The purpose for the Committee of Chief Risk Officers (CCRO) recommendations, as stated in the introduction to their 198-page opus, is "to provide guidance on new methods and tools to establish a strong foundation for future growth in this (merchant energy) industry." But the reality is that the recommendations, almost without exception, fail to provide strong leadership in the areas of past and potential future abuse.
Even a cursory walk through the CCRO's four areas of recommendations-governance, valuation, credit, and disclosure-makes this clear.
Inadequate governance, perhaps more than any other issue, has been cited for the excesses and financial chicanery perpetuated by some in the energy merchant industry. The CCRO states that its governance recommendations are "patterned after that of the banking industry."1 The major points of the CCRO's governance recommendations are:
These are all reasonable and fair recommendations. But are they enough? The application of the first three recommendations did little if anything to stop bogus trades and fraudulent deals at Enron and Dynegy (point of note: Dynegy's CRO was named 2001 CRO of the year by the Global Association of Risk Professionals, at the same time Dynegy's cash-flow-inflating Project Alpha was active),2 and El Paso Energy. The industry widely viewed these three companies as having best-practice risk policies and top-notch talent.
The implementation of the forth principle, making mid-office compensation independent of trading results, would be a departure from current norms.3 Unfortunately, this recommendation is naïve, proposed without any recognition as to the likelihood of the success of its implementation, and is destined to fail.
In the past, top risk managers have been able to make 30 to 50 percent of their salary as variable pay based on merchant energy results. While not quite on par with the compensation of top front-office employees, these levels have ensured a talent pool sophisticated enough to keep up with products developed by the front office. If mid-office compensation is moved away from merchant energy results, which alternative might maintain the same level of talent?
Most logically, the alternative should be linked directly to the mid-office's output-the accuracy of its risk models and the reporting of risk. But it is culturally untenable for energy merchants to consider rewarding mid-office employees high variable pay for having accurate risk models if the overall division wallows in losses. More likely the mid-office can expect flat salaries and job cuts if the merchant energy division fares poorly.
Promised earnings from merchant energy activities have been (and, for companies expanding their operations in this area, continue to be)4 relied upon to beat Wall Street earnings per share (EPS) expectations, thus protecting equity valuations. The front office is responsible for meeting these often highly aggressive growth targets, which can have direct significant impact on the wealth of senior corporate and divisional executives.
In recognition of the importance and pressure of meeting earnings, the front office almost always exercises an inordinate amount of de facto organizational power over the mid and back office. As noted above, even in shops with strong independent CROs, the mid-office can be pressured to contribute to earnings5 or turn a blind eye to liberal valuation interpretations.
The CCRO governance recommendations are ineffectual given recent history, the issue of executive compensation, the problematic nature of mid-office independence, and the proposed structure that continues to rely on internal risk groups to be the sole policeman of merchant energy risk. Although the CCRO notes that this structure has worked successfully in the banking sector6, it fails to note that the independence of a bank's risk-management group is augmented by the scrutiny and threat of audit by banking regulators. By ignoring the importance of an independent third party to guard the autonomy of the risk function, the CCRO governance recommendations do not come close to its objective of providing guidelines that will instill the integrity in merchant energy risk practices on par with the banking industry.7
While the issue of flawed governance has received the most press and legislative response ,8 the favorite weapon for earnings manipulation has been the loose standards governing energy instrument valuation. Unfortunately, the CCRO did not rise to the challenge of recommending actionable industry standards but instead issued rather general guidelines that propose little that's new. The CCRO recommendations are to:
Like the governance recommendations, these recommendations are valid and should be employed, but they break no new ground; the majority of companies are already complying. These recommendations are unsatisfactory in that they do not endorse a VaR methodology.
While the CCRO describes the characteristics of the three primary VaR models, it stops short of endorsing a Monte Carlo as the preferred model. Of the three, only Monte Carlo can take into account forward market information and account for the non-linear risk from options.
[Editor’s Note: Monte Carlo simulation is a method of pricing derivatives by simulating the evolution of the underlying variable (or variables) many times over. Monte Carlo is useful in the valuation of complex derivatives for which the exact analytical solutions have not been found, but it can be computationally intensive. Monte Carlo simulation can also be applied to a portfolio of instruments, rather than a single instrument, to estimate the value-at-risk of that portfolio.]
In addition, the CCRO provides no guidance on how to set appropriate VaR limits. The banking rules on which the CCRO modeled its efforts speak at length to risk capital adequacy and how to set VaR limits. The CCRO provides a description of "risk capital" but fails to go to the next level and state recommendations for what should constitute minimum safe levels of risk-capital-to-VaR .9
[Editor’s Note: The value-at-risk of a portfolio is the worst loss expected to be suffered over a given period of time with a given probability. The time period is known as the holding period and the probability is known as the confidence interval. Value-at-risk is not an estimate of the worst possible loss, but the largest likely loss.]
