A View From The Exchange: Should the FCA redefine its fining policy?

If a policy is routinely ignored by a public body and shown not to be fit for purpose, shouldn't that policy be shelved?  In relation to the FCA's penalty setting framework it feels that that time has come.

Up until 2010, penalties for regulatory breaches under FSMA were simply set by reference to "all of the circumstances of the case".  This enabled the regulator to take into account all key factors that it considered relevant to the level of punishment.  

However the Upper Tribunal felt uneasy with the lack of guidelines on appropriate fines and so a five-step framework to calculate financial penalties was put in place.  This involves assessing the firm's relevant revenue (or an individual's remuneration), applying a pre-set multiplier depending on the seriousness of the breach and then applying further multipliers or dividers for aggravating or mitigating factors.  As of right the subject of the action receives a 30% discount for settling the action at an early stage.

This framework gives the regulator an express discretion to change the penalty where it would otherwise be disproportionately large or where a higher penalty is necessary as a future deterrence.  This discretion was historically used very rarely – one notable example being when five banks were disciplined for FX failings in 2014, with the fines calculated under the standard framework being increased very significantly by the FCA for credible deterrence reasons.

Current approach

However this position has certainly changed – it feels as if Final Notices published over recent months have much more frequently included significant adjustments (upwards or downwards) to the fine calculated under the standard framework.  In practice in these cases the FCA has chosen to ignore the number produced by the framework and instead substitute its own number. 

For example, in the recent financial crime systems cases against Starling Bank and Metro Bank, the fine imposed on both banks was reduced from the figure under the standard approach.  In the Starling Bank case the step 2 figure of £168m was substituted by the FCA with a figure of £40m.  In the Metro Bank case the step 2 figure was arbitrarily reduced by the FCA from over £317m down to under £24m to produce a financial penalty (after discount) of £17m.  Both of these are massive adjustments away from the figure calculated by the standard framework.

The TSB arrears handling case did not involve a step two adjustment but instead included an unprecedented 40% reduction to take into account a mix of both aggravating and mitigating factors, which significantly reduced the overall level of penalty.

Similarly in the PwC / LCF case, the FCA declined to use "relevant revenue" at all and simply inserted its own notional revenue figure of £15m as the appropriate level of penalty.

Conversely, in the CB Payments the standard framework came to a total step 3 financial penalty of just £377 – which the FCA felt was too small and therefore decided to substitute its own arbitrary figure of £5m.

The FCA did the same in two recent cases against individuals, Leigh Mackey (whose penalty figure was increased at step 4 from £24k to £192k) and Kristo Kaarman (whose penalty figure was increased at step 4 from £41k to £500k). 

In all these cases the FCA has deviated from its standard policy as a method of calculating the financial penalty, and instead substituted the figure that it considers is appropriate in all the circumstances of the case.  

Back to the future?

The frequency with which the FCA is now doing this raises two questions.  First, why is the standard framework routinely coming up with financial penalties that are so far away from what is reasonable?  And second, what is the point of having a framework for calculating financial penalties if you almost universally ignore it and instead insert the number you first thought of?  

As with the FCA's new approach to selection of cases for enforcement, perhaps it would be sensible for the regulator now to return to the pre-2010 approach by formally shelving its five-step framework and instead being open about the fact that it is applying its old test of what is fair and reasonable in all of the circumstances.

Author: Joanna Middlemass