Login

User Name:  
Password::  
Forgot Password?
Sign up with

Join now (it's free)

User Name:
First Name:
Last Name:
Email:
Password::

Finding a Reliably Fast and Honest Broker

Trading forex is extremely difficult, so difficult that close to 95% us retail traders end up as net losers for a multitude of reasons. What makes it more difficult, if not impossible, is when we end up trading with a broker who is unreliable and dishonest. A dishonest broker can employ multiple methods to make even the best trader or EA into a loser.

What makes a reliable broker? 
  • Fast execution time in opening and closing trades with minimal slippage
An unreliable broker will have inexcusable delays in the opening and closing of positions, which often leads to negative slippage. An unreliable broker will also have inexcusable lost connections that prevent trades that should have opened or orphan trades that should have closed. 

What makes a honest broker? 
  • Low spreads and low commission rate
  • True Swap Rates
  • No re-quotes or very few
  • No slippage when closing trades.
  • Low spread spikes during news events
  • Not running the infamous Virtual Plug-in.
A dishonest broker will  fatten the spread and commission beyond tolerable levels, present costly swap rates, and will turn on the the infamous virtual dealer plugin in order to create spread spikes, execution delay, requoting and negative slippage on client accounts.  No broker needs to do any of this, for there is plenty of money on the table in a modest spread markup alone, but greed + dishonesty motivates too many brokers to engage in the above cheating tactics. 

I have already pointed in Look For Razor Thin Transaction Costs you should be on the look out for brokers the lowest spreads + commissions, as well as brokers with the closest true swap rate (closest to the bank's swap rate without gross manipulation).

In this article I will focus on execution speed and slippage, looking for ways to find the broker with fast executions without delay, requote and negative slippage.  

What makes one broker faster than another? 

It is hard to know for sure what goes on behind the scenes, but I can reasonably speculate that broker execution speed has to do with four factors: 
  1. Supply and Demand Factors
  2. ISP-to-Broker Distance Latency 
  3. Execution Technology
  4. Virtual Dealer Plugin Manipulation
Permit me look at each in turn. 

1. Supply and demand factors

Newbies mistakenly liken their broker to a super market where you pay the price you see rather than a free market where the broker needs to finds someone to take your order. Market markers will themselves be the counterparty to your order but STP/ECN brokers will need to find the best available prices from their liquidity partners (or in the case of ECN, the anonymous liquidity pools). Electronic routing technologies have made this price matching process incredibly quick, but it can never be as fast as instant and seamless auto-execution of the prices as seen with a demo. S/D type delays and resulting slippage are most readily seen during high demand periods (such as after a news event when thousands of orders compete at the same time for the same price, causing most trades to enter or exit too late), or during low supply periods (such as the Sydney session when most banks are closed and most traders are not trading, causing your order to wait for a bit as the broker looks for a match). To avoid S/D delays and resulting slippage, try to avoid trading during high or low volume sessions and just after a news release. Note: to add the problem of delay and slippage during news announcement, there is the problem of gaps, and the array of features built inside the virtual dealer plugin that can cancel, disallow, or severely slip pending orders during gaps and news releases.  

2. ISP-to-Broker Distance Latency

The distance between your ISP and broker ISP can create latency in execution. Reason? When you click on a buy order, that packet of data must travel via the internet from your ISP all the way over to your broker's ISP.  It is measured in milliseconds, or 1/1000th of a second. The closer your respective ISPs, the faster is the delivery of the packets, and the further away, the slower. Latency in online trading can be crucial because data must travel round trip to make a single transaction. A price quote must travel from the broker's server to the trader's computer to be processed and then must travel the same distance back to broker's server to enter, exit or adjust an order. Markets will continue to move while price quotes are in transit, and if market is volatile and latency is high, slippage can occur. When the markets move quickly, delays in milliseconds make the difference between execution at a requested price versus heavy slippage.