Furthermore, the CCRO does not address the reliability and accuracy of model inputs. The best VaR models and strictly enforced VaR limits are rendered useless if there is not a framework to guarantee the independence and integrity of the data that feeds these models. Given the subjective decisions that go into developing and selecting the volatilities, correlations and illiquid forward curves that are used to calculate VaR, a skilled front or mid-office employee can easily manipulate VaR and MTM upward or downward .10 The CCRO should have emulated the BIS 98 Qualitative Requirements, which recommend model parameters be estimated independently of trading desks.11
Meanwhile, the CCRO provides no endorsement of valuation best practices. Interestingly, the valuation section of the CCRO recommendations provides little detail about energy instrument valuation methodology. For instance, while it is generally understood that energy prices are mean reverting12 and that power prices are subject to price "jumps," the CCRO provides no recommendation that companies use mean reversion and/or jump diffusion as part of their valuation. This is an area ripe for manipulation, since energy companies can significantly manipulate the value of certain derivatives through the selection of valuation methodology.
[Editor’s Note: Jump diffusion is the process proposed by Robert Merton whereby the price of the underlying neither simply jumps nor follows a pure diffusion process but moves by a combination of a jump followed by continuous diffusion. Option pricing models have been extended to incorporate these kinds of jump price dynamics with directional bias, but there are still theoretical problems associated with jump diffusion models. For example, the underlying asset in a foreign exchange option is an exchange rate which can be denominated in either of two currencies. However, jump diffusion models do not give the same prices when compared in a common currency.]
The CCRO report does not provide a standard for stress testing, scenario, and sensitivity analysis. The CCRO spends considerable space in its recommendations to endorse and discuss stress-testing portfolios as a valuable supplement to VaR for risk measurement. In its recurring pattern of stopping short of progressive recommendations, the CCRO does not recommend running a standard set of stress tests, leaving the energy industry behind even the insurance industry, which has defined seven standard scenarios.13
The credit and disclosure sections are somewhat better conceived, although not without imperfections. In the credit section the CCRO does a good job describing the basic Credit 101 mitigation techniques (e.g. netting contracts, collateral, etc.) but marginalizes the importance of information provided by bond market pricing, relying instead on credit rating agencies and internal financial analysis. The CCRO does mention that market bond pricing can be used to price credit risk but makes no mention of the fact that it can also be used to calculate default probability.
Inexplicably, there also is no mention in the credit section of the importance of integrating credit risk with market risk frameworks.14 As is well known, market and credit risk are positively correlated in the energy industry. Thus the evaluation of each independently (or as if they had a correlation of zero) will always overstate a merchant energy company's total risk.
The disclosure section is little more than a set of template forms that can be filled out for internal and/or financial reporting. As discussed above in the valuation section, the composition of VaR under current practices is extremely malleable based on intent. Knowing VaR without an understanding of the underlying market values (e.g. forward price curves, volatilities, and correlations) can be misleading. These reported values can be shifted significantly depending on which VaR model is employed, which forward curves are used, and how volatilities and correlations are calculated. While the CCRO does suggest that the type of VaR model used be divulged, this is not sufficient disclosure. At a minimum, disclosure should include:
If the industry believes that releasing this information is detrimental to its competitive advantage, then a set of common forward curves and associated market data can be developed and used for reporting purposes, as is done in Australia.15 Either way, investors and regulators require an understanding of the underlying data and assumptions that comprise VaR to be able to use this information fully.
Additionally, the CCRO did not go far enough. By focusing only on the disclosure of VaR and mark-to-market values, investors and regulators should also receive:
[Editor’s Note: To mark-to-market is to calculate the value of a financial instrument (or portfolio of such instruments) at current market rates or prices of the underlying.]
Eight of the top 10 energy traders have either gone out of business, left the business, or are seriously cutting back. Liquidity is down 50 percent to 75 percent,16 and the last six months have seen bankruptcies and 78 debt downgrades.17 Thousands have been laid off, and retirements have been destroyed. The existing norms, practices, and regulations for the merchant energy industry need to be reformed. By limiting its recommendations to the most basic and uncontroversial, the CCRO dodged the critical issues that desperately needed to be addressed.
The CCRO had a unique opportunity and impetus to develop and commit to a set of strong recommendations that moved beyond old mindsets, drawing upon the considerable amount of work done in the financial services industries to address these same issues .18 But apparently the process of getting all 31 participants to agree resulted in the lowest common denominator, which unfortunately will not go far in restoring investor and regulator confidence in beaten-down energy merchants.
If the CCRO is serious about its mandate to recommend leading risk management practices to facilitate the redemption of energy merchant sector, it must move beyond the timid cherry-picking it has shown in its current recommendations and address: