|
|
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
NGI's Code of Conduct and Price Index Methodology
(Updated January 2008)
Code of Conduct
Intelligence Press Inc., more generally known as Natural Gas Intelligence or NGI, operates under the journalism code of ethics, which requires that the personnel creating and compiling the content of our publications and management have no financial interest in the companies on which they report. NGI also warrants that its editorial and price survey employees and management do not trade in the natural gas commodity market, nor in any paper market relating to the natural gas commodity market.
Also, as journalists, NGI employees strive to maintain the highest standards of truth, accuracy, fairness and impartiality. This means double-checking facts, correcting errors as soon as possible, and striving to tell all sides of a story as equally as possible.
NGI also has a record of meeting deadlines and never failing over more than 20 years to publish a scheduled daily or weekly issue. In the real-time world, NGI strives to stay abreast of the news with timely reports over the Internet.
Confidentiality
In order to gain the broadest possible market participation and to protect
our sources' competitive standing, NGI pledges not to reveal the
confidential data that it receives, nor the source of any price information,
nor will it reveal the parties involved in any transaction to any outside
organization, except to the extent that it is legally required to do so (i)
in response to any summons, subpoena, request of governmental or regulatory
authority, or otherwise, or (ii) in order to comply with any applicable law,
order or regulation, (hereafter, a "Required Disclosure"). NGI, before
seeking to disclose Data in a Required Disclosure, shall to the extent
legally permissible notify the Data Provider prior to making such disclosure
in order to permit the Data Provider an opportunity to seek an appropriate
protective order or grant a waiver of compliance with the provisions of this
agreement.
Except for any Required Disclosure, without the prior written consent of the Data Provider, NGI will not disclose to any person either the fact that Data has been made available to it, that it has inspected any portion of the Data, the fact that discussions with respect to the price index surveys are taking place with the Data Provider or other facts with respect to these discussions, including the status thereof; provided, however, that the Data Provider hereby consents to the public identification by NGI of Data Provider as providing information for use in NGI's publication of price index surveys.
Data sent to NGI price editors at the prices@intelligencepress.com e-mail address is archived and encrypted on a daily basis. Only the price survey staff and management have access to the confidential data submitted. Those personnel with access to the data are required to sign pledges of confidentiality restricting the use to purposes of evaluation, compilation, or editorial review of various price index surveys for publication (in aggregated form only and without disclosing the source thereof).
A copy of NGI's proposed Confidentiality Agreement is available at http://intelligencepress.com/features/ngi_confidentialityagreement.pdf
Price Index Methodology
Guide to Providing Data to Natural Gas Intelligence
To improve the quality of the natural gas price survey data and to comply with the FERC's July 2003 policy statement, NGI is asking that data providers adhere to the following guidelines:
- Report all negotiated, fixed-price, non-affiliate natural gas transactions for both daily "incremental" and monthly (bidweek) "baseload" delivery.
- Daily data includes all fixed price deals done each business day (where a business day is any day the IntercontinentalExchange offers trading of daily/next day physical gas) before the 12:30 p.m. EPT pipeline nomination deadline for gas to flow the next day or over the weekend, as is the case Fridays.
- Monthly bidweek data includes transactions done on the last five business days of the month (where a business day is any day Nymex is trading) for gas to be delivered in the following month.
- Nymex physical basis deals done during bidweek should be denoted as such and included.
- Each transaction should be listed separately and should include the following:
- trading location
- transaction date
- beginning flow date
- ending flow date (may be the same as beginning flow date)
- volume
- price ($/MMBtu or C$/Gigajoule for Canadian trading locations)
- buy/sell indicator*
- counterparty*
- Provide data from a central, mid- or back- office reporting source within the company.
- Data providers should make all reasonable efforts to send price data to prices@intelligencepress.com by 4:00 p.m. (EPT) on the business day before the beginning flow date of the data being submitted. Data submitted after 4:00 p.m. (EPT) will be included at the editor's discretion. (Because we understand that data providers may be submitting data to more than one index provider, NGI does not impose any specific file format or specificity pertaining to how the data is arranged within the file.)
