Trading Risk Management Glossary: Position Sizing, Margin and Drawdowns
Trading risk management is the process of defining how much can be lost, what could make losses larger and how exposures behave together. This glossary explains position sizing, leverage, margin, drawdowns, volatility, expectancy, portfolio correlation and other terms used when evaluating risk.
Risk measures are not interchangeable. A stop distance is a plan, margin is collateral, notional is exposure, and a historical loss statistic is not an insurance limit. Understanding their differences is particularly important in leveraged Forex, futures, options and CFD markets.
Trading risk management glossary
Account equity
The net value of an account after liabilities and relevant open-position valuations under the provider’s method. It is the starting point for many risk limits but is not identical to immediately withdrawable cash.
Available margin
Funds or collateral capacity available for new positions under a broker’s rules. It can contract rapidly when prices move, volatility rises or the provider increases requirements.
Basis risk
The possibility that a hedging instrument and the underlying exposure do not move together as expected. A hedge can appear sensible yet perform poorly when contract specifications, maturities or market locations differ.
Beta
A statistical measure of an asset’s sensitivity to a chosen benchmark. It depends on sample and method, so it should not substitute for evaluating sector concentration, liquidity or downside scenarios.
Capital at risk
A term that must be defined before use: it can refer to planned loss at an exit level, collateral committed, or worst-case economic exposure. These quantities often differ sharply.
Conditional value at risk
Also called expected shortfall under specified methodologies, an estimate of average losses beyond a chosen loss percentile. Its usefulness depends on the model, horizon and data; it is not a guaranteed extreme-loss bound.
Concentration risk
Exposure to a small number of issuers, instruments, sectors, countries or common economic drivers. Many positions may still represent one concentrated risk if they respond similarly to the same catalyst.
Correlation
A sample measure of co-movement between return series. It changes across markets and time; correlations can rise during stress even when the historical portfolio appeared diversified.
Counterparty risk
The danger that a contractual counterparty fails to fulfil obligations. Clearing, collateral and legal protections can reduce particular risks without eliminating settlement or operational uncertainties.
Credit risk
The chance of loss from a borrower or issuer failing to meet obligations or suffering a deterioration in creditworthiness. Credit spreads can widen even without an actual default.
Diversification
Combining holdings with different exposures in an attempt to reduce dependence on a single risk factor. It does not guarantee gains or protect against market-wide losses.
Drawdown
A fall from a previous equity peak to a subsequent trough. It measures an observed path and can be more decision-relevant than average volatility because it shows the depth of capital impairment.
Expected shortfall
A risk measure focused on the severity of outcomes in the tail beyond a selected threshold, given modelling assumptions. Tail estimation can be especially fragile when severe events are rare in the dataset.
Expectancy
Average outcome per trade or other repeatable decision, based on probabilities and payoff magnitudes. Positive estimated expectancy may vanish after costs, regime changes or inadequate sample size.
Exposure
Economic sensitivity to a market factor. Gross exposure adds absolute longs and shorts while net exposure offsets them; neither measure alone captures options convexity or financing risks.
Forced liquidation
Closure or reduction of a position by a broker, exchange or clearing system when rules are breached or collateral becomes insufficient. Execution may occur at adverse prices and without a trader’s preferred timing.
Gap risk
The possibility of prices jumping over anticipated trading levels without executable liquidity in between. Ordinary stop orders do not eliminate this risk.
Gross leverage
Gross notional exposure divided by equity under a stated convention. It is a rough scale measure; identical gross leverage can conceal very different volatility, liquidity and tail-risk profiles.
Hedge ratio
The quantity of hedging exposure relative to the position intended to be offset. An appropriate ratio depends on contract size, price sensitivity, currencies and basis risk, not simply matching headline prices.
Initial margin
The collateral needed to establish a position according to product-specific rules. It is not a fee, a stop level or the maximum amount that a futures or leveraged position can lose.
Kelly criterion
A mathematical position-sizing framework based on assumed probabilities and payoffs. Small estimation errors can produce large sizing errors, and full-Kelly sizing can lead to substantial drawdowns.
Leverage
The scaling of asset exposure relative to committed capital, obtained through borrowing or derivatives. It amplifies the effect of price changes on account equity and can shorten the time available to respond to losses.
Liquidity risk
The possibility that an exposure cannot be sold or hedged promptly without substantial price concessions. Quoted liquidity may disappear exactly when risk reduction is most urgent.
Maintenance margin
An ongoing collateral threshold or requirement for a margined position. Falling below it may trigger collateral demands or liquidation depending on provider and clearing rules.
Margin call
A notification or demand to restore sufficient collateral or equity. Some electronically managed products may liquidate positions automatically rather than waiting for a manual response.
