Retirement is three problems, not one

The three phases

A retirement plan passes through three distinct periods, and each rewards different behaviour.

Most planning content addresses only the first. The second is barely discussed and the third is treated as an afterthought — which is backwards, because a plan can survive a mediocre accumulation phase and cannot survive a badly handled distribution one.

Accumulation: the boring phase that does the work

Long horizon, regular contributions, and the arithmetic of compounding doing most of the lifting. The behaviour required is unglamorous: contribute, raise the contribution with income, and do not interrupt.

Two things dominate outcomes here, and neither is fund selection.

Starting. Because value depends on the number of doubling periods, a rupee invested early is worth a multiple of the same rupee later. Someone starting at thirty rather than twenty-two has given up two doublings, and no amount of later contribution fully substitutes for that.

Not interrupting. Every exit during a fall converts a temporary drawdown into a permanent loss and forfeits the recovery. This is the behaviour gap, and over a thirty-year accumulation it costs more than any plausible difference between good funds.

The unavoidable structural point: for a horizon this long, an allocation that cannot outpace inflation is not the cautious choice. It is a different way of failing, and it is a quiet one.

Preservation: the phase nobody plans for

The decade before retirement is genuinely different, and the reason is arithmetic rather than temperament.

A 40% fall at thirty-five is unpleasant and has twenty-five years to recover. The same fall two years before retirement is not the same event — there is no time, the corpus is at its largest so the rupee loss is at its largest, and withdrawals are about to begin against a diminished base.

This is the point of maximum vulnerability in the entire plan, and it arrives exactly when most people stop paying attention because the target looks close.

The conventional response is to shift gradually towards stability as the date approaches — not all at once, which sacrifices the growth still needed, but as a glide over years. What matters is that it is deliberate and scheduled rather than reactive, because a reactive shift happens after the fall rather than before it.

Distribution: the rules invert

On the day income stops, three things change at once.

Volatility stops being noise and becomes damage. While accumulating, a fall is an opportunity — your contributions buy more units. While withdrawing, a fall forces you to sell more units to raise the same rupees, permanently shrinking the base. Sequence-of-returns risk exists only in this phase.

Inflation stops being an abstraction. A thirty-year retirement means the income you need at the end is far larger than the one you start with. A plan that funds today's expenses and stops there has funded roughly the first third of the problem.

You cannot wait it out. The accumulator's answer to a bad market — do nothing, keep contributing — is unavailable, because the withdrawals do not pause.

Which is why the distribution phase needs its own design rather than an assumption that the accumulation portfolio will simply keep working.

Where plans actually fail

1. Starting late. The largest and least recoverable error, because the missing input is time and it cannot be bought.
2. Planning in today's rupees. A target computed at current prices understates the requirement badly over a multi-decade horizon.
3. Too cautious for too long. Decades in instruments that barely match inflation is a slow, near-certain shortfall that feels prudent throughout.
4. Ignoring the preservation phase. A large drawdown immediately before retirement, with no glide path in place.
5. No plan for withdrawal. Arriving at retirement with a corpus and no method for turning it into income.
6. Underestimating healthcare. It rises faster than general inflation and lands hardest late, which is exactly when the corpus is most depleted.
7. Counting the family home. A residence you live in funds nothing — the distinction between net worth and investable net worth.
8. Retiring the plan when the person retires. A thirty-year distribution phase needs reviewing, not setting once.

What is specific to India

Three structural features worth planning around rather than assuming away.

Retirement is largely self-funded. There is no broad state pension for most workers, so the corpus carries essentially the whole load — which makes the required multiple of expenses larger than in the countries most retirement content is written about.

Healthcare is bought privately. Cover becomes harder and more expensive to obtain with age and with any diagnosis, which is the argument for holding it continuously from well before retirement rather than acquiring it at the point of need.

Family support runs both ways and is changing. Plans that quietly assume children will provide are planning on someone else's finances, and increasingly on a different household structure. Worth making explicit rather than assumed.

The specific instruments — EPF, PPF, NPS and the rest — carry rules and tax treatment set by statute and revised periodically. Each deserves its own treatment against verified current figures rather than a summary here.

Where to start

1. Estimate annual expenses in retirement, not today's. Some fall; healthcare rises; the total is usually not much lower.
2. Inflate to your retirement date, and remember they keep rising after it.
3. Choose a withdrawal rate you can defend for the length of your retirement.
4. Compute the corpus, subtract other income sources, and work back to a monthly contribution.
5. Write down the glide path — when you will begin shifting allocation, and to what.
6. Decide the withdrawal method before you need it, not in the first month of retirement.

Steps five and six are the ones almost always missing, and they are the two that govern the phase where plans actually break.

Testing a plan against bad sequences

A retirement plan tested against an average return has not been tested, because the distribution phase is governed by the order of returns rather than their mean.

FNOTrader's Mutual Funds app computes rolling returns across every start date in the full AMFI history — around 34 million NAV rows — alongside maximum drawdown. Read the worst window rather than the median, and check what a drawdown of that size would have done arriving in the two years before your intended retirement date. That single test finds more broken plans than any amount of corpus arithmetic.

FNOTrader is not a SEBI-registered investment adviser and this is not retirement advice.

Common questions

Continue reading

More in Retirement Planning · App: Mutual Funds · Definitions: glossary · Free tools: calculators · All: every article