May 2007

VARIOUS THOUGHTS ON JAPAN


Economic Growth and the Yen carry trade

It is widely believed that global hedge funds play the leading role in the ‘Japanese Yen carry trade’ as they borrow low yielding Yen and diversify into other higher yielding or capital appreciating assets. What is little realised is the role played by Japanese individuals. Japanese household financial assets were ¥1,540tn ($12.8tn) at the end of December 2006 of which ¥700tn ($8.4tn) was held as Yen denominated deposits. Japanese investors receive effectively no yield on these deposits on account of the low level of short-term interest rates that has prevailed over the last eight years. Over the last three years Japanese investors have doubled their investments in foreign currency assets (mainly bonds and currency) from ¥20tn to ¥40tn. Most of this investment is channelled into mutual funds that account for as much as 70% of the foreign currency investment. Many of these funds pay monthly dividends. Current inflows into overseas invested mutual funds are running at ¥1tn per month (or a ¥12tn annualised rate) which, if continued, will offset the capital inflow implicit from Japan’s trade surplus with the rest of the world. It is likely therefore that Japanese individuals have been as influential as hedge funds and other investors in using the ‘Japanese Yen carry trade’. As the post war baby boomer generation in Japan begins to retire with big lump sum pension payments, we expect these funds to be targeting high yield products such as foreign currency products, arguably accelerating a trend already in place. So far only 2.5% of household financial assets are invested in foreign currencies. As long as Japanese interest rates remain low, we think, this percentage is likely to rise significantly.

Much has been written about the revival of the Japanese economy since 2003. To put it in perspective, from December 2003 until December 2006 nominal GDP increased at an annualised rate of 1.2%. Over that same period US nominal GDP increased at an annualised rate of 6.3%, almost six times as fast. Furthermore, the main contributor to growth in Japan over this period has been exports and its associated capital investment. Pure domestic demand growth has been negligible.

Why is it that consumption, the biggest component of GDP, has not recovered when quoted company corporate profitability has hit new all-time highs? Quoted companies represent a relatively small part of the domestic Japanese economy and in any case are disproportionately orientated towards the manufacturing export economy, the only source of strong growth in recent years. The vast bulk, 74%, of the nearly 48 million non-financial corporate sector workers in Japan work in unquoted companies with a paid-in capital of less than £500,000. From FY1997, when nominal GDP peaked in Japan, to FY 2005 personnel costs per employee in these companies fell by more than 10% to approximately £15,000 per annum. At the other end of the scale, personnel costs per employee for large manufacturers rose by 8% to approximately £35,000 per annum. However, this rise only benefits around 3 million workers, or 6% of the total. Thus the recent recovery in quoted company corporate profits has been achieved via cuts in personnel expenses in smaller companies, which has reduced overall household consumption and will in turn act to prolong the consumption slump.

As the shrinking Japanese population will limit demand for goods in future years, growth will rely on the expansion of demand for services. A growing service economy tends to be dominated by small companies - as the expansion of the economy in the USA illustrated from 2002 to 2006, when the smallest companies created the most new jobs. From 2002 to 2006 companies employing up to 100 workers grew their payrolls by 6.4%, yet over the same period the large companies employing 1,000 or more people shed 3.1% of their staff. The problem for overall demand is that these newly created jobs tend to be the lowest paid and, although the wage disparities between big and small companies in the USA are large (the wage levels of those in very small companies in the USA are 58% of those in large companies), they are much greater in Japan (where wages of the very small company employees are just 38% of those in large companies). Therefore the boost to employment in the services sector needs to be especially large to have an influence on the economy as a whole. Such a boost in Japan will likely come from two factors: a wealth transfer from the corporate and government sectors to consumers, if interest rates are raised and savers begin to earn decent returns on accumulated savings; and deregulation especially of large government controlled service industries such as healthcare, nursing and education.

It is just as likely that short-term interest rates might in due course have to rise to support the weak Yen and quell the outflow of capital, as more accumulated household savings search for yield in foreign currencies, rather than in response to the strength of domestic demand per se. This would be paradoxical as such a move might conveniently and coincidentally begin that wealth transfer to consumers and, ultimately provide a catalyst for the longed-for recovery in consumption.

Japanese Banks

Ostensibly the prospects for Japanese banks have improved since the early 2000’s, a time when government money was invested in preference share issues to shore up balance sheets. First, there were numerous mergers. In an industry such as banking, advances in technology make mergers far more likely to yield economies of scale than in most other businesses. However, in Japan, especially in such a traditional industry as banking, significant personnel reductions that are a necessary part of any such restructuring proved culturally impossible. Thus, the efficiency benefits of these mergers will most probably take years to seep through. Second, lower provisioning and falling bad debt write-offs coincided with an improving economy and, most importantly, property prices, the main collateral to bank loans, recovered. Indeed the bulk of the profits improvement in banks since 2003 came from this source. Third, banks have experienced an upsurge in fee income resulting from securitisation of loans and greater sales of fee-related products such as mutual funds and mortgages. And finally there has been a contribution to profits from growth in new lending, especially to property companies (including real estate investment trusts and developers). However, lending growth has lagged behind expectations as the recovery in profitability and cash flow in the large company corporate sector, which accounts for a significant proportion of the banks’ loan portfolios, has been so good that such companies have been paying back loans and/or resorting to the capital markets instead. Alas, these changes above imply a worsening of asset quality. Not only will banks’ loan portfolios be proportionately more exposed to small and medium sized companies but also the increase in exposure to increased property lending at such low property yields brings increased specific risk as well.

