Surely only properly informed advice can lead to the client's properly informed consent to the risks in their financial plan. More specifically, a flawed understanding of, and consequent inaccurate explanation to a client of sequence risk may result in them believing they can de-risk their portfolio by reducing their equity exposures. Over the last 40 years, as the numbers show, such a strategy would have resulted in sub-optimal portfolio outcomes. We would have diminished retirees' lifestyles for no logical reason.
The past has an annoying habit of repeating itself, particularly in relation to investments. That's why we argue the case that each and every professional adviser has to individually come to an informed view on sequence risk.
But first we must thank Abraham Okusanya for his detailed response to our first article. Abraham is vitally engaged in what we argue is an important issue for advisers as they develop suitable investment strategies for the de-accumulation market. The planning profession needs more like him.
Retirement portfolio planning is perhaps the last great frontier for real advisers.
We bring no portfolio solution biases to this discussion. We were surprised when we first saw the sequence risk numbers a year ago. We have spent a goodly amount of time and energy trying to disprove them before publishing. We are just helping the numbers to talk, showing them the light of day.
Where we are agreed with Abraham: At the end of 10 years the historical data balances for all five positions from best to worst showed that the 80% growth asset exposed portfolio were either better or no worse than the 40% growth asset portfolio. This applies across all three withdrawal rates: £3,000 real per annum, £5,000 real per annum and £7,000 per annum.
Why is this so? (Here we are pretty sure Abraham does not agree with our explanation). We said portfolios exposed to higher levels of growth assets benefit more in good times, and recovered more rigorously after every correction in the last 40 years. There is nothing new in that. Similar patterns emerged, as we showed with the data in our original article, in both the US and Australia. What drives the outcomes, in our view, is simply long-term equity risk premium.
1) We think Abraham has mistaken sequence risk for capacity for loss. The larger the amount withdrawn, the quicker the portfolio runs out of money when markets turn south. This is true for all portfolios, including the examples we share in the original article (here).
The outcomes are no worse in the 80% exposed portfolio than the 40%. Very similar balances in the poor and worst cases and better in the average, good and best cases.
While sequence risk clearly gives a wide range of outcomes from one portfolio, when there are two portfolio outcomes compared, the critical driver is actually the amounts withdrawn. This converts the analysis into a generic capacity for loss discussion. Once again there is little surprising in this. This has nothing to do with the sequence of investment returns and everything to do with spending too fast.
2) Abraham argues that a longer data history than our 40 plus years would be more representative of what will happen in the future. There is an ongoing debate on whether longer time period are more accurate or not. From our perspective, we know that the accumulation indexes we used are reliable and reflect recent political engagements in the market. Earlier data may be less reliable and a product of non-repeating economic and political forces.
3) Abraham argues that balances of accounts would reveal more if projected out to 30 years rather than the 10 we used. We chose 10 years because it represents a reasonable period to stay with an asset allocation. It also clearly shows the range of investment outcomes.
The selection of a longer period would show a nil balance for poor and worst outcomes which would be unhelpful in explaining comparative portfolio behaviour. Average, good and best returns which look pretty good at 10 years would look outstanding over 30 years. Our concern has always been in preparing clients for the worst, not giving them false hopes about the best. We have provided advisers with detailed portfolio histories to help better frame clients' investment expectations for more than 12 years.
4) We are not sure we understand Abraham's point in relation to portfolio longevity. Portfolio 40 according to our numbers will last on average 31.8 years. This is not significantly different from Wade Pfau's claim that: 'at 5% withdrawal rate, a portfolio with similar asset allocation will last less than 30 years in 55% of the dataset - less than half will last 30 years.' This may be a coincidence. It certainly deserves further work.
5) Abraham argues that we need to look at overall portfolio volatility not just the part that is driven by growth asset exposures. We believe we do meet his objection because as we increase portfolio equity exposures, we decrease defensive asset exposures. Our analysis is at the portfolio level.
We do agree with Abraham that the management of sequence risk is an important issue that needs to be aired and shared. Retirees need to be able to give their properly informed commitment to the advice they are given. It would be more than a shame if that consent was based on improperly informed advice.
Paul Resnik is director and co-founder and Peter Worcester is a consulting actuary at FinaMetrica.