Every batch I teach in Ahmedabad, the same thing happens in the first week. A student, usually someone already working as an MFD or a bank RM, asks me some version of: “Sir, which is the best fund right now?” I ask them back: “Best for whom? For what goal? Held in what proportion, next to what else?”
That exchange is the whole subject of this article. The industry spends enormous energy on fund selection: star ratings, past-return leaderboards and the annual crop of “top 5 funds for 2026” videos. The decision that actually moves the outcome, how much of the money sits in equity versus debt versus gold versus cash, gets decided casually, sometimes by accident, more often by whatever was sold last.
I want to show you, with arithmetic you can reproduce in a spreadsheet, why the mix decides more than the picks. Then we will go through four practitioner-grade ideas that sit on top of that fact: strategic versus tactical allocation, the mechanics of rebalancing, glide paths for goals, and allocation as a behavioural tool. This is written for budding planners and serious DIY investors alike. Nothing here needs more than school maths.
One housekeeping note before we start. Every rupee figure and return assumption in this article is an illustration, constructed so you can check the working. Real markets will not oblige with smooth 12 percent years, but the logic holds up when they misbehave, which is why I am comfortable teaching it.
The arithmetic that settles the argument
Let us compare the two decisions head-on: the allocation decision and the selection decision, on the same portfolio.
Assumptions (illustrative only): equity averages 12 percent a year, debt averages 7 percent a year, over a 15-year horizon, on a lump sum of ₹50 lakh. No taxes or costs for now; we will add those later.
Decision A: the allocation call. Investor One holds 70:30 equity-to-debt. Investor Two holds 30:70. Same funds, same discipline, different mix.
The blended expected return is just the weighted average:
- 70:30 mix: (0.70 × 12%) + (0.30 × 7%) = 8.4% + 2.1% = 10.5% a year
- 30:70 mix: (0.30 × 12%) + (0.70 × 7%) = 3.6% + 4.9% = 8.5% a year
Compound each over 15 years using FV = amount × (1 + r)^n:
- At 10.5%: ₹50,00,000 × (1.105)^15 = ₹50,00,000 × 4.472 = ₹2.24 crore
- At 8.5%: ₹50,00,000 × (1.085)^15 = ₹50,00,000 × 3.400 = ₹1.70 crore
The allocation decision alone is worth about ₹54 lakh on this portfolio.
Decision B: the selection call. Now hold the allocation fixed at 70:30 and suppose our investor is genuinely skilled (or lucky) at fund selection, and picks an equity fund that beats the market assumption by a full 1 percent every single year for 15 years, so 13 percent instead of 12.
- Blended return: (0.70 × 13%) + (0.30 × 7%) = 9.1% + 2.1% = 11.2% a year
- FV: ₹50,00,000 × (1.112)^15 = ₹50,00,000 × 4.917 = ₹2.46 crore
That is worth about ₹22 lakh over the base case. Real money, no question. But set the two decisions side by side before you decide where your evenings go:
| Decision | What you control | Swing in outcome (this example) |
|---|---|---|
| Allocation: 70:30 vs 30:70 | Entirely yours, decided on day one | About ₹54 lakh |
| Selection: fund beats peers by 1% every year for 15 years | Mostly not yours; persistent outperformance is rare and unknowable in advance | About ₹22 lakh, and it can just as easily be minus ₹22 lakh |
The allocation edge is certain to apply, in whichever direction you set it. The selection edge is a hope. You cannot know in advance which fund will lead its category for the next 15 years, and the fund you pick after studying past returns can lag by 1 percent just as easily as it leads by 1 percent. That puts the outcome swing of selection at roughly ₹22 lakh either way, around a mix you chose. The mix is the steering wheel; the fund pick is which petrol pump you stop at.
If you have heard advisors quote a famous “asset allocation explains 90 percent of returns” line from American pension-fund research (the studies by Gary Brinson and colleagues, later revisited by Roger Ibbotson and Paul Kaplan), be careful with it. I am deliberately not quoting the percentages here, partly because I want you to trust arithmetic you can reproduce rather than a slogan, and partly because that research is almost always misquoted: it measured how much of a portfolio’s movement over time is explained by its policy mix, which is a different question from how much better one investor’s return will be than another’s. The defensible claim is the narrower one the worked example above demonstrates. The mix you choose sets the neighbourhood of your outcome, and fund selection shuffles you around within that neighbourhood. Choose the neighbourhood first.
None of this means fund selection deserves zero effort. It means it deserves effort in proportion, and of the right kind: screening for consistency and risk behaviour rather than chasing last year’s topper. We have covered that discipline separately in how CFPs evaluate mutual funds, so I will not repeat it here.
