Asset allocation and the bucket framework

Educational only. Not investment advice, not SEBI-registered research. All percentages are illustrations of a framework, not a recommendation for your situation. No allocation guarantees a return.

Asset allocation is the decision that explains roughly 90% of the variation in your portfolio's outcome. Everything else — fund selection, timing, which fund outperformed last year — is a rounding error by comparison.

And almost nobody makes it deliberately. Most people's allocation is a by-product of what they happened to buy first, in which order they got each SIP set up, and which asset was performing when they started.

The goal of this article is to make your allocation a decision rather than an accident.

The core idea: the date is the risk

The most useful principle in this entire box:

Risk is not "how much does this investment move." Risk is "how much does this investment move in the five years before I need the money."

A 60% equity portfolio is a great investment for a 25-year-old's retirement. It is a reckless one for a 29-year-old's 2-year-old child's school fees, because if equities fall 40% in that year, the fee does not wait.

The same asset carries a completely different risk depending on whose goal it is attached to. That is the whole framework. Everything below is a consequence of it.

The three buckets

Every rupee of your money belongs to a bucket, and the bucket is determined by the date it is needed.

Bucket 1 — Safety (0 to 3 years)

IN:    savings account, liquid fund, short FD, PPF (if 15y+ available)
OUT:   equity funds, long FDs, corporate bonds, anything volatile
WHY:  a 30% fall in the wrong year can permanently break a goal
      whose date does not move

Money needed within about 3 years has essentially zero tolerance for a −30% year, because there is no recovery time before the deadline.

The single most damaging mistake in Indian investing is having retirement money sitting in a mid-cap fund. The money is not going to be needed for 25 years — but the first withdrawal is. This single mismatch destroys more retirement plans in India than any market crash.

Bucket 2 — Transition (3 to 10 years)

IN:    50-70% equity funds, 30-50% debt funds / FDs
OUT:   concentrated equity, sector funds, all-in equity
WHY:  there is time to recover, but not unlimited time

Mid-point goals get a balanced allocation. This is where most of a family's money actually sits, and it is where the discipline is hardest, because the pressure to go "for growth" is strongest exactly when the date is closest.

Bucket 3 — Growth (10+ years)

IN:    70-100% equity funds, low-cost and broad
OUT:   cash, FDs, "safe" options that will not beat inflation over 25 years
WHY:  time absorbs volatility; this is where compounding actually works

This bucket is for money genuinely 10-25 years away. Here, equity risk is not a risk to be avoided — it is the reason the money grows enough to matter. A 25-year horizon in cash is a guaranteed failure.

Assigning by goal, not by preference

The most important operational step. Take every goal you have and put it in a bucket.

Your goal	Years	Bucket	Suggested equity	What it should be in
Child's school fees	4	Safety	0%	Short FD ladder, liquid fund
Child's college	9	Transition	~40%	Equity + debt mix
Car purchase	2	Safety	0%	Savings or short FD
House deposit	6	Transition	~30%	Balanced, moving to debt
Retirement at 60	25	Growth	80-90%	Equity, global if permitted
Emergency fund	0	Safety	0%	Savings, liquid fund
"Money for a house in 3 years"	3	Safety	0%	FD ladder

Then count the money in each bucket and check that it is consistent. Almost everyone discovers the problem here: the retirement bucket is well funded and the safety bucket is not, because the safety bucket is unfunded and therefore invisible.

A goal with no funded bucket is not a goal. It is a hope, and it is competing with a funded retirement for the same surplus.

The equity percentage formula — and why it is wrong as a rule

The most quoted rule in personal finance:

Equity %  =  100 − your age

It is a decent rough starting point and a bad rule. It treats a 30-year-old with ₹10 lakh and a 30-year-old with ₹10 crore as having identical risk, which is absurd — one has a time problem, the other does not.

The actual variables that should set your equity percentage are:

1. Time horizon — the dominant factor.
2. Ability to bear loss — could you sell at −40% and rebuy, or would you panic?
3. Stability of income — a salaried person with a mortgage is very different from a freelancer with a 6-month notice period.
4. Existing fixed liabilities — a home loan is a bond you issued to a bank at 8.5%. You are already levered.
5. The size of the loss you would tolerate in rupees — a concrete number, not a percentage.

A better version of the rule:

Equity % ≈ f(time horizon) × f(income stability) × f(cash needs)

And a 30-year-old in a ₹10 lakh portfolio should usually hold a higher equity percentage than a 60-year-old in a ₹10 crore portfolio with a home loan, because the smaller portfolio has more needs and less diversification. The "100 − age" rule gets this backwards.

