Cambridge Symposium 2016
5 September, 2016 | Speeches
David Green CB QC, Director, at the Cambridge Symposium on Economic Crime 2016, Jesus College, Cambridge.
This is my 5th report to the Symposium as Director of the Serious Fraud Office.
Turning immediately to some recent notable results:
In the Barclays Libor trial, the verdicts were significant because this particular jury broadly rejected the suggestion that alleged “market practice” somehow absolved individual dishonesty; two of the defendants were senior employees of Barclays; and the trial in London of two US nationals for their conduct whilst based in New York reflects the close cooperation between SFO and DoJ that has characterised the Libor investigation.
The Sweett Group pleaded guilty to failing to prevent bribery by persons associated with it in relation to a construction project in the Gulf. That was the first conviction for a Section 7 Bribery Act offence; and an example of a case where, reflecting the company’s conduct and the degree of cooperation, prosecution rather than a DPA was appropriate.
We have completed our first two Deferred Prosecution Agreements. Each of the judgments provides valuable guidance to all concerned, more on this in a moment.
Operation Canon, a rather more traditional fraud, was an £80m Ponzi/investment fraud running over four years and centred on an electrical wholesale business in North London. All three defendants were convicted.
In terms of confiscation and asset recovery, our Proceeds of Crime team obtained orders totalling £3.4m and recovered £19.6m in the year to April 2016. Since then to the end of June, we have obtained orders of £11.6m and recovered £2.4m. It is in the nature of our complex casework that these totals are subject to substantial fluctuation year on year.
Currently, looking ahead, we have 40 defendants awaiting their trial in 8 different cases. These cases include the Euribor trial (with six defendants), FH Bertling (eight defendants, one of which is the corporate), Alstom and two major fraud trials. Pressures on court listing can lead to gaps of 12-18 months between charge and contested trial.
Our current caseload is split roughly 50/50 between fraud and bribery. Some results, frustratingly, cannot yet be discussed due to reporting restrictions. Be patient – I have to be.
When scandals which commentators believe must involve major fraud or bribery erupt, people expect the SFO to “do something about it”. Where we can, we have, and we will. We are not averse to taking on the most difficult cases (just look at our current case load), but basic criteria must be met first. For example, with FIFA, we had no jurisdiction; the US and Switzerland did.
By statute I can only open a criminal investigation if I have reasonable grounds to suspect an offence of serious or complex fraud or bribery. In all cases, our intelligence section (comprising lawyers, investigators and analysts, in which we have made significant investment) will take the necessary time to dig into the facts and produce an evaluation. This pre-investigation phase is vital, and we don’t supply running commentaries.
We do not announce new investigations unless a company, for instance, has to inform the market or there is some other good reason, such as encouraging witnesses to come forward. Those we have announced include Airbus, Soma Oil, UnaOil and Quindell.
Well-known active investigations including Libor, Tesco, Rolls-Royce, GlaxoSmithKline, Barclays Bank, ENRC and GPT continue, and in each case we have made significant progress; some are on the verge of a charging decision.
The real message from these very large and complex investigations is that the SFO will use every lawful lever to reach the right conclusion, and we will not give up and go away even if our well-resourced suspects choose to batten down the hatches and shun cooperation, as some do. We are happy to fight skirmishes along the way, as we did in Lord, the unsuccessful judicial review challenge to the application of our policy on the presence of lawyers at Section 2 interviews.
The role of the prosecutor is misunderstood in some quarters, and such misunderstanding ultimately tends to undermine prosecutorial independence and public confidence.
We are here to assemble the best case on the available evidence. If we conclude that the evidence would never pass the code test, then we will close an investigation, as in Forex. If the code test is passed, we present the case as strongly as possible.
If a trial is halted before the jury get to consider verdicts, then the prosecutor may be at fault. But once a case goes to the jury, it is entirely in the hands of that particular body of 12.
Some prosecutions result in an acquittal; that does not mean that the case should not have been brought in the first place.
In the Libor-related trial of the 6 cash brokers, the jury acquitted after a day’s deliberation. In the Barclays Libor trial, the jury deliberated for 11 days and convicted three (unanimously and on varying majorities) and failed to reach verdicts on two defendants who will be retried.
Jury verdicts can be surprising to all parties involved, but it is my firm view that the jury system is the least imperfect system available.
Our trials tend to be few in number and can last 3-4 months. Like our confiscation totals, our conviction rate can show marked variation year on year. The overall trend remains respectable, but always capable of improvement. Over the four years 2012-2016, the SFO’s conviction rate was 65% by defendant and 81% by case, with 49 defendants in 25 cases convicted.
The Deferred Prosecution Agreement is the new tool available to us in our dealings with corporates, and we are learning how to use it.
We have completed two so far, Standard Bank and XYZ; others are in the pipeline. As I said earlier, be patient: I have to be.