The distance delay can be measured in milliseconds by doing a ping test. A step by step test for discovering this type of ping latency is found here: http://www.onestepremoved.com/how-to-test-your-brokers-latency/

Generally, you want your ping test to your broker's server to be low in milliseconds: 50 or less is ideal and 150 or more can be too much. Bear in mind, if the distance between your broker and your server is 9536 miles (that is the distance between your server in Tokyo and your broker in London), it is not possible for latency to be below 32 milliseconds, as nothing can go faster than the speed of light. Here is are two tables that shows latency time to popular brokers based on different locations (use for reference only; don't rush into buying either company's VPS subscription): 
  • Table1 (San Diego, NYC and UK to various brokers)
  • Table2 (US and UK to various brokers)
If you are running a scalping EA and every millisecond delay can be costly, you might want to think about closing the ISP distance gap.

Solution: If was to reduce the ISP-to-Broker latency as much as possible, you can call your broker and ask if they have a server close to your own ISP, or you can set up a Virtual Private Server (VPS) that is closer to the server of your preferred broker. 

3. Broker execution technology

Broker execution technology in retail forex can be broadly divided in two types: the majority market maker brokers (MM-Brokers) who ARE the liquidity providers, and the minority STP/ECN brokers who route your orders to a pool of liquidity providers (banks). We will be discussing each type seperately only in respect to execution speed.

MM-Broker and The Possibility of Fastest Execution

Theoretically, the Market Maker Broker (MM-Broker) can be the fastest form of execution, because the broker takes the other side of the trade in-house, instead of routing it out (or streaming prices in) from liquidity providers as would STP/ECN brokers. By definition, a Market Maker is the counterparty to all its client's positions, and as such he offers a two-sided quote (two rates: BUY and SELL) that mirrors the bank rate. While the MM-Broker can offset between client's opposite positions (or hedge via their LP) in the back office, they are usually the first to take the other side of client's trade (note: this can be very lucrative for them as the odds are that most traders are going to lose 95% of the time). Increasingly MM-brokers (like the MM-banks themselves) have been moving to executable streaming rates instead of request for quote transaction models; this enables clients to click on live rates which are updated in real-time, rather than requesting to deal on a rate which was made "good" for  period of time and may be old. MM-Brokers with executable streaming rates can process trades in milliseconds, if they turned off the delay setting within the virtual dealer plugin. Larger orders greater than 5 standard lots might experience more delay as the MM-broker decides how to offset this greater risk, internally deciding if it can be offset against other client positions, hedged via a partner LP, or hedged against its own capital. Generally, however, most small-time traders trading micro lots might find that they can be filled faster with an honest MM-Broker than with an STP-ECN Broker (though bridge technology out there is quickly closing the gap).   

Caveat:
 The MM-Broker can be the fastest form of execution so long as you can find one that is honest. There is huge temptation out there for MM-Brokers to employ the virtual dealer plugins (discussed below) that create a predetermined execution delay of 1-5 seconds, resulting in slippage in their favor. 

STP/ECN Brokers and the greater likelihood of Reliable and Honest Execution

With an STP/ECN broker, they negotiate the best available bid and ask price from the liquidity provider(s), instead of acting as one themselves, and this can be a plus and negative.

On the plus side, there is less likelihood that they will using the virtual plugin against you, though they can still can. Their business model is to act as a mediator to your order, trying to find and deliver the best prices from the pool of liquidity providers, and for this they charge spread markup and/or commission. As they don't take the other side of your trade, they are less interested in seeing you lose so they can win. 

On the negative side, because the STP/ECN broker is not the direct counterparty to prices they create for you (as with the MM-Broker model), but instead acts as a mediator of prices delivered from the liquidity providers, it can be subject to more real world delay conditions. At times a markets can move very fast, and in these fast times your market order can suffer from millisecond delays and fractual slippage as it struggles to keep up with the rapidly changing prices streamed in from the liquidity providers. It cannot just fill you at the price you see as with a MM-Broker who makes the artificial market (albeit a market that is mirrored from the bank rates).