Human capital — the part almost nobody considers

This is the most valuable idea in the article, and it is not in any calculator.

Your salary is your largest asset. A ₹15 lakh-per-year software engineer with a job security that lasts 5 years has an enormous share of their net worth already invested in the IT sector, in a global economy, in equity-like risk.

A 34-year-old IT professional, ₹18L salary, ₹20L net worth:

  Human capital (5-8 years of earnings, risk-adjusted):   ~₹60-80 lakh
  Financial capital:                                        ₹20 lakh

  → Roughly 75-80% of total net worth is already
    exposed to equity-like risk, through their job.

And it is negatively correlated with what they are doing at the desk. A software engineer's income is most at risk in precisely the scenario where their equity portfolio is worst — a global tech downturn. Their salary and their portfolio fall together, and they are correlated.

For that person, an aggressive equity allocation is a concentration bet disguised as diversification.

Situation	Implication
IT / tech employee in a global downturn-sensitive role	Reduce equity; salary is already equity-like
Government employee, pension, highly stable income	Can hold more equity
Business owner — income depends on their own market	Very high existing concentration
Freelancer with irregular income	Needs a bigger cash buffer, and less equity

Before you decide your equity percentage, ask: if my salary disappeared for two years, would I need to sell my portfolio at the worst moment? If yes, your equity allocation is already too high, and your human capital is the reason.

Sequence of returns risk — the retirement problem

This is the retirement-specific version, and it is the reason Bucket 1 matters so much.

The order of returns matters enormously for a portfolio you are drawing down, and not at all for a portfolio you are accumulating into.

Accumulating (Bucket 3): returns arrive in any order. A −40% in year 1 followed by +50% in year 2 leaves you ahead versus the reverse, but the ending balance is nearly identical. Order barely matters.

Drawdown (Bucket 1 in retirement): order dominates. Two identical sequences:

Path	Year 1	Year 2	End
A — bad first	−40%	+50%	100 → 60 → 90 (−10%)
B — good first	+50%	−40%	100 → 150 → 90 (same)

Same ending value here — but now withdraw ₹25,000 a year:

Path A:  year 1  −40%  → 60.  Withdraw 25  → 35
         year 2  +50%  → 52.5  Withdraw 25 → 27.5   ← badly eroded

Path B:  year 1  +50%  → 150. Withdraw 25 → 125
         year 2  −40%  → 75.  Withdraw 25 → 50     ← much better

The ending balance is nearly identical. The portfolio that survived is not. The early withdrawals in Path A come out of a smaller base and get permanently compounded down.

This is the entire argument for the bucket framework. The growth bucket can be 100% equity because it never has to sell. The safety bucket must be conservative because it sells in the first bad year, and the first bad year might be year one.

The safe withdrawal question for India

How much can you take out of a retirement corpus each year? The 4% rule is a well-known heuristic from US data, and applying it in India needs two adjustments.

4% rule, US, 30-year horizon.

India adjustment 1 — inflation:
  US historical inflation ~3%.  India ~6%.
  A ₹1 crore corpus in India must fund the same lifestyle
  ~18 years sooner than in the US, because prices rise faster.
  → the real number is lower, because you are draining the
    real corpus faster.

India adjustment 2 — returns:
  Indian long-run equity returns have been strong, but the
  30-year forward return is a forecast, not a fact.
  → assume less, and hold more cash.

Realistic planning range:
  3.0-3.5% of the corpus in the first 10-15 years,
  then reassess, with a floor you never breach.

The most important structural decision is this: the first decade or two of retirement should draw from Bucket 1 (safe) plus Bucket 2 (transition), so that Bucket 3 is never sold in a crash. The growth bucket is for the back half of retirement, when the date is far enough away to survive a drawdown.

This is the whole point of the framework: align the selling with the date, not with the market.

Rebalancing — the discipline that requires no skill

Over time, assets drift. Equity that was 70% becomes 80% or 60% without you doing anything.

Rebalancing is selling a little of what went up to buy what went down — mechanically, without a view. It is the one investment activity that requires no market knowledge and is guaranteed to look wrong in the moment, because you are always buying the thing that just fell.

Allocation:  70% equity / 30% debt
After 3 good equity years: 82% equity / 18% debt

Rebalance:  sell equity down to 70%, buy debt back to 30%

Two rules make this bearable:

1. Do it on a schedule — annually, or when a bucket drifts more than 5 percentage points. Not daily.
2. Rebalance with new money where possible. Direct the next SIP instalment to the underweight asset instead of selling. This removes the tax event entirely, which in India makes rebalancing much cheaper.

