Collection and Storage of Price Data
Calculation of Reference Prices
Advisory Committee

Collection & Storage of Price Data

Process for Collection of Price Data

The Steel Index is working with a broad group of companies that have agreed to submit steel price data based on actual transactions.  The 'data provider group' includes mills, traders, stockholders, fabricators and end-users in all relevant regions.

Each data provider enters into an agreement with The Steel Index.  This guarantees that the identity of each and all data provided will be kept completely confidential.  

Each data provider submits price and other data related to actual transactions to The Steel Index via simple forms on a secure, password-protected Internet site.  Confidential communication is ensured by the use of a GlobalSign ServerSign SSL certificate using 128-bit encryption. 

In return, each data provider receives a weekly e-mail with the aggregated reference price data and a free subscription to The Steel Index website where a price archive is available.  The site also offers the Price Analyser tool, enabling each data provider to analyse different reference prices and compare its own prices with those reference prices. 

In parallel with the collection of price data, The Steel Index collects additional ‘softer' information of interest to the industry - such as stock level movements and sentiment on the market outlook.  This is also disseminated to data providers via e-mail and web-site services.

Data Collected

The Steel Index collects from each data provider its informed view of the average transaction prices in a given week for a selection of specified basis products, based on the actual transactions it has completed that week for relevant and similar products for the next delivery period.

Each data provider also submits the following information alongside the average transaction price for each specified product:

Total volume of relevant transactions completed that week

Average order size

Average delivery lead time (weeks)

This enables The Steel Index to understand the context and assess the integrity of the data provided as well as calculating the necessary volume-weighted average price.

Storage of Price Data

The Steel Index uses the latest IT security technology to ensure the confidential ‘raw' data submitted is protected from any misuse or manipulation.  All data providers are assigned a confidential identification number known only to a very limited number of senior level staff within The Steel Index.  All ‘raw' data is stored against the confidential identification number, with no record of the company name. 

Back to Top

Calculation of Reference Prices

The calculation methodologies employed to compile The Steel Index reference prices are in line with those used in the formulation of other established and respected indices. 

The reference prices for each product/region are calculated as the volume-weighted average of the price data points submitted, subject to the following procedures: 

• Firstly, the mean average price of the entire price data sample for a given product/region is calculated; 
• Secondly, any price points more than +/- 20% from the mean average price are considered outliers and excluded from the price data sample;
• The standard deviation of the remaining price data sample is calculated;
• The highest and lowest price points submitted are considered outliers if they are unique and excluded from the sample;
• Any other price points beyond +/- 1.0 times the standard deviation from the average price are considered outliers and also excluded from the sample;  
• The maximum volume weighting of any individual price data point in the remaining sample is then limited to 40%.

Providing the data sample for each product/region is of a sufficient size to be considered robust, the volume-weighted average price is then calculated.  This is published as the reference price for each product/region.  

If the data sample does not meet the necessary robustness criteria for any product/region, fall-back procedures are employed to compile the reference price for that product/region.

As a first level of fall-back, the data sample collected for the current period is supplemented by "rolling over" data submissions made during the previous period by Data Providers that have not submitted any data during the current period. Any data rolled forward carries a reduced volume weighting in the calculation of the final reference price.  If, after this measure, there is still insufficient data to meet the necessary robustness criteria, the data sample is supplemented with firm bid and offer price data for the relevant period. This also carries a reduced volume weighting relative to any transaction data submitted for the current period.

The methodology also includes procedures for:

• the ongoing verification of sample data
• identifying and countering any possible statistical bias or attempts at manipulation
• monitoring the sample population for ongoing consistency.

As part of the reference price formulation process, for each index a statistical analysis of the other information collected from data providers with the price data is also prepared - including total volume traded, number of data points, price range and delivery time range.  This data is analysed as part of the validity and robustness tests carried out on each index calculated.  It remains confidential and is securely stored.  

Back to Top

Advisory Committee

An advisory committee has been established to review the price definition, data collection and index formulation processes.  This committee focuses on reviewing the definitions and processes established with a view to improving the accuracy of the reference prices formulated.  It cross-references the reference price indices developed with the opinions/reports of market commentators and industry participants to test credibility.  Where divergence occurs, the committee proposes remedial actions to the processes to improve the accuracy/credibility of the relevant price index.  

Back to Top