Position Sizing Cheat Sheet

A single-page reference to the main position sizing methods, their formulas placed side by side, a worked NSE example, and an honest comparison of what each one does well and where it fails.

Position Sizing Cheat Sheet: Position sizing decides how many units to trade, and the main methods are: fixed lots (always the same quantity), fixed-fractional or percentage-risk (risk a set fraction of equity, then derive quantity from the stop), Kelly and half-Kelly (size to a formula-driven optimal fraction of a known edge), and volatility or ATR sizing (size inversely to recent volatility). The most widely used honest default is the percentage-risk model: Risk per trade = Capital times Risk%, and Quantity = risk divided by (stop distance times point value). Every method here is a standard formula or a labelled heuristic for education only, and none promises profit.

Position sizing usually matters more to long-run results and to the shape of your drawdowns than the entry rule itself. This page places the main methods side by side with their formulas, works a full NSE example, and compares them honestly, including where each one misleads. The percentages and the Kelly output are estimates and heuristics, not guarantees; sizing controls risk, it does not create an edge. For the surrounding rules see the Risk Management Cheat Sheet.

The methods and their formulas

Method	Formula	Variables
Fixed lots	Quantity = a constant (e.g. 1 lot)	Fixed by the trader regardless of capital or stop.
Fixed-fractional / percentage-risk	Risk = Capital × Risk% ; Quantity = Risk ÷ (Stop distance × Point value)	Capital = equity in ₹; Risk% = fraction risked; Stop distance = entry − stop in points; Point value = ₹ per point per unit.
Kelly	f* = W − (1 − W) ÷ R	f* = fraction of capital to risk; W = win probability; R = payoff ratio (avg win ÷ avg loss).
Half-Kelly	f = 0.5 × f*	Half the Kelly fraction; quarter-Kelly uses 0.25 × f*.
Volatility / ATR sizing	Quantity = Risk budget ÷ (ATR × Point value)	Risk budget = ₹ allotted to the trade; ATR = average true range in points; Point value = ₹ per point per unit.

The percentage-risk model in detail

This is the workhorse of honest position sizing, because it holds the money at risk constant across trades of different stop widths. Two steps:

1. Risk per trade = Capital × Risk%
2. Quantity = Risk per trade ÷ (Stop distance × Point value)

Capital is account equity in ₹; Risk% is the fraction you accept losing on the trade (0.5% to 2% is a common heuristic); Stop distance is entry minus stop in points; Point value is the ₹ per point per unit, which for index derivatives is the lot multiplier. A wider stop yields a smaller quantity, keeping the rupee risk fixed.

Worked NSE example (illustrative only)

Suppose capital is ₹5,00,000 and you risk 1% per trade. You trade Nifty, where one lot is 75 units, so the point value is ₹75 per point. Your stop is 40 points from entry.

Risk per trade = ₹5,00,000 × 1% = ₹5,000.
Risk per lot = 40 points × ₹75 = ₹3,000.
Quantity = ₹5,000 ÷ ₹3,000 ≈ 1.67 lots.
Round down to 1 lot, because 2 lots would risk ₹6,000, breaching the ₹5,000 budget.

At 1 lot your actual risk is ₹3,000, or 0.6% of capital, comfortably inside the limit. If the stop were tighter, say 20 points (₹1,500 per lot), the same ₹5,000 budget would allow 3 lots. These are illustrative figures, not a recommendation, and they exclude brokerage, STT, GST and slippage, which raise the true cost of the trade.

Comparison of methods

Method	What it does	Pro	Con	Blind spot
Fixed lots	Trades a constant quantity every time.	Simple; easy to execute and track.	Per-trade risk drifts with stop width and price.	Ignores the stop entirely, so a wide-stop trade risks far more than a tight one.
Percentage-risk	Holds rupee risk constant as a set fraction of equity.	Compounds after wins, shrinks after losses; risk stays uniform.	Needs a defined stop; small accounts hit the one-lot floor.	Assumes the stop reflects true risk; a gap can blow past it.
Kelly	Sizes to the growth-optimal fraction of a known edge.	Maximises long-run growth if inputs are exact.	Produces severe drawdowns; very sensitive to input error.	Assumes W and R are known precisely, which live trading never delivers.
Half-Kelly	Bets half the Kelly fraction.	Cuts volatility and drawdown sharply for little lost growth.	Still relies on estimating the edge.	A wrong edge estimate still oversizes, just less than full Kelly.
Volatility / ATR	Sizes inversely to recent volatility.	Equalises risk contribution across instruments and regimes.	Reacts to volatility changes with a lag.	Volatility can jump faster than the ATR updates, understating fresh risk.

