COMPREHENSIVE DATA INTELLIGENCE: empowering the energy marketplace

  • 668

    COMPREHENSIVE DATA INTELLIGENCE: empowering the energy marketplace

    Energy companies today face an unprecedented level of market flux and regulatory complexity. To make informed strategic decisions in such an environment, these companies need access to richer and more comprehensive market data than is currently available. In this article, Arun Karur explains how companies can improve trading decisions and better manage their portfolio risk and compliance requirements by utilizing actual and significant market data.

    In recent years, market understanding in the energy sector has been complicated by major global shifts in supply and demand, as well as increasing supply chain complexity and more stringent regulatory limitations and environmental protection requirements. Given such unprecedented market shifts, the need for energy intelligence—market information based on actual data that is detailed and transparently available—has never been higher.

    Currently, market transaction information is often scattered across inconsistent sources such as swap data repositories and price index publishing services, and can be opaque in terms of underlying data and calculations. Data might be aggregated across arbitrary regions and limited transaction types, and may incorporate assessments rather than actuals. Erroneous market models based on such information can cause significant financial and basis risk for the companies utilizing them.

    Companies in the energy marketplace vary in their level of data access, analysis and usage depending on the maturity of their transaction systems. But all market players are impacted by the dearth of relevant information in the energy marketplace. Trading is often conducted using in-house rules of thumb that have evolved over time, and arbitrage opportunities continue to become smaller and more fleeting. Generally, current transaction approaches in the market have created a speculative and mostly level playing field.

    Market Opportunity

    The energy sector has an opportunity to tap into the potential of a host of energy intelligence that is not yet captured by available market solutions. Consider a utility to which market participants contribute comprehensive transaction level data that includes:

    • Locations—hundreds of which are not currently indexed by any publisher—including physical gas and power, as well as all financial trade index locations
    • Trade types, both financial as well as physical, including those not typically reported such as swaps, same day physicals, physical forwards and bilateral options

    Increased data richness could include market-wide trade counts by properties such as commodity, date, tenor and trade type. Further depth of the data would include information on volumes and prices by various cross-sections such as locations, regions and trade type.

    Early participants in a potential market utility for aggregating market intelligence may face certain concerns that later participants may not. For example, data standardization and associated mapping processes would have to be ironed out. Reliable approaches for data security and anonymization would also need to be developed. Although addressing such concerns would require some investment, current advances in data management technology show that these issues do not pose an insurmountable barrier to a beneficial solution for participants.

    If such market-wide energy data could be validated, standardized and anonymized by a central utility, it could form the basis of significant insights into prices, volatilities, tenors, liquidity, seasonal correlations and operational supply-demand views.

    Energy Intelligence: Filling the Data Gap

    The examples below illustrate the potential of such rich data gathered from a significant number of market participants. They represent situations not readily addressed by currently accessible data.

    Example 1: Market Liquidity
    Consider Company XYZ that has market power and liquidity-related concerns around certain physical gas calls. Current market data available to XYZ includes total volumes traded and some limited information on forward positions. Data for natural gas options is only available on a limited basis for some locations, and is virtually unavailable for physical gas options.

    Figures 1A and 1B below illustrate actual volume data for the company in comparison to its peers, including both physical and financial call options. The volumes in Figure 1A reveal that on this particular day, XYZ is the largest participant for two of the strike buckets in the physical market. Figure 1B isolates the comparative daily physical volumes for one of the dominant strike buckets, confirming a pattern in XYZ’s recent market activity in which XYZ appears to be the dominant player on several days.

    Such robust and actual comparative information slices could provide Company XYZ with a reliable basis for considering alternative trading strategies for this product, such as rebalancing their activity in the physical markets versus the financial markets.

    Figures 1A and 1B

    Example 2: Price Discovery
    Consider Company XYZ that is hedging a significant financial exposure to certain spark spreads. The company wants to understand the level of liquidity in the marketplace and is also concerned that its sizable positions might be exposing it to an informational disadvantage. Market metrics that XYZ might typically utilize include total spot volumes and average spot prices. Market data is limited for forward volumes, as is price information at specific power nodes or physical gas delivery locations. Information on option-based hedging alternatives and actual spread trades is not available currently.

    Figures 2A and 2B below show trade type data that make it clear that Company XYZ formed a sizable part of the exact match for the spark spread swap being considered and may not have a lot of alternatives in terms of other counterparties because of its position. We also see that XYZ did not engage in any costless collars exact matches, or in anything that was nearby, revealing opportunities for it to move into other strategies. Company XYZ also appears to have a presence sizable enough to expose its positions to others in the market. In fact, the prices in Figure 2B show that XYZ paid a premium on its trades relative to alternatives.

    Similar analysis on a range of comparisons could allow the company to understand its options more clearly, while taking other risk considerations into account. One particular strength of the data used in this analysis is that the volumes in 2A are the volumes for the actual trades reflected in 2B. Currently, there is no guarantee that aggregated market data for volume and aggregated market data for price reflect the same or similar trades, making it difficult for market players to draw meaningful inferences.