- List the contact information of data provider representatives that can answer questions about the data.
- If the data provider becomes aware of errors or omissions to submitted data within three business days, the data provider should make reasonable efforts to notify NGI of those errors or omissions. To submit a correction, email prices@intelligencepress.com with a corrected data submission in the same or similar format as the original submission.
*In the interests of receiving data from the broadest selection of data providers in the near term, NGI currently is not requiring that buy/sell or counterparty information be included. Going forward however, NGI believes that this information, along with the other criteria outlined in this guide are in the best interest of market transparency and price index accuracy.
What is Included in NGI Indexes
- Data from companies who are principles to the trade
NGI receives data, subject to the guidelines above, from companies spanning multiple sectors of the natural gas industry and its customers. This includes companies who produce, market, transport, store, distribute, buy and sell natural gas.
As described above in greater detail, NGI asks these "data providers" to report all negotiated, fixed-price, non-affiliate natural gas transactions for both daily "incremental" and monthly (bidweek) "baseload" delivery.
- Data from IntercontinentalExchange
In March of 2007 Intelligence Press Inc. (IPI) reached an agreement with the IntercontinentalExchange(ICE) that increased the volume of data available for use in the formulation of NGI's natural gas price indexes. Under this agreement, which took effect January 1, 2008, IPI and ICE have devised a protocol that maximizes the amount of relevant data for use in NGI indexes while only including a particular trading company's data once.
Specifically, this protocol enables IPI to include in its index determination process, natural gas transactions from the ICE trading platform for those companies not already reporting to NGI's price survey directly. Most importantly, the process meets the stringent confidentiality concerns of both ICE participants and NGI's data providers. No change in reporting processes or procedures is necessary for those companies who already participate in NGI's price survey.
Indications of Volume of Trading and Number of Deals
In September 2004 NGI began publishing volumes and number of deals along with the ranges and averages of our indexes, which are volumetric weighted averages. For both the daily series which appears in NGI's Daily Gas Price Index and the bidweek series which appears in NGI's Bidweek Survey the deals column represents the number of transaction reports used by NGI in the calculation of the price and the volume column represents the sum of the volume of those deals, expressed in thousand MMBtus/day.
In the case of the weekly series published in NGI's Weekly Gas Price Index, the deals column represents the sum of deals published with the daily series for the appropriate survey dates. Likewise, the volume column represents the sum of the volumes published in our daily series covering the listed survey dates. Please note that the volume published in the weekly series is neither a per week volume nor a per day volume but rather a sum of daily volume collected during the survey period and used in the calculation of the individual daily indexes.
A Note About Volumes
The nature of our survey is such that we attempt to collect,
and include in our index calculations, reports of both the buy and
sell side of an individual transaction. The figures published in the
volume column of our price table represent the volume used in the
index calculation process and should not be construed as the volume
transacted or flowing at that location. Theoretically, if we were to
receive perfect data for a given location, and had received reports
from every buyer and every seller, our volume figure would be exactly
twice the volume of gas transacted. We understand that other
publishing companies tracking the natural gas marketplace report volume
in a similar manner.
With the inclusion of trading data from ICE in January, 2008 the quantity of data available for use in the determination of NGI's price indexes has, in many cases resulted in marked increases in both the volumes and number of deals published alongside our ranges and averages as compared with prior to January 1, 2008. However, these increases are not due to counting trading data from the same company twice. Rather, these increases are a reflection of the data from the ICE trading platform for those companies not already reporting to NGI directly.
How Data is Used to Set Indexes
Data submissions are first separated into their component parts or
trades and initial determinations are made as to which, if any, of the
pricing locations NGI publishes each component applies to. Also an
initial determination is made of whether each component meets our
requirements for applicable trade date and flow date(s) for the survey and a
cursory check is made on the reasonableness of the report (to ferret out
typos and obvious mistakes of transcription). Individual components or entire
submissions may be dropped from our survey at this stage or they may be
flagged for further investigation. Also components may be flagged because of
questions as to the appropriateness of the component or provenance of the
data.
