Most retail investors get diversification backwards. Either they own 1-3 stocks (concentration disaster waiting) or 50+ stocks (returns ≈ index after tracking + tax). The math says 15-25 stocks is the sweet spot. This guide explains why.
The math of diversification
Portfolio variance falls as you add stocks — but with sharply diminishing returns. Academic studies (Markowitz, modern updates) show:
| Stocks | % Specific risk eliminated |
|---|---|
| 1 | 0% |
| 5 | ~70% |
| 10 | ~85% |
| 15 | ~92% |
| 20 | ~95% |
| 30 | ~97% |
| 50 | ~98% |
| 100+ | 99% (= index fund) |
Adding stocks 16-50 captures only 5% extra diversification benefit. Below 15 = real concentration risk.
Why concentration kills
Indian large-cap stocks have ~25-30% annual volatility individually. Index has ~15-18%. A 5-stock portfolio in same sector has ~25% volatility. A 20-stock multi-sector portfolio has ~16-18%.
Translation: more diversification = smoother ride = better behavioural outcomes (you don't panic-sell in drawdowns).
But correlation matters MORE than count
20 IT stocks ≠ diversification. They're all USD-linked, all client-concentrated, all hit by AI disruption simultaneously. Real diversification requires LOW-CORRELATION holdings.
Indian sector correlation (1-year rolling)
- High correlation (0.7+): Banks + NBFCs; Auto + Auto Ancillary; Steel + Cement
- Medium correlation (0.4-0.7): IT + Pharma; FMCG + Consumer Durables
- Low correlation (<0.4): FMCG + Banks; IT + Cement; Pharma + Real Estate
- Negative correlation: Gold + Equity (in stress periods); Defensive FMCG + Cyclicals
The 15-stock model portfolio
Diversified across sectors + market caps:
- 3 banks/financials (HDFC Bank, ICICI Bank, Bajaj Finance)
- 2 IT services (TCS, Infosys)
- 2 FMCG (HUL, Nestle)
- 2 pharma (Sun Pharma, Cipla)
- 2 auto/capex (Maruti, L&T)
- 2 mid-cap quality compounders (Pidilite, Asian Paints type)
- 1 utility / defensive (NTPC or PowerGrid)
- 1 special situation / value bet
Each stock ~6-7% weight at entry. Rebalance annually if any drifts above 12% or below 3%.
The sector cap rule
No single sector > 30% of portfolio. Even if banks are your best conviction, capping at 30% prevents sector-specific blow-ups (2018 NBFC crisis, 2008 housing crisis).
Market cap diversification
Indian markets reward small + mid cap over long horizons but with brutal drawdowns. Allocation by age:
- Under 35: 50% large + 30% mid + 20% small
- 35-50: 60% large + 25% mid + 15% small
- 50+: 75% large + 15% mid + 10% small + add debt
The international + gold layer
For 5%+ diversification beyond Indian equity:
- US/global equity ETFs: 5-15% of portfolio (rupee depreciation hedge)
- Gold / SGB: 5-10% (crisis hedge; see Gold ETFs guide)
- Debt funds / FDs: 10-30% (age-based)
The over-diversification trap
Owning 40+ stocks creates these problems:
- Returns converge to index → why not just buy index at 0.2% expense?
- Can't track quarterly results of 40 companies properly
- Annual report reading load makes you skip quality checks
- Tax + transaction costs add up on each rebalance
If you can't articulate the thesis for each holding in < 1 minute, you own too many.
The decision tree
- How much time can you dedicate? < 3 hr/month = use index funds. > 5 hr/month = direct stocks viable.
- Capital ≥ ₹5 lakh: aim for 15 stocks (~₹33k each). Below that, just SIP into 2-3 index funds.
- Across sectors: 5-7 sectors minimum.
- Across market caps: include 2-4 mid + 1-2 small cap.
- Annual review: rebalance to target weights if drift > 5%.
Use the Position Sizing calculator for entry sizing within your diversification framework. Combine with bluechip universe for safe stock selection.