Recent Posts

Wednesday, February 4, 2015

What Are Government Promises Worth (Part 3) - Sovereign Debt

This is the final instalment of a series of articles regarding government promises. (The first part discussed gold and currency pegs, and the second inflation and nominal GDP targets.) I will now turn to the promise by governments to repay their bonds. This promise is inherently more plausible than the earlier types of promises I discussed, at least for some central governments. As the ongoing Greek government debt negotiations show, not all types of government debt are equal.

Recap: Flexibility Rules

Thus an army without flexibility never wins a battle.
A tree that is unbending is easily broken. 
The hard and strong will fall.
The soft and weak will overcome.
Tao Te Ching (Lao Tzu) - chapter 76 

In order for a government promise to be credible, it has to be flexible. As in the Chinese proverb, a tree has to bend in the wind. Or to use a North American football analogy, you often want a defence that will “bend, but don’t break”.

Obviously a flexible promise offers less. But giving a promise that cannot be fulfilled – such as attempting to keep a foreign currency or gold parity fixed – will eventually trigger a crisis.

But what about government debt? Debt obligations appear inflexible. This appearance is misleading for governments that have control over their currency.

The Promises Backing Government Debt

For a currency sovereign, the promise to repay debt actually consists of two promises.
  1. Willingness to pay. The government reassures bond holders that it will not repudiate its debts, and somehow create an institutional bias towards repayment that binds successor governments.
  2. Capacity to pay. The government will arrange its affairs so that there will be no effective financial constraint against repayment.

Willingness To Pay: Not The Key

The first promise – the willingness to pay – is the same for any issuer. Other than extremely dysfunctional subprime lending strategies, every borrower promises to repay debts. This seems to put governments on the same footing as other borrowers.

To be fair, there is a very strong political constituency in favour of government debt repayment. Banks, pension funds, and insurance companies would have their liquidity portfolios mauled if a government defaults. The loss of liquidity would likely cause a chain reaction of defaults in the private sector, including businesses outside of finance. And of course, a default would hurt retail investors, such as conservative middle class seniors (who have a very high turnout at elections).

The constituency behind policies like currency pegs or gold parities is much weaker. Only a small group take (direct) losses. And was the case when Nixon closed the Gold Window, foreigners often take the losses (who conveniently are supposed to have a 0% turnout at elections).

Nonetheless, I would not put too much faith on the “willingness to pay” promise. Even though there is normally a strong constituency to repay government debt, the national interest shifts over time. Meanwhile, all sensible central governments control the legal apparatus for their local currency borrowing. Therefore, it is much easier for them to restructure their obligations than it is for private borrowers without “defaulting”. (For example, take the “Gold Clause” in the United States.)

Capacity To Pay

Credibility mainly comes from the ease of debt repayment for central governments.

Since different governments have different debt operational procedures, I will first discuss a simple idealised case. Imagine a government which owns the central bank, and it does all of its borrowing in the form of an overdraft with it. As the Treasury spends money, it transfers deposits to the banking system’s accounts at the central bank. The only government liabilities are those reserves, which are a form of money. (I will follow common usage and call these deposits “reserves”, even though that is an incorrect description in places like Canada. Additionally, I will refer to the governmental spending authority as “the Treasury”, although it may be called the “Ministry of Finance” in some countries.)

Furthermore, if the government wishes to create a yield curve, the central bank could then issue bonds to drain reserves. We can view reserves as being “money”, and the central bank bonds act as “forward money”. (This is described in my primer, “What is a Government Bond?”. This observation is a defining characteristic of Modern Monetary Theory (MMT).)

It is obvious that this government cannot default on “money”. The same applies to “forward money”; all that happens is that at a certain date, the “forward money” is converted into (spot) money, which cannot be defaulted upon.

My argument is that government finances are functionally equivalent to this model, even though the details are different. One could raise many detailed reasons from statutes why real world government borrowing departs from the ideal described above. That said, we need a strict framework like a mathematical model to understand the economy; we cannot just rip random institutional factors out of context and hope to understand what their behavioural effects are.

If we create a model of government finances, we will see that the accounting identities describing government finance for the idealised system described above is essentially identical to that for real world central governments (in places like the United States, Canada, the United Kingdom and Japan). Since the accounting frameworks are the same, the model dynamics end up the same.