Once data from all or nearly all submissions has been initially processed NGI is left with a set of data for each location we publish. Each of these sets are then taken individually and determinations are made about outliers within the set and determinations are made to identify any irregular data.
Irregular data may be either price or volume levels that are not confirmed by more than one source. When there is a sufficient number of reports in a given set, then statistical measures are often used to help make these determinations, such as reports that are more than three (3) standard deviations from the mean and are unconfirmed by other companies. In addition, reports may be excluded if they are not confirmed and similar trading was not observed by other sources at related locations, at the discretion of the editor.
When irregular data is observed it is often because NGI has made errors in the initial determination of appropriate pricing location or applicable trade or flow date - these are then corrected, either by moving the report to the appropriate location or removing it from our survey. Occasionally the source of an irregularity can not be explained and NGI may exclude a report, at the discretion of the editor. Reports that were flagged in the initial processing of the data are reviewed for their appropriateness, applicability, and reasonableness and may be excluded at the discretion of the editor. Reports may also be excluded because of extremely low volume (less than 1,000 MMBtu/d ) or if there is an indication that there were special provisions tied to the transaction or if the transaction was not conducted at arms length, again at the discretion of the editor.
If NGI notices a pattern of persistent irregular reporting, NGI may contact the company involved. If the problem is not resolved, NGI may decide to exclude all information submitted by that company, but will take that step only after a formal complaint has been sent to the company's directors.
Setting the Prices
Once irregular, inappropriate or non-applicable reports have been excluded
and outliers removed, then NGI is left with a set data set at each
location from which a low, high, and average price can be published. A
volumetric weighted average is calculated using the volumetric weighted
average method, which is expressed by the following formula:
where A = average price, P = price per individual transaction, V =
volume of each individual transaction, 
P.V
cumulative total of each transaction multiplied by itsvolume, TV =
total volume.
Example:
Price | Volume | Factor (P * V)
___________________________________________
$6.26 10 62.60
$6.47 5 32.35
$6.20 15 93.00
$6.31 2.5 15.78
TV=32.5 ∑P*V = 203.73
A = 203.73/32.5 = $6.268
A rounds to $6.27
There is one caveat: If the total volume of reports at a given location for the bidweek survey is below 25,000 MMBtu/d and NGI determines, given other available market information, that the resulting weighted average is an inadequate representation of market activity then an assessment is made using that other available market information. These locations are published marked with an asterisk (*) and a footnote explains that such an assessment was made. If the totality of information is inadequate, NGI does not publish an index.
The weekly series that appears in NGI's Weekly Gas Price Index uses a simple average of the indexes published in our daily series covering the listed survey dates.
A Note About Rounding Rules
For averages, NGI rounds to the nearest cent (one-hundredth of a dollar). When
rounding a half-cent ($0.005), NGI systems will round up or down in a
random fashion. For example, in a large number of trials, $6.235 will
round to $6.23 half of the time and $6.24 half of the time.
For ranges, NGI rounds down on the low and up on the high. For example, a low observation of $6.219 would produce a low for the range of $6.21 while a high observation of $6.281 would produce a high for the range of $6.29. This rounding rule will not impact the computation of the average.
Published volumes are always rounded up to the nearest 1,000 MMBtu. For example, a total volume of 67,200 MMBtu will be rounded to 68.
Error Corrections
Because we understand that the market relies on the certainty provided by
NGI indices and that trading decisions are being made based on these
indices the moment they are published, NGI will only correct errors
within two business days of the posting of the original index. And although
the ultimate decision will be made after taking into account factors such as
the nature of the error and who was responsible, it would be unlikely that
NGI would issue a correction unless the error results in an index that
diverges by at least 1% from the originally published index posting.