Model audit

From Wikipedia, the free encyclopedia
Jump to: navigation, search

A model audit is the colloquial term for the tasks performed when conducting due diligence on a financial model, in order to eliminate spreadsheet error. Model audits are sometimes referred to as model reviews, primarily to avoid confusion with financial audit. A study in 1998 concluded that even MBA students with over 250 hours of spreadsheet development experience had a 24% chance of introducing spreadsheet error.[1] Model audits are typically requested by banking organisations, in order to reassure lenders and investors alike that the calculations and assumptions within the model are correct, and that the results produced by the model can be relied upon. When a comprehensive review of the model is required, the scope of review is often extended to include tax and accounting, sensitivity testing and the checking of data contained within the model back to the original financing and legal documentation.

Purpose[edit]

The purpose of a model audit is to provide assurance that the results can be relied upon. For this reason, the party conducting the review will provide a level of reliance on the form of an amount of liability. This may range from a multiple of the fee (2×, 3×, 4× fee, etc.) to a fixed amount, often up to US$20 million[citation needed]. In the event an error or omission is found in the model due to the model auditor's negligence, the organisation relying on the report may choose to sue the model auditor in order to recover any loss.

The objective of the model audit should be to the reduce financial risk that is being taken on by under[clarification needed] the transaction to which the financial model relates. As such, it is more important to ensure that the model audit has the proper scope, and is undertaken using a robust methodology, to identify material errors than to negotiate a liability cap if material errors are not identified.[citation needed] The model audit is not, or at least not as a primary purpose, an insurance policy, it is for reducing the financial risk that is being taken on.[citation needed]

Typical scope[edit]

For a full-scope model audit, the following elements would usually be included:

  • A review of the model's logic;
  • A review of the model's consistency with financial and contractual documentation;
  • A review of the model's consistency with relevant accounting and tax requirements;
  • A sensitivity review.

Application[edit]

Model audit has predominately been related to project finance and infrastructure finance, including Public Private Partnership ("PPP") transactions (including PFI in the UK and P3 in the USA).

Model audits are applicable to any financial model that is used to support the taking on of any financial risk, e.g. valuation models, operational models, refinancing models, portfolio model, M&A models etc.

How is it conducted?[edit]

Most financial models are produced using Microsoft Excel. The model will routinely contain sheets for input data, formulas (the 'workings') which drive the model, and outputs, which are usually in the form of financial statements (balance sheet, income statement, cash flow statement, etc.).

Model auditors may undertake a detailed 'bottom-up' review (cell-by-cell checks) of each unique formula, and/or combine a 'top-down' analysis such as the reperformance of calculations based upon the project's documentation

There is some debate as to whether a "cell-by-cell" or reperformance approach is most appropriate. A model auditor may emphasise one or the other or use a combination of both approaches.

Changed environment of the credit crunch[edit]

Banks are now more risk averse.[citation needed] Given that they do not pay for the model audit (the sponsors do), but they rely upon the model audit report, they are able to dictate the scope of the model audit. Given that the banks will now insist on a properly scoped model audit, the model audit will finally be seen as a critical financial risk management task, rather than simply have someone to sue if things go wrong.[citation needed]

Competitive dialogue for PPPs[edit]

Procurement of public-private projects is often undertaken under a competitive dialogue regime. This requires bidders to commit earlier in the process, and the implication of not identifying errors in a financial model at any early stage could be that a bidder is unable to rectify the error.

As a result, financial models are being checked earlier in the bid process, not just at financial close – these are known as pre-preferred bidder reviews. In some cases, these will have almost the same scope as a financial close model audit (but with documentation review limited as this will still be being drafted), but in other cases, the pre-preferred bidder review will be limited to an agreed scope of procedures with the objective of maximising risk mitigation whilst minimising the fee. The benefits of a pre-preferred bidder review is that it should lead to a reduced model audit fee at financial close.

The introduction of this process to North America has been controversial. The City of Brampton, for instance, has faced lawsuits[2] and controversy[3][4] about use of the process.

Cost and duration[edit]

A model audit may take between 1 and 5 weeks[citation needed], but this does not include the time taken by the model author to rectify the errors identified by the model auditor. The fee is largely dependent upon the scope of review, the number and complexity of the unique formulae in the model, the volume and complexity of the documentation and the number of versions of the model/documentation to be reviewed.

The cost will also depend on the seniority of staff undertaking the work[citation needed]. Planning of the model audit, as for a statutory audit, is vital to mitigating risk,[citation needed] and thus this needs to be undertaken by senior staff. In planning, the elements of the model audit should be allocated to staff with the appropriate level of experience, technical expertise (e.g. tax or accounting) and seniority. The senior staff needs to look at the big picture of the model to ensure that it makes sense as a whole.[citation needed]

References[edit]

  1. ^ Panko, Raymond (1997–2005). "Errors During Spreadsheet Development Experiments". Study. Shidler College of Business, Hawaii. 
  2. ^ Grewal, San (13 July 2011). "He built city hall. Now he's suing it". The Brampton Guardian. Toronto ON. Retrieved 14 July 2011. 
  3. ^ Douglas, Pam (25 March 2010). "City plans under wraps". The Brampton Guardian. Brampton ON. Retrieved 14 July 2011. 
  4. ^ Douglas, Pam (29 March 2011). "Conflict allegations raised at city". The Brampton Guardian. Brampton ON. Retrieved 14 July 2011.