Market risk
Potential loss arising from changes in prices, rates, volatility and other traded risk factors. It is distinct from operational, legal, liquidity and counterparty risks, although they can interact.
Maximum drawdown
The largest peak-to-trough decline observed in a specified equity or price series. It is a backward-looking statistic, not a maximum future loss guarantee.
Net exposure
Directional position exposure remaining after long and short amounts are offset using a specified convention. A portfolio with little net exposure can still have substantial gross leverage and basis risk.
Notional value
The reference amount used to quantify a derivative or leveraged instrument’s economic scale. It is not necessarily the initial payment, current market value or worst conceivable loss.
Operational risk
Loss or disruption from failures of systems, people, processes or external events. Order errors, platform outages, incorrect contract specifications and cyber incidents are practical examples.
Position sizing
Choosing a number of shares or contracts after considering capital, invalidation levels, cash flow, market liquidity and risk limits. A correct market thesis can still produce disastrous outcomes when sized poorly.
Probability of ruin
A model-dependent estimate of the chance that trading capital falls below a defined threshold. It depends on payoff distributions, position sizing and dependence between outcomes; real-world regimes may invalidate the assumptions.
R-multiple
Trade result expressed relative to the originally defined risk unit, typically a planned loss under specific assumptions. A realised -2R loss is possible when slippage or gaps make execution worse than the planned stop.
Risk budget
A pre-allocated allowance for a defined risk measure across strategies or exposures. Budgets are only comparable if measurement conventions and correlations have been specified.
Risk–reward ratio
A comparison between a planned adverse move and favourable move under a scenario. Without probabilities, fees and execution assumptions it cannot determine whether a strategy is profitable.
Risk parity
A portfolio construction concept allocating contributions to an estimated risk measure rather than equal money amounts. Leverage, changing covariance and tail dependence can complicate implementation.
Risk per trade
A planned amount of capital that could be lost if a specified exit is achieved under assumed conditions. It is not a guaranteed loss ceiling in markets prone to gaps or counterparty failure.
Stop-loss order
An order intended to reduce exposure after a trigger condition. A traditional stop becoming a market order may execute substantially away from its trigger; stop-limit orders risk non-execution.
Stress test
An examination of how an exposure might behave under an adverse historical or hypothetical scenario. It can reveal vulnerabilities but cannot enumerate every possible future shock.
Tail risk
Exposure to comparatively rare and severe outcomes, including gaps, crashes and nonlinear derivative losses. Historical standard deviation often gives an incomplete picture of such risks.
Value at risk
A modelled loss quantile for a stated probability, horizon and return methodology. VaR does not describe the size of losses beyond the quantile and should never be interpreted as the maximum loss possible.
Volatility
A measure of return variability over a sample or an estimate inferred from prices. Low recent volatility does not mean that an instrument is safe or cannot suffer a sudden large move.
Win rate
The fraction of trades that finish profitable according to the strategy’s chosen accounting convention. A high win rate can coexist with negative expectancy if occasional losses are large.
Worked example: position size from a risk budget
Assume a hypothetical account contains £20,000 and the investor chooses a planned risk budget of 1% of account equity for a specific unleveraged share trade. That is £200. The assumed entry price is £50 and the planned exit is £47.50, a £2.50 adverse price move per share. Ignoring commissions, taxes, slippage and gaps, £200 divided by £2.50 gives 80 shares. The nominal purchase value would be 80 × £50 = £4,000.
The calculations answer different questions. £200 is the hypothetical planned loss if the shares can be sold exactly £2.50 lower; £4,000 is the value of shares purchased; and neither number by itself describes every possible market loss. If bad news causes the market to open at £43, the 80 shares could suffer a £7 loss each, or £560 before costs, despite the planned £200 risk budget. A position-size formula helps discipline, but does not eliminate gap risk.
Worked example: expectancy and risk–reward
Consider a hypothetical strategy with a 40% winning-trade probability, an average gain of two risk units on winners and an average loss of one risk unit on losers. Its simple expected payoff is (0.40 × 2R) − (0.60 × 1R) = 0.20R per trade before costs. This is a mathematical illustration rather than evidence that the strategy possesses a real edge.
If commissions and execution effects average 0.25R per trade, the same hypothetical expectancy becomes negative at -0.05R. Estimating win rate from a small backtest can also be highly unreliable. If losing trades cluster during a single market regime, the arithmetic average may conceal difficult drawdowns, margin pressure or the inability to keep taking trades.
