The Stop Loss in Forex Trading : Use It or Avoid It ?

The stop loss is definitely one the most widely used orders in Forex automated and manual trading. This type of order instructs the broker to take you out of your position at the next tick which is worst or equal than the price value which you have assigned to the stop loss. However – since this order-type has existed – there has always been significant skepticism around it as many do not consider a good thing for the broker to know the exact place where a trader wants to exit the market at a loss. The fact is that sharing this information with the broker allows, not only possible foul-play , but it enables the broker to figure out your intended risk per position and other such characteristics of your strategy which you might not be willing to share. On today’s post I will talk about the issue of stops and the inherent disadvantages and advantages to using them on your Forex manual and automated trading strategies.

One of the most important aspects of trading is undoubtedly risk control. All trades must have a predetermined exit before they are even entered in order to adequately control the amount of money which is put at risk on every trade. The easiest way to control this risk is through the use of a stop loss order. There are several advantages to the use of a stop loss order such as a quick exit of the trade when it goes below or touches the SL and the peace of mind that comes from knowing that your risk-control mechanism is stored within your broker, making it invulnerable to platform disconnections and such other nasty things which could make you unable to handle your currently opened position.

However it is also true that there may be a darker side to the stop loss issue. When you enter your SL with your broker they become aware of how many pips the market needs to move before you are taken out. If there are enough stop loss orders near a price level they can fake through regular feed dependency -and the broker is a market maker – then they will most likely take it in order to get the money from your SL (since a market maker could potentially benefit from taking out its own clients if its books are not perfectly balanced). This happens because there is no central exchange on the forex market and a broker can easily fake a deviation of a few pips if it means to take out clients and pocket a good amount of profit.

Definitely it would be very naive to believe that market makers (which are all MT4 brokers, including ECNs — due to the fact that there is no way in which you can truly verify the ECN nature of their orders) will not take advantage of this possible source of profit and it is very likely that they take out stop levels routinely when they have a good reason to do so and they can make the stop-hunting behavior appear like regular broker dependency. You should then assume that every time a broker can get to your stop through a mechanism like the above, it will take advantage of your “known SL” and take you out “to the pip”.

Does this mean that we should never use an SL on our systems ? No. The truth is that not using an SL with MT4 broker is a very dangerous practice since there are many things that could go wrong which could make your system unable to exit a position due to disconnections, server issues, etc. Entering an order which already has the exit in place is a very good practice which will make your broker honor your exit with much more accuracy. Using internal stops makes you subject to disconnection problems and obviously to possible slippage when closing your orders which doesn’t happen when you have an SL in place.

Does this mean we need to “live” with brokers “cheating” us ? No. It doesn’t mean this either. You need to consider how much your strategy is prone to being “cheated” and this will allow you to know how much money you may be losing because of your broker playing “dirty tricks”. Broker dependency tends to be in the region of +/- 5-10 pips most of the time so if your stop loss is below 20-30 pips it is very likely that your broker will be able to cheat you a large percentage of the time. The smaller your SL is the more your broker will be able to cheat you. It is true that even if you have a 100 pip SL your broker may be able to cheat you in the 90-100 pip region but this is the same difference as you would get through regular broker dependency.

It’s also worth considering here that your broker will not take you out this way on ALL of your positions since doing this “faking” is a risky thing for the broker and they may only do this whenever there is a significant incentive to do so, meaning that they will only “stop hunt” when a very large amount of stop orders are “bundled” around a certain price region. So you can only expect them to do this a small percentage of the time meaning that – depending on your SL size – you will rarely get affected by this practice.

Another common practice which is done by EA programmers is to set a “worst case” SL with an internal closing mechanism doing most of the closing on a pre-determinate “internal SL”. This allows you to have “peace of mind” when there are disconnections and also make sure that your broker will not know where your SL is. However it is also naive to believe this will work as your broker does not hunt YOUR STOP (as I have mentioned above) but they hunt stop loss accumulation. If your strategy uses a small internal stop loss it will suffer as much from stop hunting as a properly set stop with the added problem of the worst case stop being hit sometimes (At a larger loss) due to disconnection and other such issues.

So the answer to the question of whether or not to use stops is pretty simple. It is better to use stops – to avoid disconnection problems and such – but large stops should be used in order to avoid the largest part of the stop hunting influence (however always use adequate risk to reward ratios !). As I have said before an internal SL won’t protect you from stop hunting (as brokers hunt for accumulation) and outcomes in the long term might be overall much worse due to added closing slippage and other similar issues. Certainly if you need to trade with low SL values and stop hunting is simply larger than the edge of your strategy you should consider opening up a Currenex account (minium is about 20K USD on most Currenex brokers) where you do not get affected by such issues because the brokers are truly executing positions on an Inter-bank network. However while you trade MT4/5, you should assume that stop hunting will be there and you should therefore build strategies which have edges that do not depend on its absence.

If you would like to learn more about my work in automated trading and how you too can design systems with robust and sound strategies please consider joining Asirikuy.com, a website filled with educational videos, trading systems, development and a sound, honest and transparent approach towards automated trading in general . I hope you enjoyed this article ! :o)

Print Friendly, PDF & Email
You can leave a response, or trackback from your own site.

7 Responses to “The Stop Loss in Forex Trading : Use It or Avoid It ?”

  1. Bruno says:

    Hi Daniel and thanks for this post!

