How a margin requirement is calculated

The numbers the calculation needs 

Contract size is the quantity of the underlying that one contract covers, and it is a published convention rather than a choice. A standard currency lot covers one hundred thousand units of the currency written first in the pair. A gold contract conventionally covers one hundred troy ounces. An index contract is stated as a money amount per index point. A shares contract is written on a stated number of shares. The convention differs by market and by venue, and the operative one for any instrument is published in its contract specification alongside the requirement itself.

Price is the second number, and which price is used is a detail worth being precise about. The calculation reads the prevailing market price rather than the price the position opened at, on the instruments where the requirement is worked out from price at all. Contract size multiplied by price gives the notional value of the contract, also called the contract value or the face value. That figure is not what has to be funded. It is what the profit and loss calculation is measured on, and it is what the requirement percentage is taken from.

The requirement percentage is the third number, and it is set by the counterparty per instrument rather than negotiated per account or chosen per position. It is published in the contract specification, it can differ between two instruments that look similar, and it can be changed, including on positions that are already open. Its whole job is to state what proportion of a contract's value has to be held while that contract exists.

The fourth number is a conversion rate, and it applies whenever the contract is valued in a currency other than the one the account reports in. The requirement is worked out in the contract's own currency first and converted afterwards, so the figure the panel prints depends on a rate that has nothing to do with the instrument on the ticket. That step is invisible on an account whose currency happens to match, which is why it surprises people on the accounts where it does not.

A currency pair 

A currency contract is the case where the order of the steps matters most, because the quantity is denominated in one currency and the requirement is usually reported in another. The contract covers a quantity of the base currency, and that quantity is fixed for as long as the position is open, so the notional value in base currency terms does not move at all. The requirement percentage is applied to it, and only then is the result converted into the account's currency at the prevailing rate.

Two features of that sequence carry over to every other example here. The percentage is applied to the notional value and never to the account, so nothing about the size of the account appears anywhere in the calculation. And the last row is the only one that is not a property of the contract: the same position, held in an account reporting in a different currency, produces a different figure on the panel.

Gold 

A metal contract is the simplest of the four, because the quantity is a physical unit and the price is quoted per that unit. Ounces multiplied by the price per ounce gives the notional value directly, with no base currency step in between. The consequence is that the notional value moves whenever the metal moves, and so does the requirement taken from it.

An index 

An index has no units to buy, so the contract states a money amount per index point instead. The index level plays the part price plays elsewhere, and the money per point plays the part contract size plays. Multiplying them gives a notional value in exactly the same way, and the requirement percentage is taken from it in exactly the same way.

The money per point is where index contracts differ from each other most, and it is the number most often misread. Two contracts written on indices at similar levels can carry notional values that are multiples apart purely because one specifies a larger amount per point. The index level on its own says nothing about the size of the contract.

A shares CFD, and the conversion step 

A shares contract is written on a number of shares, so the notional value is that number multiplied by the share price. It is also the case where the conversion step is most often live, because a share is quoted in the currency of the market it is listed on, and that currency frequently is not the one the account reports in.

The last two rows are a separate calculation rather than a tail on the first one. The rate applied is the prevailing one at the moment the figure is computed, not the one that applied when the position opened, so the requirement on a foreign currency contract moves with a rate the position itself has no relationship to. An exchange traded fund contract behaves exactly like the shares contract above: a number of units at a quoted price, in the currency of its listing.

Why the figure is recalculated 

Reading the four blocks together makes one thing obvious that is easy to miss when a requirement is met as a fixed money amount attached to a position. Price is an input to the multiplication. On every instrument where the notional value is price multiplied by quantity, a moving price produces a moving notional value, and a percentage of a moving number moves with it. Nothing is traded when that happens and no ticket is sent. The arithmetic simply runs again.

The currency case is the exception that proves the structure. The quantity of base currency is fixed, so the notional value in base currency terms is fixed too, and the requirement in those terms does not move when the pair moves. What moves is the conversion into the account's currency. The result is a requirement that changes for a reason outside the instrument on the ticket, which is why a currency position on an account reporting in a third currency can show a requirement drifting while the pair itself has not moved.

