There are key components in the current approach to financial advice that may produce poor outcomes for clients and the adviser.
Firstly, the risk-profiling process for determining the client's asset allocation is generally inadquate. It's focus is almost entirely on the short-term (one year) variability of investment performance (described as "risk"). Investments producing more variable short-term returns (eg equities) are described as more "risky" than those producing more stable short-term returns (eg cash). But this is true only if the client has a short time horizon. But risk, properly defined, is the chance that the client will fail to meet their goals. In reality, most clients have much longer time horizons and in that situation the higher-yielding, more variable investments (eg equities) are actually less risky than lower yielding more stable investments (eg cash). Yes the client will experience more variability with higher-yielding investments, but even allowing for that they may still have a higher chance of achieving their goals than with lower-yielding more stable returns. The current approach is really addressing fear, not risk.
Secondly, the current approach to financial advice, where financial projections use assumed average investment returns for a portfolio, show outcomes that have only a 50% chance of being met. For example, when a client is near or in retirement and advice is provided around what retirement income they can drawdown from their investments, there is a 50% chance the client will run out of funds earlier than expected. And that is just too much risk for most people - the toss of a coin that they will end up on the age pension before their target shortfall age!
To address both of the above shortcomings, financial advice tools and processes need to include stochastic modelling of the client's future combined with portfolio optimisation and presented in a way that the client understands and therefore accepts as being the best advice.