EDITOR: | October 9th, 2015 | 10 Comments

Dr. Duchesne launches Sustainability Index for Rare Earth Companies

| October 09, 2015 | 10 Comments

A decision support system for peer-to-peer comparison of rare earths projects

We developed a sustainability index (SREE) to permit peer-to-peer comparisons among rare earths projects.

Traditionally the rare earths industry has been using heavy vs light rare earth ratios as predictors of success. In practice this approach is best suited for precious metal projects where success is a correlate of cost of production against the price of gold.

But limited market demand and complex ore composition bring challenges for traditional peer-to-peer comparison among rare earths projects. There is a great deal of variance in the ore composition of deposits, and their sustainability is further affected by location, capital costs and a plurality of other factors that cannot be consolidated into a simple dependant variable such as “cost per ounce” as in the case of precious metals.

We have a burden of responsibility because the livelihood of many and investor’s money depends on our analyses. And so, we have developed a sustainability index based on multiple regression analyses to permit a weighed comparison among rare earths projects. Multiple regressions are routinely used in social sciences and in ecology but not in mining.

The general purpose of multiple regressions is to learn more about the relationship between several independent or predictor variables and a dependent or criterion variable. For example, a real estate agent might record for each listing the size of the house (in square feet), the number of bedrooms, the average income in the respective neighborhood according to census data, and a subjective rating of appeal of the house. Once this information has been compiled for various houses it would be interesting to see whether and how these measures relate to the price for which a house is sold. For example, you might learn that the number of bedrooms is a better predictor of the price for which a house sells in a particular neighborhood than how “pretty” the house is (subjective rating). You may also detect “outliers,” that is, houses that should really sell for more, given their location and characteristics.

In the social and natural sciences multiple regression procedures are very widely used in research. In general, multiple regression allows the researcher to ask (and hopefully answer) the general question “what is the best predictor of …”. For example, educational researchers might want to learn what are the best predictors of success in high school. Psychologists may want to determine which personality variable best predicts social adjustment. Sociologists may want to find out which of the multiple social indicators best predict whether or not a new immigrant group will adapt and be absorbed into society. In short, predicting the success of rare earths project is quite similar because of the complexity of variables at play.

The trick to multiple regressions is to build a sturdy mathematical model that can be tweaked through trial and errors, and supported by caffeine in industrial amounts.

We arrived to the conclusion that the sustainability of rare earth projects is a correlation of the contribution of Neodymium, Praseodymium, Terbium and Dysprosium to revenue models. We included a subfunction to account for revenues per tonne of ore. We added subfunctions to take into account financial metrics that influence the probability of success such as market cap, enterprise value cash at hand, which are determinants of success in the rare earth industry.

We based the sustainability index (SREE) on the following formula:

SREE= [A Ÿ (Nd&Pr)] +   [B Ÿ (Nd&Pr&Tb&Dy)] + [C Ÿ Revenues per tonne]+ [D Ÿ (Nd&Pr)/revenues] + [E Ÿ (Nd&Pr&Tb&Dy)/revenues] + [F Ÿ Market cap] + [G Ÿ Enterprise values] + [H Ÿ Cash in the bank] where A, B, C, D, E, D, F, and H are numerical determinants that control the relative weight of each of the variables.

To help understand the structure of SREE the table below (SAMPLE) shows data from different rare earth projects as of September 1, 2015 — this data was collated from public sources. The different colours shows qualitative weighing of data point. Each line represents one rare earth project.


The sustainability index will give you a ranking of the companies with the greatest chance of success. Due to the sensitivity of the conclusions from the sustainability index, InvestorIntel is not publishing these results, however, we are taking orders from serious parties only. Please send email inquiries to info@investorintel.com.

Dr. Duchesne will be a panelist for the Future High Tech uses and sources of Technology Metals: Coming Attractions panel and he will be moderating the Innovative processing for the technology metals market at the Global Technology Metals Market Summit in Toronto, on Wednesday, October 14th.

Dr. Luc Duchesne


Dr. Luc C. Duchesne is a Speaker and Author with a PhD in Biochemistry. With three decades of scientific and business experience, he has published ... <Read more about Dr. Luc Duchesne>

Copyright © 2019 InvestorIntel Corp. All rights reserved. More & Disclaimer »


  • NGS

    Hi Dr. Duchesne,

    This is multiple regression? How can one create the predictor equation without a training set of both predictor variables AND the dependent variable, sustainability? Yes, the predictors are publically available, the data from REE companies. But the dependent variable, sustainability, does not exist.

    Further, sustainability is a binary concept – you survive or you don’t. 0 or 1?

    For those not versed in multivariate statistics, multiple regression is an extension of simple regression. In simple regression, one takes a set of known data, values of variables x and y (e.g., tests, analyses) for a number of samples. Those data are used to describe the mathematical relationship between x and y. Then one can predict a value for y based on a hypothetical value of x. If you don’t have a training set of x’s and y’s, you can’t do a regression. In the case presented above, there are no known sustainability values, the y’s, so no multiple regression is possible.

    In the real estate example, the y is the selling price. That IS KNOWN, along with what will be the predictors, e.g., squre feet, number of rooms. etc. So you can do a multiple regression.

    Where do I look up the sustainability values? Are they in the 10Ks or the investor presentations?

    Take care,

    October 9, 2015 - 6:11 PM

  • JJ Beswick

    NGS: you’re right that “sustainability is a binary concept”; the appropriate regression models use odds ratios or logits, measures of the probability of success. Very common in epidemiology where (for example) they estimate the probability of surviving some disease.
    I expect Dr Luc didn’t want to get too technical.
    What can’t be done yet is to “fit” a model with 8 parameters (A-H) when there are only 3 data points: GWM & Moly fail, Lynas (provisional) success.