To compound the issue, the poor overall demand for loans has prevented the banks expanding their loan margins as much as they, or other investors, might have hoped. The average contracted lending rate for banks has improved from 1.3% at the end of September 2005 to 1.65% today but not as much as the change in official rates over that same period, which rose from 0% to 0.5%. More significantly, as most banks fund the majority of loans from deposits, loan deposit spreads for major banks, having fallen from 1.6% to under 1.5% from 2004 to 2006, are now edging up to over 1.6% again. However, these changes need to be much bigger to affect banks’ profits materially. Whether spreads widen significantly in the future is open to question. Conventional wisdom would suggest a widening of spreads as loan rates are re-priced faster than deposit rates but crucially it was savers who, by sacrificing all income from their deposits in the last five years, implicitly helped rescue the banks from the brink. For the banks to gouge extra profits out of savers while interest rates rise may not only be politically unacceptable but also commercial folly as it might well result in a further surge of investments abroad in search of higher yield.

We do not own any Japanese banks currently on account of some of the concerns expressed above and additionally because the transparency of reporting remains poor and returns to shareholders are inadequate, with dividend yields at 1% or less (i.e. below the market average). In addition, bank valuations reflect the widely held expectation that profits will improve as spreads expand, so are not especially cheap. This is not to say we would never own them. Japanese banks have one of the largest deposit bases in the world. To put this into perspective, the deposits of Mitsubishi UFJ Financial Group are today equivalent to £533bn, a full 50% more than Citigroup’s, and its retail deposits are 50% more than HBOS’s, the UK’s largest retail bank. For us the deposits, especially retail deposits, make the banking business model so alluring, because we believe there is generally a great inertia on the part of depositors. Only in the most extreme circumstances do depositors bother to move banks in search of better deals or a better service – as is the case currently with Japanese investors migrating to higher yielding foreign currency denominated funds. Theorectically, this gives the banks that run a good service an especially predictable revenue stream from which to plan and execute their business. In reality, having plentiful deposits is a great start but for a bank to make a good investment it needs to achieve a decent return on those deposits. For many years in Japan that has not been possible as interest rates have been so low that achieving an acceptable loan deposit spread has been challenging and remains so today. Indeed the lack of demand for loans is reflected in the loan to deposit ratios of the major Japanese banks, which average 0.76 versus 1.00 for Citigroup and HSBC. A corollary of this and the traditional employment practices is a high cost income ratio that, for the largest Japanese bank Mitsubishi UFJ, is 62% compared to 54% at HSBC and 49% at HBOS which has a comparable domestic retail franchise.

Anomalous Valuations

With Japan’s current economic torpor and threatening demographics it is probably a mistake to rely on sales growth to drive the profits and cash flow of the average Japanese domestic company. Instead, we are concentrating on identifying a number of mispriced opportunities that have arisen from the inefficiencies associated with a market that has traditionally been dominated by cross-shareholders, when business relationships were favoured over financial returns. Now that cross-shareholders are a diminishing part of the market the mispricings associated with these inefficiencies are gradually being unwound. On the basis that long-term equity investment is all about how successful companies are at investing retained earnings it is notable that many of the best Japanese companies fail in this regard. As a result, good businesses that earn a stable return on tangible fixed assets (and net working capital) of, say, 20% may actually be earning a return on equity of less than half that when it should be possible for them to earn one materially higher. But this is changing. These companies have new ‘return orientated’ shareholders - who have replaced cross-shareholders - such as foreigners and domestic pension fund managers who apply pressure to managements to improve those returns on retained earnings. To begin with, when a company’s retained earnings are too plentiful to be sensibly invested back into the business, the excess is distributed as dividends. As evidence, dividend growth for our portfolio over the last four years has been 26% per annum. Such changes further emphasise the most obvious mispricing which, as a simple proposition, is that the worst companies tend to have the lowest dividend yields – in contrast to the West, where investors often require a high starting income return from low-quality equity assets. Again, as evidence, our portfolio, which we judge is a collection of some of the best businesses in Japan, yields 1.9% versus the market’s yield of 1.1%. This is a premium of over 70% and one that has risen from around 50% four years ago as dividends from our companies have risen faster than the market.

At some juncture, investors in Japan will not only value equity dividend streams in a more rational fashion – leading to a shift in relative value from “bad” to “good” businesses, which is where our portfolios are concentrated – but may also value good companies more highly than today if retained earnings or a prudent use of leverage can enhance returns on capital. With such changes in the offing, hastened we think by the increase of corporate activity both in Japan and overseas, we are more enthusiastic than ever about the opportunities offered by investing in good Japanese companies.

Michael Lindsell
May 2007

11 May 2007 LTL 000-048-6