Alumni refresher lecture series
2009 – A tipping point
2009 – A tipping point
Highlighting the interplay between the floor-trading exchange and its on-line trading satellite in the discovery of natural gas futures prices
(pages 54-57 of printed journal)
The Global Financial Crisis has highlighted the fragility of the global financial system and has exposed financial markets, institutions and participants to intense public scrutiny. While much of that scrutiny focused on alleged market abuse, it also became painstakingly clear how fundamentally important efficient and liquid financial markets are to sustain the global economic system. Just to remind you, financial markets provide an efficient, cost-effective means of pooling current and future demand and supply for assets, and an efficient means of disseminating material information to all market participants simultaneously. In fulfilling those two functions, derivatives markets ensure fair and equitable price and a utility optimising risk transfer from hedgers to speculators. For those beneficial market outcomes to occur, you need a transparent trading process. A recent energy derivatives market 'scandal' illustrates how a lack of transparency has the potential to jeopardise the integrity of these markets. In the absence of a minimum level of regulation - or in this specific case, a leakage in the regulation - market participants are easily deceived by the partially obscured trading strategy of one major player. What follows is sustained price distortion (a speculative bubble) and the inevitable sharp market correction.
Markets trading in energy assets have only recently flourished. Earlier attempts to organise trading in key energy assets failed for a number of reasons, including a lack of market participants, an excess of regulatory intervention (governments regulating retail and wholesale energy prices), poorly designed contracts, and perhaps even an excess of price volatility due to frequent supply/demand interruptions. So what changed? First and foremost, the 'weather' changed. Climate change concerns have put the spotlight on the economic importance of energy assets and the need for their prices to better reflect scarcity. Second, as a consequence, new trading platforms were developed by private enterprise capitalising on perceived profitable market opportunities. Third, the new trading technologies embraced by these private enterprises to reduce trading costs allowed for better designed - more flexible and bespoke - contracts. Finally, the withdrawal of government regulation at wholesale, and increasingly at retail, levels allowed these markets to set prices (and volatility) that properly reflected market conditions.
Benchmark wholesale and even retail pricing of key energy assets - including crude oil and natural gas - nowadays occurs on a few key energy derivatives exchanges including traditional trading platforms like the New York Mercantile Exchange (NYMEX) and online trading platforms of which InterContinental Exchange (ICE - formerly known as Enron Online) is a prominent representative. The well-known Platt's indicative energy prices are derived from exchange based prices, but intense competition for order flow between the traditional and online trading platforms is clearly shifting in favour of the lower cost online platforms. The traditional derivatives exchanges are aware of this slow erosion in their market share and most are now also adopting automated trading systems to run in parallel with their floor trading system.
To better understand the market impact of multiple trading platforms competing for market share, we need to consider the issue of regulation. Traditional US derivatives exchanges are partly self-regulated, and partly regulated by the Commodity Futures Trading Commission (CFTC). Self-regulation consists of monitoring compliance with market rules, and monitoring trading to prevent manipulation, price distortion and disruption of delivery/settlement. A traditional exchange has the ability to impose position limits on speculators and has authority to liquidate positions. The CFTC monitors the daily trading reports and positions by individual traders of the exchange, and reviews the often price volatile expiring contracts. The CFTC also has the option to make a special data call, which is occasionally used when markets are excessively volatile. Last but not least, clearing members and exchanges have the obligation to report large trader positions.
So-called Exempt Commodity Markets (ECM), which cover most recently developed online trading platforms, face none of these 'restrictions'. ECMs were originally characterised as Over-The-Counter (OTC) markets predestined for traders 'who knew what they were doing,' and therefore could be expected to self-regulate in their best interest. The bespoke nature of their contracts would reflect commercial interests and not be suitable or attractive to speculative interests. While speculators have in fact found their way to these markets, and often provided them with necessary liquidity, regulation has not kept up with these developments. The effective online derivatives lobby has strongly argued that regulation would unduly stymie much needed market development and in any case, these online markets are unlikely to distort the price discovery mechanism. In fact, the settlement prices in online contracts are often designed to 'converge' with the traditional exchange contract settlement prices. So what we observe is a curious situation where a key derivatives exchange is fully regulated to ensure a transparent trading process, yet a satellite online derivatives exchange operates outside this regulator's scrutiny. The risk to market integrity of such dualism is manifest in natural gas derivatives trading.
