Estimating Liabilities for Child Sexual Abuse

This article provides some personal thoughts from Geoff Atkins about how an actuary might approach liability estimation for child sexual abuse in the context of the high profile Royal Commission currently underway.

Geoff has been consulting in general insurance for over 30 years, and has been an adviser to many insurers, accident compensation schemes, and governments. Geoff’s comments represent his own personal views and do not represent the views of any other person or organisation.

The Royal Commission into Institutional Responses to Child Sexual Abuse has reached the stage where there is much debate about compensation, civil litigation and redress. The Royal Commission is not due to present its final recommendations until the end of 2017, following which all levels of government will need to consider their response.

In the meantime, insurance companies and other institutions, such as government agencies, that might face an increase in the cost of claims arising from past events need to continually prepare accounts with provisions for outstanding claims liabilities. Appointed Actuaries need to come to grips with the appropriate treatment, given their responsibilities to report on insurance liabilities and financial condition.


I start with the view that there is no value for any users of the accounts, or stakeholders in the actuary trying to anticipate possible outcomes when reporting liabilities. While this may seem counter-intuitive it is where I have landed after thinking about the question for nearly two years.

The outcome of the Royal Commission is speculative at this stage, and nobody knows what will be recommended, let alone what will eventually transpire. No defendant or insurer can currently provide useful information to a user of its accounts about the impact on profitability or solvency by booking a liability figure that is no better than speculation.

Does it make sense to report that an entity has made a loss (by increasing claim provisions) because the Royal Commission is underway? Does it make sense to say that an institution is insolvent or heading that way at the present time?


If my proposition were to be accepted, then the logical consequence for an actuarial liability valuation would be:

  • don’t change a central estimate or risk margin on account of possible changes in liabilities as a result of the Royal Commission;
  • don’t reduce currently held reserves on the basis that a redress scheme would reduce insurer liabilities;
  • if there is a ‘business as usual’ reserving basis for child sexual abuse claims, continue to review and update that basis in a consistent manner (i.e. consistent claims reporting patterns and unchanged average claim costs); and
  • describe in reporting the nature of the uncertainty, the reason the approach was adopted and comment on the relative materiality of the liabilities to the company.

A further logical consequence would be that, if material, the annual financial report of an insurer or other institution would include a note as to the uncertainty surrounding the estimate of the liability.


Each year the Appointed Actuary must prepare a Financial Condition Report (FCR) for the Board of an insurer, which is provided in turn to APRA. My suggested approach to the FCR relies on the confidentiality of the report, and the consequent ability of the actuary to make a ‘full and frank’ report. This is in contrast to the insurer’s annual financial report and APRA returns.

The approach that I would think about for the FCR would be roughly as follows.

  1. Discussion of the nature and extent of potential exposures (including if there is believed to be none).
  2. Discussion of the extent to which the insurer has good knowledge of its potential exposures (e.g. insureds and policy details) or whether potential exposures exist that cannot be specifically identified.
  3. Include in the stress tests one with a plausible adverse outcome in respect of child sexual abuse claims, with perhaps
    a reverse stress test if the potential exposure is large enough for the company to be technically insolvent.
  4. In reviewing the Risk Management Framework, include commentary on how the risk management process has been dealing with the potential for changed insurance outcomes following the Royal Commission.


The two boxes opposite give a brief outline of how my suggested approach might sit with relevant standards. For an insurer these are accounting (AASB1023), APRA (GPS320) and Actuarial (PS300). For non-insurers the relevant standard is AASB137.

These comments are superficial and are not based on any authoritative source. Each insurer or institution will need to consider accounting and actuarial advice specific to their own circumstances.


While the approach described above makes sense to me now, it will not always be thus. Circumstances will clearly change in future, but in a way and in a timeframe that is not known. For this reason the approach must always be reconsidered each time liabilities are reviewed.

Each actuary (and each insurer or affected institution) will need to form his or her own views on the appropriate steps to take.



The claims liability is described as the discounted central estimate (a probability-weighted expected cost) plus a risk margin to allow for the inherent uncertainty in the central estimate. All the discussion in the standard implies that claims outcomes are some kind of tractable statistical process, albeit with external influences.

We might expect the risk margin to be more relevant in this context, but again the discussion is in statistical terms. There are no examples or mentions of ‘one-off external events’ of a contingent nature. Even if there were, an increase in risk margin has exactly the same impact on the accounts as an increase in central estimate and so is no more helpful to users of the accounts.


The 12 pages in Attachment A dealing with the insurance liability valuation have no mention of a contingent one-off external event and, like the accounting standards, imply a quasi-statistical process with past experience leading into future projections, allowing for known changes. The description of uncertainty suggested in relation to the FCR responds directly to the requirement in Attachment A para 11(a) of GPS320.

Actuaries Institute PS300

Sections 9 and 10, dealing with methods and assumptions, do not mention a contingent external event, but would not restrict an actuary from allowing for one. Section 12 (uncertainty) is applicable to this situation, but 12.1.4 (sensitivity or scenario or statistics) would, in my view, be better reserved for the FCR. Section 12.2 (sensitivity) refers to ‘a reasonable variation to key assumptions’, which would not be easily applicable to this situation.

A statement about the treatment of the potential for a liability would meet the requirements of clause 6(a) on materiality.



The standard AASB137 Provisions, Contingent Liabilities and Contingent Assets has the following relevant definition.

A contingent liability is:

  1. a possible obligation that arises from past events and whose existence will be confirmed only by the occurrence or non-occurrence of one or more uncertain future events not wholly within the control of the entity; or
  2. a present obligation that arises from past events but is not recognised because:
  1. it is not probable that an outflow of resources embodying economic benefits will be required to settle the obligation; or
  2. the amount of the obligation cannot be measured with sufficient reliability.

A Provision (a liability of uncertain amount or timing) is to be recognised when (paraphrasing):

  • A legal or constructive obligation exists, and
  • It is probable that payment will be required, and
  • A reliable estimate of the amount can be made.

Otherwise the amount is a contingent liability and would be disclosed in a note as per paragraph 86.

Paragraphs 25 and 26 discuss whether or not a reliable estimate can be made, without providing a set of criteria that would give a clear answer in the present situation. However there is an example in paragraph 50 that may help:“The effect of possible new legislation is taken into consideration in measuring an existing obligation when sufficient objective evidence exists that the legislation is virtually certain to be enacted.” This may be a parallel to the current situation.

Immediate support for anyone affected is available at:



CPD: Actuaries Institute Members can claim two CPD points for every hour of reading articles on Actuaries Digital.