Strategic allocation: the mix is derived, not declared
This is where a planner parts company with a product-seller. A planner does not pull “60:40” out of the air or out of a risk-profiling quiz alone. The strategic (long-term policy) allocation is derived from the goal’s arithmetic, then adjusted for the human being who has to live with it.
The derivation runs backwards from the goal. Work through one.
Goal (illustrative): ₹1 crore in 12 years for a daughter’s postgraduate education. The family can invest ₹30,000 a month.
Use the SIP future value formula, FV = P × [((1 + i)^n − 1) / i], where P is the monthly instalment, i the monthly return, n the number of months (144 here). We have unpacked this formula step by step in SIP maths every advisor should know, so here I will just run the numbers at a few candidate return rates:
- At 8% a year (i = 0.00667): FV = 30,000 × [((1.00667)^144 − 1) / 0.00667] = 30,000 × 240.5 = ₹72.2 lakh
- At 10% a year (i = 0.00833): FV = 30,000 × 276.4 = ₹82.9 lakh
- At 12% a year (i = 0.01): FV = 30,000 × 319.1 = ₹95.7 lakh
- At 13% a year (i = 0.01083): FV = 30,000 × 343.3 = ₹1.03 crore
Read what the numbers are saying: to reach ₹1 crore, this family needs roughly 13 percent a year for 12 years. Under our illustrative assumptions (equity 12 percent, debt 7 percent), even a 100 percent equity portfolio is not expected to get there, and 100 percent equity for a non-negotiable education goal is a mix almost no family can actually hold through a bad market.
This is the moment where allocation thinking earns its keep. The amateur response to “the numbers do not reach” is to hunt for a hotter fund. The planner’s response is to change the inputs, because the inputs are the only things under anyone’s control:
- Raise the instalment. At a 70:30 mix earning a blended 10.5 percent (i = 0.00875), the annuity factor over 144 months works out to 286.4, so ₹30,000 a month reaches 30,000 × 286.4 = ₹85.9 lakh, roughly ₹14 lakh short. Turn the formula around and the instalment the goal actually needs is ₹1,00,00,000 ÷ 286.4 = about ₹35,000 a month. Five thousand rupees more per month closes a gap that no realistic fund selection can close.
- Or extend the horizon. At ₹30,000 a month and 10.5 percent blended, the corpus crosses ₹1 crore at roughly 13 years instead of 12. Sometimes a goal can slide a year. The conversation is worth having.
- Or right-size the goal, perhaps ₹85 lakh plus an education loan for the balance, deliberately chosen rather than discovered in year eleven.
That is strategic allocation in practice: required return, tested against risk capacity, produces the policy mix, and when the two of them conflict, you adjust the plan rather than the return assumption.
One more distinction worth carrying into every client conversation: risk capacity versus risk willingness. Capacity is arithmetic: horizon length, income stability, existing liabilities, how catastrophic a shortfall would be. Willingness is temperament: how the person actually behaves when the portfolio is down 25 percent. A young professional may have high capacity and low willingness. A retired businessman may have high willingness and low capacity. The policy mix should respect the lower of the two, and I will show you the rupee cost of ignoring that rule in the behaviour section below.
Tactical allocation: the spice, not the sabzi
Tactical allocation means deliberately deviating from the policy mix for a while because you believe markets are unusually cheap or expensive: trimming equity from 65 to 55 when valuations look stretched, or adding when there is fear in the air.
My view, as someone who runs a distribution practice and teaches this material: tactical calls are the most overrated activity in Indian retail investing, and the policy mix is the most underrated. Getting a tactical call right requires being right twice, once on the exit and once on the re-entry, and the second call is the one everyone botches, because the moment that rewards re-entry feels exactly like the moment to stay away. So my working rules for students are these.
- The strategic mix does the heavy lifting. If tactical tilts are the sabzi and not the spice in a portfolio, something has gone wrong.
- Bound the tilts in advance. A written band, for example “equity stays within 10 percentage points of the 60 percent policy weight, whatever my view,” converts tactical allocation from speculation into a controlled activity. Price that band in rupees and it stops being a slogan: on a ₹40 lakh portfolio, a 50-to-70 percent band means equity may sit between ₹20 lakh and ₹28 lakh, and nothing I happen to believe that week can push it past either number. The band is decided on a calm day, which is precisely why it works on the panicky one.
