To Be, Or Not To Be: A Persistent Question for Some SWFs

Ashby Monk

After navigating the tumultuous waters of the Great Recession and the Global Financial Crisis, I think it only natural that some SWFs, in the words of Mr. Shakespeare, “suffer the slings and arrows of outrageous fortune.” In other words, it’s normal that certain domestic constituencies in the countries that sponsor SWFs try to dismantle these funds to use the entrusted assets in other ways.

We saw this in Norway, where the very notion of active investment management was questioned (and the very existence of the NBIM was implicitly threatened). And so it has gone in a variety of jurisdictions, as SWFs are now being forced to justify their existence after some severe short-term losses following the GFC. I personally think it’s healthy to have these kinds of open and transparent debates about the value of public institutions because it ultimately adds to their legitimacy and long-term sustainability.

But, I have to say, the latest SWF to be confronted with a skeptical public is a bit surprising to me: the New Zealand Superannuation Fund. Why is this surprising? Quite simply, because this is one of the best designed and governed institutional investors in the world. Nonetheless, the NZSF has a very adamant detractor by the name of Michael Littlewood. And, to his credit, he does have some worthwhile points that merit our consideration.

In a recent paper, Littlewood takes aim at the Kiwi SWF for wasting the country’s money. In his own words:

“…in the government’s overall balance sheet, the NZSF is effectively 100% leveraged. Every dollar that is in the NZSF is a borrowed dollar, so New Zealand’s taxpayers will be better off only by returns that exceed the cost of the government’s most expensive debt. That hasn’t happened so far.”

To summarize his main argument, if you’ve got heaps of debt to service, then it is wise to pay down that debt before you set up a SWF (especially if your debt is expensive). Littlewood goes into quite a bit of detail on why this is the case – and why NZ should scrap their SWF – of which I’ll spare you the minutia. But you can get it all here if you are particularly keen.

I have to admit that I’m generally sympathetic to the “total accounting context” adopted by Littlewood. And I too think it’s a risky strategy for governments to borrow money for the specific purpose of investing in stock markets. In fact, I recently wrote a paper on pension obligation bonds in the United States that basically said as much. But, ultimately, I disagree with Littlewood’s conclusions. And here’s why:

First, I’m not convinced that the NZ government would actually pay down the country’s debt if the NZSF was dismantled. Littlewood’s thesis seems to be based on a simplified version of the world wherein the government has two choices: pay down debt or accumulate assets in a reserve fund. The reality is that if the NZSF was dismantled, the politicians would be just as likely to use the cash for pet projects or cutting taxes as they would for retiring debt. So, the situation in NZ is not analogous to a pension obligation bond, where there is an explicit tradeoff between the debt issued and the investment in the stock market.

Second, I see the NZSF as a useful commitment mechanism that forces politicians to set aside money to prevent a looming pension crisis. Politicians tend to care less about the state of the nation in 30 years and prefer to spend money (to win votes) in the current electoral cycle. Conversely, the NZSF is a useful mechanism to make politicians consider the future generations of Kiwis when making policy. In a way, it gives voice to a silent, albeit legitimate, constituency: the unborn future citizens.

Third, I don’t think the benchmark that Littlewood uses to evaluate the NZSF is appropriate. The true measure of the opportunity cost of sequestering the assets in the SWF will be the cost of raising debt in twenty years’ time to pay for the demographic costs associated with population ageing. And many would argue that this cost will far exceed the rates that Littlewood is using to make his argument today.

Fourth, the time period he uses to illustrate the futility of the NZSF is unfair: 2004-2009 is not a representative time frame to assess any institutional investor, much less a SWF. And while Littlewood does try to back-test his theory to 1992, it’s hard to give too much credence to his assumptions and claims. In my view, it’s simply impossible to say how the NZSF would have performed over this time period, especially since it outperformed its benchmarks by leaps in the years leading up to the GFC. So who’s to say it wouldn’t have done so since 1992?

Fifth, if you believe in the equity risk premium (as I do) and mean reversion (again, I do), then the next few years could be quite profitable for the NZSF. After all, the NZSF is a remarkable institution with best practice standards of governance and investment management. And, it should be noted that I’m a Canadian-American that works for a British University – I have absolutely no skin in the game here. So I can objectively say that the NZSF is among the best run in the world (sitting alongside such greats as the Canada Pension Plan and the Australia Future Fund). And given that the fund’s first withdrawals are not until 2031, this SWF has plenty of time to make up for any losses.

