Research note: This material is for educational and strategy-design purposes. It is not investment advice or a promise of future returns.
Executive summary
A combined strategy portfolio should not simply average all sleeves. It should allocate across them in a way that recognises factor cyclicality, correlation clusters, implementation cost differences and India-specific operational constraints such as taxes, shorting frictions and margin usage. The strongest high-level lesson from both the academic literature and NSE’s own multi-factor material is that no single factor dominates all regimes, and multi-factor or multi-sleeve diversification is therefore structurally sensible.
In Indian equities, the most robust combined portfolio usually has a liquid long-only core made from quality, value, momentum and low-volatility sleeves, plus smaller tactical sleeves for PEAD, revisions, sector rotation, pairs and basis arbitrage. Event-driven and relative-value strategies should normally receive lower capital weights but higher governance attention, because their operational error bars are wider.
Description
There are three practical ways to combine these strategies. The first is a sleeve model, where each strategy remains intact and the allocator changes capital weights. The second is a stock-level composite model, where value, quality and momentum signals become one blended ranking, while low-vol or sector rotation acts as a risk overlay. The third is a barbell model, where a stable long-only factor core sits beside small market-neutral and event-driven books. For India, the barbell often works best because it respects the asymmetry between easy-to-scale cash strategies and harder-to-scale shorting/event programmes.
Correlation management matters more than simply counting sleeves. Trend momentum and breakout strategies will often co-move. Cross-sectional momentum, earnings revisions and PEAD can also cluster. Value and dividend yield partially overlap. Low volatility tends to offset some cyclicality but can become a crowded defensive trade. A well-designed combined portfolio should therefore penalise redundant sleeves and favour combinations with genuinely different drivers, particularly value versus momentum and beta-light relative-value sleeves versus directional factors. Value and momentum are well known to diversify each other internationally, and NSE’s multi-factor work explicitly emphasises smoothing single-factor cyclicality through combination.
Key attribute table
The combined-portfolio expectations below are illustrative portfolio-design targets inferred from diversified factor investing and implementation research.
| Time horizon | Turnover | Typical Sharpe or return expectation | Data needs | Complexity | |---|---|---|---|---| | Multi-horizon | Medium | Sharpe roughly 0.8–1.2 for a diversified institutional portfolio; annualised return target roughly 10–16% gross depending on leverage and event sleeves | Full cross-asset data stack across all sleeves | Very high |
Details
A practical capital-allocation template for an India-focused institutional book is: 20% cross-sectional momentum, 15% value, 15% quality, 10% low volatility, 10% trend momentum, 8% sector rotation, 7% dividend yield, 5% PEAD, 5% earnings revision, 3% breakout, 1–2% short-term mean reversion, 3% pairs, 3% market-neutral multifactor, 2% cash-futures basis arbitrage and 1–2% special situations. That is only a starting point. The important structural idea is that the high-confidence scalable sleeves get most capital, while the fragile or capacity-constrained sleeves earn smaller but diversifying allocations. This weighting logic is an implementation recommendation, not an official model. It is motivated by factor cyclicality evidence, transaction-cost research and Indian practical frictions.
Portfolio construction should occur at two levels. First, size each sleeve by expected information ratio, cost and capacity. Second, aggregate sleeves through a risk model that controls total market beta, sector bets, concentration, liquidity usage and turnover overlap. Use one-month and three-month realised correlation matrices among sleeve returns, but shrink them because short histories overfit. A simple yet robust allocator is inverse-volatility sleeve weighting multiplied by a diversification score and multiplied again by a capacity score. A more advanced allocator can use expected-return forecasts and a regularised mean-variance optimiser, but only if the desk has strong governance around forecast error.
Turnover management is critical. If momentum, revisions and PEAD all want to buy the same stock after results, the combined portfolio should reconcile them centrally, not trade each sleeve independently. Similarly, if value and dividend yield both want PSU energy names, a shared concentration cap should stop accidental doubling-up. The combined engine should therefore maintain a master security ledger, aggregate desired weights, net internal crosses and then trade only the residual. That design choice is not from an exchange rulebook; it is a best-practice portfolio-engineering recommendation consistent with real-world anomaly cost evidence.
