When an encyclopaedic history of economics is written, it will be said of rating agencies: "Never in the field of economics have so few acquired so much authority with so little an application of logic."

The recent upgrade of Indian sovereign risk has led to much euphoria, but Moody's reasoning is as obscure and irrelevant as it has been in the past. The flaw in the rating agency system is that the agencies are not equipped to judge sovereign risk. The judgement and analysis is therefore conducted arbitrarily without the application of appropriate concepts.

The argument is best illustrated by considering their normal approach to rating commercial or municipal paper. An investor buying a bond issued by a corporation or municipal government is interested in the risk that he faces in the purchase of the bond.

A credit agency assesses that risk and passes judgement by rating the bond based on the quality of the risk. However it is interesting that whereas rating agencies pass judgement on sovereign international bonds, they have no rating system for US Treasury bonds, bills and notes.

"The assumption is that they're absolutely solid since they're obligations of the federal government backed by its full faith and credit. This means the government has authority to raise taxes to pay of its debts." (Security Industry Association).

But if that assessment is valid for the US government, it is of equal validity to the domestic debt of any other government. Just as it is totally implausible for the US government to repudiate its own debt because it has the authority to raise taxes, so also does the government of India or the government of Japan have that same authority to raise taxes. Yes but both these countries have suffered downgrades in the past for their excessive domestic debt.

Even on this occasion, although Moody's acknowledged that most of the Indian government debt was financed from the domestic banking industry, the analysts showed concern. But concern about what?  If the purpose is to determine whether an investment is safe, then surely the same rules apply as they do to US government paper.

The Indian government has put its full faith and credit behind its bills and bonds as the US has done so why would the consequences be any different? It seems the concern is quite different. Now the apprehension that loans from domestic banks to the government would tend to crowd out financial resources for the private sector. Yet the logic of that argument is unclear.

The requirements of the budgetary deficit arise from payments of interest for past loans not as a consequence of current expenditure exceeding current revenue substantially. The primary deficit of the budget deficit has been relatively small, so that the demand on real resources has hardly been sufficient to compete with the private sector.

In so far as additional demand on financial resources is concerned, there may well have been crowding in rather than crowding out. For the fact is that the government has been transferring substantial sums to the private sector in payment of interest on its past loans.

Commercial banks have acquired resources from government borrowings, which may not have been available from private loans.

There has therefore been little crowding out as a consequence of government demand for financial resources. On the contrary, commercial banks have chosen voluntarily and out of choice have exceeded their investment in government securitiers over their statutory obligations.

It is therefore very far from clear that Moody's expression of concern at India's growing public sector debt has any relevance to crowding out. But that apart how is it the duty of a rating agency to consider economic consequences of the government's borrowing policy?

It is claimed by Moody's that heavy public sector borrowing "limits the prospect for much faster growth. These considerations constrain all of India's ratings in spite of the country's improved external liquidity." (Moody's Press Release, London 3rd February).

It may indeed limit the speed of Indian growth but that in no way effects the credit rating of  sovereign paper, any more than the US government policy on Iraq effects the quality of its Treasury Bills. Once they are issued it must be assumed that they will be paid.

The main lacuna in the rating agencies argument is that they have not distinguished between sovereign debt financed domestically and sovereign debt financed internationally.

Recently John Cochrane of the Chicago School of Business presented a brilliant paper at the NCAER conference where he came to the startling conclusion that "nominal government debt including the monetary base is just like corporate equity.

"Real indexed or foreign debt is just like corporate debt."(Italics  Cochrane's) The distinction made by Cochrane is that a failure in repayment of a foreign debt is the corporate equivalent of bankruptcy, while dealing with difficulties in sovereign repayment of domestic debt will have consequences on the pricing of domestic debt.

Each type of debt relies upon the future earnings of the company, or future surpluses of the government but the consequence of failing to earn enough is primarily on the valuation of equity in one case and bankruptcy in the case of corporate debt.

It is this subtle but important distinction that credit agencies have failed to analyse. And it has led them to enormous confusion and anomalies.

Last year both S&P and Moody's downgraded Japan's sovereign status so that it was just above that of Chile, Hungary and Estonia. This curious result emerged because the rating agencies made no distinction between the government's debt held by domestic holders and foreign debt of the poorer countries.

The simple fact that the Japanese had a debt burden of 140 per cent of its GDP led to the conclusion that its finance was in a worse state than that of Estonia.

But the logic of that argument, apart from it being infuriating to the Japanese, is plain silly. 95 per cent of the Japanese debt is held by domestic savers; there is no prospect that the Japanese government will go bankrupt even if it increases its debt.

The cost of the debt might have to go up, the authorities might need to print a few more notes but because of its domestic debt Japan is not likely to suffer the consequence that Estonia might on failing to pay its foreign debt.

But the basic criterion that the rating agencies have developed on sovereign debt would gain considerably if they applied it to US government paper. A rating of 'AAA' for a US bond may be ideal for a US pension fund investor but may be very unsuitable for an Indian investment fund with Indian liabilities, which may require an equivalent investment payable in Indian rupees.

An asset rated 'AAA' for a dollar holder with dollar liabilities is not necessarily 'AAA' for a rupee holder with rupee liabilities. It is this difference in perception that has been masked in the present sovereign risk rating system. When this is properly exposed, the rating agencies might wisely abandon the concept of sovereign rating.