Finally, Littlewood appears to be personally against pre-funding of pensions altogether, preferring PAYG setups (see here). To me, this seems at odds with all that we have come to know about retirement security and planning in the era of population ageing. Anyway, I’ll just say that being against pre-funding runs counter to conventional thought on the issue. (Which, I should mention, is why I have a lot of respect for Littlewood – I genuinely like people who aren’t swayed by conventional wisdom, ask hard questions, and seek to inspire debate!)

Anyway, all this is to say two things: Littlewood has a valid point about considering the cost of debt before setting aside assets to invest in financial markets. You wouldn’t put money in your savings account if you had heaps of credit card debt would you? (The answer is “No!” – you can’t possibly earn enough on the savings to warrant paying high interest rates on your credit card debt.) Still, I think his conclusions in the case of New Zealand and the NZSF are too critical and based on an overly simplistic view of the world. But, again, that’s just my opinion.

I thought I’d end this post with an excerpt from a Cabinet Policy Committee Memo on the Pre-Funding of New Zealand Superannuation drafted by the Hon Dr. Michael Cullen back in 2000. I think it’s remarkably straightforward and still spot on:

“Establishing a separate dedicated Fund now rather than first paying down debt is preferred for two reasons. First, having to make explicit annual provision for funding of future obligations brings home to the present the cost of future retirement income obligations. The political reality is that, without that discipline, the demands of current expenditure would be likely to dominate, and continuing to rapidly pay down Crown debt much below the comfortable levels we already enjoy would not receive the same priority. Nevertheless, even with the pre-funding scheme envisaged in this paper, net debt levels will continue to fall as a percentage of GDP over the coming years. Second, I do not pretend that there is any “free lunch” from the higher returns likely to be obtained from investing in risky financial assets. However, there certainly are benefits for the financial position of the Crown as a whole from diversifying its portfolio of assets and liabilities. This Government has already made moves in this area with our recent decisions on the Government Superannuation Fund (CAB(00)M11/1G). While building up a general government fund of financial assets would also achieve this end, it would not give people confidence that NZS could be paid in the future. A dedicated fund would much better provide this confidence.”

I couldn’t agree more. So, in answering Shakespeare’s ‘to be, or not to be’ question, I guess I’m with Hamlet: The existence of SWFs may not be ideal, but the alternative is still worse.

4 Responses to “To Be, Or Not To Be: A Persistent Question for Some SWFs”

  1. 1 Michael Littlewood August 2, 2010 at 5:01 pm

    The six issues that Ashby Monk has with my hurdle rate/total accounting context treatment of the New Zealand Superannuation Fund (NZSF) fall into three groups: political, macro-economic and micro-economic. Before getting to those, I should say that I think New Zealand has the best designed Tier 1 pension in the world. It’s simple, effective, cost competitive and it works on just about any relevant measure. Next, if New Zealand has to have a SWF, the NZSF seems to be one of the better designed – whether the “best designed and governed institutional investor” is beyond my competence to confirm.

    But if New Zealand doesn’t need the NZSF then it doesn’t matter how well it is designed.

    1. The political context: Ashby justifies the NZSF on the grounds that we cannot trust politicians to actually pay down the debt from fiscal surpluses. This leaves aside the issue of effectively (as I suggest) borrowing money to invest in financial markets that I deal with later. He says that politicians might use the cash for pet projects instead.

    I think that this issue of trust, or distrust, is actually the only plausible explanation for the establishment of the NZSF in the first place. The government was collecting too much tax (significant fiscal surpluses). The NZSF’s author (the then Finance Minister Michael Cullen) did not trust voters to spend any tax reductions wisely. Neither, seemingly, did he trust his ministerial colleagues to spend tax revenues wisely by increasing their budget allocations. So, perhaps, he thought he would ‘remove’ about a third of the surplus and put it out of harm’s way. As an aside, I think that also justifies the extraordinary subsidies offered at the last moment to KiwiSaver in 2007 (a tax reduction Kiwis could not spend) but that’s another story.

