The Brokerage Business Model
The business model of a forex broker is allowing the client to access the market and getting paid for this service through a variety of fees. Today this access is mostly electronic, which means that the broker has to offer software tools to enable forex clients to trade. Some brokers have developed their own trading frontends such as e.g. Saxo Bank, Dukascopy or Interactive Brokers while others employ tools from third-party software vendors such as MetaQuotes (Metatrader). The business depends on winning many trading clients with a large number of small tickets and charging different kinds of fees.
The different fee types will be explained later on. The important thing about forex trading brokerage is that the business is highly scalable and profitable as long as the broker is generating enough tickets or volume. So, the forex broker is focused on keeping the clients trading as well as constantly winning new clients as a large portion of forex clients go bankrupt or stop trading. Be very careful about the FX broker you choose. Forex brokers should be properly licensed and regulated by a recognized governmental authority, check this Forex Brokers review site before investing your funds.
There are several thousand brokers and introducing brokers the competition is fierce which led to significant pressure on the pricing structure and brought down fees. This means that the public and easy to understand fees such as ticket fees or spreads have been constantly decreasing. Today some brokers are even offering zero spread. Ever wondered how they make money? You will read about that later on. Apart from the fees every trader understands (e.g. spread or ticket fee), there are more complex ones such as spreads on overnight positions or volume fees. Obviously, the brokers are providing a valuable service to the clients and have to get paid for this service. What differentiates trustworthy and reliable brokers from the rest is happens behind the curtains. There are numerous ways a retail broker can improve its own profitability without having to increase visible fees.
The Technical Infrastructure
The basic setup of a broker is a trading front-end where the client inputs orders, a platform to manage all these incoming flows as well as external price feeds for pricing and hedging.
In an easy word, the forex broker is a bridge between the retail client and the professional pricing feeds from many other market participants such as banks. This bridge has a function and delivers value. It offers a trading frontend, keeps the client accounts, manages positions, watches margins, etc. and therefore the services of the bridge have to be paid through any kind of trading fees. What differentiates the trustworthy brokers from the rest is how this “bridge” is operated or in other words how this brokerage infrastructure is configured. As the brokerage infrastructure is primarily a piece of software their many ways how client friendly or hostile it has been set up by the broker.
Forex clients mainly think that brokers take their orders and execute them against “the market” or in other words against incoming price feeds. For this plain vanilla service a service charge is paid, i.e. a spread or ticket fee.
But there is much more logic and complexity involved than just collecting many small forex flows and hedging it out to other liquidity sources in exchange for a small fee. Needless to mention that additional complexity should also be capable of significantly improving the profitability of the forex flows or the brokers P&L respectively – which might not be beneficial to the client or even directly against the client’s interest. The following chapters demonstrate how this can be done.
The Client Relationship
Retail trading clients with margin trading accounts have a very short lifespan. More than 50% of the client accounts are eliminated after less than twelve months. Approx. 90% do not survive the first 18 months without replenishment of the account. Brokers know all about it.
The most important difference between forex trading and professional trading is that the broker has the complete picture of the client. This means the broker has the trading account, the collateral, all insight about the clients trading style (e.g. trade settings, profitability, etc.) and most important: the broker knows the client’s position as all positions are kept in the brokers position keeping tool respectively client account. Read in the section below how the knowledge about the client’s position is used and abused by some less trustworthy brokers.