Important point: The quality and speed of execution can vary between STP/ECN brokers depending on two factors: 1) the depth of the liquidity pool; and 2) the sophistication of the bridge technology. The greater the number and diversity of liquidity providers (whether in the form of contracted LPs or anonymous LPs forming an ECN pool) the easier it is for the bridge technology to stream in, at any given moment, the best bid from one provider and the best ask from another. But not all software bridge technologies are created equal. Some are far more sophisticated than others in hunting for (and streaming in) the best bid from one LP and the best ask from another LP. I have been witness to software that can so quickly and intelligently differentiate and grab the best bid from one LP and ask from another LP that the resulting spread is frequently negative. Yes, you read that correct, negative!--for instance, the EURUSD bid steams in at 1.26436 and at the same moment the ask steams in lower at 1.26432 (result: -0.4 pip spread), instead of how we see it, as a bid streaming in at 1.26436 and at the same moment a higher ask of 1.26448 (result: 1.2 pip spread). Oh, I wish I had access to such expensive software to make spread-free trades; I would make a mint :)

Working with a near zero raw spreads STP/ECN brokers with deep liquidity pools and sophisticated bridging software can make plenty of money on any spread they choose to deliver to us, and so are more likely to try to lower spreads down to beat their competitors and win more clients, and less likely to employ the virtual dealer plugin grab more money from us.

Those with fewer liquidity partners and less sophisticated bridging software may be more tempted to employ the virtual dealer plugin. Any STP/ECN broker can use, with perhaps a lighter touch than their MM-broker counterparts, the virtual dealer plugin to create a predetermined delay, if even only for a second, causing you be slipped by a 0.2 pips or so, which is additional profit for them (particularly if they take what should be your profitable slippage and only deliver back to you the negative slippage). Another thought: if the STP/ECN bridging software is rather slow, it works in the broker's favor and there might be little incentive to make it any quicker : in the interim between when the client executes the market order and when it can be offset by a liquidity provider, prices do change (for better or worse), and if the price moves in the client's favor (positive slippage), the broker can quote at the client's tagged price (and pocket the better price), and if the price moves against the client, the broker can quote at the worst price. This is what the NFA discovered about FXCM (see Slippage Cheater #2 below): FXCM stole their client's positive slippage and they failed to upgrade their electronic software to make it any faster. The greater the software delay, the more potential for slippage, positive and negative, and thus more profit for the broker who wants to steal the positive slippage.  

4. Virtual Dealer Plugin

Have you ever felt that your broker is out to get you? Well he may. Do a google search on Virtual Dealer Plugin and MT4 and read up on it. It is a pretty scary. 

There is a Metaquotes PDF circulating in the internet is a manual for brokers on how to use the infamous virtual dealer plugin to cheat clients. I would post this PDF here (along with a shocking video showing the plugin in action) but I would no doubt be contacted by metaquotes (like so many other websites) for copyright infringement and forced to delete it. 

I can tell you how to get it though:
* for the PDF of the user manual: google "metatrader virtual dealer plugin 4shared" and you will be able to download the pdf document easily from that file sharing website.

* for the video of the plugin in action: type in metatrader into the search box of thepiratebay.se and you will get the torrent of a guy running Metatrader administrator software (available to brokers) along with the virtual dealer plugin, along side a client MT4,  on his own computer. Nice of him to show all the tricks in action (dealer tricks along with virtual dealer tricks)

The PDF document reveals what we all expected, that there is software out there that allows the broker to engage in a number of dirty tricks, namely: 
  1. execution delay (1-5 seconds). 
  2. in the course of the above delay, if the price moves in client favor (potential profitable slippage for client), broker can determine the low percentage (0-10%)  of client orders on volume to be re-quoted at at the better price (along with maximum profitable slippage: 0-2 pips); 90%+ of profit slips are quoted at old price or rejected [Off Quotes]. 
  3. in the course of the above delay, if the price moves against client (negative slippage for client), the client gets re-quoted at new price, with the broker selecting the upper limits of max negative slippage from 1-10 pips.  
  4. gap mode dedicated to processing orders on news (ex: NFP), with conditions to make all pending orders to be filled at worst possible slippage. 
So outside of gap mode, the major way that brokers can cheat you using the virtual dealer plugin is to automate a predetermined execution delay, usually something that they think they can get away with, such as 3 seconds. It is in that 3 seconds that the market price can change and you can get requoted and slipped. A "requote" means that the price of the order is outdated compared to the current market price; usually the price has changed by X pips. Frequently the requote will represent that fact that the market price has moved against your position and so consequently you will suffer negative slippage of anywhere from 0.1 to 10 pips. While it is possible that the price has changed in your favor (positive slippage), you will rarely see profitable slippage because it will be grabbed by the broker. Instead, if the price moves in your favor, what you will see is either that the you will be filled at the old ("tagged") price or your order will be rejected [Off quotes]. 

