Career Change
Options
Comments
-
it should be possible to deliver mass market advice tailored for a greater range of personal circumstances at a low cost.
Mass market, tailored advice, low cost. Which two do you want?
Tailored advice + low cost = small IFAs
Mass market + tailored advice = SJP (and its competitors)
Mass market + low cost = robo-sales
Robo-floggers do not give advice. Getting someone to fill in a questionnaire, passing the results through an algorithm to get X, and flogging them model portfolio X is not advice and never will be, regardless of how the technology develops.
At this point the robo-sales visionaries will probably say "something something AI", overlooking the fact that to be capable of delivering true advice, we need to pass the Singularity and invent an AI superior to ourselves. An artilect inferior to humans cannot give humans advice. (Would you take financial advice from a human who was visibly stupider than you?) At this point nobody will need financial advice because the Singularity AI will either give us everything we want for the rest of our lives, or kill us or plug us into the Matrix.0 -
It's also worth noting that for most IFA firms, the average client is not interested in the individual investments, the structure or research or even anything to do with the actual investments. They want the comfort. They want someone that knows them to tell them everything they have is ok (and with tailored advice it should be). They want someone that prompts them to do something or make them aware of issues. Some even want you to control their large purchases. i.e. each year asking if they can have an expensive holiday or a cheap one.
Not sure AI is going to be anywhere near that level for some time.
For decades, the regulator has looked at simplified advice offerings (effectively flow chart style advice) and each time there has been a too high failure rate. The failure rate can be reduced by adding in more stages to the flowchart but then it stops being simple and low cost.
One of the biggest issues is the difference between a textbook answer and a real-world answer. A text book answer would still put friendly society savings plans as good advice. Or purchased life annuities as best advice. Anyone who is learning to be an adviser has to answer questions on the basis of the text book but then in the real world have to do things differently because of commercial reality (pricing, features, niche options etc). For example, there is a provider that can pay out £30,000 using the small pots rule (which is limited upto £10,000 per pot). The text book answer is that it cannot be done. Real world answer is it can. AI will go with the text book answer.I am an Independent Financial Adviser (IFA). The comments I make are just my opinion and are for discussion purposes only. They are not financial advice and you should not treat them as such. If you feel an area discussed may be relevant to you, then please seek advice from an Independent Financial Adviser local to you.0 -
One of the biggest issues is the difference between a textbook answer and a real-world answer. A text book answer would still put friendly society savings plans as good advice. Or purchased life annuities as best advice. Anyone who is learning to be an adviser has to answer questions on the basis of the text book but then in the real world have to do things differently because of commercial reality (pricing, features, niche options etc). For example, there is a provider that can pay out £30,000 using the small pots rule (which is limited upto £10,000 per pot). The text book answer is that it cannot be done. Real world answer is it can. AI will go with the text book answer.
Presumably with something like the £30k small pots rule the question of not being able to do it is resolved by "more steps in the flow chart" ,e.g. saying "is it over £10k, no, but is it over £10k each if i create two new sister pensions and do a partial transfer into them, no, ok it's feasible. But of course whether it's sensible to actually take the £30k now rather than keep it invested a bit longer is different question...
We can see that AI at the moment isn't really "artificially intelligent" as it can only do what it's been programmed to do by someone who knows what the rules are and has kept it up to date. I expect doing the initial build of the decision tree by reference to what IFAs would seek to do in the real world is pretty tricky because every time the developer says to the IFA what would you do next if faced with this this and this, the IFA's response would start with, "well, it depends, let's explore x y z..." and the flowchart gets another few levels or layers in multiple dimensions.
Still not impossible to build as presumably there is a methodology to "teach" the rules and recommendations which is how an employee gets to be able to do client facing work himself. Just not an easy thing to teach a robot - as many businesses who have tried to save costs by outsourcing a business process to a big office full of eager but process-driven employees in cheap jurisdictions, can tell you. Many people don't like speaking to call-centres where your conversation partner is inflexibly working off a script and doesn't really *know* the problem you face on your faulty product or your parking ticket notice etc. So using robots to handle their life savings might be an issue.
A robot to deal with basic concepts and keep you within the regulations while overlaid with human intervention might be quite credible - but then, as you say, you lose the cost advantages of not having humans, and perhaps the cost advantages are not much better than what you'd get with a decent system and cheap humans and a lot of economies of scale supporting the expensive humans.0 -
The FCA have published failings with robo-advice today. Nothing of surprise.
1 - Failure to disclosure whether they were advised, non-advised, discretionary or non-discretionary.
2 - The FCA found some firms compared their fee levels against peer services in a potentially misleading way.
3 - The FCA also found problems with the way these companies conducted suitability assessments, with many firms failing to evaluate a client's knowledge or experience, investment objectives and capacity for loss. Some firms did not ask clients about their knowledge and experience at all and told the FCA they felt their service was suitable for all individuals regardless of their investment knowledge and experience.
4 - The FCA also found failings in the ongoing client relationship, with most firms providing automated fund management services unable to show they had adequate and up-to-date information about their clients when providing an ongoing service.
5 - Among automated advice firms, there was also "little consideration" given to governance processes, with a "mixed" awareness of the need for stress-testing and cyber security.
Whilst some of the above have been discussed often on this site, it does appear that what the FCA call "robo-advice" and what robo-advice providers consider it to be is different. FCA appears to be looking for that AI as some of the things it needs cannot be met without human discussion.I am an Independent Financial Adviser (IFA). The comments I make are just my opinion and are for discussion purposes only. They are not financial advice and you should not treat them as such. If you feel an area discussed may be relevant to you, then please seek advice from an Independent Financial Adviser local to you.0
This discussion has been closed.
Categories
- All Categories
- 343.6K Banking & Borrowing
- 250.2K Reduce Debt & Boost Income
- 449.9K Spending & Discounts
- 235.7K Work, Benefits & Business
- 608.7K Mortgages, Homes & Bills
- 173.3K Life & Family
- 248.3K Travel & Transport
- 1.5M Hobbies & Leisure
- 15.9K Discuss & Feedback
- 15.1K Coronavirus Support Boards