The SFO’s operational handbook chapter on DPAs – What does it tell us?

In October, the SFO made public a chapter from its internal “Operational Handbook” regarding various aspects of the Deferred Prosecution Agreement process. The document (which has been published in the interests of transparency), outlines the process and requirements applicable at each stage, broken down as follows:

• Evidential and public interest tests
• DPA negotiation process
• Parallel investigations
• Invitation to enter DPA negotiations
• Terms of negotiations
• DPA disclosure
• Statement of Facts and Agreement
• DPA terms
• Financial penalty
• Court application
• After a DPA

What’s new?

By its nature, in pulling together material from other sources, the handbook does not introduce any new requirements that lawyers or corporates need to be aware of - but it does act as a helpful guide to the above stages.

It also serves as a reminder of both the stringency of certain elements of the SFO’s operating parameters, and how these must be navigated throughout the process; but also, by the same token, the flexibility permitted in other respects.

For example, there is a helpful restatement of the lower evidential test required at the point of entry into negotiations (that is, when the full Code evidential test is not met, but there are “reasonable grounds for believing that a continued investigation would provide further admissible evidence within a reasonable period of time”) compared to the full prosecution process. Other points to note in the same vein include the reminder that no formal admissions are required; and all DPA terms agreed between the SFO and the corporate are required to be fair, reasonable and proportionate.

Some continuing difficulties highlighted: is there a defence available?

In pulling the various source documents together, the chapter does highlight some of the elements of the DPA dynamic which can be tricky for practitioners in practice. Foremost of those are: the availability of defences to the offence(s) suspected; and the extent of the ‘cooperation’ required from the corporate in the DPA process.

With regards to DPAs concerning offences under the Bribery Act 2010, there has long been a question mark ,as to whether the defence for a corporate under section 7 (the ‘failure to prevent’ offence) is or ever will be available on the facts. Where a corporate must show that it had adequate policies and procedures in place to prevent the bribery having occurred, the fact that an offence under section 1 or section 6 of the Act did, in fact, take place can make the defence very difficult to prove. Many have struggled to understand how the defendant shows procedures were adequate if it is accepted (as it must be for a DPA) that their design failed to be effective in a particular instance.

The chapter in the handbook highlights this difficulty by noting that among the public interest factors in favour of a DPA are:

• the existence of a proactive corporate compliance programme both at the time of offending and at the time of reporting but which failed to be effective in (the particular) instance; and
• (cases where) the offending represents isolated actions by individuals, for example by a rogue director.

As well as operating as factors in favour of a DPA (as opposed to prosecution), either of these points may, if made out, mean that the corporate has an effective defence to the section 7 offence. In circumstances where the defence is available, it may be that boards find it very difficult to square the benefit of the DPA process over a contested one, where the contested process may have a significant likelihood of acquittal.

Cooperation, cooperation, cooperation

The other perennial difficulty that the guidance reminds us of is the extent of the ‘cooperation’ required by the DPA process and the extent to which the corporate must, effectively, provide the SFO with information it might not otherwise have obtained in order to secure the agreement, and in circumstances where the full extent of the relevant offending (if any) may never have become known.

For example, public interest factors in favour of entry into DPA negotiations include both (as we have long known) notification of the wrongdoing within ‘reasonable time of the offending conduct coming to light’, and notification of offending otherwise unknown to the prosecutor. Those elements of the ‘cooperative’ approach do not sit easily with the fact that it will be a factor against inviting a corporate into the DPA process where it reported wrongdoing but ‘failed to verify it’, or reported knowing or believing the information to be inaccurate, misleading or (more likely) incomplete (i.e. has not investigated sufficiently). The upshot of these factors is that a corporate must investigate enough so as not to be ‘failing to verify’ or be ‘incomplete’ in its self-report; but not so much as to delay the self-report beyond whatever is ‘reasonable’ in the circumstances. That may be very finely balanced in most cases.

Moreover, with respect to notifying offending of which the SFO is not aware, a corporate must, it seems, be sufficiently confident right at the outset that it will prefer to be seen as ‘cooperative’ in order to access the outcome of a DPA-driven process (that is, at the end of the very long, very uncertain road down which it faces). It must quickly decide that such a process is more favourable for the company than the prospect of any potential conviction (and must therefore, we assume, be satisfied both an investigation will be opened, and that the investigation would result in trial).

Moreover, if it does reach that level of confidence as to the matters first identified, it will also need to weigh up the possibility of the need to disclose further issues above and beyond those which have already come to light; and the fact that those further issues may alter the risk calculation made in seeking the DPA in the first place. Once engaged in and committed to the DPA process, the risk of providing ‘incomplete’ disclosure on wrongdoing may realistically mean accepting liability for a greater extent or range of offending than originally anticipated.

Where does this leave us?

These issues have long caused headaches for corporates and advisers stuck in the weeds of the early stages of a criminal investigation. Not only is it still truly unclear as to what extent the section 7 defence is or ever will be successful; but a decision must be made very early on as to whether or not to seek a DPA, thereby requiring suspect companies that believe they have an ‘adequate procedures’, explicitly to abandon that in favour of a negotiated resolution. Furthermore, that early approach is still key to entry to the process, but at what price? It is impossible to know quite how corporates can arrive quickly enough at the SFO’s door and therefore be seen as truly ‘cooperative’, whilst nevertheless investigating matters that come to light with appropriate rigour, and thereby acquiring sufficient confidence that it is, on balance, favourable to approach the SFO both on (i) the merits of the issue(s) first identified and on (ii) the basis that such wrongdoing may not ultimately be the ‘whole picture’. The timing and value of that early self report, and effective abandonment of potentially available defence(s), has only been further thrown into doubt by some of the recent cases. With each DPA that is agreed, the market’s approach to the approach set out in documents like the Handbook (and of course, the Code and the Guidance) continues to be further nuanced; and we therefore await further cases in 2021 with great interest.