What I find fascinating about the features of the virtual dealer plugin is the overwhelming temptation it presents the broker to steal so easily, and at such a grand scale, without repercussion. The broker merely has to dial up the execution delay by an additional two or three seconds (so that client does not notice it) in order to create more slippage, the positive side of which he keeps. With this setting in effect, it would be a cash cow for the broker: he would be shaving off thousands of extra pips (and dollars) from the thousands of delayed and slipped trades conducted from his thousands of clients every business day. All this extra slippage would be in addition to the spread markup that the broker receives, so the broker profits an incredible amount from every trade (pip spread + pip slippage), multiplied by thousands of trades per day. All this windfall profit comes without work and without risk. The software does all the pip siphoning and the broker risks nothing other than his reputation, which he might not have to worry about if most clients do not know what goes on behind the scenes, and if most of his competition are doing the same thing anyway. In contrast to the broker, it is the client who bears all the hard work of trading and all the cost: the client bears the cost of trading an incredibly risky market, where the chances are he might lose his shirt, and he bears the additional transaction cost of  visible spread (+ commission) AND invisible slippage.  

Regulations Regarding Slippage

If there are unfair and dishonest means brokers can delay and slip you and earn money from this cheating tactic, then it follows that there must be Regulators out there that have taken notice.  

Yes, as slow as the NFA is to take notice of a flagrant cheating, it did finally investigate and fine two of the largest forex firms in the US, Gain Capital and FXCM, for unfair (asymmetrical) slippage, along with other manipulations. 

Slippage Cheater #1: Gain Capital

Fine by NFA: $459K
NFA Complaint (June 30,2010): here

On June 30, 2010, NFA just issued a severe complaint against Gain Capital and its CEO Glenn Stevens for violating bunch of regulations. It was already an immensely profitable market maker (earning over $150M in gross revenue, $42.5M in net, from 60,000 retail clients) but it wanted to squeeze even more from its clients. One of the topics of the complaint had to deal with dealing desk manipulations: Gain routinely and repeatedly adjusted leverage and margin requirements, causing positions to prematurely liquidate and lose nearly $425,000 (a big loss for the clients which was handsome gain for Gain). The other topic, more pertinent to the theme of this article, had to deal with Gain's use of the infamous virtual dealer plugin, for both its retail and institutional servers. Gain established the following slippage parameters for the Virtual Dealer Plug-In used for both servers: 

Virtual Dealer Plugin Settings Retail Inst'l
Delay (sec)
1
1
Maximum Volume (lots)
50 100
Maximum Losing Slippage (pips)             2
20
Maximum Profit Slippage (pips) 2
3
Maximum Profit Slippage Volume (lots)            5
5

While a 1 second delay might not seem like much, it is enough to create slippage, particularly when prices are changing rapidly in a fast market. Gain manipulated the slippage game in their favor by rigging the volume controls for both servers. Clients trading larger than 5 contracts would be negatively slipped up to 2 pips for retail and up to 20 pips for institional, while all positive slipped trades would have been rejected. Unsurprisingly, the NFA discovered that most retail clients who traded under 5 contracts suffered no slippage losses; the profit slippage evened out the negative slippage.  However, the retail clients that traded more than 5 standard contracts on the retail server experienced $169,502 in losses to to unfavorable slippage, yet never received any gains when favorable slippage occured. Like the retail server, the institutional server also limited profitable slippage to five standard lots, but it also went further in creating asymmetrical settings for maximal losing and maximal profit slippage and allowed for negative slippage up to 20 pips. From May 1, 2009 through July 31, 2009, customers ordering greater than five standard contracts on the institutional server experienced almost $100,000 in losses due to unfavorable slippage when the market moved against them, yet never received any gains when favorable slippage occurred. In sum, MT4 servers gave a lot of dials to cheat traders and Gain could not resist the temptation to play with delay (albeit modestly) and volume controls (more agggresively) to cheat their clients. 