The justification for that assertion follows from two arguments.
  1. Central banks fix the short-term rate of interest, which is the same thing as the interest rate the government borrows at. (Long-term rates are determined by the expected path of short rates.) The borrowing rate for the Treasury is the same as that faced by the central bank in my idealised model.
  2. The government owns the central bank. If yields in the Treasury Bill market rise above its desired short-term rate, the central bank has unlimited capacity to hoover up those bills, funnelling the profits to the Treasury. This prevents any rational belief in default from developing amongst private sector entities. As a result, private sector borrowers have no need to distinguish between central bank bonds versus Treasury bonds.
In practice, there are some limitations on government finance. For example, the Canadian Federal Government would not be able to spend 10% of GDP in one afternoon to cover losing bets on sports events. ("He called a pass?”) However, these constraints are not visible under sensible fiscal policies. In any event, wild gyrations by the government in its money management would destabilise the private sector, regardless of the institutional framework. As long as it is somewhat careful, the government can run quite large deficits without incident.

Differences between real world economies and my idealised model revolve around the following two possibilities.
  1. The belief that a central bank has an ability to force the Treasury to default. Since such a move is political suicide for the ruling elite, I do not see such threats as being credible.
  2. There are arbitrary accounting constraints put into place by the Treasury that can trigger a default. The “debt limit” in the United States is the key example. Obviously, these forms of constraints exist, but they do not represent real constraints that limit action. For example, there is no resource cost associated with raising the debt limit.
Although the conversion from “forward money” to money follows rigid rules, this does not matter, as the transition does not require the use of real resources. (By way of comparison, if “money” was gold, the government would need gold to convert bonds into “money”, and then there would be a real cost of conversion.) The flexibility of the overall system comes from the fact that there are no rigid promise about the value of money relative to real goods and services (as discussed in my article on inflation targets).

Sub-sovereigns and countries that borrow in foreign currencies (or the euro) do not have a tight linkage between their debt and money. In which case, their debt has to be analysed as if it were issued by a private sector borrower.

Circling back to the issue of promises, what matters is the implicit promise is that the government will not create rigidities within the financial system that cause their bonds to lose their status as “forward money”.

An Alternative View: Balance Sheet Approach

Balance sheet considerations point in the same direction. The government debt that matters is the debt that is held by other entities. Government bonds are a core holding throughout private portfolios, and they are liquid.

The implication is that in February 2015, government debt is being held because it meets the portfolio needs of investors. Barring some massive regime change in policy, there is no reason to expect that investor preferences will be completely different in March 2015.

In other words, if the debt can be placed at time t, it will probably find takers at time t+1, on terms that are “reasonably close” to that of the previous time period. (Investors who were long Treasurys in the summer of 2013 may disagree with that assessment.) Although market values will fluctuate, investors’ strategies are diverse enough to limit movements in price.

Once again, the key is to avoid policies (such as currency or bond yield pegs) that can create patterns of asset holdings that are vulnerable to massive portfolio weighting shifts (typically when the policy is about to abandoned).

Concluding Remarks

The repayment of debt by a sovereign has no inherent costs, and so there is no need to renege on the promise of repayment. The key is to avoid rigidities that will force a rupture within the financial system; under such an environment, default may be preferable to alternatives. The tendency of mainstream macroeconomics to embrace policy rules increases rigidity and risks within government finance. The ongoing implosion of the euro area will hopefully reinforce the tendency of politicians to ignore this advice, and retain their flexibility.

See Also:

(c) Brian Romanchuk 2015

1 comment:

  1. Your writing frequently inspires me and this post is another example. However, this time I think the detail in the post creates complexity that obscures the goals of the post. My reaction was that a much easier description of the creation of fiat money and debt was needed.

    I undertook to create a less complex description. The result is found at

    in a post entitled "A Simple Model for Government Borrowing".

    I think a good model must be simple but accurate. If accurate, a simple model can be enhanced with complexity until it describes the real macroeconomic world.

    The model I describe is a simple base. The model you describe can be built from this simple base by adding shortcuts (such as using 'reserves') and adding additional handlers and rules.

    Thanks for your post.


Note: Posts are manually moderated, with a varying delay. Some disappear.

The comment section here is largely dead. My Substack or Twitter are better places to have a conversation.

Given that this is largely a backup way to reach me, I am going to reject posts that annoy me. Please post lengthy essays elsewhere.