There has been a great deal of coverage in the last few days of the New Zealand Superannuation Fund, all prompted by the news that the chief executive, Adrian Orr, had been given a substantial pay increase by the Fund’s Board, over the objections of the State Services Commission and the then Minister of Finance.

I don’t have a strong view as to how much the chief executive should be paid.  In general, I also don’t have a particular problem with that amount being determined by the Board, without ministerial involvement.  Then again, this is simply a body managing a large pool of (borrowed) government money, and I couldn’t see a particular problem if the relevant Act was to be amended to make the terms and conditions of the chief executive a matter determined by the Minister of Finance, or the State Services Commission, perhaps taking advice from the Board.   After all, that is exactly the model that applies for the Governor of the Reserve Bank.

Amid the recent media coverage, there has been a lot of hyper-ventilation about the performance of the Fund, and of Orr himself.  In his Dominion-Post article, Hamish Rutherford reports that

One commentator suggested if Orr had achieved such a return in New York he might have made a billion dollars.

That seems unlikely frankly.   Orr simply isn’t –  and I wouldn’t have thought he’d claim otherwise –  some investment guru, blessed with extraordinary insights into markets, prospective returns etc etc.  He was a capable economist, and a good communicator (at least when he doesn’t lapse into vulgarity), who turned himself into a manager and seems to have done quite well at that.   He always seeemed skilled at managing upwards, and his management style (in my observation at the Reserve Bank) seemed to err towards the polarising (“are you with us, or against us”), attracting and retaining loyalists, but not exactly encouraging diversity of perspectives or styles.  He isn’t exactly a self-effacing character. (That is one reason I’m not convinced he is quite the right person to be the next Governor of the Reserve Bank.)

The New Zealand Superannuation Fund has made money, both before and since Orr took over a decade ago.  Of course, amid a trend increase in global asset markets it has been hard not to.   The NZSE50 gross index, for example, has increased at an annualised average rate of about 9.8 per cent per annum since 1 September 2003 (when the NZSF opened its doors).

As for how good the NZSF have been, it is probably too early to tell.  Don’t take my word for it: here is how they themselves put it

It is our expectation, given our long-term mandate and risk appetite, that we will return at least the Treasury Bill return + 2.7% p.a. over any 20-year moving average period.

The Fund has now been operating for only about 13.5 years.  In some respects, the returns to date look quite good –  they’ve averaged 5.6 percentage points per annum above the Treasury bill return –  but for a Fund with the sort of risk parameters they have adopted one can only really evaluate performance over very long periods.  And global asset returns have been pretty attractive over much of the last 15 years.  Will that be repeated?  Will there be a big sustained correction?  The only honest answer is that no one knows.  (And the 20 year time horizon is probably a reason why the institution’s CEO shouldn’t be remunerated to any significant extent on some investment performance formula –  unless there are clawbacks built in for the next 20 years).

But even on the returns to date, it might be reasonable to pose some questions.    The Fund puts a lot of emphasis on expected returns, and not a lot (at least in the published material) on the risk they are running.    In some respects, that is in line with Parliament’s mandate for them to be

maximising return without undue risk to the Fund as a whole;

What, we might wonder, is “undue”?   Who decides, and under what constraints?

A common measure of risk, especially on assets that are frequently marked to market, is the variability of returns.    One tool for relating returns to risk is the so-called Sharpe ratio, which compares the incremental returns obtained through the fund manager’s investment management choices (ie the margin above a risk-free rate) with the standard deviation of those returns.  If the resulting number is very low, the incremental gains might often be prudently best treated as “noise” –  good luck, perhaps, rather than the result of a consistently superior investment strategy.  On the other hand, all else equal a high Sharpe ratio, over a reasonable period of time, provides greater reassurance that the fund manager is adding value.   When I ran the Reserve Bank’s financial markets operations, we had able staff proposing all sorts of clever active management schemes to add value to our foreign reserves operation.  Sharpe ratios were one of the tools we used to evaluate prospective and actual results.

