Fund sets sail for stormy seas

By BRENT SHEATHER

Contributors and beneficiaries of the huge $3.5 billion Government Superannuation Fund recently got a letter updating them on the fund's new investment policy, which involves a shift in emphasis from bonds to shares.

The fund is the super scheme for many Government employees - the police, the armed forces, judges and MPs, among others.

It closed to new members 10 years ago, but is still one of the country's biggest super funds. All told, the fund will look after the retirement of some 74,000 people.

The fund has about $3.5 billion in assets, but it is estimated that it will have to pay benefits of $12.2 billion in future. The rest of the money will come from two places: investment returns and the taxpayer.

So the way that $3.5 billion is managed is of vital interest to every New Zealander.

Because the fund is Government guaranteed, benefit payments will not be affected by the fund's investment performance. But that performance makes a big difference to taxpayers: a 1 per cent difference in returns means $35 million more or less in taxes each year.

The fund's strategy also is of interest to the average investor because it gives an insight into the thinking of some of the world's leading fund managers, including overseas giants Frank Russell, Alliance Capital, State Street Global Advisors and, closer to home, AMP, BNZ, BT and Tower - all of whom are involved in helping the fund manage its money.

A key part of any savings plan, for $300,000 or $3000 million, is the asset allocation - what proportion to invest in bonds, versus property, versus shares.

For some time now the average New Zealand pension fund or balanced unit trust has favoured an asset allocation of about 40 per cent in bonds, 5-10 per cent in property and around 50 per cent in shares.

Until recently the Government Superannuation Fund was required to keep all its money in bonds, but that restriction has been removed and its investment advisers have recommended a massive shift to 35 per cent bonds, 65 per cent shares, and no property.

The fund portfolio will be more aggressive than the average NZ pension fund: 65 per cent in shares versus 50 per cent. Furthermore, while the average NZ pension fund has 71 per cent of its shares overseas, the GSF will have 80 per cent of its shares overseas. In the bond sector, the fund is again at the extreme of the range, with 60 per cent overseas versus 50 per cent for the average fund.

Unsurprisingly, given the tax advantage, all of the fund's $1.75 billion overseas share portfolio will be passively managed.

As for currency, the fund is hiring a specialist currency manager whose job is to hedge - that is, insure against currency movements - 100 per cent of the overseas bond portfolio and 75 per cent of the international shares.

The fund's decision not to invest in property may raise a few eyebrows given property's resurgence in popularity in the last year, not to mention its performance (best-performing asset class in the UK over one, five and 10 years).

The fund will review its asset allocation annually and may later include property.

The fund's move from 100 per cent bonds to 65 per cent shares represents a huge vote of confidence in share markets. How did its advisers arrive at this new weighting? Presumably the fund's investment advisers rolled into Wellington one day and pointed out that shares have outperformed bonds by 5 per cent a year in the United States since 1924. No problems there; the fact that shares have outperformed bonds is just that - fact.

That is one of the reasons why, all over the world, shares have been steadily displacing bonds in portfolios for more than 20 years. But, as late arrivals to the US stockmarket know to their cost, past performance is not a great guide to the future.

Although GSF management points to the long-term track record of sharemarket investments, there is a danger that the sharemarket party may be over just as the fund arrives.

A recent research article by two prominent academics and investment managers in the US Financial Analysts Journal suggests that much of the popularity of shares results from the fact that historically they have produced a real return - after inflation - of 8 per cent a year.

Shares have outperformed bonds by 5 percentage points a year. In the jargon, that 5 per cent is the "risk premium" - the extra return investors get for accepting the volatile returns that shares produce.

However, says the article, in future the risk premium will be much smaller, and could even be negative. If that is true, shares will produce a lower return than bonds, as well as being riskier.

The GSF is forecasting that its portfolio will earn 3 per cent more than Government bonds, but a negative risk premium would make that prediction look ridiculous.