I have always emphasised that cooperation is a prerequisite for a DPA. Both judgments emphatically underscore that vital consideration. Cooperation covers not just prompt self-reporting but also ongoing assistance to our criminal investigation into the conduct of the company and the individuals who drove that conduct.
The greater the cooperation, the faster the investigation phase and the more likely the DPA. Contrast the Standard Bank DPA with the Sweett Group prosecution.
We well understand that the most cooperative company may be anxious to make concurrent settlements in other jurisdictions, and will want to test and challenge our calculations of profits to be disgorged and the appropriate multiplier in the course of DPA negotiations.
But the traditional tactics of the litigator have no place in DPA negotiations. It is not about point-scoring or wearing down an opponent, it is about reaching a transparent agreement that a court will accept, reflecting the interests of justice.
Looking at the broad lessons from the first 2 DPA judgments, both from the President of the Queen’s Bench Division.
- The court conducts a detailed supervision of the exercise, detailed examination of the facts, the conduct of the company, and the application of the “interests of justice” test to those facts.
- The judgments maximise the transparency of the process.
- In Standard Bank, there was emphasis on the nature of the offending (a single, serious failure to prevent bribery), the degree of cooperation, the bank’s previous record, the change in ownership which made the current organisation different in character than at the time of the offence.
- In XYZ (concerning an eight year course of systematic bribery) those considerations were augmented by the importance of incentivising corporate exposure and self-reporting of criminality, and ensuring that innocent parties were not punished.
- XYZ showed the connection between the interests of justice and the proportionality of the DPA’s terms, in particular the ability to pay a financial penalty. XYZ could afford to pay a maximum of £352k towards the financial penalty; the innocent US parent company contributed £6.2m. This was described by the Judge as a “pragmatic answer” achieved through what he called “stepping back” and resulted in a total penalty figure equivalent to the total profits from the corrupt contracts.
- The Judge summarised in XYZ that:
- “it is important to send a clear message…that a company’s shareholders, customers and employees…are far better served by self-reporting and putting in place effective compliance structures. When it does so, that openness must be rewarded and seen to be worthwhile”
The SFO and International Cooperation
All our cases have a significant international dimension. We have invested real effort in building strong cooperative relations with foreign agencies in key financial centres across the globe. This involves secondments, rolling discussions, exchange of information and coordinated activity.
Some commentators seem to regard this as “not cricket”.
To them I would say, get used to it.
Through this cooperation, practical agreements have been reached, for instance with the US authorities on primacy over different facets of the Libor investigation.
Enhanced cooperation has facilitated coordinated searches in multiple jurisdictions around the world, happening at the same time, attended by SFO staff. It has also produced valuable intelligence and, through MLA, evidence in support of our casework.
This means, I suggest, three things for suspects: We are more likely than ever to obtain intelligence of corporate misconduct, to obtain admissible evidence from abroad, and to engage with foreign jurisdictions so that a cooperative company can obtain organised outcomes to investigations.
Corporate criminal liability
As an enthusiastic supporter of a new “failing to prevent economic crime” offence since 2012, I warmly welcome the consultation announced at the Anti-Corruption Summit.
As things stand, before a company can be prosecuted in this country, the “identification principle” requires the prosecutor to identify the “controlling mind” of the company and prove that that person was complicit in the offence under investigation.
In a world of increasingly complex corporate structures, the identification principle can hobble the prosecutor in those cases where it is right to prosecute the company.
The principle is illogical in that at present it is the only route to liability for all major economic crime offences (fraud, false accounting, money laundering) except bribery and (soon) tax evasion.
The principle operates unfairly: it is always easier to identify the controlling mind in a small company than in the case of a large corporation.
It is also creates unhelpful incentives for senior executives: on the one hand to distance themselves from knowledge of operations in fraud cases, on the other to preach compliance in bribery cases so as to demonstrate adequate procedures.
Tom Hayes was prosecuted in this country for his role in Libor manipulation. The operation of the identification principle meant that we could not touch the bank for which he worked whilst manipulating Libor. That bank was held to account for Hayes’ conduct in a New York courtroom, where vicarious liability made the prosecution a much simpler matter.
A “failure to prevent economic crime” offence would significantly increase the prosecutors’ reach in those cases where a company should be held to account for the conduct of persons associated with it. It would also underpin the anti-corruption, responsible capitalism and social justice agendas, and the prosecutors’ contribution to those efforts.
This symposium poses the question: “where should the buck stop”?
My answer as an investigator/prosecutor is: the buck stops where the evidence takes us.
In conclusion, the SFO is ready, willing and able to play its part in the fight against economic crime by taking on the topmost level of complex fraud and bribery for which it was designed and intended. We are well equipped and confident in that exacting role and mission.
Thank you for your kind attention.
*This speech has been temporarily edited for legal reasons*