Margin Higher than the Maximum Possible Loss?

Discussion in 'Options' started by dragonman, Aug 9, 2011.

  1. Hi, I understand that in the U.S. the options margin requirements, which are rule based, should not be higher than the maximum possible loss of an option spread position. So that, for example, the margin of a short vertical put of 10 points should not be higher than USD 1000, under any circumstances.

    However, there are derivative exchanges, such as Eurex, that use SPAN margin system and not the rule based margin system as in the U.S. Theoretically, this should be a favorable margin treatment, since the SPAN is a more sophisticated margin system which try to accurately evaluate the true risk of the position, and should not require margin which is higher than the maximum possible loss.

    Nevertheless, I encountered a different reality: I have entered a few positions to my account, in which I was required to add margin in addition to the premium that I received, so the aggregate amount (premium amount + additional margin) was actually higher than the maximum possible loss. For example, for a short vertical put of 10 points, for which I received a premium of 950 (it was in the money), I was required to provide an additional margin of 100, so the aggregate margin requirement was 1050, which is higher than the maximum possible loss of 1000. Another example: shorting a 12 points iron condor would require margin of 1420, and not 1200. The reason (of my broker and Eurex): the SPAN algorithm calculates the risks of the legs of the positions and may require these additional sums.

    Can someone please explain me the logic of this method? It sounds to me an inferior method to the U.S. rule based margin requirements, in which the trader can know in advance what are the margin requirements and not be subject to some algorithms that "calculate" the "true" risk of the position and actually requires margin which is higher than the maximum possible loss. Thanks.
  2. Please answer my question even if you are not familiar with Eurex margin treatment but with a U.S. portfolio margin treatment (it should be similar since both systems are SPAN based).

    It really sounds strange to me that a sophisticated SPAN or portfolio margin system will require margin which is greater than the maximum possible loss while the old rule-based margin system will require a lower margin on the same spread position which will not exceed the maximum possible loss. Thanks.

    Also, if you have any link to a reliable calculator of a U.S. portfolio margin account please post here.
  3. jef



    eurex uses RBM but it is like span.

    IM=risk + net option value = additional margin + premium margin

    What is the cost to liquidate your position?
    buy your put. It is the premium margin.
    But the prices will change of new price less settlement price
    it is the additional margin and it is equal to the worse loss.

  4. I checked and find that the problem is not unique to Eurex risk-based margin system, but all the risk-based systems, including U.S. portfolio margin, have the same problem, which is that in calculating the maximum possible loss they do not fully take into account the premium received for shorting the option spread. Since such premium should be kept in the account anyway (since it is fully offset against the option value and therefore does not increase the buying power), and in addition to the premium there is a requirement to add a "risk-based" additional margin, the result is that the actual margin (premium + additional margin) is higher, in certain circumstances, than the maximum possible loss.

    Such circumstances include, for example, options strategies such as DITM short vertical puts (in which the premium received could be almost equal to the maximum possible loss and IN ADDITION you would need to provide additional margin), iron condors which their short strikes are ATM (so that the the premium almost entirely cover the maximum loss and IN ADDITION you need to provide extra amount), and there are more strategies in which it occurs.

    The point is that ONLY as long as the total margin is equal or less than the maximum possible loss it seems reasonable to consider some "risk-based" method which would require additional amount due to risk issues, but it seems IRRATIONAL to require additional margin which, in addition to the premium received, is higher than the maximum possible loss.

    Here is an example of an iron condor:
    short SPX 09/17/2011 1175 Call
    short SPX 09/17/2011 1175 Put
    long SPX 09/17/2011 1170 Put
    long SPX 09/17/2011 1180 Call

    For this iron condor I can receive a premium which is close to 500 (minus commissions); lets say I received just 470. Since the risk in this spread is limited to 500, my maximum possible loss, considering also the premium I received, is 30 (500 minus 470), and this is exactly the margin which will be required under the rule based margin system, since it also considered the premium received.

    However, somehow, the OCC's TIMS algorithm, which calculates the margin amount which is required in portfolio margin accounts, requires to add 150 IN ADDITION to the 470 premium that I already received. So that, in effect, I need to keep a 620 margin against a short spread position which its maximum possible loss, under ANY circumstances, is only 500 (it should be noted that although the TIMS does not state exactly that I need to post 620 margin, but only states "150", this is the actual result since for the 150 amount it does not consider the premium that I already received, as opposed to the rule based system). That is, 120 EXCESS AND REDUNDANT margin amount which has no justification whatsoever. Of course, if the TIMS algorithms wants so, he can even require 200, 300, or even more as additional margin, based on some calculation which the trader cannot expect in advance, as opposed to the old good rule-based margin system in which a trader can be sure that he will not be required to post margin which is higher than the maximum possible loss.

    So are the "risk based" margin algorithms, which were initially developed by CME and OCC and were almost blindly adapted by many option exchanges throughout the world, seriously flawed in this regard? Or did I misunderstand something here?
  5. btw you can check by yourself the actual margin which is required in portfolio margin accounts in the following link:

    note that the additional margin is shown as "haircut", and this amount is IN ADDITION to any premium received for shorting an option spread (so that if you received a premium of 470 for a spread which its maximum loss is 500, as in the example above, and see a 150 haircut, this haircut IS NOT part of the 470 premium but is required IN ADDITION to the 470).
  6. jef


    Why didn't you read my answer?
    read more write less.

    you think IM = worse loss
    it is false.

    If you sell a put they have to buy it.
    otherwise the next day the loss can continue.
  7. I already read your answer, and basically the margin should not be higher than the maximum loss of the position (see also jayre post in this regard), and this loss also cannot be higher than the cost to liquidate the position.

    The risk parameter becomes more relevant in positions which are not spread orders or in positions in which the margin requirement is lower thant the maximum possible loss, since in these positions thw worse loss can really be higher than the current margin. However, in an iron condor of 500, European style, the max loss is 500, period. Not 620. This is exactly how the conservative U.S. rule-based margin system works and there is no reason that OCC's TIMS or other risk based margin system, which should be even more favorable than the rule-based system, will require margin higher than that. If anyone thinks differently please advise.