Recent weeks have inspired new fears of the 'covenant-lite' finance extensively present in highly geared private equity deals. Lenders willingly forgo conventional restraints in terms of asset cover, earnings cover, capital expenditure, financing restrictions and dividends. Fears of a credit bubble are crystallising Anthony Bolton's valedictory words in May as he left active fund management with Fidelity. For Bolton, covenant-lite can mean no covenant at all. He identifies current behaviour in both New York and London with the latter stages of a bull run. His dire warning is that lenders' indulgence has a counterpart in the easy lending behind the spate of mergers and acquisitions in the late 1980s. Ineffectual covenants and paucity of covenants deprive lenders of the ability to monitor and act on danger signs.

The opposite position to Bolton's is that covenants are obsolete, especially for established companies with predictable earnings cycles. Covenants are gratuitous constraints that wreak economic and social harm. They are straitjackets whose protectiveness undermines the very safeguards they are supposed to provide.

At their most perverse, covenants provide no protection at all while threatening corporate room for manoeuvre. At this point, heed Fawthrop and Donaldson. Who are they? Roger Fawthrop, a founding father of serious financial study in the UK, used to tell his students, including your correspondent, to open up Gordon Donaldson at least every six months. Fawthrop's purpose was to inspire due disrespect for conventional accounting. He was referring to two classic texts Corporate Debt Capacity (Harvard University Press, 1961) and Strategy for Financial Mobility (Harvard University Press, 1969). The former has been reprinted many times and both get mentions in the latest manuals and textbooks.

The merit of Corporate Debt Capacity is to pinpoint the shortcomings and dangers of conventional ratios based on industry norms, arbitrary internal rules, asset cover and income cover. He substitutes a formulation of debt capacity based on a range of estimates of the deployable resources that would be available at the nadir of a foreseeable period of adversity. Within the range, management accept an estimate of the level and duration of adversity which matches their propensity to risk. Cashflows leading to the nadir have to be estimated and Donaldson shows how to distinguish mandatory outlays from the discretionary outlays that management would wish to maintain even under distress.

These might include exercise of real options, continuity of employment and dividend levels. A skilfully designed covenant is required to reflect such realistically conceived debt capacity. The result would be more meaningful than conventional covenants, which constrain but which essentially emphasise default and failure of which there is only a small probability in the large majority of cases.

But covenant-lite lending may be better even than realistic covenanting based on Donaldson. This could apply to some companies whose value is as much founded on their rich real options as on their assets in place. Covenant-liteness enhances mobility by preserving options to delay, expand, abandon and redeploy resources, even to determine a loan's repayment schedule.

Traditional capital investment appraisal fails to account for the real options which crucially attach to projects. A project with a negative net present value may have a positive net present value in the future. The option to delay the project has value. A company which adopts the project too early to comply with a standard set by a covenant, eg, in terms of liquidity, forgoes the value of an option to delay.

A modest investment undertaken to gain a competitive toehold may contain a valuable option to expand. Such decisions often have to be taken at the shortest notice, and value would simply be destroyed if the decision had to be delayed for the sake of compliance with this or that covenanted ratio. Research and development is a call option that can be exercised or allowed to lapse and an onerous covenant should not be allowed to stifle it.

The opportunity to abandon is a put option with value. A rigid covenant could force a company to prefer a project with fewer abandonment opportunities but the right accounting numbers over a project with more escape routes but the wrong ratios. The irony here would be that such a covenant - albeit designed to enhance safety - would impose a riskier course of action.

How do you measure the financial mobility which heavy covenants impair? Enter here Donaldson's strategy for financial mobility. His formulation of financial room for manoeuvre posits a novel balance sheet that includes not only resources in place but also resources that are available and perils that are contingent. In practical terms, his 'inventory of resources of mobility' tells the ease (or difficulty) with which a company can respond to financial emergencies and opportunities. Donaldson ranks resources in descending order of mobility and includes contractual, contingent and other accessible funds and resources that can be liberated by changing operations.

In summary:

Uncommitted reserves

Instant (eg, unused credit line)

Negotiable (eg, debt issue)

Reduction of planned outflows

Volume-related (eg, change in production schedule)

Scale-related (eg, administrative overheads)

Value-related (eg, dividend payments)

Liquidation of assets

Land, plant and machinery

Business units

To this should be added the value of the risk capital that can be generated by insurance and other risk packaging. James Mackintosh (Financial Times, 5 June) describes the opportunities for selling loans and holding them with the protection of credit derivatives. Loans can be traded across habitats including hedge funds, the junk bond market and structured credit investors. The extreme argument becomes that, in a mature market where risk can be so packaged and traded, covenants are near-trivial. Companies are able to control exposure without the costly formalities and other transaction costs of dealing with maybe a multiplicity of lenders. Covenants may burden both parties through destabilisation and the costs of dismantling and revision.

Such an extreme position against covenants insufficiently allows for the reality of the agency costs that result from the potential conflict between shareholders and lenders. As soon as risky debt is created, shareholders have an incentive to transfer value from lenders to themselves. They can do this by incurring undue operational risk in the form of highly risky projects, by forgoing some projects even though they have a positive net present value, or by paying themselves excessive dividends. For example, if shareholders' value has already withered away through losses, the failure of a highly risky project leaves shareholders no worse off but impairs the value of the lenders' stake. On the other hand if the risky project succeeds, value is garnered for the shareholders but may leave the lenders with the same unimpaired claim they had beforehand.

Lenders recognise this incentive to transfer risk. They assume the worst about the shareholders and seek to compensate themselves by imposing costs on them. Covenants reassure lenders by disciplining shareholders, resulting in a lower imposition of agency costs. But we have seen how those same covenants can destructively restrict mobility.

The end has to be a quest rather than a solution. A naked loan bereft of covenants forces a company to bear the full burden of agency costs. The challenge for theoreticians and practitioners in accounting, finance and the law is to design covenants that do not impose futile or counterproductive restrictions but discipline managers in meaningful ways. Rather than standards based on asset cover, income cover and growth, the key criterion is preservation of financial mobility in the face of distress and opportunity. This recognises the mature enterprise in a modern financial system as a continuity for which default and failure are an insignificant likelihood.

Roger Lister is a chartered accountant and a visiting professor at Salford University.