Skip to main content.

Estimating limits of liability for contracts

Background

Supplier liability clauses are included in contracts to protect the customer from being exposed unnecessarily to financial loss due to a breach of contract provisions by the supplier. While a risk-averse customer may see clauses that impose unlimited liability as offering protection against a supplier breach, this comes at a cost: the supplier must take additional steps, either in the way it delivers the project or through risk-financing measures like insurance, to be in a position to fulfil its obligations, and the additional costs of doing so will be included in the contract price. Otherwise, the supplier may simply decide not to tender for the contract.

Traditionally, Australian Commonwealth and State Government agencies had a policy that required suppliers to government to assume unlimited liability in most circumstances, and they included provisions for this in their standard contract templates as the default starting point for tendering and contract negotiations.

In 2006 the Commonwealth government implemented a policy that signalled a move away from the requirement for unlimited liability. The policy proposes that a more realistic limit of liability may be derived from a risk assessment where the risks are substantially and demonstrably lower, and where unlimited liability imposes an unnecessary burden on the supplier. The rationale is that suppliers must incur additional costs if they are forced to accept unlimited liabilities, and that these costs will be passed on to the Commonwealth without generating any real benefit for either party. Information technology, communications and telecommunications (ICT) contracts are specifically targeted in the policy.

The then Department of Finance and Administration issued a guide on the application of the policy. The latest revision of the full guide is available from the Department of Industry website.

Broadleaf and law firm Ashurst prepared the guide, in a process that included extensive consultation with Commonwealth agencies, industry and other stakeholders such as the then Department of Finance and Administration. This work has given us a unique, detailed view of supplier liability in procurement organisations and agencies.

Liability models

Since the policy change, Broadleaf has developed, refined and implemented quantitative models to assist in estimating appropriate levels of liability, consistent with government policy on supplier liability in ICT contracts. They apply to a wide range of contracts and to both industry and government clients.

The process is a logical extension of more traditional risk assessments. It incorporates simple but effective simulation methods to represent sources of liability and assess liability levels.

Legal context

From a legal perspective, our approach is focused on understanding the scenarios and events where a supplier may be in breach of or negligent in its acts under a contract, leading to the client incurring damages or additional costs.

Each contract contains unique features, so the risk assessment process needs to be tested against the proposed contract and tailored to fit the specific contract requirements. Such specific requirements could include the nature of the services to be provided and the supplier’s chosen method of delivery, how product and public liability is addressed, how professional liability is dealt with, and how these liabilities interact with insurance requirements and other contractual provisions such as performance guarantees and liquidated damages.

On occasions, common law, legislation or other regulatory frameworks dictate how liability is dealt with. For example, courts determine appropriate levels of damages in instances of injury or death, so any proposed limits do not apply to these circumstances. In other instances, limits of liability may not apply to damage to property, infringement of copyright or intellectual property rights, or to the loss of confidential information. Contract terms should specify whether these areas are to be left unlimited or capped to some extent.

Modelling process

Broadleaf’s liability modelling process has been developed and refined over many assignments. It involves the following steps:

  • Conduct an assessment of risks associated with the procurement activity and their allocation under the contract
  • Identify those risks relating to supplier acts, omissions or negligence that fall within the ambit of the contract’s liability clauses, and develop associated scenarios
  • Develop quantitative estimates of the likelihood of the scenarios occurring
  • Develop quantitative estimates of the financial impacts of the scenarios on the customer, defined in terms of best case, most likely and worst case outcomes
  • Investigate the relationships between the scenarios and identify those risks that could only ever arise once, those that might arise several times during the course of the contract, and those sets of risks that are mutually exclusive or exhibit some other more complex linkage
  • Develop a scenario-based simulation model and tailor it to the specific characteristics of the contract
  • Simulate the outcome over many thousands of iterations
  • Summarise and analyse the results
  • Determine a suitable level of confidence for a limit of supplier liability.

In some cases, the contract structure may require limits for a number of liability classes, rather than one overall liability limit. Where classes of liability are required, Broadleaf’s modelling process is modified to produce a model for each class.

Liability limits per event or in aggregate may also be modelled.

Figure 1: Liability estimating process

Process for estimating contract liability limits

Model outputs

The output of the model presents a range of levels of liability the customer might require and the associated probabilities that these levels will not be exceeded. Figure 2 shows an example. This graph also shows the level of liability that would need to be set to give the customer the required level of confidence that it would not be exceeded during the course of the contract. For example, Figure 2 indicates the customer could have 90% confidence that a liability level set at $20 million would not be exceeded (or, put another way, impacts over $20 million might arise with probability less than 10%).

Figure 2: Model graphical output (indicative)

Distribution showing the level of confidence that prospective liability limits will not be exceeded

Table 1 shows some of the data underlying Figure 2 in tabular form.

Table 1: Model tabular output – summary level

Confidence level

Liability limit ($ million)

80%

$16M

90%

$20M

91%

$21M

92%

$21M

93%

$22M

94%

$23M

95%

$24M

96%

$25M

97%

$27M

98%

$34M

99%

$48M

An understanding of the limitations of the input data and of the model is necessary to assess and qualify the output, especially when dealing with the tail of the probability distribution as shown in Figure 2. The final outcome is a value indicating the level of liability that will not be exceeded during the course of the contract to the required level of confidence.

Benefits

Advantages that arise from adopting this process for estimating limits of liability include:

  • Improved insight into the potential sources of loss for the customer
  • A common understanding between customers and suppliers of the nature and magnitude of potential liabilities
  • Avoidance of costs associated with unnecessarily high limits of liability that deliver little value to the customer and do not reflect actual risks to the contract
  • Larger numbers of companies prepared to tender for work where realistic risk-based limits of liability are considered rather than unlimited liability
  • Simpler and faster negotiation of contract terms associated with supplier liability.

Explicit in this process is the need to undertake a risk assessment as the basis for the model. A key requirement and a significant benefit of this process is that customers and industry must systematically identify, assess and treat the risks involved in their procurement activities.

Developing estimates of supplier liability that are derived from detailed risk analyses is a significant shift from previous practices, where the limits may have been set arbitrarily to insurance levels or multiples of contract values without any clear understanding of the actual contract risks or justification. Limits of liability developed using quantitative analysis reflect the best understanding of the risks to the contract, their allocation and their financial consequences.

In our experience, estimates of supplier liability based on identified risks are lower than limits set as a multiple of contract value or supplier insurance levels. Lower levels of liability translate to lower insurance costs for suppliers and less financial exposure. Suppliers commonly pass the cost of liability and insurance to the customer by including it in the contract price; contracts that include liability limits derived from sound risk assessments should cost less than those with unlimited liability.

When supplier liability limits are derived from detailed risk assessments, they are generally more acceptable to all parties. This approach should lead to more successful negotiations, with less time and resources spent negotiating liability levels, often a sticking point in the past. Often, there is no negotiation necessary as the proposed supplier liability levels are immediately recognised as reasonable and justified.