Every broker wants you to think that you are a valuable client. But in fact, you aren’t. An average retail trader can’t expect to be a valuable client as the broker has too many single clients to take care of 31 and also needs some time to find out which forex trader is actually trading enough to be profitable. First of all most retail traders have a short lifespan. This means most traders produce losses and stop trading. Therefore the single client relationship is not an ongoing source of income for a broker, but rather a relationship which has to be shimmed to compensate for the client acquisition and maintenance costs. Second, most retail traders only have small trading accounts and do not generate many trades. Assuming that a retail trader opens an account with 10.000 USD collateral and then trades in average one 100.000 USD position per day and trades for only one year the calculation for the broker could be e.g.: 50-100 USD fees per million traded sums up to 100-200 USD fees per client per month. Generating a new client through Internet marketing, exhibitions, and sales employees will cost approx. 100 USD per new client. On top, there is a lot of overhead involved in account openings. Furthermore operating a professional and regulated broker in a trustworthy jurisdiction is expensive. So the client has to be active for a few months to be profitable for the broker. This is why retail traders can’t be treated as kings. The trading related Websites are full of users complaining about their brokers and the service offered. It’s an industry without personal relationships and money is only made through high volumes with low individual efforts. All processes have to work perfectly and smoothly or the broker will not be able to generate prof its.
Back Testing vs. Real Money Trading
There are several essential parameters that separate backtesting from real money trading. Many professionals today believe that backtesting is a waste of time as it does not reflect reality. There are many books about backtesting and several core issues such as overfitting or walk forward testing are constantly being discussed in the trading community. So backtesting is a science with many different opinions. Backtesting itself shall not be discussed in this chapter. One of the main differentiators between backtesting and real money trading comes down to IT operations. Backtesting is a quick process that runs for some minutes, hours or days on a PC or server.
It does not require any ongoing IT operations, does not involve any trading discipline or emotional involvement. Real money algorithmic trading requires a professional hosting, maintenance, and operation of a server system. This operations task involves many minor but highly important details such as starting and stopping of the algo for e.g. night trading, weekends, news events, scanning the market conditions for suitability of the algo, changing parameters, loading data, monitoring jobs, etc. Depending on the algo and the IT infrastructure it can be a highly complex ongoing IT operations task that requires IT-professionals and procedures to do it well. It also requires an operations manual with detailed process descriptions and panic procedures. This operations manual shall be written before starting to trade and has to be followed strictly. It shall also help to stay calm in very difficult situations.
I doubt that most traders are capable of performing this operations task well due to either limitation of available time and IT skills. But there is also a broker trick attached to backtesting when backtesting on the broker’s platform. There are tools that allow less trustworthy brokers to influence the backtesting performance. When backtesting on the brokers own platform the broker has access to the backtesting data such as market data utilized, backtests performed, backtesting performance and strategy used. Depending on the broker platform the actual strategy and parameters can be accessed. This means the broker knows the strategy and can influence the backtesting results.
Of course, a very profitable strategy is very likely to be operated in a real money account generating volume and fees. This means there is a financial incentive to manipulate the backtesting results and make the testing more profitable. So the same logic as in demo accounts applies alternating input data (price feed) by e.g. cutting out price data, smoothening volatility, optimizing executions or skewing prices. Some platforms detect heavy backtesting, monitor the outcomes and manipulate the price data according to the strategy tested. Many backtesting platforms also do not offer the feature that backtesting is performed on the bid and ask prices. So the test may run only on bid or on aks, which leads to ignoring the effect of spread and execution inefficiencies. The effect of such execution details affects heavy trading strategies more than longer-term position trading strategies.
As the broker owns and operates the platform the possibilities are limitless and it is just a matter of changing parameters in the software. Most traders – and regulators – will not be able to detect such manipulations. Compare it to the operating system of your mobile phone or your computer. Do you know which data is being collected, sent to the manufacturer and how it may be used? So backtesting on a retail broker platform is just an indicator whether or not the strategy could potentially earn money but does not deliver any proof of profitability. Neither can the parameters for optimization can be derived from such tests even when testing on unseen data/walk forward. Traders that require reliable backtesting need to invest substantial time and money.
The essential toolkit will require unbiased price data from an independent vendor together with a third party software. Both can’t be seen or influenced by the broker. The investment in market data and software tools are mostly too high for retail traders. As mentioned above backtesting is a science of its own and does require expert knowledge. Just running the algo on the broker’s data and the platform is not more than a rough guess.