The behavioural point: rebalancing forces you to sell the winner and buy the loser, which feels wrong every single time. It is correct every single time, and it is one of the few places in investing where the emotionally uncomfortable action is also the arithmetically correct one.

When to change your allocation

Trigger	Change
Every goal moves 1 year closer	Re-check buckets — the math shifts
A bucket's date arrives	Move it to the next bucket down, on schedule
Income rises materially	Revisit the equity percentage
Home loan taken or paid off	Leverage changed; reassess
Job changes	Human capital changed; reassess
Allocation drifts more than 5pp	Rebalance
A large expense appears	Fund it from a bucket, do not raid equity
Every year	Re-run the goal maths with current numbers

The important discipline: change the allocation on a schedule, not on a feeling. Adjusting equity because the market has been falling is the exact error that makes people sell low.

Failure modes

One portfolio for everything. A single allocation for a 2-year goal and a 25-year retirement. The 2-year goal then has 25-year risk, and the retirement has 2-year caution. This is the most common structural error, and it is usually invisible because it is just one set of holdings.

Treating retirement as Bucket 3 forever. The growth bucket is right for the accumulation. The first withdrawal decade needs a different structure, or sequence-of-returns risk will do the damage.

Ignoring human capital. A tech employee's portfolio is not diversified if their salary depends on the same sector. This is a real, underappreciated concentration.

All equity, always, "because I am young." Being young says you have time. It does not say you have no near-term goals, no dependents, and no salary exposure to the same market.

Rebalancing by selling every quarter. Each sale is a tax event in India. Annual or threshold-based, using new money, is better.

Rebalancing by market view. "I'll rebalance when the market recovers." This is not a rebalancing policy, it is a hope, and it usually means never rebalancing.

Changing the allocation after a big fall. Selling equity after a −40% year is the single most reliable way to convert a paper loss into a real loss.

Allocating to a "safe" asset that is not. A corporate bond fund is not a bank deposit. A 15-year PPF is not liquid. See Where to park cash.

Treating 100 − age as a decision. It is a starting point, and it ignores horizon, income stability, liabilities, and existing holdings.

Forgetting the last-mile problem. A corpus that works as a withdrawal rate may not work as a lump sum. A 4-year degree needs a defined schedule, not an annual withdrawal. See Goal-based corpus maths.

A safety bucket that is empty. The most damaging version, because it is invisible. "I will deal with it when I need it" means selling equity in the worst year.

Never rebalancing at all. Drift eventually produces an allocation nobody chose, often more concentrated than intended.

Exercise

Build the real allocation for your household. This takes an hour and is the highest-value hour in this curriculum.

1. List every goal with an amount in today's rupees and a year.
2. Assign each to a bucket — Safety (0-3y), Transition (3-10y), Growth (10y+).
3. Count the money in each bucket. Compare with the amount each goal needs.
4. Compute the equity % in each bucket versus what that horizon allows.
5. Find any goal in a bucket it does not belong in — especially a near-term goal sitting in equity.
6. Assess your human capital: what fraction of your net worth depends on one industry or employer?
7. If that is high, reduce equity and re-run step 5.
8. For the retirement bucket: what is the first withdrawal year, and what is in Bucket 1 to fund it?
9. Set a rebalancing rule: annually, or on a 5pp drift, preferably with new money.
10. Do the whole thing again next year and compare. Goals move, and the allocation must move with them.

Step 6 is the one that changes the most people's answers, and it is the one nobody does.

Checklist

1. Does every rupee I have belong to a specific goal in a specific bucket?
2. Is any goal within 3 years sitting in equity? Move it.
3. Is my retirement money structured for the first withdrawal decade, not just the accumulation?
4. Have I assessed my human capital, and reduced equity if my salary is already equity-like?
5. Do I know my equity %, and can I defend the number without referring to my age?
6. Do I have a Bucket 1 that can fund 2-3 years of withdrawals?
7. Have I set a rebalancing rule, with a date?
8. Am I changing the allocation on a schedule, or on a feeling after a fall?
9. Are my "safe" assets actually safe, and actually liquid?
10. Is every near-term goal fully funded in its correct bucket?

Educational only. Not investment, tax, or legal advice. Not SEBI-registered research. The allocations, ratios, and withdrawal rates shown are illustrations of a general framework and are not recommendations for any individual. Asset allocation cannot guarantee a return or eliminate loss, and past performance does not indicate future results. The safe withdrawal rate depends on inflation, returns, taxes, and longevity, none of which are predictable. Mutual fund investments are subject to market risks; read all scheme-related documents carefully. Consider your own objectives, risk tolerance, and horizon, and consult a SEBI-registered investment adviser before investing.