    Figures 2A and 2B

    Example 3: Valuation Risk
    Suppose Company XYZ holds a substantial portfolio of NYMEX lookalike options of varying maturities and strips and they wish to assess their valuation risk for the portfolio. The typical metric available for use in this case would be an average inferred volatility. Limited detail is available on individual trade prices or volatilities, and none on trade volumes. It is often difficult to identify whether values are based on actual trades or inferred models. Additionally, rather than using a single point for their valuation, XYZ would need a range of price information to understand whether or not they are effectively masking valuation risk by adopting a single point price.

    The chart in Figure 3A shows a traditional volatility smile available from a market service provider, with actual trade data overlaid in the yellow range showing volatilities across the market. Additionally, the bar chart in 3B shows matched volume data for the same trades as in 3A, making it clear for the particular price range how much of the volume is contributed by Company XYZ versus others.

    One revelation in 3A is that the implied volatility smile does not fall uniformly in the middle of the actual market range, and has a bias closer to the top in the negative strike buckets. XYZ’s option trades have a heavy representation in this negative wing, and appear near the top of that range indicating that the volatility smile curve could be dominated by XYZ.

    The volumes in Figure 3B make it clear that XYZ is a sizable participant both in the lower and upper wings of the range, but additionally that in the lower wing, its trades are narrowly concentrated in specific strike buckets. XYZ can now observe how the valuation of its portfolio might swing dramatically if it considered possibilities throughout the market range, and not just along the volatility smile commonly available.

    Figures 3A and 3B

    Example 4: Regulatory Compliance
    In the final example, Company XYZ has received a show cause letter from the CFTC alleging that they traded in a manner that resulted in an artificial price in an over-the-counter gas contract. XYZ currently has access to aggregate information from metric-based benchmarking services—a data set that does not allow comparisons between its activities and those of other players. However, the regulator has on-demand access to additional information, such as SDR data from financial markets. Also, the market range for physical markets is generally unavailable because of low reporting requirements for such transactions.

    Figure 4A shows the data offered by the CFTC to support its claim. It shows the average market price for financials overlaid with information about XYZ’s trades, leading to a conclusion that XYZ may be supporting an uptick in the pricing.

    However, Figure 4B reveals a more comprehensive view into actual trades during this time window, including physicals and buys, as well as sells, with a range of prices and not just an aggregated average.

    Assuming that XYZ’s strategy was to arbitrage the difference between physical and financial markets, Figure 4B paints a dramatically different picture from the conclusion based on 4A. It is apparent from the more complete data set that XYZ was both buying and selling, and that its sales in the physical market do not push the price down. Further, the market range for this data set is wider than that in 4A, diluting the CFTC’s claim about the level of XYZ’s price impact.

    Figures 4A and 4B

    A Market Utility: Additional Services and Benefits

    The four examples above illustrate how rich energy intelligence could empower market participants to make informed strategy decisions, better capitalize on their assets and minimize financial and regulatory risk. Energy companies today face immense reporting challenges due to regulatory demands and the ongoing flux in technical standards. Over time, sufficient data intelligence could lead to a standardized understanding of transactions that would allow for classification and result in more manageable regulatory tracking and declarations.

    This points to the need for a concerted market-wide effort, such as the centralized utility discussed, which could process and standardize actual transaction data captured from market participants, as illustrated in Figure 5. Such a utility could reduce the burden on individual companies by providing:

    • Trading and oversight services, including transaction data merging, monitoring reports and forward curve validations
    • Capital management such as netting, comparison and clearing services for confirmations and settlements
    • Company-level support for regulatory compliance reporting, and industry standardization around locations, transaction types, etc.

    Experience from other sectors that are more mature in the use of data intelligence shows that increased data transparency and standardization leads to higher market participation, adding value to and reducing risk for the sector overall. The efficiencies available to individual participants could benefit the broader industry by:

    • Promoting industry collaboration and process standardization
    • Removing unnecessary technical complexity, cost and risk for participants by offering a centrally hosted solution that is scalable and extensible to meet evolving regulatory requirements and market offerings
    • Providing an integrated global reporting hub, hence easing the compliance process for individual firms as well as industry-wide regulators and service providers
    Figure 5

    Looking Ahead

    Participants in the energy sector should consider the development of an industry-led data aggregation and utilization solution. The availability of actual data for a more complete market picture can help participants address the challenges in the evolving regulatory and market environment. Energy companies have already been mining internally sourced data, but effective aggregation of actual market data has so far eluded the marketplace.

    Similar advances in other markets, such as the financial sector and the retail industry, indicate an untapped but achievable opportunity for the energy sector. Those players willing to participate in industry-wide data provision and analysis efforts would be empowered to differentiate themselves in the marketplace with an increased capability to utilize the resulting energy market intelligence.

    The Author
    Arun Karur

    Arun Karur
    is Vice President and Head of Commodities at Sapient Global Markets. Within this role, Arun helps drive the company’s go-to-market strategy for downstream oil and gas companies, utilities and commodities divisions of investment banks globally. He works with clients to develop strategic initiatives around areas such as business intelligence and data governance, reporting and analysis, achieving transparency into physical and financial aspects of portfolios and capture of trading and asset data for strategic planning.

    Leave a Comment