Capital Allocation

Quick answer: Capital allocation is the deciding of how much of a trading account is committed to each strategy or instrument so that the whole book carries balanced, intended risk.

In simple words

If position sizing decides how big one trade is, capital allocation decides how the whole pot is split across strategies and markets. Putting equal rupees into two strategies does not mean equal risk, because one may be far more volatile than the other. Good allocation balances risk contributions, not just rupees, and accounts for how strategies move together.

Purpose

Allocation exists because most edges are small and fragile; spreading capital across weakly correlated sources of return is how a book earns a steadier line than any single strategy could.

Professional explanation

Rupee weight is not risk weight

Splitting capital evenly by rupees is intuitive but usually wrong, because it ignores volatility. A trend strategy on Bank Nifty and a mean-reversion strategy on a large-cap may take the same rupees yet contribute very different risk, so the volatile one dominates the book's profit and loss. Allocation should be framed in terms of each sleeve's risk contribution, most simply by scaling each to a target volatility so that no single strategy quietly drives the whole account.

Correlation changes the maths

The risk of a combined book is not the sum of the parts; it depends on how the parts co-move. Two strategies that both go long equity beta will draw down together, so their combined risk is close to additive, whereas two genuinely uncorrelated strategies partially cancel, giving a smoother equity curve for the same gross exposure. Allocation therefore needs a correlation view, and crucially an awareness that correlations are unstable and tend toward one in a crisis, exactly when diversification is most needed.

Common allocation schemes

Equal-weight assigns the same rupees to each sleeve: simple and robust but risk-blind. Volatility targeting scales each sleeve to a chosen volatility, equalising risk contribution under an independence assumption. Risk parity goes further and uses the covariance matrix so that each sleeve contributes equal marginal risk to the portfolio. Optimisation methods such as mean-variance can in theory maximise return per unit risk, but they are notoriously sensitive to estimation error in expected returns and are usually constrained heavily or avoided in favour of simpler, more stable rules.

Reserves, buffers and dry powder

Not all capital should be deployed. Allocation includes deciding how much stays uncommitted as margin buffer and opportunity reserve, so that a volatility spike does not force liquidation at the worst moment. On leveraged NSE derivatives, margins can rise sharply during stress, so a book fully committed at normal margins can be forced to cut positions precisely when spreads are widest. A deliberate cash or unused-margin reserve is itself an allocation decision, not idle money.

Rebalancing and drift

Allocations drift as strategies win and lose, so a static plan slowly becomes concentrated in whatever has been performing. Periodic rebalancing back to target weights enforces a discipline of trimming winners and topping up laggards, but it incurs cost and can hurt during strong trends. The engineering choices are the rebalance frequency, the tolerance bands that trigger a rebalance, and whether to rebalance on a calendar or on a drift threshold, each trading off transaction cost against tracking to the intended risk profile.

Allocation as a live control

In a production system, allocation is not set once but adjusted as strategies degrade or regimes change. A sleeve whose live performance decays relative to its backtest should have capital reduced or cut, and a risk engine typically enforces caps so that no single strategy or instrument can exceed a maximum share of the book. This links allocation to monitoring: allocation decisions should be driven by measured live behaviour, not by attachment to an idea.

Formula

Practical example

Advantages

• Balances risk, not just rupees, so no single strategy dominates the account
• Diversifying across weakly correlated sleeves smooths the equity curve
• Reserves and caps protect against forced liquidation during margin spikes
• Rebalancing enforces disciplined trimming of concentration

Limitations

• Correlations are unstable and tend toward one in crises, undermining the diversification exactly when it is needed
• Volatility and covariance estimates are noisy and backward-looking
• Mean-variance optimisation is extremely sensitive to expected-return errors and can produce extreme weights
• Rebalancing costs money and can lag strong trends
• Over-diversification can dilute a genuine edge into mediocrity and raise operational complexity

Why it matters in practice

• Allocation determines how a portfolio behaves in stress far more than any single strategy's average return
• It is where correlation risk is either controlled or ignored

Common mistakes

• Allocating equal rupees and assuming that means equal risk
• Estimating correlation from a calm period and trusting it to hold in a crash
• Running mean-variance optimisation on raw return estimates and getting knife-edge weights
• Deploying the entire account so a margin spike forces liquidation of good positions
• Never rebalancing, letting the book quietly concentrate into the recent winner
• Treating uncorrelated backtests as proof of live independence when both share a hidden factor such as equity beta or short volatility

Professional usage

Institutional books manage allocation with an explicit risk model: they estimate a covariance matrix, target a portfolio-level volatility, and set hard caps per strategy, per instrument and per risk factor. Many prefer robust, low-parameter schemes such as risk parity or simple volatility targeting over full optimisation, because these are less fragile to estimation error. Allocation is reviewed on a fixed cadence, tied to live performance monitoring, and enforced by the risk engine rather than left to discretion.

Frequently asked questions

Voice search & related questions

Natural-language questions people ask about Capital Allocation.

Sources & references

Last reviewed 11 July 2026. Educational content only — not investment advice. Markets and rules change; verify current conventions with SEBI, NSE/BSE and your broker.