Measuring and Assessing the Evolution of Liquidity in Forward Natural Gas Markets: The Case of the UK National Balancing Point.

• Author: de Menezes, Lilian M.
• Position: Report - Statistical table

1. INTRODUCTION

   Liquidity can be defined as the ability to match buyers and sellers at the lowest transaction costs (O'Hara, 1995). This definition impounds a dynamic feature of the markets and implies that in a liquid market, executing a transaction over a short-time horizon does not entail higher costs than spreading the same transaction over a longer horizon. It also evokes the concept of elasticity, such that small shifts in the fundamental values of demand and supply result in negligible price changes when liquidity is high. Consequently, in a liquid market trading activity affects pricing only in a transient and marginal way (Hasbrouck, 2007), and the likelihood of uncompetitive behaviors and price manipulation is reduced. By contrast, illiquidity is a barrier to market entry and a source of competitive disadvantage, mainly to smaller players. Hence, measuring liquidity is critical when assessing market quality.

   In the context of evolving European natural gas markets, how best to measure and assess liquidity has become increasingly relevant, not only to those interested in the cost of hedging and investment decisions, but also to regulators and policy makers, who need to monitor market quality. The liberalization process, by increasing the exposure of market participants to demand-supply imbalances, has fostered the development of forward markets and a move away from the traditional oil-indexed long-term contracts towards hub (spot) pricing. According to IGU (2016), the share of volumes traded indexed to hub prices rather than oil prices has quadrupled in the previous ten years. Greater competition and forward trading have also encouraged the participation of financial institutions (investment banks, hedge and pension funds, and trading companies), which have further contributed to the development of hub trading. Yet, the interest of market participants in trading an asset depends upon its value, as well as on the market mechanisms and costs underlying the trading process (Harris, 2003; Hasbrouck, 2007). As natural gas trading garners increasing interest in European markets, liquidity measurement is relevant when investigating price dynamics, and has implications for efficiency, welfare and regulation of the trading mechanisms and market structure.

   This paper measures and assesses the evolution of liquidity in the OTC forward market at the UK National Balancing Point (NBP, hereafter), which is the most mature hub for natural gas trading in Europe (Cummins and Murphy, 2015; EC, 2015). One-month-ahead forward contracts are considered, as these are the most frequently traded contracts and thus representative of the European natural gas forward market. Using tick-by-tick indicative quotes, transaction prices and volumes from 2010 to 2014, measures of spread and price impact that are drawn from the financial market microstructure literature are estimated. These measures allow for an examination of the transactional properties of the one-month-ahead NBP forward market and the likely effects of trading activity on prices. Therefore, this study also evaluates whether such measures, which have been designed and applied in financial markets, are applicable to the physical natural gas markets. A time-varying approach is adopted, with the intent of addressing the evolving hub trading and exploring changes in market liquidity that might have occurred over the sample period.

   The remainder of the paper is organized as follows. In Section 2, the literature on liquidity measurement in financial and energy markets is reviewed. Section 3 focuses on the data and the empirical methods used to assess the liquidity dynamics in the one-month-ahead NBP forward market. The results are reported in Section 4 and discussed in Section 5. Section 6 concludes the paper.

2. MEASURING LIQUIDITY IN FINANCIAL AND ENERGY MARKETS

   2.1 Liquidity Measurement in Financial Markets

   As per Kyle (1985), liquidity encompasses different transactional properties of a market: tightness, defined as the cost of turning around a position over a short period of time; depth, that is the size of a traded volume innovation required to change the price by a given amount; and resiliency, or the speed at which prices recover from a random shock. Overall, these properties highlight the dynamic feature of liquidity, which reflects the transaction costs carried by investors to complete a transaction.

   According to market microstructure theory, transaction costs include three components, namely: order-processing costs, inventory costs and asymmetric-information costs (Stoll, 1978; Harris, 2003), and can have different impacts on asset prices. Order-processing costs refer to commissions, fees, taxes and other certain costs in a transaction.

