Drawdown strategies determine how income is taken from retirement assets over time. Many approaches exist, each with different implications for income stability, tax patterns, and long-term sustainability. Modelling does not predict the future; instead, it illustrates how a strategy behaves under uncertainty.
Several withdrawal rules are widely discussed in retirement planning. They differ in how they balance income stability, sustainability, and responsiveness to market conditions:
• Fixed-amount withdrawals. A constant nominal amount is withdrawn each year. Income is stable, but purchasing power may fall, and the fund may deplete unevenly depending on market returns.
• Inflation-adjusted withdrawals. Withdrawals rise with inflation to maintain purchasing power. This can accelerate depletion if returns are weak.
• Percentage-of-pot withdrawals. A fixed percentage of the current fund value is withdrawn annually. This avoids full depletion but produces income that varies with market performance.
• Guardrail-based withdrawals. Withdrawals adjust when the fund value moves outside predefined bands. This aims to stabilise income while reacting to market conditions.
These approaches all share a common feature: they treat the pension pot as the central resource, and the withdrawal rule is defined primarily in terms of the pot itself.
Although widely used, pot centric rules have several shortcomings when viewed within a holistic financial model:
• They ignore the rest of the household balance sheet. Real retirement finances include ISAs, taxable investments, cash reserves, and guaranteed income streams. A pot centric rule cannot coordinate withdrawals across these assets.
• They determine spending indirectly. Spending becomes whatever the withdrawal rule produces, rather than being planned explicitly.
• They can create uneven tax patterns. Some rules generate spikes or troughs in taxable income, which may be undesirable.
• They may withdraw when withdrawal is unnecessary. If other income sources already meet spending needs, a pot centric rule may still trigger withdrawals.
These limitations motivate an alternative approach: spending-first drawdown, where withdrawals are driven by planned spending rather than by the pot itself.
EvolveMyRetirement uses a drawdown method designed to integrate naturally with its broader modelling philosophy. The strategy is not chosen from a menu; it is a consequence of how the model represents retirement finances.
Each member has a target for the total taxable retirement income. In any post retirement year:
• If taxable income from other sources is below the target, drawdown is made to top it up (subject to available funds).
• If taxable income is at or above the target, no drawdown is made.
This smooths taxable income across retirement and avoids unnecessary withdrawals.
Non pension assets are used only to meet cashflow shortfalls:
• Taxable investments first.
• ISAs next.
This ordering reflects tax considerations while ensuring withdrawals occur only when needed to support cashflow shortfalls.
The model’s objective is sustainable spending, not optimisation of pension withdrawals in isolation. The drawdown rule is therefore designed to:
• Prioritise spending. The target taxable income is determined based on the planned spending budget.
• Provide flexibility. Drawdown responds to actual income and cashflow needs rather than following a rigid pot-based rule.
• Smooth taxable income. The target taxable income helps avoid large fluctuations in tax liability.
• Coordinate all assets. The strategy considers the entire financial position, not just the pension pot.
This contrasts with pot-centric rules, which define withdrawals independently of spending and independently of other assets.
Although EvolveMyRetirement does not apply guardrails to drawdown, it includes an optional guardrail-style mechanism for discretionary spending. Each year, net worth is compared with what it would have been in the average case, and discretionary spending is flexed up or down accordingly. This models adaptive behaviour without altering the drawdown rule itself.
Monte Carlo simulation is used to explore how the drawdown rule behaves under uncertainty. Percentiles indicate likelihood within the distribution of simulated outcomes. Results depend on the assumptions entered, and real-world outcomes may differ from those explored in the model.