Using Kelly safely

The Kelly fraction f* = W − (1 − W) ÷ R gives the bet size that maximises long-run growth for a known edge, where W is the win probability and R the payoff ratio. Two cautions make it usable rather than dangerous:

It is an upper bound, not a target. Full Kelly produces large, uncomfortable drawdowns, and it assumes W and R are exact. In trading they are noisy estimates, so full Kelly systematically oversizes.
Bet a fraction of it. Half-Kelly or quarter-Kelly retains most of the growth while cutting volatility and drawdown substantially. This is why practitioners almost never trade full Kelly.

If the Kelly formula returns zero or a negative number, the implied edge is non-existent; the correct size is not a small bet but no bet.

Volatility and ATR sizing

Volatility sizing allots each trade a risk budget and converts it to quantity using a volatility measure, usually the ATR: Quantity = Risk budget ÷ (ATR × Point value). A volatile instrument gets a smaller position and a calm one a larger position, so each contributes similar risk to the portfolio. The trade-off is lag: because the ATR is a trailing average, a sudden volatility spike is under-counted until the average catches up, so a gap or event can deliver more risk than the sizing assumed.

Choosing a method

For most discretionary traders with defined stops, the percentage-risk model is the sensible default: uniform risk, natural compounding, simple maths.
Volatility or ATR sizing suits systematic and multi-instrument portfolios where equalising risk across positions matters.
Kelly, used at a fraction, is a useful sanity check on how aggressive a size is, not a size to adopt literally.
Fixed lots is acceptable only when capital and stop widths are stable, or as a deliberately simple starting point.

Whichever method you use, it controls risk and shapes the equity curve; it does not manufacture an edge. Every formula and figure here is educational and evergreen, and none of it promises a profit.

Frequently asked questions

What is the most reliable position sizing method?

For most traders with a defined stop, the percentage-risk (fixed-fractional) model is the sensible default. It keeps the rupee risk constant across trades, compounds naturally after wins and shrinks exposure after losses. It is not the highest-growth method in theory, but it is robust because it does not depend on precisely estimating your edge.

How do I size a Nifty position from a stop?

Fix the money at risk as Capital times Risk%, then divide by the stop distance times the point value. With ₹5,00,000 at 1% risk (₹5,000), a 40-point stop and a Nifty lot of 75 (₹75 per point), each lot risks ₹3,000, so you trade one lot and stay within the ₹5,000 budget. A tighter 20-point stop would permit three lots.

What does the Kelly criterion actually give me?

For a known edge, f* = W minus (1 minus W) divided by R gives the fraction of capital that maximises long-run growth, where W is the win probability and R the payoff ratio. It is a theoretical ceiling that assumes the inputs are exact and produces severe drawdowns, so it is a reference point rather than a size to use literally.

Why use half-Kelly instead of full Kelly?

Because a live edge is only estimated, and full Kelly assumes it is known precisely. Betting half the Kelly fraction keeps most of the long-run growth while sharply reducing volatility and drawdown, which makes the strategy survivable in practice. Many traders go further to quarter-Kelly for the same reason.

What is volatility or ATR-based position sizing?

It sizes each trade inversely to its recent volatility so that every position contributes a similar amount of risk. Using Quantity = risk budget divided by (ATR times point value), a volatile instrument gets a smaller position than a calm one. Its weakness is lag: a sudden volatility spike is under-counted until the trailing ATR catches up.

What is wrong with trading fixed lots?

Fixed lots ignore the stop, so per-trade risk drifts as capital and stop widths change. A trade with a wide stop can risk several times more than one with a tight stop, even though you traded the same quantity. It is acceptable only when capital and stop distances are stable, or as a deliberately simple starting point.

Does correct position sizing guarantee profit?

No. Sizing controls how much you can lose and shapes the equity curve, but it cannot turn a losing system into a winning one. If your expectancy is negative, better sizing only slows the losses. Position sizing manages risk and improves survival odds; the edge has to come from the strategy itself.

What if the Kelly formula gives a negative number?

A zero or negative Kelly fraction means the estimated edge is absent, so the mathematically correct position is no position rather than a small one. It is a signal that, on the win rate and payoff you have assumed, the trade has no positive expectancy and should be skipped rather than sized down.

Last reviewed 12 July 2026. Educational content only — not investment advice.