    I think you meant SL instead of spread in the following sentance?
    “Broker dependency tends to be in the region of +/- 5-10 pips most of the time so if your spread is below 20-30 pips it is very likely that your broker will be able to cheat you a large percentage of the time”

    Also, do you think that ALL MT4 brokers are stop-hunting? Not a single broker who is straight enough not to do this?

    • admin says:

      Hello Bruno,

      Thank you very much for your comment and for pointing out that mistake (which I have just fixed) :o) I believe that not all brokers might “stop hunt” mainly due to the strength of the controls they are subjected to. Perhaps you would expect more regulated brokers to stop-hunt less (as they would get caught much easier) but then it would be naive to think this is the case since we have no reason to be certain about this being a fact. The best thing to do is to assume that everyone stop hunts and to develop trading strategies that do not depend on assuming this is not the case. I hope this answers your question :o)

      Best Regards,

      Daniel

  2. fd says:

    Hi Daniel,

    I would agree in most parts, however I have a slightly different view on internal stops. It’s in how you define them. You seem to have the definition that when an order is opened a stop loss in a fixed distance is calculated which will lead to a close by EA logic as soon as it would be hit. Clearly, you are right in saying that this approach has more disadvantages than advantages. However, when you normally close a position based on indicators, then you don’t have a fixed closing level which could give you an advantage in robustness. When your signals are valid a misplaced fixed stoploss will significantly affect your performance. I therefore tend to see stoplosses as a kind of black swan protection only.

    Another aspect which could make closing a position by EA logic preferable to limit orders (it also applies to take profits) is that it avoids hidden trading costs. When a limit order is hit, you will get the next available quote. In most cases this will be very near the defined limit, but it can deviate depending on market conditions and liquidity. With a close, you define which deviation from the quote price is acceptable to you. To be honest, there can be market conditions which will prevent you to get through with your order, however it seems a bit more transparent to me.

    One nasty broker trick is to expand the spread just for one tick. It happened to me, but luckily I noticed it. The nasty thing is that the ask price is not recorded in the MT4 history. So it’s a good idea to watch for spread anomalies continously and check closed trades whether they have been subject to such attempts.

    Regards,
    fd

    • admin says:

      Hello Fd,

      Thank you for your comment :o) You’re definitely right here about several things, including the fact that there are also advantages to the use of internal stops. However the use of an internal closing logic has the same problem as a limit order when closing positions. For example if you want to close a position at a price level you need that price level to exist and neither a stop loss nor an internal closing logic will be able to close a position if such price never exists. So in every case a stop loss placed at a given value will get executed much faster and much more accurately than an internal closing logic if both are to close the position at the same level. Another important aspect – which I mentioned on the article – is that independently of where your internal closing logic (if it acts in the same fashion as an SL) is located, it will also suffer from the consequences of SL hunting which is merely a moving of price within broker dependency to take out stop loss accumulation (and your internal closing logic is therefore likely to follow the same path). Of course closing logic criteria which are not “like an SL” are the best because they close positions based on a given change on the market that forecasts a decrease in the probability of trade’s outcome to be successful (not related to a specific movement against the position like an SL).

      I have also seen this “spread expansion” trick a few times but it certainly shows that you need to use wide enough stops or internal logic to protect yourself from any of this “dirty games” a broker may be playing. As I have said on the article we need to assume that the brokers are going to stop-hunt us all the time so that we can design systems that suffer the least if this is indeed the case :o) Thank you very much again for your contribution Fd !

      Best Regards,

      Daniel

  3. JT says:

    Good post Daniel,

    It always seems where one can make an attempt and effort to some sort of gain for themself, there is some other force behind the works one step ahead of you seeking to quell that gain. Based on what ‘fd’ said the spread widening thing is very bothersome. Lets just face it, it is down right sealing. Stop hunting and similar issues should be where the regulators should jump in, and not worry so much about FIFO rules and no hedging. On the bright side though, in my logical thoughts I suppose, if you have a system that really follows movement and the direction of where the market is going, it might be a bit harder for a broker to put on the brakes and come back to stop a bunch of positions out.

    JT

    • admin says:

      Hello JT,

      Thank you for your comment :o) The problem here with regulating brokers in this regard is that the volume of positions executed is massive and therefore finding this sort of thing is terribly hard. Certainly it has happened and brokers have been fined very large amounts of money (although pennies to them) for doing this sort of thing. The NFA recently fined Gain capital a ton of money due to some of this “dirty tactics” they were executing on their clients but most likely for them it “pays more” to continue playing dirty, paying the “small fine” whenever they get caught by regulators. Assuming that regulators will prevent this from happening would therefore be naive and the best thing to do – beyond moving to a proprietary trading platform on a real inter-bank network – is to develop strategies which simply do not get affected in a significant way by these tactics. Certainly for any system that trades with wide stop orders (either internal or in the form of a limit order) this will be the case. Thank you very much again for your comment,

      Best Regards,

      Daniel

  4. Amir says:

    Hey,sorry there.Actually,the broker can hunt a person stop loss during low volatility.They target me specifically,but at that time I improvise my strategy a little bit,and voila…they cannot stand it anymore and just erase all my open trades,and when I called them,they said that I closed the trades…I lost 90% of my equity because of this s/l hunts…again and again and again…fuhh,almost 6 months I struggled with this problem,and still no way to solve it

Leave a Reply