Why can’t the regulator use its own statistics to fix the market?
It may be a distant memory given how fast the market is moving at the moment, but the FCA is still in the middle of a pretty significant review of how the Financial Services Compensation Scheme is funded.
The way the professional indemnity insurance market works for advisers – or actually, doesn’t work, as FCA boss Andrew Bailey himself has said – is central to getting everyone’s liabilities in order.
The FCA’s Gabriel return – often heavily criticised by advisers – actually collects some very useful and necessary information on PI that could help the regulator get a grip on the issues at hand. It seems not to have used this to its full, or indeed any, potential so far.
Cutting through the noise
The Gabriel return’s PI inputs are pretty basic and shouldn’t be arduous for users to fill in: start date and end date of policy; name of insurer; annual premium, which areas the policy covers (mortgage, insurance, or retail investments advising), any business lines the PI does not cover and any excesses that apply.
It is not inconceivable that, through aggregation of this data, the FCA could work out the average length of policies, the average annual cost of premiums and how this has changed over time.
It could also work out the most commonly excluded business lines. It could tell, say, if a significant number of investment and pensions advisers who also advised on mortgages had occasionally actually omitted that from cover.
If the FCA is worried the market is concentrated among too few providers, the output from the drop down list of insurers on the return could spit out what percentage of the market each holds. There’s an ‘other’ option too, to count up the total number of smaller providers that are actually out there.
If the FCA fears the market is offering deals that are too short term, the length of policy data punched into Gabriel should readily confirm this.
Yet, for reasons that have never quite been adequately explained to me, this information isn’t at the FCA’s fingertips.
A while back, I put in a Freedom of Information Act request asking the FCA if it could run off any aggregate numbers from Gabriel about the PI market. It couldn’t.
I wasn’t asking for a more sophisticated analysis (e.g. correlating the providers selected with exclusions and excesses that apply to see if it is particular providers that are causing the issue). The FOI was simply asking, from the FCA’s data, how much are advisers paying for PI?
The concerns raised regarding PI in the FCA’s consultation on FSCS funding seem quite specific: “some policies exclude the insolvency of the policyholder or the FSCS as a claimant” and “may also exclude particular types of sales and advice from their cover”. But how many policies? Which types of advice are excluded? How frequently?
Industry representatives are about to go into further talks with the regulator. That’s all well and good for collecting anecdotal evidence, but where is the concrete analysis that can really punch through both sides’ vested interests?
Of course the PI inputs are still important; compliance professionals have told me that a red flag will alert the FCA to late professional indemnity insurance renewals if the numbers don’t add up.
I understand that reporting periods might differ, and there may be nuances in the qualitative inputs for excesses for example, but most of this information is simple quantitative stuff that comes in on a six-month basis. I understand that it may incur a small cost to churn into an accessible format.
Maybe the FCA has received some other good intelligence outside of Gabriel from the market on the size and shape of the PI problem – though no one I’ve spoken to seems to have any good data to that end.
Who can blame them, with such a range of complex policies out there? That means it’s up to the regulator to use what it already has to furnish the market with the knowledge it should have stored in its databases.
Other areas of Gabriel have been used to good effect by the FCA to extract informative aggregate data, for example complaints and qualifications data. It’s about time we saw the same from the PI returns.