VoxEU Column Global governance Macroeconomic policy

Putting an accurate price tag on government credit support

Governments run the world’s largest financial institutions. The size of government activities has grown in recent decades but comprehensive estimates are unavailable. This column presents new evidence on the costs of government credit support. It argues that governments tend to understate credit costs and the consequences of that could be considerable. Cost underreporting may lead to overinvestment and capital misallocation, could encourage the over-reliance on credit, reduce transparency, and cause a build-up of financial risk.

The BRICS officially launched the $100 billion New Development Bank along with a new reserve currency pool worth over another $100 billion on 15 July. That institution joins the ranks of multilateral development banks, which in 2012 collectively had portfolio holdings of more than a trillion euros of loans for infrastructure and other assets.

Development banks are a part of a much larger phenomenon, which is that governments run the world’s largest financial institutions. Collectively, government investment and insurance operations dwarf those of the largest commercial banks. The size and scope of activities have grown over the last several decades to include the explicit and implicit guarantees of too-big-to-fail private and international financial institutions and non-financial firms, direct and guaranteed loans, and more traditional insurance and guarantee programmes such as for bank deposits.

Comprehensive estimates are unavailable, but the IMF (IMF 2012) provides some sense of the size of governments’ credit market activities. It reports that the debt of government-related enterprises and outstanding government-guaranteed bonds totalled about 20% of GDP for many OECD countries, and more than 50% for the US. Those figures exclude the contingent guarantees and national credit programmes that would significantly increase those percentages.

To efficiently allocate capital, private investors clearly need accurate information about costs and risks. The same is true for government investors, but that information is generally much less available to them or to the public. That point is underscored by a recent examination of the financial accounting and budgetary practices of OECD countries, which reveals that their government institutions systematically and significantly understate the costs of the credit support they provide (Lucas 2014).

The understatement of cost manifests itself in a variety of ways that differ across countries and jurisdictions. However, a universal mistake is that governments take their cost of capital to be their borrowing rate, irrespective of the risk of the investment under consideration. That practice ignores the fact that taxpayers are effectively the equity holders in all risky government investments. Another widespread problem is that no upfront cost is recognised for financial guarantees in most governments’ budgets.

The extent to which governments understate credit costs has gone largely unmeasured, and hence has received little attention. However, the potentially adverse consequences of overlooking the hidden costs are considerable.

  • Cost underreporting favours overinvestment and leads to capital misallocation;
  • It encourages over-reliance on credit relative to other forms of government assistance;
  • It reduces transparency about size of subsidies and the public sector; and
  • It encourages a build-up of financial risk by governments that could hinder their future ability to respond to a crisis.

Fortunately, a remedy is available that is already widely used in the private sector. A switch to fair-value accounting in budgeting for credit would recognise the costs borne by taxpayers acting as equity holders. By recognising its full cost, it would put credit support on a more level playing field with other types of government assistance (whose costs are largely measured at market prices), and eliminate the budgetary arbitrage that occurs when governments treat their borrowing rate as their cost of capital.

The feasibility of applying fair value accounting to government credit programmes, and the quantitative difference it would make for some programmes, has been demonstrated in a number of academic studies and in reports by the US Congressional Budget Office (Lucas 2012 and references therein). Of course, there are sizeable political and practical impediments to reforming budgetary practices. Further research that sheds light on the true magnitude of costs and risks of government credit support and on how fair value accounting for those costs can be practically implemented could help reduce the resistance to change.

How should a government to think about its cost of capital?

The cost of capital for a government investment is essentially the same as it would be for a large private sector firm. That proposition follows from the basic principles of financial economics—that undiversifiable or market risk has a cost; and that the cost of capital for any investment depends on its undiversifiable risk and not on the combination of debt and equity used to finance it.  

The logical problem with treating a government’s borrowing rate as its cost of capital is most easily understood by example. Consider a risky loan made by a government and funded through the issuance of government debt. If the risky loan is repaid in full, the proceeds cover the repayment of government debt and any surplus goes to taxpayers. If the risky loan defaults, the government debt still is repaid in full, with taxpayers making up any shortfall. Hence, taxpayers serve as the equity holders in risky government loans. Likewise, taxpayers are equity holders in any risky government investment, real or financial.

In practice, governments take their cost of capital to be their borrowing rate. By doing so, they implicitly treat taxpayer equity as requiring a zero rate of return. Omitting the cost of equity switches many credit and investment projects that are money losers from an economic perspective to appearing profitable on a budgetary basis. That creates budgetary arbitrage opportunities—buying securities at market prices appears to generate profits. Of course, some investments that are unprofitable may nevertheless be socially beneficial and credit provision is a legitimate function of government; the issue at hand is accurate cost measurement.

Governments understate their cost of capital

In general, understatement of cost varies across credit support programmes but it tends to be particularly acute for financial guarantees. A few examples (drawn from Lucas 2014) illustrate the magnitudes of the distortions and the potential real consequences.

The European Bank for Reconstruction and Development (EBRD) provides loans, equity investments, and guarantees to enterprises in developing regions, financed with debt issues and capital from member countries. Member countries have committed about €23 billion in a form of guarantees known as ‘callable capital’, which the EBRD can draw on should its reserves become insufficient to cover its obligations to debt holders.

Committing callable capital is costly to member countries because if it is called there is little prospect it will ever be returned. Nevertheless, most countries record no budgetary cost when they make those promises. Using an options pricing approach that incorporates the cost of market risk associated with international banks, I find the unrecognised cost to members of callable capital over 20 years to be about €7 billion. Those costs are absent from the budgets of member countries.

The EBRD’s view of its own profitability is also distorted by treating taxpayer capital as free. Using a weighted average cost of capital approach and contrasting that to how it accounts for capital costs, its profits appear to be overstated by about €700 million a year. The underreporting of cost lessens the incentives for member countries to monitor the activities of the bank and provides an incentive for the bank’s managers to invest more aggressively.

A second example is the Tennessee Valley Authority (TVA), a fully federally-owned utility that produces about 1/6 of the electricity in the US. Its physical assets, like transmission lines and power plants, are funded through retained earnings and debt issuance. The Valley Authority’s debt carries interest rates only slightly above Treasury rates thanks to the implicit guarantee from the government. Taking into account the cost of that guarantee to taxpayers, it has in recent years received unrecognised subsidies of about $600 million annually. That hidden cost, were it to appear in Authority’s financial statements, would flip its reported profits to losses. A side effect of cost understatement is that electricity is underpriced because the Tennessee Valley Authority’s electrical rates are based on officially measured costs.

A final example is the European Financial Stability Facility, which was created in 2010 as an emergency liquidity facility in response to the Eurozone crisis. Its ability to cheaply borrow large sums from international capital markets rests on the €700 billion of capital and callable capital committed by member countries. Given the nature of the risk, calls on capital will be infrequent but potentially enormous. It is hard to precisely measure the fair value cost of the exposure, but an options pricing approach suggests that it’s in the range of €20 to 80 billion. As with the EBRD, none of that cost appears on the member states’ books under current budgeting practices.


International Monetary Fund (2012), “Fiscal Monitor 2012, Balancing Fiscal Policy Risks”.

Lucas, D (2012), “Valuation of Government Policies and Projects,” Annual Review of Financial Economics, Vol 4.

Lucas, D (2014), “Evaluating the cost of government credit support: the OECD context”, Economic Policy, 29: 553–597. doi: 10.1111/1468-0327.12034

1,575 Reads