Interestingly enough, because of the invistigations into Gain, the NFA issued a statement in early 2011 saying it will begin investigation involving all US forex brokers in order to find out if any are cheating their clients. The regulator will search for signs these firms are designing computer systems to take advantage of slippage (the price move that happens between when a client orders a trade and when it is executed), looking to see if trades are being executed only when the currency price moves in the firm's favor. Then they discovered later that year that FXCM had been doing this in a very big way. 

Slippage Cheater #2: FXCM

Fined a total of $14M:
-$2M by NFA,
-$6M by the CFTC,
-$8.3M in Restition (to retroactively credit back to its customer the positive slippage it stole from them after June 18, 2008)
NFA Case Document (Aug 12,2011): here
CFTC Press Release (Oct 3, 2011): here

On August 12, 2011, NFA issued a complaint charging FXCM and CEO Niv with with asymmetrical slippage practices similar to what they charged Gain with the previous year. Count I spells out most of the case: for many years, FXCM had practiced what was called "asymmetrical positive price slippage"-- which in a nutshell meant that if there was any price change, from "tagged price" and LP market price offset within their direct access market (DMA) system, resulting from execution delay, FXCM would steal (my word, not NFA) the better price (positive slippage, favorable to client) and fill the client instead at the "tagged price."  On the other hand, if FXCM offset the customer's order at a worse price (negative slippage, unfavorable to the customer), FXCM gave the customer the worse price rather than the better price that was tagged when the customer submitted his/her order. The upshot: From January to September 2010, FXCM derived appoximately $520,000 from positive slppage, none of which it passed on to its customers. Two months later, the CFTC estimated that FXCM stole (my word, not CFTC) 8.3M in positive slippage from clients from 2008-2011, and probably in their complete disgust of this heist and the uncooperativeness of FXCM in the investigation, they issued a whopping $6M fine on FXCM on top of the $2M fined by NFA already, with an order to FXCM to return the $8.3M back to clients. Though $14M seems like alot to us, and it is probably one of the biggest fines given to any entity in the forex industry, it is probably only one month of profit for FXCM.   
Count II of the NFA case is also interesting in that it charges FXCM with "failing to adopt or carry out adequate procedures to ensure the timely and effecient execution of customer orders, regular evaluations of the capacity of its electronic trading platform to efficiently execute customer orders, and the implementation of appropriate modifications to its trading platform to increase capacity when necessary". I think this is interesting because, despite FXCM being the wealthiest FX Retail Broker in the world, it had seen little need to improve the execution speed and capacity of its electronic trading platform. NFA does not speculate as to FXCM's motive for not updating its electronic trading platform to be more fast and efficient, but I have one: FXCM greedily saw that its slower electronic platform could produce more execution delay, which would result in more slippage to clients, and given their asymmetrical positive price slippage procedure, more profit for themselves.

In their damage control press release, FXCM had trumpeted that they have now "enhanced their execution to offer price improvement on all trades," meaning that all orders are now eligible to receive positive slippage. One wonders now if they have learned their lesson and are trying hard to be more honest regarding slippage, albeit only after being smacked in the face with a $14.5M fine, losing face, and being babysat by two regulatory agencies?   

In Jan, 2012, after they had finished dealing with the flagrant slippage cheaters like Gain and FXCM, the NFA formalized its rule regarding price slippage: if you allow slippage in your favor you must allow it in the client's favor as well. The rule eliminates the well known slippage cheat we had previous examined. Let me sum up the previous slippage cheat that the rule sought to eliminate (using my words, not the polite legalize words of the NFA): in the time between when the client places a trade and the time it is accepted by the dealer/plugin, if the price moves in favor of client (positive slippage), broker rejects the order or executes at the requested price, not the better price (broker gets to steal the better price); however, if price moves against the trader, the trade is executed at the poorer price. Thus, before the ruling, broker could steal all positive slippages and client would receive only negative slippage, on all execution delays. After the ruling, NFA member brokers must allow negative AND positive slippage to client.  

I had a lot more respect for the NFA and CFTC after I saw that they had successively stood up to the two largest brokers in the US and punched them in the mouth with serious charges and fines for secretly robbing their clients of positive slippage. They claimed to be investigating all their member firms on this slippage matter, and so I can rest assured that US brokerages now will be policed for slippage malpractice. 