    My answer to Ashby is that his distrust of politicians should apply at the end of the process, just as much as at the beginning. He apparently trusts tomorrow’s politicians to reduce other taxes collected during the drawdown from the NZSF by the amount of that drawdown. I think it is just as plausible that the total cost of government will be higher during the drawdown (‘free’ money from the NZSF) by the amount of the drawdown. There is no commitment in the NZSF’s structure that binds tomorrow’s governments to take full account of the drawdown when it works out what total income it needs to meet expenditure commitments.

    The political distrust should also apply to the ongoing accumulation – just imagine the dilemma – New Zealand has a catastrophic earthquake (quite a like scenario); the government’s earthquake fund is entirely exhausted; the economy has taken a huge hit; tax revenues have fallen and the government has huge infrastructure commitments and can’t persuade financial markets to lend it the money it needs at a reasonable cost. The NZSF has realisable assets that would meet that entire cost so the government is faced with a choice – does it call in the NZSF or raise taxes? I think that’s a rhetorical question but what about the retirement income ‘bargain’ struck with baby boomers to justify the NZSF’s existence?

    The greater danger is actually pet projects of the kind that Ashby used to justify sequestering the funds in the first place. Yes, it will be more difficult than before but, as I demonstrated in the paper, the last election saw the main opposition party (now the government) promising to change the NZSF’s rules to do what it wanted, rather than let the ‘Guardians’ continue what they were doing (70% + invested overseas).

    So here’s a suggestion. Let’s keep the NZSF’s calculation process in place (the rolling 40 year calculation of the expected cost of our Tier 1 pension) but, instead of giving the money to investment managers as now, the funds are entirely used to repay government debt. That’s essentially what happens in the US with the Social Security Trust Fund. I think it’s not a sensible idea there (or here) but it would satisfy most of Ashby’s political issues (points 1 and 2). It would also go some of the way to satisfying my hurdle rate/total accounting context issues. That does not, however make it a sensible idea because of the reasons I discuss next.

    2. Macro-economic issues: The NZSF will not change the cost of our Tier 1 pension (New Zealand Superannuation – NZS) by $1. It will change slightly its incidence (its future contribution is now expected to be 10-15% of the expected NZS outgo). But the cost of NZS is, and will always be, the amounts paid in the year of payment and it doesn’t matter how money might have been arranged in the government’s accounts beforehand.

    If we are truly concerned about the expected cost of NZS, we have three main choices: first, we can reduce the benefit (either the dollar amount or the starting age); next we could increase the number of taxpayers (babies, immigration) so that the expected old will be a smaller proportion of the workforce than we now contemplate. Lastly, we could work to improve the productive capacity of the employed population so that the smaller proportion of workers will have a realistic chance of supporting the old in the style to which we have become accustomed.

    The NZSF will make either no or only a small contribution to any of these. I therefore suggest that, absent the political and micro-economic issues (discussed next) we should ditch the NZSF and start talking about things that might really help New Zealand to afford the growing aged population. The NZSF is a distraction to the issues New Zealand should really be talking about. The pension itself is only a part of the discussion we need.

    As I suggested in the paper, while it may be/is a good idea for individuals (and perhaps even their employers) to save for citizens’ retirements, it is a fallacy of composition to suggest that the country should behave like individuals. In fact, as I explained, behaving like individuals has the apparent (real at the present) risk of increasing the cost of government, and increasing New Zealand’s risks.

    3. Micro-economic issues: Ashby suggests that the current ten year bond rate (our longest dated and therefore usually the most expensive) is an inappropriate ‘hurdle rate’ and that we should rather use the expected cost of debt in 20-30 years when the pensions have to be paid. Aside from the significant difficulties of actually running that measure, I have to disagree.

    The government was (until the current contribution holiday) collecting about $2 billion (1.1% of GDP) in taxes more than it currently needed. With each dollar of that, it has a choice – invest or repay debt. With every current dollar already in the NZSF, it also has a choice – leave it there or repay debt. Given those choices, the first debt to go would be the most expensive – usually 10 year bonds. I do not understand how else that bargain could be expressed. Just how might Ashby express the counterfactual?

    We updated the direct losses to 31 May 2010 recently (only roughly because we won’t see audited accounts for another couple of months). Against the hurdle rate, the NZSF has cost New Zealand taxpayers about $1.3 billion. The political issues aside, that’s a real financial loss; real money. And, as the paper explains, even if that loss were fully recovered, that would only mean that it would be as though New Zealand had not set up the NZSF in the first place. It would not come close to compensating taxpayers for the losses against a return that recognised the equity risk premium.