A third convention removes the movement altogether. Some specifications state the requirement as a fixed money amount per contract rather than as a percentage of anything, in which case there is no notional value step and no recalculation. Which convention applies is a property of the instrument and the venue, published per instrument, and not something an account holder selects.

Across the five classes 

The four blocks above deliberately held the requirement percentage constant, so the only thing that varied between them was how a notional value is built. In practice the percentage varies too, and it varies by class in a direction that is fairly consistent from firm to firm because it is derived from the same underlying property everywhere: how far and how fast the notional value can move, and how reliably a contract of that size can be closed when it does.

• Major currency pairs sit at the low end. They are the most heavily traded contracts in existence and they move, in percentage terms, less than most other markets over a given day.
• Major indices sit near the currency pairs. An index is an average of many companies, so the movements of its constituents partly cancel and the average moves less than the shares inside it.
• Commodities and metals sit above the indices as a group and vary widely within it, because a metal and an energy contract behave very differently.
• Exchange traded funds sit across a wide band, and where one lands follows from the breadth of what it holds: a broad market fund behaves like an index, a narrow sector fund more like the shares inside it.
• Individual company shares sit at the high end. A single company can move a long way on a single announcement, with no averaging of any kind to absorb it.

That ordering describes how the numbers are usually arranged. It is not a schedule. The percentage attached to any specific instrument is published in that instrument's own contract specification, requirements within a single class routinely differ by a wide margin, and a percentage read once is not a percentage that holds indefinitely.

Why the same equity stands behind different sizes 

The requirement percentage is a divisor as well as a multiplier, and that is the fact behind a question that comes up constantly: why an identical account can hold a much larger contract value in one class than in another. A lower requirement percentage means a given amount of equity stands behind a larger notional value, and because profit and loss are calculated on the notional value and not on the amount held, an adverse move of a given percentage against that larger notional produces a proportionally larger money loss, one that is not limited to the amount deposited. A favourable move of the same percentage is measured on exactly the same basis and produces the mirror figure.

The two adverse rows are the point of the block. The same amount held produces a debit four times the size in the lower requirement case, from a move of identical percentage, because the notional value it stands behind is four times larger. A requirement percentage is a statement about how much notional value a given amount of equity is capable of standing behind, and about what the arithmetic on that notional value produces in either direction. It is not a measure of how much can be lost.

Where the figure appears 

A YAL account runs on one of two platforms, MetaTrader 5, and both perform this calculation on the server before a position exists. The order ticket reports the amount that would be held for the size entered, recalculating as the size is changed. On MetaTrader 5the instrument's specification window lists the contract size and the requirement alongside the other terms. The arithmetic in this lesson is the arithmetic behind those numbers, and a figure worked out by hand that lands a little away from the platform's own is almost always a rounding difference in the conversion step rather than a disagreement about the method.

Where practitioners disagree 

The first argument is about whether a requirement percentage should be flat across an instrument or tiered by the size of the position. One tradition holds that a larger position is harder to close in adverse conditions, so the proportion held against it should rise in steps as the size rises, and that a flat percentage understates what a large contract costs to unwind. The other holds that tiering makes the number unpredictable at exactly the moment predictability matters, because a requirement can jump when a size band is crossed, and that a published flat figure is the one a holder can verify. Both are in use, tiering more commonly on single name instruments than on major currency pairs.

The second argument is about where the percentage comes from at all. One approach sets it from observed volatility and revises it as conditions change, on the grounds that a requirement standing against a contract that has become more volatile should be larger. The other sets a published figure and leaves it alone except in exceptional circumstances, on the grounds that a requirement which moves on its own can rise at the precise moment an account is least able to absorb it, and that a term revised frequently is not really a published term. The trade between responsiveness and predictability is genuine, neither side has retired the other, and the operative convention is a property of the venue rather than of the calculation described here.