Below are the questions I’ve been asked to address next month at the Pension and Benefit Conference.
I hope that when robots go to grammar school, they’ll be able to do a little better framing the question!
The framing of the question suggests that I’m expected to come down on the side of humans, but this supposes that there are people ready and willing to give that advice.
The answer to that question is rather different today than it was three years ago. In the heyday if advice, before the RDR was implemented, there were over 51,000 pension advisers registered by the FCA, that number has more than halved as transparent charging has reduced the numbers willing to pay for advice. Whether the number of us needing advice has decreased has reduced is another story. The pension freedoms are an adviser’s fly-trap.
Some would answer that, regardless of technology, distance learning would have been needed to plug the gap created by the RDR. But the last five years have seen another significant change, the rise of “search”.
We are now used to getting answers to our questions at a click or swipe. If not google, youtube, if not webchat – the bot. The web not only has the gen, but it is increasingly easy to navigate. If you want information, it is now available, it’s free, it’s easy and it’s open 24/7.
From search to guidance
Searching gets us information, but it does not present that information in a way that helps us take a decision. Indeed, the commercial instincts of those who own the social media sites distorts our searches so that we now enter into a game of cat and mouse with the advertisers , our judgement being exercised to sort editorial from advertorial, representative from selective data.
From guidance to advice
Advice in this context is defined as the provision of a definitive course of action, guidance no more than a fair representation of information necessary to make an informed choice.
With advice comes liability and with liability comes an entirely more vigorous inspection of the circumstances surrounding advice.
The stakeholder pension decision tree was a fairly cumbersome representation of the simple decision taken by a company, a decision tree representing the Nutmeg algorithm would be as different as an aged oak is from the sapling.
The questions posed above are simply scratching at the surface of the complexity of a decision as complex as the investment of a pension pot in drawdown.
But here we have an important decision to take, one that is critical to the FAMR and to the be application of Financial Technology to our financial decision making.
The liability for how we use our robot is far from clear. If the robot has been to the FCA’s sandpit and been deemed a fit and proper robot, should it be let loose on the public without fear of reprisal? Whose is the responsibility for maintenance of the algorithm and the application of the research that informs it ? Is liability for advice time bound or is the benefit of hindsight always with the user?
Too hard to call?
There is inertia in financial services towards the status quo. This bias is understandable. We have had it so good for so long that the threat of a new way of doing things is treated with disquiet. However, the move from search to guidance to advice that is envisaged by robo-advice seems to have been adopted not just outside of financial services but in general insurance, mortgage broking and in the purchase of life insurance. Investments are of course longer term and more capital is at stake than an insurance premium. It shouldn’t be this hard.
The biggest threat to pension providers from robo-advice is probably through the transparency that digital information brings. The pension dashboard will be a reality in 2019, from it will emerge a system of pot follows member. For the system to work, we will need to know not just how to switch but the cost of switching. This will require new levels of transparency from pension providers and asset managers. Concepts such as the “free switch” will need to be tested against before and after scenarios where transaction costs will be laid bare. As these costs emerge, so will a call for greater transparency on the day to day costs of running our pots.
Tom Tugendhat at a recent meeting pointed out that the data needed to operate robo advice is no different from that that powers the pension dashboard or helps us understand our funds on a “true and fair” basis.
The real threat of robo-advice is not that it gives the consumer too little, but that it gives them too much. The challenge is in the organisation of information to enable informed choices and in managing liability so that robots can be empowered to give advice.
I don’t think this is far away. The biggest obstacle to these changes happening – is fear of change. Advice has not change – but the way it is delivered may be about to change for ever.
If you want to see the slides I will be using at PBUK, here they are