We all do it. You need to understand something and the first thing you do is google it and invariably one of the first hits you get is Wikipedia. The problem is that Wikipedia isn’t always correct. When it comes to Margin, it can lead to a lot of confusion.

Initial Margin

If you google Initial Margin, this is what you get:

“The initial margin requirement is the amount of collateral required to open a position. Thereafter, the collateral required until the position is closed is the maintenance requirement. The maintenance requirement is the minimum amount of collateral required to keep the position open and is generally lower than the initial requirement. This allows the price to move against the margin without forcing a margin call immediately after the initial transaction. When the total value of collateral after haircuts dips below the maintenance margin requirement, the position holder must pledge additional collateral to bring their total balance after haircuts back up to or above the initial margin requirement.”

Even I’m confused by this! The problem is that this description is US-centric and probably about 30 years out of date for exchange traded products. Initial Margin as a name doesn’t help. There is no specific margin calculated for adding a new position to an existing ETD portfolio; margin algorithms such as SPAN are a portfolio measure, taking into account the risk of new and existing positions. And any increase in margin caused by a new trade will generally only be calculated after the trade has been accepted for clearing.

Non-US clearing houses tend not to have the concept of Initial and Maintenance Margin — the 1.1 multiplier applied by CME to some accounts. They may apply a multiplier to specific accounts, but the margin calculated is called Initial Margin at all times. And anyway, even if one position is closed, you will probably still have open positions within your portfolio and hence still have an Initial Margin requirement.

As for the need to post additional collateral, it will either be because the market has moved against you and your portfolio has made a loss or because the risk of your portfolio — and hence the Initial Margin requirement — has increased, so more collateral is needed to cover it.  The concept of getting it back up to or above an initial requirement based on a single position just doesn’t make sense.


So if Wikipedia confuses in its attempts to describe Initial Margin, how does it do on, everybody’s favourite margin algorithm, SPAN? This is what you get:

“The Standard Portfolio Analysis of Risk, or SPAN, is a system for calculating margin requirements for futures and options on futures. It was developed by the Chicago Mercantile Exchange in 1988.

SPAN is a portfolio margining method that uses grid simulation. It calculates the likely loss in a set of derivative positions (also called a portfolio) and sets this value as the initial margin payable by the firm holding the portfolio. In this manner, SPAN provides for offsets between correlated positions and enhances margining efficiency.”

At least the first paragraph here is correct! The second paragraph is just not very helpful. And I have no idea what it means by “grid simulation” — SPAN is not a methodology based on a “grid” of parameters and nor is it a VaR algorithm with thousands of scenarios that needs grid computing to be able to calculate the results in a reasonable amount of time. At its heart, SPAN is a simple scenario-based algorithm with a mere 16 scenarios.


Initial Margin isn’t the only type of margin, for example, there is Variation Margin and Delivery Margin. So what do you get on Wikipedia for Margin?

“In finance, margin is collateral that the holder of a financial instrument has to deposit with a counterparty (most often their broker or an exchange) to cover some or all of the credit risk the holder poses for the counterparty. This risk can arise if the holder has done any of the following:

  • Borrowed cash from the counterparty to buy financial instruments
  • Borrowed financial instruments to sell them short
  • Entered into a derivative contract.

The collateral for a margin account can be the cash deposited in the account or securities provided, and represents the funds available to the account holder for further share trading. On United States futures exchanges, margins were formerly called performance bonds. Most of the exchanges today use SPAN (“Standard Portfolio Analysis of Risk”) methodology, which was developed by the Chicago Mercantile Exchange in 1988, for calculating margins for options and futures.”

Let’s look at the first sentence. That’s actually not too bad — although the margin is paid to the CCP (otherwise known as Central Counterparty or Clearing House) rather than the Exchange. And although margin is theoretically there to cover credit risk — having enough funds available to sort out the mess if the margin provider defaults — most margin calculations are based on market risk. For example, CCP Initial Margin algorithms are intended to cover potential losses incurred between a clearing member defaulting and the CCP closing out or auctioning off their position.

The bullet list is very cash and security–centric. It feels as though derivatives have been added as an afterthought — this is not helped by the description of a margin account focusing on share trading —  and there is no distinction made between Exchange Traded Derivatives and OTC.

If you check the CME Website they still tend to talk about performance bonds rather than margins — so the name hasn’t changed and we are still “two nations divided by a common language”!

It will be interesting to see when the final line gets updated. I’m old enough to know that the statement is historically correct, but more and more CCPs are moving away from SPAN towards VaR based methodologies (including CME themselves — although they are calling their algorithm SPAN 2).

On the Margin Wikipedia page you will also find the following:

Types of margin requirements

The current liquidating margin is the value of a security’s position if the position were liquidated now. In other words, if the holder has a short position, this is the money needed to buy back; if they are long, it is the money they can raise by selling it.

The variation margin or mark to market is not collateral, but a daily payment of profits and losses. Futures are marked-to-market every day, so the current price is compared to the previous day’s price. The profit or loss on the day of a position is then paid to or debited from the holder by the futures exchange. This is possible, because the exchange is the central counterparty to all contracts, and the number of long contracts equals the number of short contracts. Certain other exchange traded derivatives, such as options on futures contracts, are marked-to-market in the same way.

The seller of an option has the obligation to deliver the underlying of the option if it is exercised. To ensure they can fulfill this obligation, they have to deposit collateral. This premium margin is equal to the premium that they would need to pay to buy back the option and close out their position.

Additional margin is intended to cover a potential fall in the value of the position on the following trading day. This is calculated as the potential loss in a worst-case scenario.”

There are plenty of misleading statements in this little list.

The main difficulty here is the first and third bullet points. Basically, they are describing exactly the same thing, the only difference being the thing that has been traded — a security versus an option. In fact, Premium Margin isn’t the only term used for this margin for options, Net Liquidating Value (or NLV for short) is also often used.

Variation Margin is a mark to market. The only problem is that it isn’t always realised. For futures and future style options the description is correct, but for forward contracts — like those trade on LME — Contingent Variation Margin is used. Here, the profit or loss is calculated based on the difference between the current market price and the original trade price. If this is a loss then collateral must be provided, but if it is a profit it can be used to offset other margin, in exactly the same way that Premium Margin (NLV) can be used.

As for the description of Additional Margin, this reads more like a description of Initial Margin to me. In the world of clearing, Additional Margin is usually used as a coverall for Margin requested from Clearing Members when the CCP has a risk concern of some sort, but there isn’t currently an official margin component to cover it. For example, Additional Margin may be required if it is felt that a portfolio carries too much risk compared with the capital of the clearing member.

Just this small peak into the world of online descriptions of margin has made me understand why some of our clients can be confused about the margin they are being asked to pay. It’s given me an idea to write an “OpenGamma Guide to Margin”. Hopefully next time you google “margin” you will find this instead and everything will be much clearer!