Electronic trading increasingly uses programs in which a computer algorithm makes autonomous decisions according to predefined rules and determines, adapts and transmits the related order parameters in line with these rules (algorithmic trading). Some algorithmic trading programs are capable of generating, amending, or cancelling a large number of buy and sell orders within extremely short time intervals. This is referred to as high-frequency trading. As a rule, market participants engaging in high-frequency trading only enter into positions in financial instruments for a short time.

High-frequency trading has increased the speed and complexity of trading. This is associated with risks: for example, large order volumes may place a heavy burden on trading systems. Algorithms may also react to market events and trigger additional algorithms as a result, which may in turn trigger even more algorithms (cascade effect), leading to an increase in volatility.

In order to curb the potential risks associated with algorithmic high-frequency trading, the German federal government agreed on a draft High Frequency Trading Act (HFT Act) at the end of September. TheBundesrat and the Bundestag must still consider the draft before the final act is passed. This article provides an overview of the content of the planned requirements.

High-frequency traders to be supervised
High-frequency traders who are currently not supervised by BaFineither as a credit institution or as a financial services institution will in future need authorisation from BaFin. To date, high-frequency traders were not subject to any authorisation requirement if they only traded financial instruments for their own account and did not provide financial services or conduct banking business.

The future authorisation requirement will apply not only to high-frequency traders who are admitted to trading on a trading venue as trading participants, but also to those firms to which trading participants grant direct electronic access to the trading venue. Direct electronic access exists when a trading participant allows another person to use its identification (trading ID) for directly and electronically transmitting orders to the trading venue. Unfiltered access in which the order does not pass through the trading participants' pre-trade controls is prohibited.

The authorisation requirement will be introduced as part of the planned expansion of the definition of proprietary trading in section 1 (1a) sentence 2 no. 4 of the German Banking Act (Kreditwesengesetz -KWG). As long as a firm is domiciled in another EU or EEA member state and has approval there that includes trading on own account, it does not need any additional authorisation in Germany because of the European passport under the Markets in Financial Instruments Directive (MiFID).

Effective system and risk controls
Under the draft act, investment services enterprises, asset management companies (Kapitalanlagegesellschaften) and self-managed investment stock corporations (Investmentaktiengesellschaften)engaged in algorithmic trading must structure their trading systems so that they do not disrupt the market. This will be required by section 33 of the revised German Securities Trading Act (Wertpapierhandelsgesetz - WpHG). This section will also be referred to by section 9a of the German Investment Act (Investmentgesetz - InvG), which will be amended correspondingly. Algorithmic trading is trading in financial instruments for which a computer algorithm automatically determines the specific order parameters. Order parameters are in particular the decision to initiate an order, its timing, price and quality, as well as how the order will be processed after entry with limited or no human involvement. Systems that are only used for routing orders to one or more trading venues or are used to confirm orders are excluded from the definition of algorithmic trading.

Firms engaged in algorithmic trading must ensure in particular that

• their trading systems are resilient, have sufficient capacity and are subject to appropriate trading thresholds and limits;
• no erroneous orders are transmitted and that the system's functioning in a way that may create or contribute to a disorderly market is prevented;
• their trading systems cannot be used for a purpose that violates market abuse regulations or trading venues' regulations.

The firms must additionally have in place effective business continuity arrangements to deal with unforeseen failures of the trading system. They must fully review and properly monitor their systems. Finally, they must document every modification of computer algorithms that they use in trading.

Certain trade practices are market manipulation
Some trade practices that can be used in algorithmic trading have the potential to manipulate the market. The European Securities and Markets Authority (ESMA) published guidelines on systems and controls in an automated trading environment for trading platforms, investment firms and competent authorities in February 2012 in which it named, among others, the following cases giving rise to concerns:

• Quote stuffing: entering a large number of orders and/or cancellations/updates to orders so as to create uncertainty for other participants, slowing down their process and to camouflage their own strategy.
• Momentum ignition: entry of orders or a series of orders intended to start or exacerbate a trend, and to encourage other participants to accelerate or extend the trend in order to create an opportunity to unwind/open a position at a favourable price.
• Layering and Spoofing: submitting multiple orders often away from the touch on one side of the order book with the intention of executing a trade on the other side of the order book. Once that trade has taken place, the manipulative orders will be removed.

The Market Manipulation Definition Regulation (Marktmanipulations-Konkretisierungsverordnung - MaKonV) will explicitly state in future that certain trade practices executed using computer algorithms are to be viewed as market manipulation. It is irrelevant whether the strategy is executed by means of algorithmic or high-frequency trading. In accordance with section 3 (1) no. 4 of the new version of the MaKonV, purchase or sell orders can be an indication of false or misleading signals that are sent to a market by means of a computer algorithm and are not placed with the intention to trade, but rather in order to

• disrupt or delay the functioning of the trading system;
• make it difficult for third parties to identify genuine purchase or sell orders in the trading system; or
• create a false or misleading impression about the supply of or the demand for a financial instrument.

