The Importance of Order Execution in Intraday Continuous

Order Execution in Intraday Continuous

Order execution — i.e. the ROI-optimized execution of volume and price decisions in continuous markets — is the third step of the short-term power trading value chain. 

In contrast to pay-as-clear trading (e.g. typical of Day-Ahead or Intraday auctions), continuous trading does not assign the same price to all accepted orders within a particular delivery period. Instead, all orders are anonymously displayed in highly dynamic order books, continuously matched and executed whenever a buy order price exceeds a sell order price and vice versa.

Before diving deeper into the role of smart order execution strategies in power trading, let’s take a closer look at how continuous markets function.

Mechanics of Trade Execution in Continuous Markets

All orders submitted in a continuous order book are grouped by direction (i.e. buy or sell) and ranked according to the price-time priority principle. While buy/sell orders with the best price are listed first, orders with the same price are ranked from oldest to newest. Continuous order books are commonly characterized by the following variables:

• Best bid price — the highest price of a buy order
• Best ask price — the lowest price of a sell order
• Bid-ask spread — the difference between the best ask price and the best bid price, which is a strong indicator of market liquidity
• Mid price — the average of the best ask price and the best bid price
• Last traded price — the price of the last executed trade
• Bid price levels 
• Ask price levels 
• Volume available at each bid price level
• Volume available at each ask price level

Below you can find an illustrative example of a typical order book and the process of order execution.

Let’s assume one wants to buy 1MW for a given product. Entering a price higher than the best ask price — e.g. 109 €/MW (a so-called “aggressive order”) — means that the order will be matched and executed immediately at the best ask price (in this case, at 108 €/MWh) as follows:

Alternatively, the buy order can be submitted at a price lower than the best ask price — e.g. 90 €/MWh (a so-called “passive order”) — and “float” until the order is matched. In that case, a trade would be executed at 90 €/MWh. The closer one’s buy price is to the best sell price, the higher the chance of an executed order.

As the example shows, trade prices are set by “passive” orders — i.e. the orders which are floating in the order book. 

The above-mentioned principles are relevant to most order book-based trading systems including stocks, foreign exchanges and cryptocurrencies. However, some characteristics of Intraday Continuous order books in power markets increase complexity:

• Market coupling (XBID). Cross-border Intraday coupling allows for matching orders across different bidding zones from more than 20 European countries. The volume traded across each zone is determined by a market coupling algorithm which attempts to converge prices while respecting physical power transfer constraints. XBID orders disappear from Intraday order books one hour before the start of product delivery.
• One order book per product. Every hourly, half-hourly and quarter-hourly product is traded in a designated order book with respective gate opening and gate closure times — the period during which orders for a particular delivery period can be submitted. At the same time, gate closure times differ between markets (e.g. 5 minutes before delivery in Germany, Netherlands, Belgium and Austria and 30 minutes before delivery in France).
• Lack of liquidity. Insufficient trading activity in some European Intraday markets (e.g. Ireland) can result in large bid-ask spreads and price volatilities.

Development of Order Execution Strategies

Algorithmic trading softwares (or algo-traders) are designed to accommodate the peculiarities of power markets and maximize the efficiency of volume and price decisions. 

Next to commonly known order execution parameters such as stop loss or take profit limits, some specific power trading parameters can be clustered into the following categories:

• Volume distribution. When dealing with a large order volume, traders usually employ one of the following three options:

1. evenly distribute volumes throughout a given trading day via a simple iceberg strategy
2. distribute volumes according to a historical volume distribution
3. buy/sell the whole volume at once

• Pricing (i.e. maximum/minimum price of a buy/sell order). Based on a user-defined limit price (highest/lowest tolerated buy/sell price), algo-traders commonly include a range of sophisticated parameters to continuously optimize and adjust the price of an order.  
• Position closure (i.e. when and how to close a position during a given trading day). The ultimate aim of an algo-trader is to optimize the ROI of a given trade. However, there will always be orders which are “out of the money” — i.e. not within the range of the user-defined limit price. As a result, at a certain amount of minutes/seconds before gate closure, algo-traders are commonly configured to gradually raise order prices above the limit price in order to assure execution.

To illustrate the importance of order execution strategies in Intraday Continuous markets, let’s take a look at the real-life example below.

During the trading day portrayed in the chart, power prices varied between 215 €/MWh and 295 €/MWh for a given product, peaking shortly before gate closure (a common pattern in continuous markets). In light of the growth of renewables, such levels of volatility have become the norm rather than the exception. To maximize ROI potentials and cost-savings of a trading decision, algo-trader parameters are tailored to navigate complex volatility levels to avoid negative “slippage” (i.e. an out-of-the-money deviation from a trading index like VWAP). 

The common algo-trader parameter categorization in power markets primarily caters to asset-backed traders. However, modern automated trading strategies can have execution requirements which make the idea of a single set of parameters superfluous. Especially in the context of high-frequency trading — a complex algorithm-based method of trading that executes a large number of orders in a very short time — the line between decision and execution becomes blurry.

The next generation of algo-traders will therefore need to align intelligent decision making and order execution to enable full-automation approaches in Intraday markets.