Amaranth – How to Lose $5.0 Billion
without Even Trying
How Did It Happen?
By Cal Payne
Senior Director, Energy Risk Management
This article first appeared in Utilipoint’s IssueAlert of September 25, 2006 and is reprinted with the permission of Utilipoint. Sign up for Utilipoint’s free IssueAlerts ... which give an analysis of the most critical issues in the energy industry ... by visiting www.Utilipoint.com
It wasn't too long ago that critics of the energy merchant sector were applauding the entrance into energy commodities market by investment banks, commercial banks, and hedge funds. These new entrants were sophisticated players with world class risk management capabilities and an almost unlimited access to capital, right? Well it appears in the past two weeks; just one hedge fund has lost more money in a month than the energy merchant sector blow-ups did in aggregate in over 10 years of trading activity. What happened here is difficult to comprehend. How did it happen?
Swinging for the Fence
The losses incurred were in the volatile natural gas market as the result of strategies executed by Brian Hunter, a 32 year old gas trader who operated out of Amaranth's Calgary office. Mr. Hunter had experienced tremendous success for Amaranth lately that represented a large percentage of their recent portfolio returns. While successful, he has a reputation for executing what might be construed as risky trade strategies with both the general market and a previous employer. Traders come in all shapes and sizes with different styles and personalities. My personal preference is for traders who are thoughtful and disciplined in their approach to the market. The kind of trader that cuts their losses when a position moves against them and banks their gains when conditions are favorable. The ones where if you were to measure their results over time, you would see a somewhat jagged line moving consistently upward and to the right on a performance chart.
In opposition to that is the home run hitter. This individual swings for the fence every time they come up to bat. When they connect it is magic but when they miss the results can be devastating. Their graph typically looks pretty scary with higher peaks and deeper valleys. In some cases the term “betting the company” can be applied to these types of traders if the proper control infrastructure is not in place. On the surface, this appears to be the case in the Amaranth situation. Their public disclosures regarding the losses do not indicate that they were the result of unapproved transactions or activities outside an established limitation, but from an extraordinary large position that put the firm at great risk. Mr. Hunter, the head of the energy trading desk, swung for the fence and has put his firm's future in jeopardy.
Without the benefit of seeing Amaranth's portfolio, I can only imagine how large a position was taken to create the potential for such an enormous loss. Various publications this past week have indicated that the company was executing strategies that involved primarily spread type transactions between future months and another involving the purchase of options. The spread strategy is essentially a buy in one period and an offsetting sale in another as a bet on the widening or narrowing of the price differential between the two. The options strategy apparently involved the purchase of what appears to be substantially out of the money options create an opportunistic upside if large price swings were to occur. The nature of those types of trades would seem to indicate a less risky profile. If their public assertions are accurate, one would have to assume that to realize the losses they have incurred, the size of the positions that were taken by their energy trading desk were absolutely huge.
Size Matters
When you bet big, you win or lose big. That goes without saying, but holding an extremely large position can offer other challenges especially when you are trying to exit on an accelerated basis. How long will it take to liquidate the position? Is it a single day, two days, a week, or more? That is a serious issue when you are trying to exit a position gone bad. Market liquidity and its relationship to the position tenure, and the relative size of the position itself are important factors to consider regarding the appropriateness of the size of any position held. We know that Amaranth's positions were big, but how big were they relative to the market in the time periods the positions were held? How much additional damage was done as the result of extra time required to exit the position.
I worked with an investor that was holding a significant position in a relatively thinly traded stock. He had been aggressively accumulating a position in that firm in an environment where the company's share price was consistently on the rise. The investor was convinced that the ever higher price was evidence of the validity of the investment strategy and was looking at a large paper gain over his average cost of acquisition. I felt differently, my analysis indicated that perhaps the primary reason the share price was rising was that his actions were driving the market and that if he were to attempt to harvest that value, his sales would most likely drive what I thought was an artificially high price downward. He essentially was the market with a percentage of ownership around 40% of the shares outstanding. In this case the investor's activity eventually resulted in a sizeable loss; a big departure from the paper gain he thought was there for the taking.
