sharma23yash-oss Portfolio Review (India)
Before any opinion, build the picture. Most Indian retail portfolios have three problems the owner cannot see: they are one bet wearing fifteen names, they own the same stocks twice through funds, and they have never counted the tax cost of fixing it.
Step 1 — Normalise the holdings
Get everything into one table, at current prices:
| Instrument | Type | Qty | Avg cost | CMP | Value ₹ | Weight % | Bought on | Holding period | Unrealised ₹ |
Types: direct equity, equity MF, debt MF, ETF, gold (SGB / gold ETF / physical), EPF/PPF/NPS, fixed deposits, real estate, cash. If the user gives only equities, ask what else exists before commenting on asset allocation — a "100% equity portfolio" is often 40% equity once EPF and property are counted.
Compute total portfolio value and each position's weight against it.
Step 2 — Look-through the funds
This is the step almost everyone skips. Every equity mutual fund is a bundle of stocks; the user's true stock exposure is direct holdings plus the fund look-through.
- Pull each fund's latest monthly portfolio (AMFI / AMC disclosure).
- Multiply each fund holding weight by the fund's weight in the portfolio.
- Add to the direct equity weights.
The output is the true single-stock exposure. A user with 6% in a bank directly, plus four funds each holding 8% of the same bank, typically has 10–12% in one name and thinks they have 6%.
Also compute fund-to-fund overlap (see mutual-fund-analysis-india). Two large-cap-leaning funds overlapping 70% are one fund charged twice.
Step 3 — Concentration and diversification
| Measure | How | Reading |
|---|---|---|
| Top-1 weight | >15% is a single-name risk decision, not diversification | |
| Top-5 weight | >45% means five decisions drive the outcome | |
| Effective number of holdings | 1 / Σ w² (inverse Herfindahl) |
A 30-stock portfolio with an effective N of 6 is a 6-stock portfolio |
| Sector concentration | vs Nifty 500 weights | Note deliberate vs accidental tilts |
| Market-cap split | large / mid / small using AMFI's semi-annual list | Small-cap weight above ~25% needs to be a choice, not a drift |
Effective N is the most useful single number in a retail portfolio review. Compute and report it every time.
Also flag correlation clusters, which look like diversification but are not: PSU banks together; all-NBFC exposure; IT services names that share the same USD-revenue and US-discretionary-spend driver; commodity producers sharing one input price; and "India consumption" names that are really one macro bet.
Step 4 — Factor and style read
Without running a formal regression, characterise the portfolio:
- Growth vs value: weighted P/E and P/B versus the Nifty 500
- Quality: weighted ROCE, net debt/EBITDA, and the share of the portfolio in companies with negative free cash flow
- Size: weighted market cap and the mid/small share
- Momentum: share of holdings within 10% of a 52-week high versus more than 30% below
- Yield: weighted dividend yield
Then say plainly what the portfolio is. "This is a mid-cap growth portfolio with a financials overweight and no defensives" is a useful sentence. "You are well diversified" is not.
Step 5 — Position-level review
For each holding, classify:
| Class | Definition | Action framework |
|---|---|---|
| Core | Thesis intact, position sized correctly | Hold; rebalance only on weight drift |
| Oversized winner | Thesis intact, weight above the sizing rule | Trim to target, with tax cost computed |
| Thesis broken | The reason for owning it no longer holds | Exit regardless of price paid |
| Orphan | No stated reason for owning it | Either write the thesis or sell |
| Anchored loser | Held only because it is down | The purchase price is not a fact about the company |
| Legacy / inherited | Never chosen | Review on merits, note the cost basis and grandfathering |
Ask the user for the one-line thesis on each position. The ones they cannot articulate are the review's real output.
Step 6 — Risk checks specific to India
- Promoter pledge on any direct holding — run
shareholding-pattern-analysis - Governance flags on the top five weights — run
forensic-redflags-india - Liquidity: for each holding, position size versus 20-day average traded value. If exiting a position would take more than a few days' volume, it is not a liquid position at that size.
- SME-platform and micro-cap names: mark separately; they behave differently in a drawdown
- Unlisted or pre-IPO holdings: illiquid, and taxed on a 24-month long-term clock
- Single-employer risk: ESOPs plus salary plus employer's stock in a fund is a concentrated bet on one payroll
Step 7 — Tax-aware rebalancing
Never propose a trim or exit without the tax number. Pull it from capital-gains-tax-india:
- Long-term equity gains are taxed at 12.5% above ₹1.25 lakh per tax year; short-term at 20%.
- Positions within days or weeks of completing 12 months should usually wait. Compute the date and say so.
- Realised losses this tax year can offset gains — short-term losses against both, long-term losses against long-term only.
- Pre-1 February 2018 holdings carry grandfathered cost; the taxable gain is often far smaller than the unrealised gain shown by the broker.
- Sequence the rebalance across two tax years where the saving is material and the market risk of waiting is acceptable — and say what that risk is.
Present rebalancing as: target weight → trade → tax cost → net benefit. A rebalance that costs more in tax than it reduces in risk should not be recommended.
Step 8 — Asset allocation, if the user wants it
Only after the equity picture is clear, and only against a stated objective and horizon. Cover:
- Equity / debt / gold / cash split versus the user's stated horizon and drawdown tolerance
- Emergency reserve in liquid or overnight funds, sized in months of expenses
- EPF, PPF and NPS as the debt allocation most Indian investors already hold and forget to count
- Gold as a rupee-hedge and diversifier, typically a single-digit to low-teens allocation, with SGB versus gold ETF chosen on tax and liquidity grounds
- International exposure, subject to the industry-level RBI/SEBI limits on overseas investment by mutual funds, which have been capped and reopened more than once — verify the current position before assuming a fund can accept fresh flows
Do not produce a specific allocation as a recommendation. Produce the trade-offs and let the user choose, or route them to a SEBI-registered investment adviser.
Output format
# Portfolio Review — <date>
**Total: ₹X | Positions: N | Effective N: n.n**
## What this portfolio is
<Two sentences. Style, size, sector tilt, and the dominant risk.>
## Composition
| Asset class | Value | Weight | Target? |
## True stock exposure (direct + fund look-through)
| Stock | Direct % | Via funds % | **Total %** |
## Concentration
| Top 1 | Top 5 | Effective N | Largest sector | Mid+small % |
## Correlation clusters
- <cluster>: X% combined
## Position review
| Holding | Wt | Class | Note | Holding period | Unrealised | Tax if sold |
## Fund overlap
| Fund A | Fund B | Overlap % |
## Issues, ranked
1. ...
## Rebalancing options (with tax)
| Action | From → To | Tax cost | Rationale |
## Questions for you
1. <thesis questions on the positions with no stated reason>
Hard rules
- Never recommend buying or selling a specific security. Frame everything as observations, trade-offs, and options with their costs.
- Always do the fund look-through before commenting on diversification.
- Always attach the tax cost to any suggested trim.
- Use the user's own stated objective and horizon. If they have not given one, ask before commenting on allocation.
- State clearly that this is analysis, not investment advice, and that a SEBI-registered investment adviser is the right route for a recommendation.