Dear Prudence: Welcome to my Nightmare

My Newsroom column this week worries about the government's abandoning of normal measures of fiscal prudence. 

Prior to all of this mess, the government's targeted prudent debt levels were 15-25% of GDP. It made a lot of sense in a country subject to substantial risks. Treasury had always figured there had to be room to accommodate one big crisis or two minor ones within a maximum debt to GDP ratio of 60%. We're going to be hitting 53.6% in 2023 and 2024 - low by some standards, very high for a small country that likes to be able to borrow in its own currency and without backstop from someone like the EU or ECB and with very high levels of private debt from foreign lenders. 

And that could be fine if every single dollar of it were absolutely needed in Covid response. But there's been a ton of spending in and since the budget that has little to do with Covid, or mitigating its economic consequences, and absolutely no effort to reprioritise spending from other areas. 

And that's putting us close to a terrifying cliff. 

Remember that we need headroom more than other countries. A big earthquake hitting one city in a country with dozens of major cities is more manageable than one hitting Wellington, or Christchurch, or both. 

Here's my nightmare. Welcome to it. I hope that it's just dreamland stuff and that Treasury's Debt Management Office has done a whipround suggesting that there'd be appetite for further debt on reasonable terms even up to much higher ratios. 

But here's the nightmare nevertheless.

In 2023 the Alpine Fault opens up, as a foreshock. Damage is substantial enough to trigger reinsurance, but GNS predicts a high probability of a bigger earthquake to come - and the Wellington Fault is also under pressure. Kinda like September 2010, but bigger because it's the Alpine Fault, and with higher probabilities of a February 2011 to come.

The debt-to-GDP ratio is already around 54% of GDP. While Treasury and others had been comfortable with those kinds of debt levels because credit agencies said everything was fine and because nothing obvious was happening in the spread between inflation-indexed bonds and standard ones (not that any such spread could open up under QE), there's a very very big difference between international agencies' assessments of creditworthiness for marginal changes around one debt level and appetite for lending another 20% or 30% of GDP. Treasury's modelling had always said to leave room for another 20% in case of a Wellington earthquake; GNS worries that Wellington and the whole South Island are ready to go - say 40% chance of a bigger quake hitting both. Reinsurance obviously becomes absolutely unavailable. The entirety of the predicted shock will hit the government's books because the government backstops EQC and EQC has basically nothing in reserve consequent to the Christchurch earthquakes - or at least nothing relative to the scale of what's expected. Either the government will be taking on more debt to cover EQC's liabilities, or it will find ways of short-changing claimants in ways that would make EQC's Christchurch record look positively generous. 

The government puts a good face on it, reminding everyone of NZ's strong reputation for fiscal prudence and that we're a sound borrower. But quietly, in the background, all of the usual larger buyers of NZ government debt are telling the government, a bit sheepishly, that they're just not able to help us out this time. The risk is too great if the government needs access to another 25% or 30% of GDP on top of its current borrowing. It's been able to get through the foreshock, but if the bigger one comes, things are going to be rather dicey.

In 2024 the bigger one comes. The government needs to take on the debt but knows it has no buyers. Its options are not good. JP Morgan, or a comparable agency, comes round and says "Look. We know you're in a tight spot and we can help. Here are the terms. I expect you'll find them hard to refuse. Your extra borrowing is now denominated in other peoples' currencies because we all know you'll be too tempted to devalue your way out of this mess at these levels. And the interest rates will be high because of the risk. You'll find those interest rates become even higher in real terms for you if your dollar drops in the ways you might need it to to sort things out. You can likely expect that any old debt you want to roll over will likely have to take on similar conditions, at least until you've gotten things back in line. You're going to be back into the world of the 1980s in which the government was spending over 6% of GDP as financing costs on debt. But you really have no other option. It will take you a good couple of decades to get out of it, even with tight control on your spending, and you and I both know that you've forgotten how to do that. You're going to have to learn it again. The Gods of the Copybook Headings have limped up to explain it once more. They always do. They're reliable like that.

The odds of the scenario are low because the odds of the Alpine Fault opening up in any particular year are low. Last time I asked GNS about the risk of the Wellington Fault opening up, it was on the order of 0.83% per year. But it's overdue. It could happen tomorrow; it could happen in fifty years. The longer it takes to get debt-to-GDP ratios down to levels where the government can stand to take on an additional 20%+ of GDP in borrowing, the more likely it is that the earthquake comes before the books are ready for it. If they're projecting ratios of 42% by 2034, well, that's not prudent.  

We must be feeling awfully lucky, if the government is running its accounts the way it currently is. Again - none of this is argument against taking on debt to deal with the pandemic. This is rather argument against the current very lax spending controls in areas unrelated to pandemic response, and the push from many quarters to implement measures that will make it harder rather than easier to grow our way out of this.  

I desperately hope the Debt Management Office at Treasury has had its chats with the usual suspects and that I'm entirely crazy to be worried about this. If they're all very happy to continue with normal kinds of practices up to 80% or 85% of GDP, then I'd sleep easier. 

Anyway, here's a snippet from the Newsroom column:

The Government’s projected path for paying off new Covid debt is very slow, and predicated on some perhaps optimistic assumptions about core government spending returning to more normal levels after 2024. Combined with its willingness to take on more debt to fund ongoing initiatives like higher pay in childcare centres and expanded school lunch programmes, it appears the solid bipartisan consensus about prudent debt levels has been broken.

Before Covid-19, the targeted net core Crown debt range was 15 to 25 percent of GDP. During the budget lockup, Finance Minister Grant Robertson was asked if his projected debt levels in 2034 represented a “new normal” for prudence, or if he expects a continued slow path back to pre-pandemic levels. Unfortunately, he declined to answer – simply defending the new debt as necessary and prudent.

But prudence in a crisis like this also requires making it easier to win back the necessary headroom for dealing with future misfortunes. Treasury’s modelling was based on the likely need to increase debt by about 20 percent of GDP should Wellington be hit by a major earthquake. Since no one knows when a geological fault might open up, taking a decade or two longer than necessary to win back that headroom seems just a little imprudent.