Webinar and Q&A: The quest for permanence: are registry buffer pools the right solution?

You can view the webinar below and read our answers to questions attendees asked during the discussion. For a deep dive on registry buffer pools, and how insurance can serve as a more effective risk mitigation mechanism, you can find our white paper here.

1. How can carbon insurance provide a solution to permanence that is long enough in duration to provide 100+ years of storage?

In-kind insurance can ‘guarantee’ permanence by replacing a reversed credit with a fresh unretired credit, and it can do that for as long as someone pays the insurance premium, whether that’s 10 years or 200 years. Of course, insurance cannot prevent reversals from occurring in the first place. No existing solution can do that!

2. If the insurance premium is priced every year for your reversal product, wouldn’t it be more expensive in the early years, making it more difficult for project developers to start a project? Also, wouldn’t higher insurance premiums make it difficult for developing country projects – shouldn’t there be an adjustment for economic inequality?

We expect that switching from the buffer pool to insurance is likely to promote project development and benefit project developers of good projects, since the insurance premium is being paid for one year up front – not, like the buffer pool contribution, which is charged for many years up front, and which is therefore much more expensive. In addition, instead of forcing the developer to make the buffer pool contribution, our view is that the buyer of the credit should pay the insurance premium, and that the developer should be able to sell the credits that are otherwise being held in the buffer.

If the project is a low-risk project, we would expect the premium to be quite low in the initial years – for an afforestation project in a temperate area with low probabilities of reversal risk, the premium could be as low as $.10 per credit to be insured. If an insurance premium is very expensive, it is also because the project is risky. And whether that is in a developing country or not, the project should probably not happen at all.

3. How will insurers keep a pool of carbon dioxide removals, especially in the absence of fungibility? Does this mean that the insurance premium is paid in credits and not in cash?

As an in-kind insurer, CarbonPool plans to invest premiums and its own additional risk capital into carbon removal credits from different projects, geographies, and registries. The lack of fungibility presents operational challenges, but these are not insurmountable. Premiums must be paid in cash, not credits, because the insurer must invest the premiums into carbon removal credits to add to its credit pool.

In-kind insurers like CarbonPool will of course have balance sheet risk as their own pool of carbon credits will be exposed to the same risks as those of their clients’. We seek to mitigate this through rigorous project selection and diversification across project geography and type. Our portfolio will be composed of high-quality credits, and specifically CCP-ready, afforestation and IFM, highly-rated credits. In addition, we will not invest into the same projects that we insure to avoid cumulation of losses.

4. Any thoughts on using geologic removals as a backstop?

Geologic removals, where carbon is stored underground for the long term, are assumed to have much longer expected permanence than nature-based carbon removals – though they are not free of reversal risk, albeit a much lower one than with nature-based carbon removals. At present they are very expensive and not readily available, reflecting the fact that we are still early in the process of developing efficient methods of capturing and storing geologically. Therefore, they do not currently represent a commercially viable alternative as a “backstop” (an insurer’s risk capital). It would also be difficult for an insurer to guarantee a payout in geological storage credits for this reason.

5. Can insurance cover more than natural catastrophe risk?

Insurance can cover, and already covers, more than just the natural perils. For example, certain project management risks that cause shortfall can and are covered by insurance, as well as certain political risks.

6. Why isn’t insurance being used more widely? What roadblocks is the system facing to use insurance more?

Insurers have not really been present in the carbon market until relatively recently, around the last two years. This is one of the reasons buffer pools were created to begin with. Now that insurance is more widely available, the registries would need to agree to accept insurance instead of buffer pool contributions, but their agreement is the only roadblock to starting to use insurance. And more insurers need to enter the market so that there is enough ‘capacity’ – meaning, enough insurance underwriting capital – available to cover the projected volumes of carbon credits to be produced in the next years.

7. How can in-kind insurance ensure “solvency” to regulators?

CarbonPool has built a unique solvency model to calculate its solvency based on principles used in the Swiss Solvency Test that takes into account cumulation risks and diversification benefits, as well as the unique nature of carbon credits and the many factors that influence their value.

8. Have there been assessments of the feasibility for like-for-like, in-kind compensation in insurance mechanisms that may have been done?

We have conducted extensive due diligence to ensure that we would be able to secure enough carbon removal credits to cover the projected volume of credits that we intend to insure. We believe that the transition from buffer pools to insurance will need to be done gradually and that there will therefore be some time for the supply of removal credits that are needed to pay claims to scale up. If buffer pools are transitioned to insurance, more credits should also become available for sale as they will not be locked up in buffer pool contributions. In short, we believe that in-kind insurance is a feasible alternative to buffer pools.

9. When do you consider buffer pools to work and where not, considering project activity and permanence?

We believe that buffer pools don’t work to ensure permanence. They are better than nothing, as they require a margin on top of the issued credits, but that’s all they do. But they cannot possibly assess a risk upfront for 40 or for 100 years, and therefore they do not guarantee permanence; furthermore, once the developer stops actively engaging in the project, contributing credits and updating risk assessments, there appears to be little or no monitoring to detect reversals, so even if they are occurring they are not being compensated for.

10. On which type of carbon credits do buffer pools work, and are they enough on each types of carbon credits or not?

As far as is publicly known, buffer pools mix avoidance and removal credits into one pool. Current buffer pools treat all credits as interchangeable: once the project’s contribution is depleted, any credits in the pool can be used to compensate for additional reversals. This means that if a removal credit is reversed, it will get compensated for mostly by avoidance credits, which are the majority in buffer pools. This means higher-cost removal projects do not get compensated with credits of the same quality in case of a reversal event. As for whether there are enough credits in buffer pools, we cannot know for sure since there is not enough information about them, but there are very strong indications that the pools do not have enough credits in them; there have been several scientific papers that suggest that buffer pools don’t have enough credits in them to last for the whole permanence duration that they promise to cover (for more on this see our white paper).

