In this second longer than usual post friend of thebriberyact.com, David Lawler, gives us a forensic accountants eye view on the state of play in relation to penalties for corruption violations.

David is a forensic accountant and partner at Forensic Risk Alliance, a forensic accounting and e-discovery consultancy with offices in UK, US, France and Switzerland. He has been involved with many FCPA and Bribery Act settlements, audits, and compliance assessments. His book, Frequently Asked Questions in Anti-Bribery and Corruption has just been published by John Wiley. He can be contacted on dlawler@forensicrisk.com.

As we saw in Part 1 of this article critical in any settlement or case taken all the way to trial, is the assessment of the gain/benefit/profit made on the contract(s) alleged to have been tainted by bribery. But before this figure can be communicated to the prosecutor, even on a preliminary basis, the defendant and its legal team must fully understand the financial effects that the bribe had on the cashflows and profitability of the defendant.

Penalties in the US depend on ‘gain’. This does differ from an accounting measure of ‘profit’. Again, there is little settled law, but generally the starting point will be :

Gain = Incremental revenues gained from the contract – all direct costs

Although both US agencies rely on gain, there is no need for the DOJ and SEC to agree on the amount of the gain, although in practice they usually do.

Here in the UK, whilst Confiscation Orders under Part 2 of POCA are generally based on revenues, and can bite on any and all property of the defendant, Part 5 proceedings are concerned only with property which has been obtained through conduct which is unlawful under the criminal law. Revenue subsequently and legitimately spent with innocent third parties does not form part of the amount to be disgorged, and this in effect means that costs can be deducted from revenue. There is little settled precedent for how Civil Recovery Orders under Part 5 of POCA for major bribery cases will be calculated, but experience to date suggests that the approach will be similar to the US, with the starting point being equivalent to ‘gain’.

Incremental revenue

The first thing to consider is the period on which the relevant law bites. In the UK this can go back as far as 20 years, although the date on which the OECD Anti-Bribery Convention became effective in the relevant jurisdiction is sometimes taken into consideration, with some prosecutors accepting that revenue earned before this date was not subject to disgorgement.

A US ‘gain’ calculation requires an estimate of what effect the bribe has in securing any new business or project, and therefore what the revenues would have been ‘but for’ the bribe. Deducting this hypothetical ‘but for’ revenue from the actual revenue gives the incremental revenue.[1] There must be a direct link between the illegal activity and the amount to be disgorged. This means that disgorgement calculations need to differentiate between illegally obtained gains, which are potentially subject to disgorgement, and legally obtained gains, which are not. So, in the case of a bribery scheme, the disgorgement potentially bites only on the actual gains that were generated from contracts won by virtue of the violations subject to the charges. Contracts untainted by violations of the law (or other alleged bribery schemes not subject to the charges) can properly be excluded from the gain calculations.

In the right circumstances it can be argued that in the case of a world-class company found to have bribed (which has tended to be the case in recent prosecutions), the company would have had a chance of getting some revenue even if no bribe had been paid. This implies that the incremental revenue is less than the full revenue earned on the project. Take, for example, a project worth £20m. A company assesses that it would have a 30% chance of winning it in any event.[2] However it pays a bribe, in order to make this a certainty. An economic analysis of expected values would look at the hypothetical revenue ‘but for’ the bribe as being £6m (30% x £20m). The actual revenue was £20m, leading to an incremental revenue of £14m, instead of the ‘full’ £20m. Similar arguments are possible when companies are operating at full capacity. If a company pays a bribe in order to win a project in one location, this is done at the expense of a project in another location: either way the company still potentially remains at full capacity. Therefore in such a case, the incremental revenue might again be far lower than the full revenue on the project.

In cases where a bribery scheme was not as successful as anticipated, the Prosecutor may consider ‘gain’ to be the benefit that the defendant sought to obtain by making the corrupt payment, and not the actual amount realised. Or, looking at it another way, a defendant will not be allowed to benefit simply because their plan failed to work as anticipated.

What costs can be deducted when calculating gain?

Expenditure that would have been avoided if the project had not been performed should be deducted from the incremental revenue when assessing profitability. So the next big question is what, if any, costs can properly be deducted from revenue when assessing gain or benefit?