We first take a closer look at the unusual events that shaped natural gas derivatives trading throughout 20061. Following the disastrous hurricane Katrina - and the disruptions it caused to natural gas supply from offshore exploration to the onshore Henry Hub storage facility in Louisiana - natural gas prices increased to unprecedented levels. The Henry Hub price is the reference point for retail gas pricing on the Eastern seaboard of the US. Henry Hub price volatility has, therefore, significant public and regulatory implications. Retailers use the Henry Hub NG futures price as a signal and a hedging tool. This extreme price volatility offered new trading opportunities for hedge funds that thrive on the opportunities for complicated trading strategies that this situation provided. The Amaranth hedge fund, in fact, made the NG derivatives trade its core focus of trading activity in early 2006. From January to April 2006, Amaranth established significant futures spread positions - predominantly on the NYMEX platform. These spreads involved long positions in winter months' maturities, offset by short positions in spring months' maturities. If, as Amaranth believed, a repeat hurricane season were to hit Louisiana again in August of 2006, the following winter-months contracts would skyrocket in price as they had the year before, while the spring-months contracts would remain relatively unaffected. The resulting widening of the spread between winter and spring would create tremendous profits for Amaranth. Throughout the first half of 2006, this strategy seemed to work as the spread value did indeed increase, although this was entirely due to the demand/supply pressure from Amaranth itself. The fragility of the position became evident in May when Amaranth started to have trouble finding counterparties to trade with, and liquidity virtually disappeared. To avoid the spread value from collapsing and thereby causing cash problems due to the marking-to-market, Amaranth had to keep extending its positions. By late July NYMEX notified Amaranth of its position limits and indicated that it had to curtail its trading activities. At that stage, Amaranth started to move a significant part of its positions to the ICE platform and further extensions of these positions occurred almost exclusively on ICE. NYMEX (and the CFTC) were unaware of this by now huge concentration in Open Interest by a single trader. They could not be aware, as the ECM-exempt ICE trading platform did not report trader positions to the CFTC. While Amaranth had been mostly successful in sustaining its spread position value, by mid-August it started to unravel. Betting market prices on the prospect of a 2006 Louisiana hurricane indicate that, by then, the likelihood of a repeat hurricane had dropped from even odds throughout the first half of 2006 to below ten per cent and falling. It was clear that winter prices would not reach the previous year heights and Amaranth's future position precipitously collapsed. Amaranth formally defaulted in the first week of September and, somewhat interestingly, its positions were assumed by a rival hedge fund that subsequently made substantial profits.
What were the consequences of this collapse? Customers of Amaranth lost their investments. There were allegations that Amaranth had deceived its customers by changing and concentrating its trading activities. NYMEX and the CFTC suffered from bad public relations and a series of government and regulatory inquiries were set up with the potential to ramp up the level of regulation. Perhaps the most important consequence was public mistrust in the price signals provided by derivatives trading in natural gas futures.
To investigate the premise that online trading is merely a satellite to floor trading - and therefore could not feasibly distort the price discovery process - a recent paper2 has taken a closer look at price leadership during Amaranth's trading activities in 2006. Figure 1 neatly summarises its findings.Figure 1: The NYMEX Information Share
The graph displays the inferred information share (the percentage contribution to price discovery) for the NYMEX exchange trading in the Natural Gas futures contract, maturing in September 2006. This contract was first listed for trading in 2001, but any serious liquidity only occurred in January 2006. What we observe from the graph is the dominant NYMEX information share until the end of April 2006, when the ICE information (100 per cent minus the NYMEX information share) rapidly becomes dominant. While this is only a single contract maturity (at any one time up to 72 contract maturities can be traded), Kofman et al call this a signature pattern. It occurs where NYMEX dominates price discovery when a contract is 'far-from-maturity', and ICE dominates price discovery when a contract is 'near-maturity'. The explanation is that liquidity is low in far-from-maturity contracts, and whereas floor platforms have to stand ready to offer liquidity, online platforms do not and effectively withdraw from this low-volume market. For the high-volume near-maturity contracts, liquidity is cheap and speculators will opt for the lower-cost online platform.
Using proprietary trading information, Kofman et al also investigate the impact of Amaranth's trading activity on the ICE platform. They find that Amaranth's ICE transactions significantly increase the ICE information share. Lastly, when the Amaranth-induced artificial futures prices at last collapsed in early September, there was a dramatic reversal in information shares back to NYMEX. Nonetheless, after a short adjustment period, ICE has retained its dominant information share in the near-maturity contracts - undoubtedly due to the low-cost argument mentioned above. These findings certainly put paid to the notion that ICE is merely a satellite exchange with no real impact on price discovery.
The Amaranth affair has exposed the dangerous interplay between regulated and unregulated energy derivatives markets. It is interesting to note that one of the reasons for originally exempting OTC markets from regulation was that their products would cater for a sophisticated clientele. This idea has, of course, been superseded by today's electronic trading platforms used by many OTC markets. Electronic (online) trading has undoubtedly improved trading efficiency in OTC products and thereby reduced the cost of trading, often making these markets accessible for retail traders and speculators. The no-regulation proponents suggest that this 'low-cost bespoke' nature of electronic OTC markets makes them essentially satellites orbiting the central regulated exchange. There is therefore no need to have a fully transparent audit trail that exposes every speculator's trading positions - potentially driving them and the liquidity they offer from the market. Preserving market-wide liquidity is certainly a powerful argument. Yet, if these satellites have the potential to 'distort or influence' prices on the exchange, then a trade-off needs to be made. The 2008 Congress Farm Bill notes that 'if [exempt commodity] markets. play a significant role in setting energy prices, they will be required to register with the CFTC and comply with several regulatory core principles.'
Our investigation of information shares has highlighted the interplay between the exchange and its satellite in the discovery of natural gas futures prices. To avoid a repeat of the Amaranth 'black-out' by the derivatives regulator, two measures are therefore required. First, the regulator should establish and enforce position limits of traders on the exchange - including these traders' positions on satellite markets. Second, the exchange and satellites should be required to publish an Open Interest (OI) concentration measure. This would give market participants some confidence in the 'market-reflecting' nature of the price. Or, if this OI concentration measure is particularly high, it would indicate the possibility that current prices are not truly reflecting the market forces of demand and supply for the underlying asset. Neither of these two measures requires real-time or even public disclosure of trader identity, thereby safeguarding the anonymity so craved by speculators.