The paper, "What Risk Premium Is Normal?", by Robert Arnott and Peter Bernstein, concludes that the excess returns that shares produced, compared to bonds, over the past 75 years were not "normal" and the size of the excess was largely a result of a number of non-recurring developments.

Those include a dividend yield which was 5.5 per cent 75 years ago (today it is about 2 per cent) and the fact that the P/E ratio - the price investors will pay for a share, relative to a company's earnings - has steadily risen.

The authors suggest that shares can now be expected to return no more than bonds, and that it is sensible to expect either shares or bonds to produce a real return of about 2 to 4 per cent a year.

The fund's high weighting in shares is not surprising. Yet there are unmistakable signs that the trend towards more shares in portfolios is beginning to turn.

fxdrop, 3, 60 A CCORDING to the US Federal Reserve Board, the proportion of shares in US pension funds peaked in 1999 at 65 per cent, fell to 62 per cent in 2000 and is heading lower. For British pension funds, the share weighting peaked at 81 per cent in 1993 and had fallen to 71 per cent in 2000, with Merrill Lynch forecasting that UK pension fund share weightings could eventually fall by a further third.

Some strategists say the US and UK equity sharemarkets are weak today because pension funds have been switching to bonds. Late last year, the $7.4 billion Boots Chemist pension fund caused a stir when it swapped its entire share portfolio for bonds.

Apparently the Boots decision was not a verdict on the relative value of shares versus equities. It was a matter of risk: Boots effectively guarantees the pension fund's future payments so by swapping risky shares for 30-year AAA-rated bonds, the company can be more certain about its future pension obligations.

Like the Boots fund, the GSF is a "defined benefit" scheme. That means employees contribute a percentage of their salaries and in return receive an annual payment which is based on their salary over the last five years, years of service and age.

These days defined benefit schemes are about as popular with finance directors as WorldCom debt, because if the investments in the pension plan go bad, the sponsoring company has to make up the difference.

With the other type of super scheme - a "defined contribution" plan - the size of the payout is not set, so the contributors take all the investment risk.

With the GSF, the sponsoring "company" is the Government. That means the buck stops with you and me. By opting for lots of shares in a defined benefit scheme, the fund's managers are taking a different approach to that taken by Boots, and one contrary to the strategy advocated by a number of prominent academics.

In a recent article in the British Financial Times, two J P Morgan executives quote Boston University Professor of Finance Dr Zvi Bodie as arguing that close to 100 per cent of a defined benefit plan's assets should be invested in highly rated fixed interest securities. "As long as a company is solvent, shareholders bear the risks and rewards of allocating plan assets to equities, since the fund beneficiaries enjoy a defined benefit with no upside". The shareholders in the case of the GSF are the taxpayers of New Zealand.

As well as raising questions over the asset allocation, GSF's management and advisers may also be overstating the returns the fund can expect from shares.

One story we are hearing repeatedly in leading financial papers such as the Financial Times, the Economist and from sharemarket gurus like Warren Buffet and John Bogle is that the days of double-digit equity returns are well and truly over.

In the long run we can expect returns from international shares to be equal to the dividend yield (2 per cent) plus the long term growth rate of dividends (4-5 per cent a year), implying total returns of 6-7 per cent a year.

John Bogle, founder of the Vanguard Group, a large US managed fund company, reckons 6.5 per cent a year for shares - before tax, before fees and before inflation. The bad news does not seem to have reached the GSF's advisers, who have projected that their passively managed international shares will return a truly fantastic 9.8 per cent - after tax.

While this explains the fund's commitment to shares, it seems highly likely that its management - and ultimately, taxpayers - will be disappointed.

One important side-effect of GSF's new asset allocation will be a huge increase in total costs, including management fees, most of which will end up overseas. The cost of running the Boots scheme was 0.5 per cent of total assets each year, which would imply that the GSF will be paying about $17.5 million annually, although its management won't be specific because of "commercial sensitivity".

* Brent Sheather is a Whakatane investment adviser.