Cost Categories and Why They Matter – Allowable Deductions

Again, Courts in many jurisdictions disagree over the question of which costs (if any) are properly deductible. In any event, when approaching a gain or benefit calculation, it is vital to understand the cost base of the company and how its costs are related to the revenues generated from the alleged bribes.

Direct costs can usually, at least in the US, be deducted from revenue to arrive at gain. Direct costs are the costs of raw materials and other resources that are closely and necessarily associated with generating the relevant revenue. Most of these costs are variable in nature, in that they vary with the level of sales activity. For example, the costs of buying the steel to manufacture an oil facility are direct and variable: these costs are avoided if the contract does not proceed.

Businesses also have various kinds of indirect costs, including the cost of utilities, property, and the wages for administrative and back-office staff. Pure overheads cannot generally be deducted if they would have been incurred by the company irrespective of whether the project was performed or not. Overheads also include non-cash costs, including the depreciation of assets. These costs are merely accounting entries which slowly charge the acquisition cost of assets against profit over the life of the asset. They do not represent the transfer of cash to anyone, and again are generally disallowable.

One term that is often seen in accounting circles is Earnings Before Interest, Tax, Depreciation and Amortisation, or EBITDA. EBITDA is a good starting point for a calculation of gain. It is greater than net profit, and can be derived by adding back interest, financing charges, tax and deprecation to net profit given in the profit and loss account. As it does not deduct non-cash items – such as depreciation – in its estimate of earnings, it provides a better estimate than net profit of the amount of cash generated from a particular activity.

Indirect and Non-Allocated Costs

The Prosecutors will often take EBITDA and then look to add back any indirect costs already deducted from revenue, thus increasing gain even further. Defendants will look to resist this.

A defendant is unlikely ever to win persuade a prosecutor that ‘overheads’ should be deducted from revenue when calculating gain. There is, however, scope for examining closely exactly where the boundary lies between direct and indirect/variable and fixed costs. The way we classify these costs is not mere semantics. ‘Indirect costs’ – both fixed and variable – are often a large share of a firm’s total cost base and are typically many times larger than a firm’s net profit. If it can be successfully argued that some of these costs are, in reality, directly related to contracts, then this can reduce the disgorgement substantially (and if a multiplier effect is taken into account in the fine calculation, could reduce the fine even more). Defendants will also argue (again assisted by their forensic accountants) that although overheads and so-called ‘fixed costs’ might be fixed in the short term, the majority are actually variable in the medium to longer term: property costs can be reduced if the level of business drops, because the company would be able to move to new, cheaper premises, or extra space can be leased if business increases. The longer the time horizon, the more costs are able to be varied, and in the real world, they vary in a stepped manner. Using these sorts of arguments, defendants can try to maximise allowable cost deductions, and thus reduce gain.

Causal Links

A further difficulty arises in the real world when estimating gain, as a company normally has several business lines or undertakes several contracts, only some of which are tainted by bribery. Let’s take the example of a US oil services company active in the US, Brazil, Nigeria, and Iran. The contract in Iran has been found to have been obtained by bribery, and it is therefore necessary to estimate the profitability of just this contract for settlement talks with the SFO.

Certain costs can be directly associated with the Iranian operations. The wage costs of employees involved on the Iran contract and the cost of materials used on the project, for example, would not have been incurred if the contract had not been undertaken, and it is entirely sensible to deduct these. There is likely to be an Iranian office, and the question of whether these costs are direct, variable or fixed overheads is fact-specific. In practice, some are likely to be allowed. More difficult, though, is the question of the general costs of the organisation which are not associated uniquely with any one contract or country. Take, for example, the costs of the Human Resources and IT departments, which are centralised resources based in the US head office, but provide services to all the employees and projects. How do we treat these costs – are they fixed overheads, or are there some allowable deductions here? Of further interest is the cost of finance: although traditionally viewed as a cost that cannot be deducted from revenue to arrive at gain, there is some limited precedent in the Courts for prosecutors to allow an element of direct finance costs to be deducted.

Some other lessons from the US // factors to bear in mind

Bribes Must Be Added Back

Any bribes paid are not allowable deductions for the purposes of gain calculations. To the extent that the alleged bribes were paid by the party against whom fines or disgorgement is being sought, then they need to be added back to the measure of accounting profit, and not serve as a cost deduction. If the bribes were paid by a third party and have not been deducted in the accounts of the settling party, then no adjustment is necessary.