How has the NZSF done on that metric?  Since it opened the doors, the average annualised return has been 9.9 per cent (recall that NZSE50 return of 9.8 per cent).  Treasury bills –  the Fund’s risk-free benchmark –  provided an average return of 4.3 per cent, so the average margin over the Treasury bill return was 5.6 per cent.

But the standard deviation of those annual excess returns over the full period since September 2003 is around 13.5 per cent, for a Sharpe ratio of just over 0.4 per cent (and these are all pre-tax numbers).  That is pretty low.  In other words, while the headline returns –  through a period of strong asset price growth –  may have looked impressive, the risks they have been running have been (deliberately and consciously) high.   I checked, by way of comparison, the returns on the low-risk (low return) superannuation fund I’m a member (and trustee) of: since 2003 the standard deviation of the annual returns on that fund since 2003 have been around 4.5 per cent.

Adrian Orr has now been CEO of the NZSF for almost a decade.  In that decade, annual returns (above Treasury bill) look to have averaged just over 5 per cent, but the standard deviation of those annual returns has been higher at around 17 per cent.  In other words, the Sharpe ratio for the Orr years, is even lower than that for the full period of operation.  But, as a reminder, the Fund itself reckons one needs a 20 year run of data to evaluate their investment management performance.

Based on the NZSF’s own data the monthly returns are also pretty volatile.  The standard deviation of monthly returns (over the risk-free rate) over the life of the Fund has been around 3.3 per cent.    Given that many of the Fund’s holdings are quite illiquid, one probably shouldn’t put too much weight on the monthly return numbers, but it is a reminder of just how much risk the NZSF is incurring –  not for itself, but for the taxpayers of New Zealand.  At best, they might just have been getting compensation for the risk they’ve taken, but there doesn’t seem to be anything exceptional about their performance given that level of risk.    That, in itself, isn’t intended as a criticism: why would we expect a public agency in New Zealand to be able to add much (risk-adjusted) value, whether through asset allocation, or tactical departures from their own internal benchmarks?  But it is a bit of a reality check.  And as Hamish Rutherford noted, on deals like Kiwibank, the super fund’s returns are, over time, likely to be flattered by the privileged position NZSF had going into negotiations –  there were very very few buyers acceptable to the government, and ACC and NZSF will have known that, and reflected it in the price they offered NZ Post.

My own unease about NZSF is rather more fundamental, and doesn’t reflect on any of the individuals involved in managing the funds or the organisation.   The NZSF is often loosely described as a sovereign wealth fund.  In fact, it is nothing of the sort.    Norway and Abu Dhabi have sovereign wealth funds –  accumulated from the proceeds of the sale of state-owned natural resources (oil and gas).   It is real wealth, and needs to be managed somehow.  Of course, it could all be passed on to citizens to do with as they please, but there are plausible –  not necessarily 100 per cent compelling –  reasons for managing the flow of the proceeds of the sale of a large non-renewable natural resource over time.    If so, the money is there and has to be managed somehow.

By contrast, the New Zealand Superannuation Fund arose because successive governments took more in taxation from New Zealanders than they needed to fund their operations.  At one stage at looked as though the New Zealand government would manage to build up a large financial asset position.  But, except briefly just prior to the 2008/09 recession, they didn’t even manage to do that.  Instead, we now have a quite large stock of government debt outstanding, $33 billion of which is used to run a state-sponsored and managed quite-risky hedge fund.   It is a discretionary commercial operation, and it should be evaluated on the same sorts of grounds Treasury and the government lay down for other investment projects.  And given that risk imposed on us by the government is risk (capacity) we could ourselves otherwise choose to utilise elsewhere, it should also be evaluated by looking at the sorts of returns private sector businesses require in analysing possible uses of capital.

