Australia’s Sovereign Rating: Is it an Issue?

Background

Moody’s has recently commented that Australia’s “AAA” sovereign rating could be under pressure, notwithstanding the current stable outlook.  Regardless of whether the pressure is immediate or longer term, this comment is neither a surprise nor newly made.  With additional middle class welfare embedded earlier in the mining boom and then significant debt-funded stimulus during the GFC, the Government has failed to respond to the reversal in commodity prices and associated investment activity that once filled its coffers.   Australia has a structural deficit.  In summary, the Australian Government’s balance sheet is very well footed – if swiftly headed in the wrong direction.  Moreover, burgeoning household debt, a lack of economic diversity and a large China exposure mean that Australia cannot tolerate as much debt as its peers for any given rating.

So is a possible downgrade for the Commonwealth to the “AA” range something to concern ourselves with?  With respect to the credit risk to the Commonwealth bondholder, obviously not.  The number of countries with consistent “AAA” ratings is much smaller than it used to be.  It does not include for instance, Japan, France, UK or the United States.  Even smaller, more indebted countries such as New Zealand can continue to fund themselves reasonably effectively in times of stress with an “AA” moniker.

Default Risk vs Technical Risk

The reality is that in the very high quality portion of the bond universe, default experience is extremely remote and the observed differential in default rates between AAA to AA assets is fairly marginal.   Accordingly, credit risk itself is a very small component of what drives Australian government bond prices.  For instance, using the corporate default statistics we think the additional compensation required from a downgrade from “AAA” to “AA” is perhaps worth, at most, 0.05%-0.07% per annum for a 10 year bond.  Referencing the sovereign, 5year credit default swap market, the difference in cost between Australian (“AAA”) and New Zealand (“AA”) government credit has averaged about 0.14% p.a. in favour of Australia but at times, such as now, favours New Zealand.

Yet we cannot ignore that rating changes can also influence bond returns from a technical perspective.  For instance, a rating downgrade can lead to a reduction in the capacity of, or even prohibit, certain investors from holding Commonwealth bonds.  These investors tend to be offshore sovereign wealth funds or central banks which currently dominate the register.  However, noting the amount of “AAA” sovereign downgrade activity already experienced in recent years (allowing sufficient time for these investors to relax their mandates to accommodate the US bond universe, amongst others) and the minimal impact we have seen on those bond prices, precedent suggests that a downgrade based on technical factors would not cause material price effects.

Impact on State Government Bonds

You may be aware that Ardea IM’s long held view is that Australia’s unique (tortured?) brand of federalism and historical precedent means that the credit risk of the states and the Commonwealth is essentially the same.  This view of interdependence is much greater than that of the agencies, who distinguish state versus federal risk with, in some cases, multiple notch rating differences.  Yet a significant premium exists between federal and state government bonds – more so than the rating default statistics suggest, if you do not ascribe to our view.  This is because many potential Commonwealth investors, such as central banks, are restricted from buying state government bonds and/or are more restricted holders of less than “AAA” assets.  This is borne out by much higher offshore holdings of Commonwealth versus state government bonds.  This technical element, liquidity premiums plus issuance volumes drive relativities between Commonwealth and state bonds.

After adjusting for liquidity differences, we think investors are generally more than compensated for holding state government bonds relative to Commonwealth bonds.

A downgrade of the Commonwealth will have a flow-on impact on the States, since the credit risk of the two parts of government is clearly related.   Victoria and NSW, both “AAA”, would both be downgraded and it is likely the other states would also be downgraded to ensure the current rating relativities remain.

If you do not take our view of state and commonwealth risk, it is intuitive to think that the downgrades will have a greater impact on state government bonds than Commonwealth bonds.  This is because the implied default risk differential from rating point to rating point grows exponentially as ratings decline, however for very high quality ratings e.g. downgrade from “AA” to “AA-“ the observed default experience is still not significant.  And in a sense, as the deterioration in credit quality is driven by the Commonwealth rather than the States, arguably, in a relative sense the fundamental view of the States has improved versus the Commonwealth.  This we think from a credit perspective, the deterioration of the finances of the sovereign as reflected by a downgrade, is unlikely to have a material impact on the credit quality of States, even if one ascribes to the rating agency view of the world.

An opportunity in State Government Bonds

As with Commonwealth bonds, the more likely driver of returns from a ratings downgrade is not the higher credit risk compensation but a further technical reduction in demand for state bonds.  However, it is our view that this risk is mitigated by the fact that state bonds are held more by domestic than offshore investors.  Because the potential downgrade is related more so to the Commonwealth, and the typical offshore investor is more sensitive to rating changes, state bonds should outperform the Commonwealth in this case.   This view is offset by lower liquidity in the state bond market i.e. smaller levels of forced selling are amplified in less liquid markets.

Partly because of this uncertainty and value in semi government bonds, we have been gradually reducing our overweight position to state bonds relative to the Commonwealth but on balance, spread differentials are still compelling.

We think the risk of technical demand reduction will most likely be felt by NSW and Victoria as they are “AAA” rated.  However again, the historical evidence is not great. Whilst Queensland initially did experience some underperformance versus its “AAA” peers around the time of its downgrade in 2009, Western Australia did not in 2014. This can be explained potentially by spread widening happening in relation to an increase in debt issuance rather than the downgrade itself.

To reiterate, we believe the key driver of performance between Commonwealth and State bonds is the relative issuance levels of each market and restrictions on offshore investors not
their ratings themselves. Barring Western Australia, the States are predicted to rein in issuance and balance budgets whereas, as Moody’s forewarns, the Commonwealth debt burden looks set to increase without a drastic period fiscal rectitude. More recent defence procurement spending and infrastructure investment by the Commonwealth (pointedly not the states) will only amplify issuance trends. This supports our preference for state bonds, particularly for inflation linked securities where the new supply projections are increasingly more favourable.

Impact on Australian Bank Exposure

We think a downgrade of the sovereign is more likely to impact the major banks who are ultimately the funders of the Australian current account deficit in offshore markets. Whilst moves are afoot to distance federal government implicit support for the majors, at this stage the implicit (and during the GFC, explicit) support to the banks is akin to 2-3 notches of rating differential in our view. Accordingly, major bank ratings are likely to also be downgraded by at least a notch if the sovereign is similarly downgraded in the near term.

Offsetting this downgrade threat is the greater oversight of APRA leading to tighter lending standards and gradual moves, since the GFC, to sure-up liquidity and capital. Ignoring any change in the outlook for the banks e.g. bad debts, a sovereign downgrade, if not near term,may be partially offset by an improved standalone business profile.

Overall, in this somewhat fluid environment for banks and their relationship with the sovereign, we think the Australian bank exposure will generally under-perform for a number of reasons over the medium term including, ironically, that sovereign support will wane even if the quality of it is less. And a more near term rating downgrade based on the sovereign rating will only add to that negative sentiment. In isolation, we think that a rating downgrade driven by the sovereign support to “A+” from “AA-“ should require an additional 0.10%-0.15% p.a. of credit compensation for 10 year exposure but noting technical and liquidity factors e.g. forced selling due to buyer limit constraints for a lower rating, a downgrade could result in at least twice that amount. However we believe some of this sovereign risk is potentially priced into current bank spreads. We note Australian banks are already trading wider than the lower “A/A-” rated European and even “BBB+” US bank holding companies because of views around the toppy residential housing market in particular, which we share.

We are generally underweight Australian banks.

Andrew Bartlett
Portfolio Manager, Credit and Spreads