Netting Gains against Losses

In securities actions, there is usually netting of gains made on certain share transactions against losses on others. Depending on the timing and nature of these losses, there may be no net gain because the losses entirely wipe out any gain. Netting of losses is seldom allowed in bribery claims. However, there can be situations where a ‘net gain’ calculation is appropriate. Take, for example, the following business unit, which undertook two contracts, both of which were obtained by the same bribery scheme:

When assessing the gain, can the losses of £3m on Contract B be deducted from the gain of £5m on Contract A? Is the amount of the gain £5m, or £2m?

US prosecutors generally start from the position that netting losses against gains is not permitted, and the onus is always going to be on the defendant to demonstrate why it is appropriate. Where the contracts themselves are separate, discrete profit centres, the prosecutor’s position is going to be difficult to shift, because it could properly limit its prosecution to simply Contract A, and accordingly the court would only consider the factual matrix and financial position of Contract A when making its award. In cases like this, the prosecutor is unlikely to accept the netting of A’s profits against B’s losses. If, however, the prosecutor could not without difficulty limit its case to one contract, and would have to prosecute instead the activities of a business unit or entire company, each comprising the consolidated results of several, if not many, contracts, then netting might be possible. This would be more likely if the results from each contract were not generally accounted for or managed separately, but instead their revenue and costs were always combined into the results of a larger unit. A complementary way of looking at this is to argue that if the loss-making Contract B had to be undertaken as a consequence of entering into Contract A, the losses on Contract B were inevitable and necessary to gain the benefit of the gains on Contract A. In a case such as this, the results from the two contracts could not properly be split.

It must be borne in mind, though, that this is a potentially risky position for a defendant to take, because it also combines the gross revenues. If the courts do not in fact accept that the full amount of the operating costs here is an allowable deduction and wants to calculate disgorgement based simply on revenue, then the defendant could be looking a far higher gain figure – here £16m as the starting point (which, incidentally, as we have seen would be the starting point in a UK confiscation order.)

Prejudgement Interest

Disgorgement in respect of an FCPA offence is generally taken to be “the amount with interest by which the defendant profited from his wrongdoing”. Prejudgement interest is intended to capture the time value of money. It is the notional interest that the defendant could have earned by investing the funds to be disgorged between the date they were received and the date of judgment or settlement. For example, if a bribe won a contract which gave rise to a gain of $10m in 2006, then a settlement in 2011 would include interest on the $10m for five years. Prejudgment interest can add very significant sums to the total amount to be disgorged, but is discretionary, and Courts have routinely awarded prejudgment interest in SEC enforcement actions where the defendant’s scheme evidences a high degree of bad intent. Conversely, they have also declined to award pre-judgment interest on the disgorgement amount when there has been extensive cooperation on the part of the defendant.[3]

Rough Justice

The burden is on the defendant to show that the Prosecutors’ calculations are unreasonable, and to rebut its disgorgement calculations, however the initial calculations will come from the company, so it is vital that companies control this process.

Inevitably, a certain amount of reverse-engineering goes on in major settlements, when the parties come up with a final settlement that ‘feels right’ and work back to the ‘gain’. Defendants must be alive to this.

Assessing the profitability of a contract or business unit for the purposes of a fine discussion is not a straightforward process, and to corporate accountants ‘profitability’ seldom equates to the allowable gain or benefit. All defendants who are asked to provide prosecutors with estimates of the profitability of contracts or business units should ensure that they talk to experienced forensic accountants first: it is difficult to backtrack when you want to add lots of genuinely allowable costs to hastily given estimates.

[1] http://scr.bi/nPCVEZ http://www.scribd.com/doc/37367261/World-Bribery-and-Corruption-Compliance-Forum-Opening-Remarks-of-Professor-Mike-Koehler

[2] This can be estimated by looking at the historical likelihood of winning contracts on tenders uninfluenced by bribery.

[3] The US Court of Appeal in SEC v First Jersey Securities confirmed the disgorgement of $22.2m in unlawful gains, plus $52.6m in prejudgment interest. http://bit.ly/pE7Cxz http://www.columbia.edu/~hcs14/FIRSTJERSEY.htm