- If a client wants tactical management, consider outsourcing it. Balanced advantage and multi-asset funds exist to move the equity level by model or mandate inside the fund, with no action, and no capital gains event, at the investor’s level. Whether a specific fund does this well is a selection question. That the category structure exists is an allocation option worth knowing.
Notice that even here the practitioner’s tools are allocation tools, policy weights and written bands and review dates, rather than predictions about where the Nifty goes next.
Rebalancing: the only free discipline in investing
Set a mix and leave it alone, and the market will quietly change it for you. That drift is what rebalancing corrects, and the arithmetic deserves to be seen once in full.
Setup (illustrative): ₹40 lakh portfolio at a 60:40 policy mix, so ₹24 lakh equity, ₹16 lakh debt. Suppose a good year: equity rises 25 percent, debt earns 7 percent.
- Equity: ₹24,00,000 × 1.25 = ₹30,00,000
- Debt: ₹16,00,000 × 1.07 = ₹17,12,000
- Total: ₹47,12,000; equity weight is now 30,00,000 ÷ 47,12,000 = 63.7 percent
Nobody did anything, yet the portfolio is riskier than the one that was designed. To restore 60:40:
- Target equity = 0.60 × 47,12,000 = ₹28,27,200
- Sell ₹1,72,800 of equity (30,00,000 − 28,27,200) and move it to debt, which takes debt to 17,12,000 + 1,72,800 = ₹18,84,800
Now watch what that one mechanical act does in the following year, under two illustrative paths:
| Next year’s market | Rebalanced portfolio | Drifted (no rebalancing) | Difference |
|---|---|---|---|
| Equity falls 20%, debt earns 7% | 28,27,200 × 0.80 + 18,84,800 × 1.07 = ₹42.79 lakh | 30,00,000 × 0.80 + 17,12,000 × 1.07 = ₹42.32 lakh | Rebalancing ahead by about ₹47,000, with less risk carried |
| Equity rises 20%, debt earns 7% | 28,27,200 × 1.20 + 18,84,800 × 1.07 = ₹54.09 lakh | 30,00,000 × 1.20 + 17,12,000 × 1.07 = ₹54.32 lakh | Drift ahead by about ₹23,000, with more risk carried |
Be straight with clients about what this table shows, because most sales pitches are not. Rebalancing is not primarily a return booster. In a long one-way bull run, the drifted portfolio will beat it, as the second row shows. What rebalancing does reliably is force you to sell a little of what has run up and buy what has lagged, on a schedule, with no forecast required, and keep the portfolio’s risk equal to the risk that was actually designed for the goal. It wins money when markets mean-revert and it earns its keep every single year in risk control. Discipline you can write down and be held to is rare in this business, and that, more than the ₹47,000, is what the client is really buying.
Three practitioner refinements follow.
When to act. Rebalancing every month is churn; never rebalancing is drift. A common practitioner approach, and the one I use, is a band rule: act only when an asset class drifts a set distance from target, for example 5 percentage points (so a 60 percent equity target is left alone between 55 and 65), checked at a fixed review date. The 63.7 percent above breaches a 5-point band, so the ₹1,72,800 sale goes through. A 61 percent reading would not, and the correct action there would be nothing at all.
Rebalance with fresh money first. Selling equity in a taxable portfolio can trigger capital gains tax, which is a real cost of the discipline. Before selling anything, redirect new flows: in the example above, the ₹1,72,800 shift could instead be achieved by pointing the family’s ₹40,000 monthly SIP entirely at debt, since 1,72,800 ÷ 40,000 is a little over four instalments, so four to five months of redirected SIP does the same job (approximately, since prices keep moving). Slower, but tax-free. We have worked through the current capital gains rules and the actual tax arithmetic on equity fund sales separately in LTCG tax on mutual funds, with worked examples; read that before you rebalance by selling, because the tax cost belongs in the decision.
Use the tax-sheltered corners of the balance sheet. Rebalancing inside NPS, or letting EPF and PPF stand as the stable debt-like layer of the household mix while equity mutual funds carry the growth layer, moves the rebalancing burden to places where switches are not taxable events at the investor level. The comparative mechanics of those three vehicles are in EPF vs NPS vs PPF: the maths. The household’s allocation is the allocation across everything, not just the mutual fund folio, and planners who forget the EPF balance routinely run families more conservatively than anyone realises.
Glide paths: the allocation is a function of time, not a constant
A policy mix that is right at fifteen years from a goal is wrong at two years from it. The structured answer is a glide path: a pre-agreed schedule by which the mix de-risks as the goal approaches.
Why it matters is best seen in rupees. Take that education goal, now grown to an illustrative ₹80 lakh corpus with one year to go, and suppose equity falls 30 percent that final year.