11. Could there be some form of integrated system, where buffer pools and carbon insurance are supplementary to each other?

In an integrated system, buffer pool contributions could provide for a small margin of error (e.g., due to over-issuance, limited reversals), with insurance then complementing those buffer pools by ensuring permanence of the carbon removal from the atmosphere (which buffer pools cannot do in all cases). Insurance actuarial methodologies can also help assess the risk of each project and calculate the correct contributions to cover those risks—although we stress that there is no tool or approach that would allow an accurate 100-year contribution to be calculated up front because up front contributions cannot cover losses many years into the future, since the risk profile of a project changes regularly because of weather, climate and other impacts.

You could also imagine a scenario where insurance could be used on top of the buffer pools, to protect companies from the collapse (and other flaws) of the buffer pools. Any company that is concerned about reputational risk should be concerned about relying on buffer pools as they are operated today, and in-kind insurance can help protect against reversal risk which is not properly compensated for by the buffer pool. However, we believe that, in the medium term, the more efficient solution is to replace the buffer pool with insurance in its entirety, because it will free up credits that are already in short supply from being contributed to a pool (since insurance premiums will be paid in cash).

12. One difference between a buffer pool and insurance is that if a reversal happens, then a buffer pool helps ensure that the atmosphere is no worse off (by compensating in-kind), whereas insurance makes only a monetary compensation. If the problem is that buffer pools are not working well, isn’t the solution to make sure they work well?

Indeed, classical monetary insurance provides money, not credits, and it takes time and effort to convert that into credits. There is also no guarantee the money will be sufficient to replace the credits that were lost. In-kind insurance, like the product offered by CarbonPool, provides high-quality credits to replace the lost credits, one for one. This ensures that the carbon that was re-released into the atmosphere by a reversal event is made up for by carbon removed from the atmosphere.

We also do not believe that buffer pools can be made to work well at all, given their fundamental flaws. The real solution is to mutualize the risk through insurance, contribute based on a proper actuarial risk assessment, backstop it with risk capital and reinsurance, get it externally audited on a regular basis, and adjust the premium annually on the basis of an annual risk assessment.

13. What if you had larger buffer pools—wouldn’t this allow for a longer durability period?

Larger buffer pool contributions will certainly ensure longer durations, but it’s very difficult to decide what the right size would be. It’s as if you’re using a piggy bank in lieu of medical insurance, and you decide to increase your weekly contributions to the piggy bank because you understand it might be insufficient if you get sick in the future. By how much do you need to increase the contributions? You might get diabetes in a year, or you might get cancer in fifteen years, or both – and the cost of treating those conditions will not be the same as it is today.

14. Existing registry tools do make ineligible projects that are past certain thresholds of risk, and they do require re-assessment of non-permanence risk at regular intervals. How is this different from how insurance tools would assess risk dynamically and disincentivize high-risk projects?

There are several differences. For one, these tools do not make a proper actuarial calculation that can produce an accurate assessment of risk for the credit’s 40- to 100-year durability period (depending on the registry). Additionally, these tools do not work after the end of the project’s crediting period, because there is no mechanism to charge additional contributions; this means the last credits issued cannot be guaranteed permanence for any significant time.

15. Can’t registries mitigate their risk due to the “portfolio effect” of a diversity and volume of projects in a pool?

A portfolio of independent risks will, up to a point, have diversification benefits. After a certain point, adding further projects does not bring any further diversification benefit (think of the stock market, where it is well established that once you have 30 securities, adding any more doesn’t provide additional diversification). The precise probabilities depend on the single projects and the correlations between projects of the different risks each is running, so it’s important to analyze the specific portfolio to understand how much the diversification benefit is of adding another project.

16. What about buffer pools that truly set aside just credits by a project, from that project, for that project?

That would not work, because the asset and liability would be fully correlated (meaning: if the project burns down, then the credits set aside from the project to cover a loss would also be damaged). There would be no risk pooling and diversification, which are fundamental to insurance.

17. How do these mechanisms align with ICVCM, VCMI, and other integrity principles and guidance?

We believe that in-kind insurance in particular fulfills a function that is very much aligned with the goals of the movements to bring integrity to the VCM, by managing permanence risk in a transparent, auditable way using a common, regulated, and well-known risk mitigation tool.

18. Please compare and contrast registry buffer pools, as well as buffer pools used in other industries

The buffer pools of the carbon credit registries are slightly different from each other in terms of the calculation of contributions, as well as the rules driving cancellations and freeing-up of credits; however, they all suffer from the same problems when promising permanence: they are generally insufficient, they have no reinsurance or risk capital backstop to account for worst-case scenarios, they are not transparent, and it is simple impossible to properly assess 40- to 100-year risks upfront. In other industries, certain types of short-term credit (e.g., microcredit) have some similarities where a collateral pool plays the role of the buffer pool; however, it doesn’t cover permanence but just the (short) duration of the credit, and pools are resized continuously to cover the current risk exposure.

19. What is the proposal for handling existing credits in a buffer pool?

We believe that, once insurance is the agreed solution, buffer pools should no longer accept any new credits or projects.

As to the credits already in the pool, there could be several solutions. One idea is that projects that are still active could be refunded their buffer contribution upon providing proof that all of the credits the projects that have been issued have been insured. Another idea is that the existing buffer pools are professionally managed by an insurance company or consultancy with insurance expertise with an emphasis on monitoring, regular reporting of reversal events and transparency about reversal event compensation.