Technological progress has brought us automated algorithms and computer systems replacing floor trading, in which stockbrokers and exchange traders conclude transaction by shouting or gesturing.
These days, the image that pops up when speaking of a stock exchange is probably a huge, grey server room with unbelievable amounts of data being processed in fractions of sec-onds. Technological progress has brought us automated algorithms and computer systems replacing floor trading, in which stockbrokers and exchange traders conclude transaction by shouting or gesturing. For instance, the largest German stock exchange, the Frankfurt Stock Exchange, abolished floor trading already on 23.5.2011, moving all trading activities to a fully electronic trading system called XETRA. This electronic stock exchange XETRA is accessible from anywhere in the world, with market participants placing their orders digitally. The algorithm underpinning the trading system then collects all orders in a central and open order book and executes orders that correspond to each other in terms of price.
Once electronic trading venues and the automated execution of orders became the norm, this also changed the trading behaviour of market participants. While it used to be laborious to execute orders, financial institutions and companies now use this technological progress to their advantage. Buy and sell orders are often no longer executed by a human being but based on programmed, fixed rules by an independently acting trading algorithm. The advantage of such trading algorithms is the system’s speed and rationality. An efficiently designed trading system can process new data, commercial news and other information from the stock exchange’s order book in micro-seconds (one millionth of a second) and make a decision based on it. The trading algorithm has no emotions and strictly follows the pre-programmed rules.
Because algorithms are so much faster than humans, this kind of exchange trading has become very popular in recent years. Already in 2012, algorithmic trading accounted for about 85 % of all trading volume (https://www.experfy.com/blog/the-future-of-algorithmic-trading). Having said that, algorithmic trading is a rather broad term, which has various more specific sub-areas. The following types of automated securities trading can be distinguished:
More and more companies that are not in the energy or financial sector are discovering the advantages of automated trading systems. In comparison to 2015, 2 % more corporations are using algorithmic trading for FX trading, bringing the number up to 10 % in 2016 (https://www.euromoney.com/article/b13k0rb4qz9d4t/corporates-drive-impressive-growth-in-fx-algo-use). For corporations with a trading volume exceeding USD 50m, algorithmic order execution is as high as 24 % for FX trading. The majority of this relates to FX spot trades. The algorithms used most likely fall into the category of Slower Algorithmic Trading. The speed with which the orders are executed is important but in comparison to the ones used by HFT traders, rather slow.
Especially two types of algorithms are popular in corporations and their Treasury departments. On the one hand, it is algorithms that break down large orders into smaller ones. These are called trade execution algorithms. Especially very large corporations frequently require this for larger financial transactions. Entering an order that is large in relation to the order book would have a strong impact on the market. The price of the financial instrument in question would rise or fall due to the high buying or selling interest. Moreover, trading participants could take advantage if such a large order placed in the market were observable and place an order at a slightly better price, knowing that there is market participant backing it up with a large order. The three most popular types of trade execution algorithms are the Time Weighted Average Price (TWAP), the Volume Weighted Average Price (VWAP) and the Per-cent of Volume (PoV) (https://blazeportfolio.com/blog/introduction-to-trade-execution-algorithms-2/).
The Time Weighted Average Price algorithm splits up the order into smaller orders of equal size, trading these over a predefined time period. With the Volume Weighted Average Price Execution trading system, not only time is of the essence but also the traded volume of a financial instrument. The historic trading volume is also taken into account when calculating the order volume. If a greater trading interest can observed in contrast to the historical trading volume, the algorithm can also trade larger portions of the order. The trading volume is also of importance with the Percent of Volume algorithm. With this method, the participation rate of the algorithm in the volume traded on the stock exchange in a financial instruments is determined. With a trading volume of 1,000 shares per minute and a fixed participation rate of 1 %, the algorithm trades 10 shares. In practice, these basic algorithms are heavily tweaked. Often, a certain price spread is programmed which means that the algorithm places an order only if the price lies within this spread. Since there are no limits to what could be programmed into an algorithm, there are a large number of other possibilities to adapt such basic concepts.
The other type of algorithm often used by Treasury departments are trading systems that are tied into various trading venues in order to allow the execution at the best price during a trade. By including various stock exchanges, the prices of certain financial instruments can be compared, allowing for a financial transaction to be carried out at the best market price.
Experts estimate that already 60% of all DAX corporations use such algorithms (Algorithmen im FX-Handel bleiben umstritten, DerTreasurer). The following advantages make using automated trading systems so interesting for large industrial corporations.
Companies not from the financial sector use algorithms mainly for their FX trading. The reason for this is the copious foreign business of large German corporations and the resulting wish to hedge against currency risks. Oftentimes, this makes for very large orders, which is another reason why algorithms that place orders over a certain period of time and distribute them across different stock exchanges are so popular. It is generally banks and larger financial institutions that offer such algorithms. These get together with the corporation to determine which type of algorithm would make sense for the corporation and which adjustments have to be made to existing algorithms. Already existing algorithms could also be used on the trading platform 360T. Various banks offer their trading systems for use there.
Unfortunately, there are no reliable sources to what extent automated trading systems are used in other asset classes traded by treasury departments. They could also be used for commodities trading or for companies concluding a large number of forward transactions.
Having said all this, using algorithms is still the domain of large corporations. Small companies’ trading volume is generally very small and have no influence on market pricing. Because of this, their trades can simply be placed on the stock exchange and do not need to split into smaller orders. Corporations are also increasingly critical of the dependency on banks this creates when using algorithmic trading systems (Algorithmen im FX-Handel bleiben umstritten, DerTreasurer).
However, technological progress has whet the appetite for more. For instance, it would be interesting to have an algorithm communicate with the Treasury Management System and analyze the flow of transactions. Should hedging be required, the algorithm would place the order by itself. Such algorithms would then become attractive also for smaller companies. However, it remains to be seen to what degree such or similar algorithms will be used by companies in the future.