They may also indicate the fixing of an artificial price level.

Trading venues must also have in place precautionary measures
Exchanges and multilateral trading facilities (MTFs) must have in place appropriate precautionary measures to ensure that exchange prices are orderly determined even in the event of significant price fluctuations. In Germany, some exchange rules and regulations already contain corresponding provisions, for example for briefly interrupting trading in the event of volatility, for switching the market model at short notice and for installing limit systems that require trading participants responsible for price fixing to comply with defined volume and price barriers. In future, these measures will be stipulated by law in section 24 of the German Exchange Act (Börsengesetz - BörsG).

Exchange operators and MTF operators will also have to charge separate fees for excessive use of the trading venue systems, in particular if disproportionate amounts of orders are entered, changed, or cancelled. This planned amendment to section 17 of the BörsG and section 31f of the WpHG aims to reduce threats from high-frequency trading to system stability and market integrity. This is because high levels of rapid entries, changes to and cancellations of orders can strain the stock exchange infrastructure.

The law provides exchanges with discretionary leeway in how they structure the fees, in particular in order to reflect the interests of the trading participants responsible for fixing the prices. However, the fees must be structured in such a way as to effectively prevent excessive use of the systems and their negative impact on system stability.

Order-to-trade ratio and minimum tick sizes
Furthermore, in accordance with the draft of a new section 26a of theBörsG, trading participants will be required to ensure an appropriate ratio between their order entries, modifications and cancellations, and trades that are actually executed (appropriate order-to-trade ratio). This also aims to prevent risks to orderly exchange trading. An order-to-trade ratio is deemed to be appropriate in particular if it is economically reasonable based on the liquidity of the financial instrument concerned, the specific market situation, or the function of the trading enterprise. The exchange rules and regulations must stipulate more detailed provisions on the appropriate order-to-trade ratio. In the case of MTFs, statutory orders can be issued to determine more detailed provisions on the amount of fees and an appropriate order-to-trade ratio.

Under the draft act, exchanges, MTFs and systematic internalisers will additionally have to set an appropriate level for the smallest possible price change (minimum tick size) for the financial instruments being traded. To this end, section 26a will be added to the BörsG and sections 31f and 32c of the WpHG will be correspondingly amended. The background to this rule is the trend towards increasingly small minimum tick sizes. The reduction in minimum tick sizes has led to orders being divided into progressively smaller orders due to the intensified activities of high-frequency traders. This is because high-frequency traders profit from even the smallest price fluctuations as a result of their large trading volumes. This has caused the order-to-trade ratio to increase. Furthermore, minimum tick sizes that are too small can have a negative effect on the price discovery process. When setting the appropriate minimum tick sizes, the relevant self-regulatory initiatives by the Federation of European Securities Exchanges (FESE), the relevant trading venue's market model and the composition of the trading participants can be taken into consideration.

Electronic identification of algorithmic trading
When trading surveillance units monitor daily trading, they cannot discern whether or not a particular order was created by an algorithm. It is equally impossible to match individual orders to a particular trading algorithm if a trading participant uses multiple algorithms. However, a unique match is needed to exclude an algorithm that was incorrectly parameterised or programmed conclusively, quickly and without liability risks from exchange trading in order to prevent risks to the exchange infrastructure.

For this reason, the draft act seeks to expand section 16 (2) of theBörsG to introduce electronic identification for orders generated algorithmically (flagging).

Supervisors' right to information
Finally, BaFin, exchange supervisory authorities and trading surveillance units will have a special right to information under the draft act in order to enable them to conduct better surveillance of firms engaged in algorithmic trading. Section 4 of the WpHG and section 3 of the BörsGwill be correspondingly amended.

The supervisory authorities will be able in future to require information to be provided on algorithmic trading and the systems used for this trading. In particular, they will be able to require a description of the algorithmic trading strategies and the specifics of the trade parameters or the trade limits that the system is subject to. Exchange supervisory authorities can additionally prohibit the use of a specific algorithmic trading strategy if it violates exchange rules and regulations and instructions, or to remedy undesirable situations that could have an adverse effect on the orderly conduct of exchange trading.

European and international developments
Germany is not the only country driving forward the regulation of high-frequency trading. At the European level, the Commission has also made proposals for regulating high-frequency trading as part of the revision of the MiFID, in addition to the ESMA guidelines already mentioned above.

The International Organization of Securities Commissions (IOSCO) had already addressed electronic trading in October 2011. It published a report on the impact of technological changes on market integrity and efficiency on the basis of a G20 mandate.