Search
  • Stuart Fowler

Truth hurts


Speaking at the same conference for financial advisers in London and Manchester this week, I was struck by a mantra repeated over and over: ‘people are turned off by investment, they are only interested in outcomes’. The outcomes they can and do relate to are things like spending power, educating children, not running out of money and leaving something for the kids. Their experience of financial planning and investment is therefore likely to be related to their continuous confidence about their outcomes. How strange, then, that very few financial planners or portfolio managers are able to discuss specific goal-related outcomes using explicit probabilities, as percentage chances or real amounts of money associated with each probability.

Life lines

A show of hands revealed that virtually all the IFAs attending each conference used ‘cash flow modelling’. Many probably use the same as the speaker who asked the question: a spreadsheet-like application called Truth, from Prestwood Software. These ‘deterministic’ tools, including own-design spreadsheets, exist to compound assumptions out into the future for income and expenditure and relate these flow outcomes to stocks of financial resources (existing or incremental) which are also compounding at linear growth rate assumptions. The resulting graphical displays show the changing income statement and balance sheet for the household for the assumed length of the planning period. What results is changing balances. Advisers will use these as measures of sufficiency or ‘financial independence’ that will be relied on to provide the clarity and confidence that (we know) individuals value as the immediate payoff from planning.

If the linear projection rates for each item are ‘reasonable’ as a reflection of what might be thought of as ‘normal’ (which means they have to be normal for the starting point as well as the end point), I suppose the programme could be termed Half Truth. As a measure of actual predictive accuracy, as the chance of being right rather than just better or worse, it should (being generous with rounding) be called 1% Truth.

I can well understand why advisers’ clients would see these graphical representations of their changing household economy as both relevant and useful and why it would be tempting to rely on them for confidence. It is impossible, even, to imagine any planning for the household (or a business) without forward projections. Unfortunately, these simplistic projection tools do not contain the information clients really need because they are deterministic, not probabilistic. They conceal the uncertainty in both sets of inputs: their household cash flows and the payoffs from participating in capital markets by borrowing and investing.

It might even be argued that, by encouraging false confidence, they do more harm than good. Our own model for projecting outcomes in real terms suggests that, even with quite low risk tolerance, worst-case outcomes at different time horizons might be between 25% and 50% below the mean – and that assumes a model-driven approach of tightening up risk as horizons shorten, rather than ‘buy-and-hold’. That is a margin that, left unaddressed, could lead to errors in spending and saving that can never be put right without hardship. The same silence or ambiguity about the possible payoffs from with-profits endowments recently led to similar drastic consequences, so the industry has no excuse for not learning to deal with its own shortcomings.

The information clients need is about what can go wrong but they cannot respond to that information without also knowing how much better it could be with different decisions they have the power to make.

Some of the biggest sources of disaster are insurable and the advisory industry is generally good at selling catastrophe insurance, such as the impact of early death or disability. But as soon as people participate in financial markets to achieve objectives like replacing employment earnings in retirement, they take on inflation, investment and (possibly) longevity risks that together make their financial outcomes highly uncertain. The range of uncertainty can be narrowed, but at a cost. So the information people need is

  • How bad can it be – as in the worst-case or ‘floor’ outcome

  • How much does it cost to ‘raise the floor’

  • What am I giving up in terms of better outcomes.

Only probabilistic models can answer these questions with hard numbers.

Quoting the odds

My talk to the advisers was on the differences made when using probabilistic approaches to modelling outcomes. This is the Defined Outcome goal-based portfolio approach described on this website under What we do. What it is about this approach that changes the client’s experience of investment is our ability to quantify ranges of relevant outcomes so they can relate to the process even if they are not interested in investment. With choices differentiated by hard numbers, they can make more confident decisions about the things that really do matter to them: earning, spending, saving and risk taking.

What controls the effect of risk taking on the probable outcomes is also something people can readily relate to: insuring risk versus bearing risk – bets off the table or bets on the table. Insuring risks in an investment sense means avoiding or immunising them, by holding something that perfectly matches (or hedges) the required outcome. But (as they also know) insurance costs money, so raising the floor (or worst-case outcome) by hedging will also lower the entire range of probable outcomes. This is a better measure of Truth than a single linear projection that treats investment risk and cash flows as if they were independent.

Diversification is not enough

Taking bets off the table was (I argued) a much better way to manage outcome risks than investment diversification, and not just because diversification takes them into the world of investment maths they are not interested in.

Most of my fellow speakers were proposing solutions that depend on diversification between types of asset to control risk. But they were in defensive mood. This is because diversification conspicuously failed to deliver the expected benefits in this bear market, as all forms of risk premium collapsed together, under the weight of liquidity-induced risk aversion. As illiquid assets could not be sold, high quality assets had to be sold instead. Nothing escaped.

Knowing what was coming up, I was able to slip in a reminder that diversification was never supposed to be a form of risk control, but rather a process for getting rid of unrewarded risk exposures and maximising the combined risk-adjusted returns (or utility benefits) of the undiversifiable or ‘systematic’ asset exposures that were left – these being the ones that are rewarded, with ‘risk premiums’. Risk control is always likely to require limiting those exposures: bets off the table.

Why is No Monkey Business on such a different page from our industry peers? I think the best explanation is that, for all that they read and all the conferences they attend, they really don’t understand just how uncertain the world is. Or, if they are worldly enough to understand that, they are perhaps not able to translate it into realistic numbers. You can sense this from what pops out of their mouths: the 1% or 50% truths presented as ‘The Truth’.

#cashflowmodelling #diversifaction #probabilities #risk

0 views

© 2016-2020 Fowler Drew Limited

  • Black Facebook Icon
  • Black Twitter Icon
  • Black LinkedIn Icon

Authorised and regulated by the

Financial Conduct Authority