Australian states are losing their AAA ratings. Should we panic?

That’s the title of my latest piece in Crikey. Paywalled, but I’ve reposted over the fold

The recent downgrading of Western Australian state government debt by ratings agency Moody’s (following a similar step by Standard and Poor’s last year) leaves NSW and Victoria as the only Australian states with a AAA credit rating. The responses have been predictable. The WA Opposition called on the Treasurer to resign, while business groups called for a program of asset sales to restore the ‘prized’ AAA rating. But what does a AAA rating mean, and is it a sensible goal for governments to pursue?

A credit rating is not a seal of approval for the quality of government in general. As the example of NSW shows, a government can be riddled with corruption and still hold a AAA rating on its debt. By contrast, New Zealand, which routinely tops ratings for governmental transparency and freedom from corruption has a rating two grades lower at AA.

Even when the focus is narrowed to economic management, there is little correlation between credit ratings and economic performance. India has been booming for the last two decades, while Japan has stagnated, but Japan still has a higher credit rating.

Credit ratings have a single purpose: to assess the probability that the interest and principal on a bond will be paid in full and on time. A government can be well or poorly run, provide excellent or terrible public services; none of this matters to the credit rating as long as the ratio of debt to revenue capacity remains low.

Credit ratings originate in the corporate sector, and reflect a conflict inherent in the very structure of a corporation. Corporate managers have a fiduciary obligation to advance the interests of their shareholders, but also a legal obligation to honor their debts to bondholders. In some respects, these obligations reinforce each other. Wasteful spending and bad investment decisions will harm both shareholders and bondholders.

When it comes to risk, however, the interests of shareholders and bondholders are opposed. A risk that pays off yields higher returns to shareholders, but gives no extra return to bondholders. On the other hand, a risk that turns out badly enough to send a company into liquidation means that both shareholders and bondholders lose their money.

So, a AAA rating is great for bondholders, but not so good for shareholders. In theory, this should all even out. The AAA rating means a lower rate of interest, which should exactly offset the reduced risk of default. The classic analysis of the corporate finance problem, by Franco Modigliani and Merton Miller reached a stronger conclusion: the value of a company is unaffected by the choice between bond and equity financing.

In reality, however, the profits foregone through the conservative financing needed to secure a AAA rating outweigh the benefits of lower rates of interest on debt. For that reason, all but a handful of corporations have abandoned AAA ratings. Of the 500 leading corporations listed by Standard and Poors in the US, only three have AAA ratings.

A government is not a corporation, but it faces the same problem of being accountable to multiple stakeholders with conflicting interests. The primary responsibility of governments is to the citizens who elect them, but it is neither possible nor desirable to ignore economically and socially important interest groups, including businesses, trade unions and, most relevantly in the current context, bondholders.

The possibility that an Australian government will default on its debt is remote. The closest we ever came was during the Great Depression, when the Lang government in NSW refused to make interest payments to (primarily British) bondholders. The Commonwealth stepped in to make the payments and Lang was dismissed by the State governor, and defeated at a subsequent election.

Unlikely though it is in Australia, overseas experience shows that when governments default on their debt, citizens usually suffer significant hardship. So, it makes sense for governments to manage their finances carefully, to ensure they have plenty of reserve capacity to manage adverse shocks. However, governments have many other concerns, just as important or even more so: they must invest in infrastructure, protect public health, provide education services, deal with natural disasters and so on.

For bondholders, by contrast, these other concerns are of no interest; the only risk that matters is that of default. So, just as with private corporations, there is a conflict between the interests of bondholders and those of other stakeholders, most importantly citizens. Bondholders will always prefer a AAA rating, but that may not be the best choice for citizens.

Also as with private corporations, the interest rate saving associated with a AAA rating may not be sufficient to offset the cost of the measures needed to achieve and maintain such a rating. A upgrade of one step in credit ratings is typically associated with an interest rate reduction of 0.2 to 0.4 percentage points (in the jargon of financial markets, 20 to 40 basis points). On a debt of $50 billion, comparable to that of NSW, that saving would be worth $100 million to $200 million per year. But the policies often advocated to secure or maintain AAA ratings, such as Public Private Partnerships and privatisations may cost far more.

To take just one illustrative example, the failed privatisation of Port Macquarie hospital, undertaken by the Greiner government in the early 1990s, cost the NSW public around $50 million a year, until the hospital was eventually taken back into public ownership. That’s just one regional hospital. The losses incurred through bungled toll road projects, largely borne by road users, are incalculably larger.

For governments, as for households and businesses, low debt levels are a good thing if they can be obtained easily. But forgoing economically and socially valuable investments, or selling assets at less then their true value, in pursuit of a higher credit rating is not fiscal prudence, it is economic mismanagement of the highest order.

14 thoughts on “Australian states are losing their AAA ratings. Should we panic?

  1. Toll road PPPs have been pretty shambolic, to put it mildly. The traffic flow estimates are too prone to gaming, and yet they are a big factor in deciding on a “fair” price. I think it was the tunnel in Sydney which had only one third the predicted traffic flow, bringing the company to its knees in double-quicktime. Utopia.

  2. Who’d have thought that such things were entirely susceptible to rational analysis?

    What hope do we have of electing the right government when such eminently understandable concepts are never explained in the regular media?

  3. I have always been puzzled by this “higher risk = higher returns”. Presumably this is just for the successful enterprise. If any risk was real then there must be enterprises who lost?

    On average then, the higher returns for a successful enterprise should just match the returns for the industry as a whole.

    Risk is probably a code word for something else.

  4. How good are credit ratings as a predictor of credit worthiness, anyway? we’ve got a few years real-world data, surely someone could run the numbers.