   The intuitive meaning of inventory costs is derived from a microstructure model where customers trade only with the market-maker, i.e. an institution or individual that quotes both bid and ask prices in a financial instrument, e.g. futures and forward contracts. Market-makers (e.g. London Stock Exchange, New York Stock Exchange) are normally required to provide sufficient liquidity to the market in order to reduce price volatility and guarantee market efficiency. They trade to make a market, rather than for their own investment reasons, and are subject to uncertainties concerning the future transaction price and volume of an asset. Hence, market-makers assume the risk of holding a certain amount of a particular asset, i.e. inventory risk, in order to provide liquidity to the market and facilitate the trading process of such an asset (Demsetz, 1968; Stoll, 1978; Amihud and Mendelson, 1980, 1986). Inventory costs, therefore, represent the market-maker's compensation for bearing the risk of supplying immediate liquidity; they are different from the physical cost of storage (i.e. carrying costs, such as building and facility maintenance costs, insurance, financing costs) and influence the asset price temporarily (Stoll, 1978).

   An adverse selection problem emerges in the presence of traders that are better informed about the asset fair value (Bagehot, 1971; Glosten and Milgrom, 1985). Informational-based trading is a risk faced by uninformed traders, since the gains of the informed traders are the losses of the uninformed traders. Consequently, asymmetric-information costs can arise in the market, and reflect a balancing of gains and losses due to the presence of informed traders (O'Hara, 1995).These costs have a permanent effect on the asset price and the quantity that can be traded at any given price (Easley and O'Hara, 1987). In contrast to order processing costs, inventory costs and asymmetric-information costs are more difficult to measure and are associated with the transactional properties of a market. As such, they are linked to market liquidity, and may be inferred from measures of market liquidity.

   Since investors consider liquidity when making investment decisions, inventory costs and asymmetric-information costs have implications for hedging and the effectiveness of portfolio-diversification strategies. Several measures of spread have been introduced in the financial literature to investigate market tightness, and discriminate between inventory and asymmetric-information costs. Different econometric approaches have been also used to make inferences about the relative contributions of transaction costs and their implications for liquidity. Furthermore, given the link between liquidity, trading activity and asset prices, several measures and approaches have been adopted to evaluate the impact of trading activity on prices, and thus assess market depth and resilience. These approaches and measures are reviewed below.

   2.1.1 Measures of spread

   Different measures of spread have been adopted in the financial literature, and are mainly devoted to capture the tightness of a market. The most commonly used measures of spread are the quoted bid-ask spread and the effective spread (e.g. Chordia et al., 2000; Bessembinder, 2003; Goyenko et al., 2009; Foucault et al., 2013). The bid-ask spread is the difference between the best ask price and the best bid price, and represents the transaction cost paid by a customer to the market-maker for a round-trip, i.e. a purchase followed by a sale of the same amount. This measure of spread is associated with inventory costs (Roll, 1984; Stoll, 1989), nonetheless it can overstate the actual transaction costs if: the market-maker (i.e. a participant who undertakes to buy and sell at specific prices in a market) adjusts the bid-ask spread to control the inventory level; or, in the presence of asymmetric-information (Stoll, 1989). Moreover, in the over-the-counter (OTC) markets, where a centralized trading platform is absent, buy and sell trades are negotiated through inter-dealer brokers. Although the bid and ask quotes are posted by the inter-dealer brokers based on actual trading orders and expressions of interest, they are not binding (Jankowitsch et al., 2011). As a result, transaction prices can be different from the bid and ask quotes, thus the quoted bid-ask spread can be a misleading measure of market tightness in OTC markets.

   The effective spread was introduced in the financial literature as an alternative to the quoted bid-ask spread and provides a more reliable measure of tightness. It reflects transaction prices that are negotiated either inside or outside the indicative quotes (Huang and Stoll, 1996), and is computed as the difference between the actual transaction prices and the average of the bid and ask quotes, namely the midquote, which is a proxy for the asset fair value (Bessembinder, 2003; Foucault et al., 2013). Since the midquote is a basis to evaluate whether the buyer is paying a high price and the seller is receiving a low price, the effective spread can be totally ascribed to the trading process, thus measuring transaction costs in the market. This fact has been...