- Still at 60 percent equity: loss = 0.60 × 80,00,000 × 0.30 = ₹14.4 lakh, with no time left to recover
- Glided down to 20 percent equity: loss = 0.20 × 80,00,000 × 0.30 = ₹4.8 lakh
The glide path was worth ₹9.6 lakh in that scenario, and it required no forecasting at all, only a calendar. This is the same logic as sequence-of-returns risk in retirement: once withdrawals begin or a deadline arrives, the order of returns matters as much as the average, a point we develop fully in the retirement corpus maths planners actually use.
A template I give students as a starting point for negotiable-deadline goals (illustrative, to be adapted to the client, never applied blindly):
| Years to goal | Equity band |
|---|---|
| More than 10 | 65 to 75% |
| 7 to 10 | 55 to 65% |
| 4 to 7 | 35 to 50% |
| 2 to 4 | 15 to 30% |
| Under 2 | 0 to 10% |
Run the education goal through it and the table stops being abstract. Twelve years out, that family sits in the top row at 65 to 75 percent equity, which is roughly the 70:30 mix the ₹35,000 instalment was priced on. Eight years later, with four years to go, the same plan belongs in the 35 to 50 percent row, and in the final year under 10 percent, which is precisely the difference between losing ₹14.4 lakh and losing ₹4.8 lakh in the crash above.
Two professional notes on using it. First, the glide path is written into the plan on day one, with approximate calendar dates, precisely so that the de-risking sale does not depend on how markets feel that year. If the step-down year arrives with equity down, you step down anyway, because the alternative is doubling the bet with the goal money. Second, glide paths already exist inside products: NPS’s auto choice option trims a subscriber’s equity exposure automatically with age along a preset schedule, and target-style hybrid structures do versions of the same. A planner should know when to build the glide path by hand and when a product’s built-in one is good enough for the client in front of them.
Allocation is also a behavioural tool, and here is its price tag
Everything so far treats allocation as arithmetic. Its second job is quieter. The mix is the main thing standing between an investor and their own worst instincts. This is where my two hats, distributor and teacher, see the same thing from both sides: in a falling market, no fund factsheet has ever kept a client invested, but a mix they were genuinely comfortable with often has.
Put a rupee figure on it. Two investors, ₹50 lakh each, same illustrative market path: equity falls 30 percent in year one, then rises 20 percent in each of the next two years; debt earns 7 percent throughout.
Investor A holds 60:40, correctly sized to her temperament, and does nothing:
- Equity: 30,00,000 → 21,00,000 → 25,20,000 → ₹30,24,000
- Debt: 20,00,000 → 21,40,000 → 22,89,800 → ₹24,50,100
- Total after three years: about ₹54.7 lakh
Investor B went 100 percent equity because a higher expected return “made sense,” panicked at the bottom of year one, and moved everything to debt:
- Equity: 50,00,000 → ₹35,00,000, then switched to debt
- Debt: 35,00,000 × 1.07 × 1.07 = about ₹40.1 lakh
The gap is roughly ₹14.6 lakh, and notice what caused it. Not fund quality: Investor B may well have owned the “better” fund. The loss came from an allocation mis-sized to the human holding it, which converted a temporary decline into a permanent one. On paper B’s mix had the higher expected return, and in a spreadsheet B wins comfortably. Nobody holds a portfolio in a spreadsheet, though. This is exactly why FPSB now treats the psychology of financial planning as core curriculum rather than soft-skills garnish. We have written about that module in detail in Psychology in Financial Planning: the CFP behavioural module, explained.
The practical rule I teach: the best allocation is the most aggressive one the client will actually hold through a bad year, and the way to find it is not a quiz score alone but rupee conversations. “If this ₹50 lakh reads ₹35 lakh on your screen for eighteen months, what will you do?” is worth more than any risk-profiling questionnaire. That ₹35 lakh is not a scary number I invented for effect; it is exactly what an all-equity ₹50 lakh showed at the bottom of year one in the working above. Price the mix in the currency the client will actually feel, and you find the real constraint before the market does.
A planner’s allocation workflow, start to finish
The four ideas fit into one repeatable sequence, the kind you should be able to run for any client, or for yourself:
- State the goal in rupees and years. No mix can be judged without both.
- Compute the required return from the goal amount, horizon and investable surplus, as in the SIP working above.
- Assess risk capacity (arithmetic: horizon, income stability, liabilities, cost of shortfall) and risk willingness (temperament, tested in rupee terms). The binding constraint is the lower one.