  5. @Ivor

    The mathematics looks simple to me. If Option A offers the prospect of a 70% return on success but a 70% chance of failure, while Option B offers the prospect of a 30% return on success but only a 30% chance of failure, isn’t the summary ‘higher risk, higher return’ a reasonable one? What do you think ‘risk’ means?

  6. @Ivor

    The risk-reward concept is simple enough. Risk is an unrealised and often unspecified cost. One does not bear cost without the prospect of reward. The higher the costs, the higher the prospective reward must be to make the trade worthwhile.

    Imagine betting on a horse that has a 1/100 chance of winning a race. A person whose insight allows them to correctly assess the horses winning prospects will not wager on that horse at less than 100/1. If he/she is offered 200/1 and the bettor places a wager, the horse still has no better than a 1/100 chance of winning, but the bettor is taking a high risk high reward bet. The bettor will very probably lose but takes the chance because the reward is high. In a portfolio of bets including horses that will probably win but pay small rewards or have a moderate prospect of winning and pay moderate rewards, this makes sense. The bettor chooses to stake the bet at 1/200th the size of one running at even money and 1/10th that of one running at 20/1.

    If the bettor can cover the entire race with bets on average offering rewards longer than their true odds of winning then the portfolio will turn a profit over time. A higher-risk higher reward gambler would stake the bets more in favour of the longer odds horses — leading to a more volatile portfolio performance but one that might prove more profitable over time. Of course, one might lose money more quickly that way instead.

    Broadly, risk trading concerns who is prepared to accept risk, and those who are prepared and able to accept the risks that would otherwise burden others should charge a suitable premium for the service.

  7. Does the risk / return calculus more or less even out over many tests? To take some examples.

    Investment A returns 10% interest after 12 months but in any one year it has a 10% chance of complete default. So 9 years out of ten it returns $10 on $100 for a $90 “profit”. But the one year it defaults it leads to a $100 loss. Net result after 10 years is a $10 loss.

    Investment B returns 10% after 12 months but in any one year it has a 5% chance of complete default. So 19 years out of 20 it returns $10 on $100 for a $190 “profit”. But the one year it defaults it loses $100. So the profit is $90.

    There must be a relatively simple formula which equates return and risk. I will have stab at working it out. (I am rusty as all heck in maths and totally unschooled at financial or economic maths so this will be very crude).

    Profit = Promised Returns – Realised Losses.

    P = Investment x Interest rate – Realised Losses.

    P = (I x r) – (I x P) – ((I x r)xP) (where P is the probability of a loss).

    Let’s test it when a complete loss occurs so P is now known to be 1.

    P = ($100 x 5%) – ($100 x 1) – (($100 x 5%)x1)

    P = $5 – $100 – $5

    P = -$100 or a loss of $100. It seems to work.

    Now let’s look at an investment which reliably fails half the time over many tests.

    P = (I x r) – (I x P) – ((I x r)xP) (where P is the probability of a loss).

    (Average) P = ($100 x 5%) – ($100 x .5) – (($100 x 5%)x.5)

    (Average) P = $5 – $50 – $2.50

    (Average) P = -$47.50

    If I wasn’t too lazy to now get into simple algebra or draw a graph, I could work out the break-even point. J.Q. would do it in his head in about 200 milliseconds or less. Thus there must be a simple graph line that relates return to risk and which you can use to deduce if an investment is “above the line” or ” below the line”. This graph no doubt even has a standard name in simple accounting / economics. I know nothing of these subjects in formal terms but it’s easy enough to deduce that such a line must exist or be constructable.

    Of course, the risk (probability of a loss) is an estimate beforehand and only a known value after the event or after a series of events.

    If the market is more or less efficient (weak market efficiency hypothesis?) would it not price returns versus risks such that for many tests for many investors over a sufficiently long period the net return of safer and riskier investments along the continuum is about the same?

    I am just posing the question as layperson.

  8. @J-D

    If things look simple – that may be a warning sign. You should not confuse mathematics with the real economic behaviour.

    If an investment is successful the point is that it has received other peoples produce in return for its supply. Unless there is a drain on other markets, this can only come from the market share of the alternative (competing) producers. Assuming the demand for supply items remains the same.

    There is no success in just producing stuff – success is only in exchanging it.

    If the demand is the same, the return for one comes at a loss of market share for another.

    So a 70% success rate – gets a low return (absorbs losses from 30%)

    A 30% success rate implies a 70% failure – so returns are higher (absorb losses from 70%).

    So there is not necessarily any greater economic returns for the industry as a whole.

    Although there may be gains in utility for final consumers.

    Risk could just mean competitiveness.

  9. @Fran Barlow

    That is not a useful way of looking at it.

    Risk for one is loss for another – that is the only way the circular flow can balance – in real terms.

    Circular flow is not a racetrack.

  10. @Ivor

    No, ‘risk’ could not just mean ‘competitiveness’. This is easily demonstrated. Find a few sentences in which the word ‘risk’ is used. Now, change each of those sentences by substituting the word ‘competitiveness’ for the word ‘risk’. Do the sentences still have the same meaning? No, they don’t. Most probably they have been reduced to gibberish. This simple experiment demonstrates that ‘risk’ does not just mean ‘competitiveness’.

  11. Credit ratings have a single purpose: to assess the probability that the interest and principal on a bond will be paid in full and on time. “

    Is this true? The Commonwealth was downgraded by Moody and S&P’s in 1986 and downgraded again in 1989.

    In 1989 Commonwealth debt was 7% of GDP. I am sure the Commonwealth was capable of paying in full debt and interest.

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 )

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