But what about other forex firms operating outside of the NFA and CFTC? Aren't most of them to this day getting away with routinely cheating their clients out of positive slippage, and wouldn't MM-brokers be making additional money given that they act as counterparty to the trade? I cannot know for sure, but given my experience with brokers, I would think so. 

It is unfortunate that there are not many means to differentiate brokers based on execution delay and slippage malpractice in the same way we can differentiate brokers based on spread. However, I have found one company that provides software that you can run to conduct your own investigation on delay, slippage and requotes. 

Tool to Monitor for Delay, Slippage and Requote

Product Name: 4xGuardian (Cost: $149)

It is a product that claims to have great new detection features to find dishonest brokers. 

Features include:

4xGuardian detects and alerts you of the following suspicious activity:
  • Re-quotes
  • Price slippages
  • Delays in execution times
  • Margin changes
  • Spread analysis
  • Broker server disconnections
  • Internet server disconnections
  • Metatrader 4 conflicts detection
  • And much more
4XSentinal (Cost: $99.95), from the same company, is a cheaper alternative and monitors for many of the above as well. 

Note: I am not affiliated with the above company, I just want others to know about a service that tracks for delay, requote, slippage and broker/server disconnections. Perhaps if many of us are using this software and sharing our results on different brokers on the ForexRazor forum, we can together better distinguish the honest from the dishonest. 

Conclusion

Most of us are aware that brokers charge a transaction fee in the form of a slippage and/or commission, but what few of us are aware that they could be robbing us blind on deliberate execution delay and slippage malpractice.  

Slippage is the result of execution delay and it can be positive or negative. If there is a 1-2 second delay from one you click the market order and when it is filled, the market price can move in your favor (positive slippage) or out of your favor (negative slippage). Unfortunately, as we have seen, brokers can delay the execution, which results in more slippages, the positive side of which can be routinely stolen.   

Execution delay is the time between when you click on your order and it is filled. Execution delay can result from a number of factors, such as the natural forces of supply/demand, server-to-broker distance latency, the speed and efficiency of the execution technology, and virtual dealer plugin delay manipulation. You might be able to sidestep some of the delay resulting from high demand/low supply market times, such as avoiding fast moving markets or news announcements, and you might be able to close the server latency gap by locating your ISP closer to the broker server. But the last two delay forces are broker specific: different brokers have different execution technologies with differing speeds, and some may deliberately control the delay via the infamous virtual dealer plugin. We saw how Gain Capital deliberately set a 1 second delay in their virtual dealer plugin to create slippage, the positive side of which they kept if clients traded larger than 5 standard lots. I would imagine many brokers out there setting the delay to 2-3 seconds, and keeping the positive slippage irregardless of the lot size being traded. We saw how FXCM created delay by not updating their execution software to be as efficient and fast as it could be, and their slower software undoubtedly created more slippage, the positive side of which they always kept, regardless of volume. Thankfully the NFA stepped in to regulate these misdeeds and police for slippage malpractice, which has undoubtedly forced all US brokers to seriously improve execution quality and curtail slippage stealing activity.  

As for most other brokers located outside the reach of the NFA, many would be engaged in deliberate execution delay and flagrant slippage stealing as a means of additional profit (in addition to the fat spread they already charge). It is up to each of us to be aware of this more insidious and invisible transaction cost, and set up on real brokers watchdog technology similar to 4xGuardian or 4xSentinel to monitor the delay, slippage, requote activity of different MT4 brokers. By sharing and pooling results of different brokers, we can reward the fastest and most honest with our money and shun the slowest and most dishonest. Brokers who strive for the fastest execution with minimal (and fair) slippage will climb to the top of broker pile, while brokers who delay the execution, deliver frequent requotes and keep positive slippage will fall to the bottom of the heap, to be shunned by all. 

Top
An e-mail with your verification code has been sent to your e-mail address. Please access your in-box and use the verification button or verification code to complete your registration.
You already have an account linked with this E-mail (it maybe standard or social login). Please, sign in with it.
Please, provide us your e-mail so we can verify your account.
Email: