Introduction
MetalMiner’s price reporting adopts a number of recognised PRA models depending on the particular price point, the availability of data and geographic location. The objective is to report prices that represent fair market value for the markets they pertain to. What the firm eschews is the widely used but industry distrusted approach of calling a limited number of market participants for an opinion on an appropriate current market price. Although IOSCO methodologies do not specifically audit the phone polling process, the IOSCO rules do set out specific requirements in terms of volume weighted average. However, these processes for secondary metal markets remain opaque and thus unclear whether or not VWA is used in the setting of say secondary aluminum alloys.
For this reason, many large OEM’s distrust, but are forced to use, existing PRA’s. MetalMiner seeks to challenge this process by using publicly available data (in most cases, government published trade data), with appropriate chemical compositions to create “should-cost” models validated extensively through statistical modeling, backtesting and OEM price validation. In other words, MetalMiner’s strategy to bring more price points onto the market involves testing and validation by large OEMs.
Regardless of the data source or derivative process, the objective is to create a transparent, honest and reliable price that allows clients to price against, use as a benchmark, verify market movements and ensure contract compliance with confidence.
With a data research and reporting team based in New Delhi, India and a UK-based control office, prices are reported ahead of US opening times on a daily basis. The broad methodologies are detailed below, but the MetalMiner Handbook details the geographic location, material description, data source and the frequency of price updating for every line item in the firm’s database.
Methodology
Description of data gathering approaches:
Licensed data sources
These are prices MM takes under commercial license from local reporting agencies around the world under subscription and often cover some of the lesser reported price points such as rare earths, alloys and semi finished metals. Usually, the prices are taken by the reporting agency from recorded trades on local markets. Where prices for like products are found globally from other sources, MM runs correlation analysis to ensure the data fairly reflects not just domestic but global price metrics.
Prices are an average of transactions on local markets. In some cases they are local exchanges and in others they are derived from records of buy-sell records.
MetalMiner is a licensee for example of: CME, LME, MSA (RMDAS), among several other exchanges.
Prices derived from publicly available data
MetalMiner has collected price points since January, 2012 providing an extensive database with thousands of prices as well as historical data. Due to the size and nature of the MetalMiner database, MM has acquired the ability to easily bring new prices to market.
Many developed markets record to considerable detail, volumes and values of 8 or 10 digit HS code commodities. This allows MM to generate product prices which accurately reflect prices paid without the need for phone based polling which necessarily inject individuals opinions or biases. For markets such as North America, Europe and Asia, the volumes of monthly imports and exports are so large that very accurate prices can be generated to reflect an average of small to large trades in a given commodity. Where a time lag exists between the gathering and reporting of such public data, MM uses a commodity inflation adjuster to extrapolate current prices. Such time delays are typically between 4 and 6 weeks as public data is usually published after the month end in which it is gathered.
Manufacturers posted prices
Certain industries post prices for their products on a weekly or monthly basis and while discounts are offered for larger volume trades, the posted prices are a fair reflection of the current market price. Generally, MM takes several producers’ posted prices and volume weights the prices to generate a MM averaged price. This avoids the possibility of one mill leaving price rises or drops unchanged because they are down for maintenance or have a production issue that temporarily skews their price list setting.
MetalMiner Index Prices
MM produces several Indexes to reflect the cost dynamics of various industries such as Automotive, Renewables, Construction and Lithium-ion Batteries. The Indexes are derived from price points in the database using proprietary algorithms developed by MM in collaboration with industry stakeholders. MM Indexes are not a transaction price but rather a benchmark against which buyers and sellers can track their transaction prices against a fixed basket of products that accurately reflect the industry cost dynamics.
These “baskets of metals” or MMIs use a methodology that mirrors that of the U.S. Department of Labor Producer Price Indexes. Thus a buying organization can use these baskets to mark month over month inflation/deflation. Some organizations also contract off of these price movements.
Each MMI contains a different metal but in all cases, the headline MMI contains a basket of metals. As an example, the Stainless Steel MMI contains LME Nickel prices, a mix of NAS surcharges, 304 coil pricing in China, the US, scrap prices etc. The index tracks the actual unit cost changes month over month and the algorithm weights each element to create the overall index value.
Data Sources
Detailed in the Handbook by price point as sources are amended and updated in response to providers timeliness and accuracy of reporting. MM reserves the right to change data sources on a like for like basis to ensure the accuracy, timeliness and reliability of price data. Any changes are documented in the Handbook and records held for subscribers review.
In addition, MetalMiner has a Database Health application that runs over every single price point in the system. This capability sends alerts and notifications to each of the core analysts and the data team in India on a daily basis. The alerts and notifications cover both data error input as well as timing errors. In other words, a trigger is pushed if a price point entered (either via API or manual entry) exceeds a value threshold (ie. a price is too high or too low based on prior prices) as well as if a price has not been entered in a timely fashion (ie. each price is set in the database with a frequency setting). This capability allows the company to proactively manage overall database health.
Timeliness of data
Procedures for checking the timeliness of data reporting.
Prices are updated periodically depending on the price point, the timing of which is stated in the Handbook. For high volume trading exchanges like the LME, CME, SHFE and local exchanges in Asia, prices change hourly and the MM prices are updated daily with the close of business or settlement prices. For other metals, the prices typically change every few days although the MM site is updated daily whether there is a price change or not. Those prices based on monthly reported publicly available census data, both specific price points and derivatives, are updated monthly within 24 hrs of the new month’s data being released.
The MM programs have an automated end of day report that flags prices that have not moved on a predetermined cycle. This report is picked up by the back-office function overnight US time and non-moving items investigated to ensure the issue is resolved by US opening the following day. Occasionally the data on which MM’s prices are based is delayed due to national holidays in which case prices are updated the first working day after normal reporting is resumed.
Correlation and verification of data points
The entire MetalMiner database has a built-in correlation capability which the company draws upon when adding new price points to the system. This capability allows the firm to test new data points to ensure data integrity and accuracy. This capability is also built into the SaaS platform, MetalMiner Insights.
Contract Specifications
Every price point has its own unique description, depending on the geography and market place from which the prices are derived. The description is as detailed as MM is able to secure from the data provider. In some cases, product descriptions are more generic than others, particularly for lower volume markets where too specific a grade or product description would preclude very similar product trades and generate very few results. The judgment is that it is better to have a reliable responsive price for copper bars, than to have a price that only changes rarely for specific sizes of copper bars or for unique grades of copper bar.
The MM database allows the display of prices in multiple currencies updated daily from a live feed of exchange rates. The database also allows the display of prices in both metric and imperial units, but the source data is recorded in the same format every day. For most parts of the world, this is per metric ton or KG, but for US markets some products are recorded in short tons or lbs. Regardless, the internal calculator allows data to be displayed in any combination of currency and unit of weight the user requires.
Audit Process
Changes to price, date, person recording, reason for change.
Prices are changed for a variety of reasons. Sometimes a source ceases to report a particular commodity, more usually we find the price is not responsive enough for our needs and we develop or seek out an alternative more dynamic price point, sometimes our clients come to us and ask for a price on a new commodity that is of specific relevance to their industry or business.
When changes are made it is critical the details are recorded in our handbook for traceability in the future.
The commodity description and geographic location are critical, if they are different in any way from what was reported previously. All changes to commodity descriptions appear in the document “Data Full Metals Catalog.”
The reason for the change is noted
The date and person making the change, and the person auditing the change to double check the veracity of the details recorded.