- Set the strategic mix where required return and the binding risk constraint overlap. If they do not overlap, change the instalment, the horizon or the goal, and say so plainly.
- Write the policy down: target weights, rebalancing bands, review dates, the glide path calendar. One page. In practice this is a simple Investment Policy Statement, and the act of writing it is what makes rule 6 possible.
- Rebalance by the written rule, fresh flows first, tax-sheltered accounts next, taxable sales last, with the tax cost computed before the trade.
- Step down the glide path on schedule, regardless of how the market feels that year.
- Revisit the strategic mix only when life changes: a new goal, a windfall, a job loss, a marriage. Not when markets change.
Notice how little of that list is about funds. That is deliberate, and it is more or less what the profession consists of.
If working through this article felt like a different way of thinking than the fund-picking content you usually see, that is essentially the difference between selling products and planning. Topics like these, from policy allocation to glide path design, are covered in depth in the CFP certification programme.
Frequently asked questions
Is asset allocation really more important than choosing the best mutual fund?
For most investors, yes, and the arithmetic in this article shows why: in the worked example, the choice between a 70:30 and a 30:70 equity-debt mix swung the 15-year outcome by about ₹54 lakh, while even an unusually good fund-selection edge of 1 percent a year swung it by about ₹22 lakh. The allocation decision is fully in your control and certain to apply, while persistent fund outperformance cannot be identified in advance. Choose the mix first, then select funds carefully within it.
How often should I rebalance my portfolio?
A calendar-plus-band approach works well in practice: review on a fixed date, say once or twice a year, but act only if an asset class has drifted beyond a pre-set band such as 5 percentage points from its target weight. This avoids both constant churn and unlimited drift. Prefer rebalancing with fresh investments or inside tax-sheltered accounts like NPS before selling taxable equity, because capital gains tax is a real cost of the discipline.
Does rebalancing increase my returns?
Not reliably, and an honest advisor should say so. Rebalancing tends to add value when markets swing and mean-revert, because it mechanically sells high and buys low, but it will lag a buy-and-drift portfolio during a long one-way rally. Its dependable job is risk control: it keeps the portfolio’s actual risk equal to the risk that was designed for the goal, which is what protects the plan when the bad year eventually comes.
What is the difference between strategic and tactical asset allocation?
Strategic allocation is the long-term policy mix derived from your goals, horizon and risk capacity, and it changes only when your life changes. Tactical allocation means temporarily deviating from that mix based on a market view, such as trimming equity when valuations look expensive. Strategic allocation should do almost all the work. If you use tactical tilts at all, bound them with pre-written bands, because tactical calls require being right on both the exit and the re-entry.
What is a glide path and do I need one for every goal?
A glide path is a pre-agreed schedule for reducing equity exposure as a goal’s deadline approaches, so that a late market crash cannot destroy money you no longer have time to rebuild. In the worked example, gliding from 60 percent to 20 percent equity before the final year reduced the damage of a 30 percent crash by ₹9.6 lakh on an ₹80 lakh corpus. Any goal with a hard deadline, education fees, a house purchase, a retirement date, deserves one. Open-ended wealth building can hold a steadier mix.
How do I decide my own equity-debt split?
Derive it rather than declare it: compute the return your goal requires from your surplus and horizon, then test that mix against both your risk capacity (the arithmetic of your situation) and your risk willingness (how you actually behave in a 30 percent fall), and let the lower of the two bind. Count everything in the household, including EPF, PPF and NPS, since they may already be a large debt-like allocation. If the required return and your risk constraint do not meet, change the instalment, the horizon or the goal, not your honesty about returns.
Sources
All rupee figures, return assumptions and market scenarios in this article are illustrative examples constructed for the worked calculations, and are labelled as such in the text; no external return statistics, product figures or tax rates are quoted in this piece. The classic research mentioned by name (Brinson, Hood and Beebower’s work on pension portfolio performance in the Financial Analysts Journal, and Ibbotson and Kaplan’s later re-examination of it) is referenced only to caution against misquoting it; deliberately, no figures from it are used here.
Related HOFP resources linked in this article:
- How CFPs look at mutual funds differently: https://houseoffinancialplanners.com/how-cfps-evaluate-mutual-funds/
- Retirement corpus maths planners use: https://houseoffinancialplanners.com/retirement-corpus-maths-planners/
- Psychology in Financial Planning, the CFP behavioural module: https://houseoffinancialplanners.com/psychology-in-financial-planning-cfp/
- How to become a CFP in India: https://houseoffinancialplanners.com/how-to-become-cfp-india/