    October 10, 2015 - 12:16 AM

  • hackenzac

    It is an epidemiology problem in a sense. Revenues would have to be in the net form to account for overhead and risk.

    October 10, 2015 - 12:46 AM

  • Tim Ainsworth

    Lol, give Blind Freddy an end market “predictor” and he’ll neatly rank the likely survivors 1 – 4.

    October 10, 2015 - 2:25 AM

  • Bill Keenes

    Tim that’s uncalled for and totally inappropriate.

    October 10, 2015 - 5:09 AM

  • Tim Ainsworth

    Lol Bill, given your status as a fully paid up member of the Dysy 1.01 Club I can totally understand your affection for the rear view mirror perspective.
    High Dy Ionic con on sale now for $34kg, incl 27% tax, you can even get a truckload or two of oxides ROW for a bit over $200kg ATM, and GM’s down to just 40 grams in the Volt drive train.
    Walk me thru the efficacy of < 1% TREO hard rock deposits again could you?

    October 10, 2015 - 6:23 AM

  • JJ Beswick

    Too funny Tim/Freddy.
    Perhaps, for the next 4 years or so, the best predictor of ROW RE production success is very simple:
    – If its name is Lynas it’s got a good chance
    – Otherwise, forget it.

    October 11, 2015 - 11:48 AM

  • allan wing

    I see Dr. Luc Duchesne has added another position to his impressive career. GO Foundation

    Dr. Luc C. Duchesne

    Executive Director

    Dr. Duchesne has worked as a scientist and/or an executive in various private, public and government organizations where he has combined business skills with a deep understanding of the scientific method for technology execution in renewable energy, and complex manufacturing processes [PhD Biochemistry (1988), MSc Forest Engineering (1985), BSc Forest Engineering (1983)]; authored or co-authored over 85 peer-reviewed articles, book chapters or books; associate editor of two scientific journals. He has taught as an adjunct professor in 8 universities and has been a member of the editorial board of 2 scientific journals.

    October 11, 2015 - 12:47 PM

  • Luc C Duchesne

    We are receiving a multitude of comments about the Sustainability Index.

    My intent was to consolidate a significant body of knowledge to build a living tool that permits you/us/everyone to make investment decisions by weighing different criteria into a single dependant variable. I’ve attached sensitivity analyses to each variable to create a dialectic tool, And further I’ve discarded possible variables which I investigated and determined they would be poor measure of success. I’ve developed the basic approach for the Sustainability Index over my 30-year career as a scientist. For example I developed models for predicting ecosystem dynamics following disturbance, including for example climatic changes. These were complex models with thousands of variables. Once I published a model for the determination of old growth ecosystem abundance from complex forest inventory databases with over a million datapoints–these were extremely contentious from a political perspective and the dataset immensely more complex than the REE database. But after much reticence it became and still is the one model to use. I’ve done this several times in my career, at times my work was used to make multimillion dollar decisions by both businesses and governments.

    And so, own personal contribution to the REE sector is to adapt this type of expertise to a field where the traditional mining knowledge would benefit from this type of analytical approach.

    But I also want to point out that I have designed the Sustainability Index is also a living document so that others may improve my thought process. For example, in building up the engine for the index, I have attached weightings to each parameters (independent variables) but I recognize that not all users may agree with these weightings or also that new parameters/understanding of the industry may emerge such that the weightings may change over time. Therefore I have structured the engine that computes the algorithm such that anyone with a key to the engine can make their decision about the weighting of the variables.

    Ultimately every user who is skilled in the REE sector can tweak the algorithm to develop their own Sustainability Index to conduct peer-to-peer comparisons.

    For project developers it permits an objective parametization of peer-to-peer comparisons. For investors it permits to determine who offers the lowest risk profile in a systematic way.

    During the upcoming TMS in Toronto I will be speaking to stakeholders in the field to tweak my weightings to integrate the best possible hierarchical ordination (weightings) of the independent variables of the algorithm.

    It is not my intention to provide investment advise but I provide the tool to permit users to make decisions with a full understanding of how they are making decisions based on a combination of hard science and best-guesses. And I have created a clear delineation between empirical data and guesswork. I repeat and emphasize again that my intent is not to provide prognostication app. I have no intend to try and read the future. History is not particularly kind to those who predict the future and fail.

    October 11, 2015 - 11:15 PM

  • Denis

    I would consider this thought and Sustainability Index like “an analytical approach for making investment decisions” for securities market\stock exchange only.
    The specific weight of each determinant has to be defined in every particular case\task, based on the investor’s priority. FOR EXAMPLE: someone has a technology for monazite concentrate processing and prefers the Australian companies with a specific HREO\TREO ratio. Therefore the investment decision of development one of the Junior projects will be based on the highest weight of these three determinants –type of the mineral, deposit location, HREO\TREO ratio. And vice versa the market cap of the REE company will not play a major role in such specific task.
    When I was making a similar calculation for my own purposes, I have collected the initial REE project’s information from PEA of each public company. Now when you can use “TMR Advanced Rare-Earth Projects Index” there is no necessity to fish out the REE project data. You can just define your own specific weight of the determinants and create your own ranking of REE companies in accordance with your particular purpose.

    October 12, 2015 - 7:15 AM

Leave a Reply

Your email address will not be published. Required fields are marked *