Assess carbon delivery risk

We launched the BeZero Carbon Markets (BCM) platform in April this year, initiating ratings coverage of over 200 projects. Since then our coverage has grown to over 260 projects, making up over half of credits currently outstanding in the market. The BeZero Carbon Rating (BCR) is our view on the likelihood that an issued credit achieves a tonne of CO₂e avoided or removed. It is a tool that enables market participants to make a risk adjusted assessment of ex post credits available in the market today.

Why measure carbon delivery risk?

What the BCR does not do is assess forward looking issuance risk. While our database of rated credits can be used to see what drives variations in carbon performance and risk across project types, countries etc., it does not provide a view on the likelihood that a project will actually deliver a given number of credits.

This can be crucial information for a number of stakeholders in the Voluntary Carbon Market (VCM) value chain, from project investors carrying out due diligence on a project, to project developers looking to understand the drivers of variations in issuance, to market participants negotiating forward agreements in the market.

Figure 1: In this illustrative example, the vintage of this project has under-delivered credits (ex post) against its initial forecast (ex ante). The Issuance Risk Monitor (IRM) is designed to help monitor this carbon delivery risk, while the BeZero Carbon Rating (BCR) measures the carbon efficacy of issued, ex post credits.

Launched earlier this year, the IRM helps address this need for a carbon delivery risk tool. It enables users to leverage our database of projects to see the distribution of historical performance of issued credits versus forecast. This can be used to risk-adjust forecasted credit issuance by a given project, with the users able to filter the sample by a number of factors such as sector, country, accreditor and others.

Database & use cases

Currently the IRM leverages data from just over 200 rated projects in 49 different countries. Chart 1 shows the distribution of these projects by sector group.

Chart 1: Distribution of >200 BeZero rated projects by sector group. Sample does not reflect the full universe of BeZero rated projects.

We collect observations of the number of ex post credits issued in each vintage by a project and compare this to the number of credits the project forecast it would issue ex ante when it first started (collected from the registry’s Project Design Document). Comparing the two, we create a percentage metric of credits issued versus forecast. This ratio represents the project’s issuance rate: a figure of less than 100% indicates the project has under-delivered, while a figure above 100% means the project has over-delivered. The figure is then aggregated across all projects in the sample and displayed to IRM users.

While deviations in individual projects’ issuance from what was forecast may be driven purely by idiosyncratic factors, the tool gives users the context of what has happened historically for projects in the market to input into models and assessments as they see fit.

Interestingly, while the average project issued the same number of credits as forecast at the project start date (mean = 100%, median = 80%), we find this distribution to have a considerable range. The projects in our sample have delivered credits between 0% to over 250% of what was forecast at the start of the project (Figure 2).

Figure 2: screenshot of the Issuance Risk Monitor from within the BeZero Carbon Markets platform

Having access to this distribution enables our users to leverage the historical performance of projects in the VCM when risk-adjusting forecast issuance. The displayed sample can be filtered by any of the data labels on the BCM platform; sector, country, accreditor etc, allowing for additional layers of granularity. However there is a tradeoff between greater specificity on the sample being used and the sample size. As a statistical exercise, the bigger the sample size the better.

Since launching the IRM in August, we have added a new feature recently that allows users to see where specific projects sit within the distribution. This empowers them with a more in depth look at individual projects, linking the issuance performance with the broader ratings analysis available on the platform to assess the risk profile and characteristics.

Feedback on the IRM amongst our early users of the tool has been positive. One user from a major energy company stated:

"It’s very relevant to us in terms of how we think about this risk, and how we price it into our portfolio.”

Future developments

We will continually be adding more and more projects to the IRM analysis, enabling our users to leverage an ever growing sample size. This includes adding unrated projects in the near future. 

There are a number of additional features we will be launching for the IRM in the coming months. Firstly showing a temporal breakdown of the data underlying the aggregation in order for users to see how performance varied throughout the project lifetime.

We will also be adding further depth by linking the IRM performance with our BeZero data template to enable users to see which factor drove variations in performance. For example, did the project issue less credits than expected due to witnessing greater leakage or project emissions than forecast? 

Finally, as we continue to build up the sample size underlying the IRM, we will be launching a bespoke tool that enables users to input details of a theoretical project, such as country, project type, accreditor etc, to see what the potential distribution of issuance versus forecast could be according to historical performance of similar projects.