Treasury has recently revised downwards the pre-tax discount rates it recommends government agencies use in evaluating projects.  Their default recommended rate is now 6 per cent real (or around 8 per cent nominal), but over most of the period of the life of the NZSF they were recommending a real discount rate of nearer 8 per cent.  They continue to assume an equity risk premium of 7 per cent.  Against those sorts of asssumptions, average annual nominal returns of 9.9 per cent just don’t look that attractive, especially when subject to huge variability (that 13.5 per cent annual standard deviation).    I don’t know what assumptions NZSF are making about expected absolute returns over the next decade, but it would be a bit surprising if they were forecasting/assuming returns as high as those on offer for the last 14 years.

Another way of looking at whether the NZSF is a good business for the Crown to be in, on behalf of taxpayers, is to look at the returns private sector businesses require.  I’ve linked previously to a nice article from the Reserve Bank of Australia, drawing on a survey of private sector businesses asked about what hurdle rates they used in approving/declining investment decisions.  I summarised it previously thus:

They report survey results suggesting that most firms in Australia use pre-tax nominal hurdle rates of return in a range of 10-16 per cent (the largest group fell in the range10-13 per cent, and the second largest in a 13-16 per cent band). Recall that nominal interest rates in Australia are typically a little lower than those in New Zealand, and their inflation target is a little higher than ours.   In other words, it would surprising if New Zealand firms didn’t use hurdle rates at least as high in nominal terms as those used by their Australia peers.     The RBA reports a standard finding that required rates of return were typically a little above the firms’ estimated weighted average cost of capital. The literature suggests a variety of reasons why firms might adopt that approach, including as a buffer against potential biases in the estimated benefits used in evaluating projects.

And here is one of their charts

Bottom line: private citizens shouldn’t want governments getting into businesses –  especially not relatively risky businesses –  where the returns are less than 10 per cent.

There are other reasons to be concerned about the economics of the NZSF:

• putting money into NZSF required tax rates to be higher than otherwise (as would the shared commitment to resume contributions at some point).  Higher tax rates discourage some economic activity that would otherwise occur here, and New Zealand tax rates are not now unusually low by international standards (our company tax rate is quite high),
• the scheme involves all New Zealanders in direct financial exposures to companies/industries they may disapprove of. NZSF attempts to get round that with their ‘socially responsible’ investment policy, but your view of “socially responsible” companies/activities may well differ from that of your neighbour.  Personally, I’d be quite happy to have money invested in whale fishing companies.  Many others might not.    Making those choices simply isn’t a natural or necessary business of government.
• large pools of government financial assets encourage the misuse of those funds in the event that the country/government comes under financial stress at some point in the future.  Those sorts of tail risks aren’t captured in the monthly or annual standard deviation numbers.
• NZSF, being a quite high risk fund, tends to perform well in periods when the government’s finances are not under stress, and to perform badly (very badly in 2008/09) when government finances come under most stress. Because the assets are quite widely held, it provides some protection against some sorts of shocks, but in any severe global economic and asset market downturn –  the sort of event New Zealand is never immune to – the NZSF investment strategy simply ensures that when problems hit they are compounded by investment losses.  As the government is already, in effect, an equity holder in all New Zealand business (through the tax system), it isn’t obvious quite why it should be attractive for New Zealanders to have the government further compound their exposures.  To take those risks might be reasonable for the prospect of exceptional returns, but the NZSF strategies look to do little more than cover a bare minimum cost of capital –  while aggravating our problems when things turn bad.

The NZSF may have been a sensible practical political option back at the start of the 2000s.  Governments were running large surpluses, positive net financial assets were in prospect, and the retirement of the babyboomers was still a decade away.   It makes little sense now, and if anything is a distraction from the necessary discussion about adjusting the NZS eligibility age in line with the longer-term trend improvements in life expectancy.  Rather than debate how to remunerate the CEO, or whether Board members should be replaced, we’d be better to look seriously at winding up the Fund now,  reducing the risks taxpayers ar exposed to and using the proceeds to repay government debt.