    I agree that the period since the NZSF’s establishment has been an unfortunate one for the NZSF’s proponents against which to test my hurdle rate measure. That does not, however, make me feel better as a taxpayer. The 2001 decision has cost $1.3 billion in real money when, for the reasons I described in points 2. And 1 (in that order), it was so unnecessary. The GFC has illustrated the real risks involved in the last government’s fallacy of composition. I would still be unhappy even if the Guardians had exceeded the hurdle rate because the NZSF is an unfortunate, unnecessary distraction.

    I agree that, in the long run, a portfolio of equity investments should return more than bonds. That’s why about half of my own portfolio is in equities. But I have not borrowed money to buy those shares in the way that the NZSF’s existence necessarily implies. Neither have I borrowed money to invest in bonds – the NZSF has effectively done that as well. That is an even riskier proposition as the paper points out and effectively increases the hurdle rate for the equity-based part of the portfolio.

    In a total accounting context, the government has a choice – keep the NZSF or realise it to reduce its debt by about 23% (from $71 billion – see ). Had it done that then, in each subsequent year, its (our) interest bill will be reduced by the current cost of that debt (about $960 million a year). I think that today’s long-term bonds is the better hurdle rate.

    Ashby’s final point concerns my preference for PAYG pensions. I am by no means against “the pre-funding of pensions altogether”. Employers must do that because they may disappear. Individuals must do that because they have no choice. Governments should not do it just because they will never disappear.

    For governments, pre-funding offers no extra security to beneficiaries as individuals are now discovering around the world. For Tier 1 pensions, PAYG is the cheapest, most effective, most flexible, fairest way to pay for them.

    Governments should not be involved in Tier 2 (compulsory, work-related schemes such as the Canadian CPP/QPP) so we do not need to discuss whether they should be pre-funded. They should not even exist.

    Thanks for drawing attention to my paper. I hope it provokes a productive discussion.

    • 2 Ashby Monk August 2, 2010 at 5:26 pm

      Michael: Thank you so much for your detailed comments on the post…and for clarifying some of your views / positions. Again, it’s all fascinating stuff that will keep the gears of my brain grinding away for some time. Without going back through each and every one of your points (…I bet we could fill a book with the various positions and counter-positions…), I’ll just say that you’ve written a very interesting paper (and comment) that will undoubtedly provoke discussion in NZ and beyond. And, as I said in the post, I think we actually agree on quite a bit. So, next time you’re in Oxford, let’s keep the discussion going over a beer. Cheers!

  2. 3 Rien Huizer August 3, 2010 at 1:13 am

    I’m firmly on the side of Mr Littlewood (a strange name for a Kiwi though) for the reasons he mentions plus two:
    1.SWFs introduce unnecessary transaction costs (even if one believes in the equity premium -which is difficult to harvest as we all know)compared to either paying down debt or lowering taxes in order to let citizens/voters/taxpayers do the saving themselves. This is a general comment in all situations where gvts have debts that are much cheaper (higher interest) than the OECD average rate, indicating that the local economy/currency are considered a lot more risky than the average.
    2. New Zealand is a country with chronic c/a deficits (one of the reasons its inflation targeting central bank needs to keep interest rates high in order to minimize imported inflation)it is nonsensical that such a country’s gvt funds an entity that exports capital. The only sensible strategy for a NZ SWF would be to simply subscribe to NZ treasury issues on a rolling basis, which (ignore the 0.5 clerical FTE this would require in the Treasury) would produce exactly Mr Littlewood’s benchmark, be current account neutral and conform to the previous gvt’s undemocratic policy to keep a portion of future revenue outside the spending authority of their (elected!)successors. Singaporeans would consider this good government, but Kiwi’s in general probably not.

  1. 1 The Compassionate Investor: Norway’s GPF-G « Oxford SWF Project Trackback on September 8, 2010 at 1:21 pm

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s


This website is a project of Professor Gordon L. Clark and Dr. Ashby Monk of the School of Geography and the Environment at the University of Oxford. Their research on sovereign wealth funds is funded by the Leverhulme Trust and The Rotman International Centre for Pension Management.

RSS Feed


Enter your email address to subscribe to this blog and receive notifications of new posts by email.

Join 370 other followers

Latest SWF News

Visitors Since August 2010

%d bloggers like this: