modify policies, such as paying higher dividends, as its own.

The influence of cross-shareholdings has worked in other ways, which with their fall will further reinforce the case for higher dividends. Many poorly performing companies, usually cyclical, capital intensive businesses have been protected in the past from the worst ravages of a downturn by their relationship investors (cross-shareholders). In practice this often means that banks have continued to extent credit at favourable terms in spite of a material deterioration of the business. However, in the future, as a greater percentage of these companies are owned by portfolio investors they may, quite logically, demand more ongoing compensation, through higher dividends, for the higher risk profile of the business as compared to others in the market. It is not clear that these companies have the capacity to deliver such a return without a wholesale restructuring or refinancing of the business. So far little, if any, of this type of restructuring has occurred as these businesses have performed exceptionally well recently, as profits have recovered over the last 3 years. However when the cycle turns, as it surely will, these companies are likely to have to respond to such changes without the same degree of protection and recourse to their cross-shareholders.

Portfolio investors are not alone in their quest for more ongoing yield from their investments. There is a general need for higher income from investments in Japan to satisfy the requirements of all Japanese savers. Over the next 10 years the Japanese population will age faster than any other developed country. By 2015, 25% of the population will be elderly (aged 65 and over) up from 20% today4 . The elderly account for the lions share of Japanese household financial assets, currently ¥1,400T (£6.8T). These investors have two main requirements from their savings: safety and income. Safety is provided by bank and postal savings deposits that are effectively guaranteed by the government and account for over 50% of household financial assets. Investment in equities is low at 15%, certainly compared to other developed economies like the USA where as much as 44% of household financial assets are invested in equities. This is not surprising given the historically poor returns, lack of safety, and lack of income. Some income is provided by the 19% invested in bonds, but since zero interest rates were introduced in 2001, little or no income is earned from deposits5. As a result it is not surprising that individuals have been looking for other sources of stable income. One such opportunity came with the birth of the Japan Real Estate Investment Trust ('JREIT') market in 2003. When the first JREITs were launched yields were 5.5% on an unleveraged portfolio of properties. Unfortunately, now yields have fallen to 3.0% and leverage has risen to 70%. Most popular has been direct investment in foreign bonds and, in particular, foreign bond funds. These have mushroomed in size over the last five years, seemingly unaffected by the volatility in exchange rates, as customers are most concerned about monthly income payments, rather than short-term capitals gains or losses. Now, 35% of all investment trusts (Japan's equivalent to mutual funds) are invested in foreign securities, mostly foreign bonds, up from 5% in 20006. In addition, this search for income has been aided by legislative change, with the government in 2003 equalising the tax on capital gains and dividends. What is clear from these examples is that there is a vibrant appetite for income, provided that the yield compensates for the risk inherent in the investment. It is also clear that the yield in the stock market today does not yet pass muster from this perspective, as individual investors' ownership of the market remains near its all time low of 20%1. However it is our belief that if the unwinding of cross shareholdings continues, it will in turn force more companies to improve shareholder returns and coupled with the requirement for income as the population ages, this will lead to higher payouts and dividend yields in Japan.

Unfortunately, looking at the market as a whole it is still debatable whether a secular change in payouts or dividend paying propensity has occurred. On the one hand, in contrast with the late 1980's, the annualised growth in dividends of 18%, exceeded the annualised growth in operating profits of 16% from 2000-20051. But on the other, with the FY2005 dividend payout likely to be 24%, it is well below the 60% payout ratios in the USA and UK7. Arguably, the lower growth potential of the economy, both in comparison to history and other developed economies like the USA (where population growth is positive rather than negative in Japan) should mean a lower proportion of retained earnings invested in capital expenditure and a commensurately higher proportion available for shareholders. Still, today, business investment as a percentage of GDP remains at 15%, the same level as from 1991, a full 5%8 higher than both the USA and the UK reinforcing the need for a restructuring of priorities mentioned earlier.

Even if, on an overall perspective, change is slow, there is, at least, evidence that a number of cash generative companies are allocating capital far more efficiently than in the past. These include such companies as Kao Corporation, the 'Proctor and Gamble' of Japan, that in the last 5 years has raised return on equity from 8% to 16%, raised dividends by 150%, its payout ratio to 40%, bought back 12% of its equity at accretive prices and acquired Kanebo, the second largest cosmetics brand in Japan, expanding its cosmetics business by 3x for debt, which, we think, should take only 6 years to pay back7. In our view Kao is a leading exponent of efficient capital management in Japan and a model for others to emulate. It and Nintendo, Takefuji and Canon and other such companies form natural candidates for our long only portfolio. It means that the portfolio's weighted average free cash flow yield is 4.4% (which, theoretically, could be all paid out in dividends as the companies have already accounted for all investment requirements for growing and maintaining their businesses) and its dividend yield is 1.6% (more than 70% higher than the market's at 0.9%) equating to 36% payout ratio, much higher than the market9. Even this portfolio of companies will not attract individual investors for yield alone today, but should these companies continue to raise dividends by double digit percentage gains (the growth rate has been 29% per annum over the last 3 years) as well as to buy back shares, we believe such equity investments will, in time, prove to be just what the individual wants. Companies such as these are leading the way in paying higher dividends and improving shareholder returns for investors in Japan, a trend that we believe will continue for many years to come.

1 Tokyo Stock Exchange, Merrill Lynch, Lindsell Train
2 Dimson, Marsh, Staunton, Triumph of the Optimists, 2002
3 CLSA, Tokyo Stock Exchnage Fact Book
4 Ministry of Health, Labour & Welfare, Japan
5 Bank of Japan, Lindsell Train
6 Investment Trust Association, Japan
7 Bloomberg, Lindsell Train
8 Smithers & Co
9 Lindsell Train

LTL000-039-3 22 August 2006