Margin Call Mistakes

As most of you know, I spent about half this year travelling the world. During this time I was fairly disconnected from the Internet, but that didn’t stop me from leaving an experimental strategy trading. This experimental strategy was higher risk – a high leverage reversion strategy – an offshoot from my grid work done earlier but not a grid at all, and definitely not as risky. However, it was highly leveraged and had a very small chance of blowing up. If the timing is right and I withdrew from the account regularly, I believed I could make it work.

I put a little bit of money in a sub-account and let it trade while I was travelling. Given the dedicated account and smaller balance, should it blow up then the losses would be limited. It was a risk I was willing to take. So I did it, and for a while it was working.

Margin Call

The above is a screenshot of MyFxBook for that account – a real account. I had multiple instances of the strategy running for different currency pairs. Initially, as you can see the results were good. Within 6 months, it had made 23% … 

However, I’m sure you’ll notice the large loss on October 16. What happened? It almost lost 50% of the account…. As the title of this post would suggest, I was Margin Called. However, it wasn’t the strategy’s fault, it was my fault.

A number of things took place for this crucial mistake to happen.

How well do you understand Margin Call anyway?

Firstly, it turns out my understanding of how margin is used, and how one could get margin called was incorrect. A newbie mistake, coming about from ignorance and being too eager to “make money” when I first started in Forex, and never going back and correcting the wrong.

Prior to this strategy I had never engaged in any trading that could result in having great losses. I always limited my losses with a Stop Loss, and never risked more than a couple of % per trade. Since I always stuck to this “safe” trading, I never really cared about Margin Calls because I never came close to it. However, this strategy was different, and I should have been more careful.

Margin call is not just about unrealized losses…

Clients that request customized strategy implementations from MooMooForex sometimes have their own “margin call” requirements, whereby the strategy should close all positions when a certain amount of loss is reached. This implementation certainly would not stop a real Margin Call should their position size be sufficiently large…

Margin, as it turns out, is not just used up by your losses but it is of course used up by your trade size. When placing a trade, the UI shows you the amount of Margin that will be used up by this trade – do you ever pay attention to this? It is possible to have a Margin Call even when you have very little unrealized loss. This happens when your trade sizes are so big that you use up all your margin as soon as you open the positions.

Given the re-entry style of this strategy, combined with the scaling, I really should have considered this. Not only for risky strategy, but this should also be considered for any strategy that you are planning on scaling up.

If one has a $10,000 account, and their max draw-down is predicted to be $2000, then one could easily believe they can scale up their strategy 4 or 5 times before they risk “losing all their money”. However, that does not work for Margin calculations. It might well be that if you traded 5 times as much, you would blow your margin before incurring any losses!

The back-tester will show me the error of my ways…

Even if I did not understand and consider the impact of Margin properly, I knew the back-tester would. I had seen risky strategies completely bomb out in back-testing because of the dreaded Margin Call which lead me to believe that if the back-test results are good without margin call, and the max draw down is tolerable, then I would be fine if I had traded accordingly during those periods.

Well, this is what the back-tester showed me when I tested this scenario later…

Margin Call Backtesting 1

There was indeed a loss, but overall the strategy almost broke-even, but this doesn’t match the actual trading results? The point of course is in the equity drop there. A large equity dip like that should not be ignored. The real question is, why did this not trigger the margin call?

Are you back-testing with the right trade size and account details?

Normally when I back-test, I usually just use my default trade amount configuration. This is because I’m merely looking at the mathematical metrics of the results and since losses and gains become adjusted proportionally to your trade size, it has no impact on the data. However, that is not the case for a margin call.

Additionally, the account balance I use reflects my main trading account balance, and it was not set to the balance of this new sub-account I was using for this strategy – which had considerably less funds.

If I re-run the back-test using the actual amount sizes of my trades and the account balance of my account, this is is what I see.

Margin Call Backtesting 2

Now the realized loss is much more significant – 42.1% of the account to be precise. Looking back to the previous years, there were a number of times where this happened and I would have noticed if I had been back-testing on the right account balance and the right trade sizes.

Conclusion

So in the end, the strategy would have made an acceptable loss of a few hundred dollars instead of several thousand. It was an expensive mistake to make, but I’m glad I made it now before I really start ramping it up. If you have a strategy that ends up holding large positions at any point, then remember that margin calculations cannot be ignored even if you have a winning strategy.

 

Posted in Backtesting, Research, Reversion and tagged , , , .

Leave a Reply

Your email address will not be published. Required fields are marked *