Backtesting the combined portfolio requires more than sleeve-level backtests added together. You must model cross-sleeve name overlap, common execution windows, shared liquidity constraints, borrow conflicts and portfolio-level tax drag. Also test partial implementation scenarios: for many Indian shops, stock-borrow availability is the bottleneck, so the live combined portfolio may need a long-only factor core with only modest market-neutral and arbitrage sleeves. Any combined-strategy document should therefore include at least three backtests: ideal full implementation, restricted-short implementation and long-only-plus-index-hedge implementation.
flowchart TD
A[Daily data pipeline] --> B[Compute sleeve signals]
B --> C[Create sleeve target weights]
C --> D[Aggregate at security and sector level]
D --> E[Apply portfolio risk and liquidity constraints]
E --> F[Net internal overlaps and hedge beta]
F --> G[Trade residual orders]
G --> H[Post-trade attribution by sleeve]
Implementation guide
- Build each sleeve independently and validate it on a point-in-time basis.
- Estimate sleeve-level volatility, turnover, correlation and capacity.
- Assign strategic capital weights that favour the most robust and scalable sleeves.
- Aggregate all sleeve orders into one master optimiser.
- Apply portfolio-wide beta, sector, liquidity and turnover controls.
- Attribute realised P&L back to sleeves monthly so weak sleeves can be resized or retired.
India-specific example
Assume a ₹100 crore institutional portfolio. A practical deployment could allocate ₹60 crore to the scalable long-only sleeves, ₹20 crore to slower diversified factor sleeves such as trend, sector rotation and dividend, ₹10 crore to event-driven sleeves such as PEAD and revisions, and ₹10 crore to relative-value or hedge sleeves such as pairs, market-neutral overlay and basis arbitrage. If the long-only factor book ends up with a realised beta of 0.88, the desk can short Nifty futures against part of the book to bring net beta toward the mandate target, provided derivatives margin headroom is available. Exchange contract specifications, margin rules and statutory charges must be applied at portfolio level rather than sleeve level.
Now imagine that cross-sectional momentum, PEAD and earnings revisions all want to add INFY after results, while low-volatility wants to hold it and value does not care. Instead of sending four independent buy orders, the combined engine should calculate the single net desired change, compare that to liquidity and participation constraints, and stage one order. That reduces turnover, market impact and operational noise. The same logic applies on the short side, where borrow is scarce and should never be wasted on redundant sleeve duplication.
Source note and file-index summary
The file pack purposefully prioritises official Indian sources and classic primary research. For the India-specific layer, the key load-bearing sources are NSE market timings, historical data archives, contract information, SLBS pages, margin pages, tax and stamp-duty schedules, Nifty Indices methodology documents and factsheets, sector-index factsheets and SEBI regulations for buybacks and margin collection. For the underlying factor and anomaly logic, the core academic sources are Fama and French on value and size, Jegadeesh and Titman on momentum, Moskowitz–Ooi–Pedersen on time-series momentum, Novy-Marx on profitability, Asness–Frazzini–Pedersen on quality, Blitz–van Vliet on low volatility, Moskowitz–Grinblatt on industry momentum, Bernard–Thomas on PEAD and Gatev–Goetzmann–Rouwenhorst on pairs trading.
The sixteen assumed strategy files in this pack are: nifty_trend_momentum.mdx, cross_sectional_momentum.mdx, value_equity.mdx, quality_equity.mdx, low_volatility_equity.mdx, size_small_cap.mdx, dividend_yield_equity.mdx, sector_rotation_momentum.mdx, earnings_revision.mdx, post_earnings_announcement_drift.mdx, breakout_trend_equity.mdx, short_term_mean_reversion.mdx, pairs_trading_equity.mdx, market_neutral_multifactor.mdx, cash_futures_basis_arbitrage.mdx, and special_situations_buyback.mdx, plus the portfolio-level combined_strategy.mdx.