Why the risk–reward ratio alone does not predict profitability
A setup offering a theoretical £300 gain for a £100 planned loss has a 3:1 potential reward-to-risk ratio, but this says nothing about how often either outcome is likely. If the chance of realising the £300 is only 15% and the chance of losing £100 is 85%, its simple pre-cost expectation is (0.15 × £300) − (0.85 × £100) = -£40. Trading decisions require probabilities, execution assumptions, position sizing and the possibility of outcomes outside planned levels.
Target and stop placement also do not create probabilities mechanically. A target may be far away but rarely reached; a stop may be near but frequently triggered by ordinary volatility. Any estimate derived from historical results needs robustness testing, including slippage, out-of-sample periods and adverse market environments.
How leverage and margin interact
Leverage changes the impact of a price movement on capital. If a trader has £5,000 in account equity and holds £25,000 of market exposure, gross leverage is 5:1 under the simplified convention. A 2% adverse move in the exposure implies approximately £500 of market loss before funding and other effects, equivalent to 10% of initial equity. The market movement did not need to be 10%; leverage magnified it relative to the account.
Required margin is a collateral framework, not a definition of the maximum loss. Margin requirements can rise when volatility spikes, and some products can be closed automatically once account equity reaches specified thresholds. Protections for certain UK retail CFD accounts differ from the obligations and risks of futures, borrowed securities and professional-client arrangements. Check the actual product and regulatory terms.
Historical drawdown and the recovery problem
Suppose a £10,000 portfolio falls 30% to £7,000. The recovery needed to regain £10,000 is £3,000 divided by £7,000, about 42.86%. A 50% decline requires a subsequent 100% gain to recover. That arithmetic is independent of trading style and helps explain why preserving capital matters even when average historical return looks attractive.
Maximum drawdown measures the worst observed path in the available dataset, which may omit crises or instrument failure. It should be accompanied by drawdown duration, liquidity conditions, margin calls, concentration and risk of further loss. A low historical maximum is not a guarantee that a larger decline cannot occur later.
Portfolio risk is more than the number of positions
Owning ten securities is not necessarily diversified if they all depend heavily on the same interest-rate expectations or funding conditions. A technology stock basket, an equity index future and a high-beta cryptocurrency holding may react to common shifts in risk appetite. Conversely, two exposures with different asset labels may contain an offsetting currency or rate sensitivity that is easy to overlook.
Review gross and net exposure, currency, sector, maturity, counterparty and liquidity. Correlation estimates are useful descriptions, but in a crisis their relationships can change. A sensible stress test evaluates plausible joint shocks, such as higher yields, widening credit spreads, falling equities and a sudden reduction in available market depth.
Three practical risk-control distinctions
A rule is not a guarantee. Pre-set exits reduce discretionary delay but cannot guarantee a transaction price. A model is not the market. VaR and expected shortfall depend on historical data and assumptions, especially in the distribution tails. A hedge is not automatically an offset. Different expiry dates, indexes, venues or currency bases can cause a hedge to fail when needed most.
Risk decisions also involve operations. A correct position-size calculation can still be undermined by using the wrong contract multiplier, confusing a pip with a point, failing to update linked orders after a partial fill or relying on a broker system during an outage. Product mechanics are a component of financial risk management.
Explore related glossary guides
Definitions are most useful when connected to neighbouring concepts. See Trading Glossary A–Z, Trading Costs, Forex Trading, Derivatives and Crypto for complementary market terms, risk considerations and examples.
Frequently asked questions
Is 1% risk per trade guaranteed to limit loss to 1%?
No. It is commonly a planned risk budget based on an expected exit. Gaps, slippage, fees or forced liquidation can produce larger realised losses, and the effect of simultaneous positions must also be considered.
Is initial margin the amount I can lose?
Not generally. Margin is collateral required under product rules. Losses can exceed the initial margin in products without specific regulatory or contractual protections.
What is the difference between volatility and drawdown?
Volatility measures the dispersion of returns; drawdown measures the decline from a previous high. A strategy can have moderate average volatility yet suffer a large cumulative decline or an abrupt gap.
Does diversification prevent losses?
No. It can reduce some concentration risks if exposures are genuinely different, but it cannot guarantee profits or protect against common market shocks.
Is a positive historical Sharpe ratio enough to justify a trade?
No. It is one risk-adjusted statistic and depends on the sample, reference rate and return measurement. It should be assessed with costs, drawdowns, capacity, liquidity and the robustness of the underlying hypothesis.
Related guides
Continue with Position Sizing and Leverage, Trading Expectancy Explained and Portfolio Risk and Drawdown.
Reference sources
Consult the SEC Investor.gov glossary, the FCA’s CFD consumer-protection guidance and the CME Group futures terminology. Actual margin and close-out rules are instrument- and provider-specific.
Important notice: This material is general financial education only. It does not recommend any position size, trading strategy or leveraged product, and loss may exceed a planned threshold.