“Give me a one handed adviser” exclaimed one of my very good clients a few years ago, ‘Billy you keep saying on the one hand this and the other hand that’.

Personally I think advisers need to use two hands because things are rarely black and white, especially where the new pension freedoms are concerned. If I had a pound for every client who said “Mr Burrows I know exactly what I want to do”, only to change their minds after I put an alternative solution to them, I would be rich my now.


One of the problems with the robo-advice is that by definition it doesn’t have any hands and so cannot put forward two sides of any argument.

I am concerned about the talk of using robo-advice to help solve the problems created by the huge advice gap, especially for those at retirement because there is a danger that without proper human input people may end up with the wrong solutions.

A good place to start is to define what robo-advice is. According to Wikipedia, ‘Robo-advisors are a class of financial adviser that provides portfolio management online with minimal human intervention.’ This serves to highlight the problems facing robo-advice when it is applied to retirement options.

Human intervention

You might think that my main criticism is with minimal human intervention but I am less concerned about this because in reality people reach for the phone when they get frustrated communicating with robots so I am more concerned about the computer algorithms.


The majority of algorithms – here we go using jargon that not many people understand. An algorithm is a procedure or formula for solving a problem. The word derives from the name of the mathematician, Mohammed ibn-Musa al-Khwarizmi, who was part of the royal court in Baghdad and who lived from about 780 to 850. Not many people know that!

The majority of algorithms are designed to solve the problem of how to construct efficient and risk appropriate investment models, they are not designed to solve the problem of how best to convert a pension pot into cash and income. It seems to me that most robo-advice process follow a linear path but in practice the process is often iterative.

Having tried to design an on-line advice process myself I am acutely aware that for most people retirement advice is an iterative process. By this I mean that the conversations and customer journey can go round in circles. The retirement advice process involves making a number of trade-offs; income certainty or income flexibility, no risk or some risk, annuity or drawdown. The best way to analyse these trade-offs is be clear what the client’s objectives are but these objectives are often re-defined as the advice process develops. This means the final objectives can be often very different to those stated at the outset. In practice this means that a client who starts of by saying that they want income flexibility and are prepared to take some risk may change their mind after being better informed about the risks and key issues. The reverse is also true; a risk averse investor who wants income certainty might decide they are better off trading the income certainty for flexibility because although an annuity meets the certainty criteria it is not certain to meet their income needs in the future.

Beware of no-advice – the stakes are too high

I hate no-advice with a passion because clients can be taken down a particular path without the opportunity to consider an alternative course of action. I am wary of robo-advice because many clients are simply not equipped to set out their retirement objectives without being nudged, helped or challenged by an expert.

I know my critics will point to the economics of using automated systems for delivering advice and argue that robo-advice is better than no advice but the stakes are too high to entrust such an important decision to an automated process. If we are to fill the advice gap people need much more than minimal human intervention they need a two handed adviser.