In general, the ability to liquidate a position becomes more and more difficult as the size of the position it represents becomes a significant percentage of the activity at a certain point in time. Could that have been a contributing factor for Amaranth? Holding a large position in the near month contract for natural gas is very different that holding one of the same size two years out. Could this have been a problem? Does holding such a large position itself distort the pricing for that position? At some point, those who are actively trading in the space begin to approximate your position and if you are trying to clear it, they can make the liquidation process more painful, increasing the cost to exit.
Where Were Their Controls?
But wait a second, Amaranth was supposed to have excellent risk management capabilities. That in itself is a broad statement and is subject to interpretation. Do they have the right infrastructure in place? What I mean by that are things like the proper span of control protections, strong governance characteristics, good risk capture systems, fully articulated policies, processes and procedures, a properly empowered risk group, a compliance culture, and a set of portfolio limitations that were consistent with the firm's risk appetite and philosophy. There are several tools available to control the level of portfolio risk. When applied on an integrated basis, they represent industry best practice standards with respect to the establishment of limits around the business. Among these are the applications of some form of a Value at Risk calculation, volumetric limits, durational limits, concentration limits, locational limits, stop loss limits, and the use of stress testing techniques applied to individual trading strategies and the aggregate portfolio to capture the effects of "tail" type events. Did Amaranth employ these?
Let's look at some of these controls. Value at Risk or VaR is considered a best practice standard for measuring the risk of a portfolio. It is a powerful tool that if correctly applied and interpreted, can provide management with meaningful information regarding potential movements against the portfolio and limit the risk positions the company is willing to take. VaR does have its limitations; depending on the VaR methodology utilized and how it is applied, it may not properly reflect a portfolio's embedded risk under situations like periods of accelerating volatility or a rapid breakdown of correlations. It certainly does not quantify what is often referred to as tail type events or risks. So let's assume that their VaR calculation might have been showing an increasing risk in their portfolio but may not have reflected what appears to have been a rapid deterioration of the relationships between time periods in their spread positions.
What about their volumetric limits? In a spread transaction, the net open position in is essentially flat, an equal sale and purchase volumetrically so no catch there if the portfolio is looked at in the aggregate only. However, if that limit was further applied to individual time periods or to spread type transactions, that application could have restricted the size of the position.
If loss limits were part of their control environment, the reduction in portfolio value would have likely triggered a liquidation of the position to limit further damage. If those controls were in place, they didn't seem to work. There is some speculation that rather than take the losses, they may have doubled down and increased their position with the hope of a favorable movement in the market in their direction. What about stress testing? If a robust stress testing regiment existed, there may have been sensitivities run that demonstrated the potential loss of maintaining the position that could be communicated to those involved in managing the company's overall results. If stress testing was done, it appears that the results were ignored. Critical to the effectiveness of any controls was there accurate, timely, and actionable management reporting generated and distributed? Who received the data and was it acted on?
If all of those components of a comprehensive risk management program were in place, what happened at Amarath, short of employee fraud or a violation of policy, should not have occurred. So that begs the question what was Amaranth's risk appetite? Could it have been that big? Were their controls supportive and consistent with established risk tolerances? Were those risk factors disclosed in their offering prospectus to the investment community who entrusted them with their funds? Or is there a story we have yet to hear?
Conclusion
I do not think we have heard everything regarding Amaranth's activity associated with this enormous loss. There are still several unanswered questions, questions that were not effectively addressed by their CEO during a call to investors on Friday. Was the necessary infrastructure put in place and controls adapted that took into consider the nuances of the energy commodities market? In the quest for returns, did the senior management of the fund make provisions for a trader on a good run that were less than prudent? While apologies have been issued, there has been no coming clean on what exactly happened and what the absolute size of the positions that were held by the firm. In a market hungry for greater returns, we should look to this example and others like the failure of MotherRock L.P. that there is no substitute for a strong risk management effort and we as investors should do due diligence before trusting investment dollars to others.
Cal Payne is the Senior Director, Energy Risk Management for UtiliPoint International and has over 25 years of corporate experience in the areas of risk management (including market, credit, and operational risk), as well as treasury, finance, and planning. He can be contacted at cpayne@utilipoint.com or directly at 913-825-6353.