Could VCM buffer pool design risk a "bank run"?

An article recently published in Bloomberg Green — Carbon Offset Market Faces Chaos as African Mega-Project Collapses — implies some important questions about carbon credit buffer pools; how they are structured, their governance, and risk management. 

In financial markets, clearinghouses have emerged as the primary means by which counterparty risk, in the form of default risk or settlement timing risk, is managed. Default risk is the risk that one counterparty fails to make good on its obligations in a transaction. This risk is often called “Herstatt risk” in reference to a famous example where the German Herstatt Bank had received payment of its foreign currency receipts in Europe but failed to make its corresponding US dollar payments, leaving its counterparties with losses. 

When two counterparties transact bi-laterally each must evaluate how likely it is that their counterparty will make good on its end of the transaction. With repeat transactions, the costs of assessing counterparty risk can be justified. In cases of infrequent transactions between counterparties, or where there is information asymmetry that makes it hard for one counterparty to assess the other, settlement risk can be mitigated using a trusted third-party escrow agent to ensure both parties fulfill their obligations. However, in situations where there is a many-to-many market (like a regulated futures exchange), it would be massively inefficient (if not simply impossible) for all exchange participants to know and assess counterparty risk for all other participants. Even if they could, they would have to adjust the prices they were willing to buy or sell at to reflect the default risk of each counterparty. 

The clearinghouse’s role 

Modern futures exchanges function as well as they do because all participants can trade confidently with all other participants at the quoted prices no matter who is behind the quote. This is made possible by another important piece of financial market infrastructure: the clearinghouse. The clearinghouse stands between buyers and sellers and guarantees that trades will settle. Every buyer and seller only has to trust the clearinghouse, not other exchange participants. This is accomplished by the establishment of a clearinghouse’s default financial resources, typically in the form of a guarantee fund. This has some similarities to the buffer pool concept in the carbon markets but with some critically important differences. 

In a US-based clearinghouse, clearing members are required to contribute collateral in the form of US dollars or equivalent (typically Treasurys) to a guarantee fund, in proportion to their individual risk (typically the size of their trading volume). The funds are aggregated, but they are still segregated by contributor to enable the implementation of a risk waterfall. For simplicity of explanation this waterfall looks something like this: 1) funds of the defaulting clearing member, 2) funds from other clearing members, 3) funds of the clearinghouse. So the first loss is applied to the defaulting member. If the loss exceeds that member’s funds then other clearing members share in the excess loss pro-rata. If their funds are wiped out and there is still an excess loss, then funds from the clearinghouse are applied. Everyone has skin in the game. If the loss exceeds the funds of all members and the clearinghouse, only at that point does the trading counterparty to the defaulter suffer a loss. 

Clearinghouses go one step further in modeling their risk management. They typically operate under a “cover 2” model whereby the default financial resources are sufficient to cover the simultaneous default of their two largest members. This ensures that even in a catastrophic loss scenario, trades can be settled. When a clearinghouse requires clearing members to contribute to the guarantee fund, those members are exposed to the risk of other members so they care about the membership requirements that must be met. They want to be confident that high-risk members who may act less responsibly and increase the (moral hazard) risk of loss to the guarantee fund are not permitted to join. One more important point, because the collateral contributed by the clearing members is US dollars, it is liquid and valuable no matter the state of the clearing member. If a member is in default, its guarantee fund dollars don’t lose value. Buffer pools appear to operate quite differently than clearinghouses. 

Potential risks with the carbon credit buffer pool design 

The structure of a carbon credit buffer pool where all credits are commingled raises some questions and potentially introduces some risks. If projects contribute collateral to the pool in the form of their own credits, that may work fine in the scenario where some percentage of the project’s sold credits suffer reversal. For example, if a forestry project made up of 10,000 acres that contributed 1,000 acres worth of credits to the buffer pool suffers reversal from a fire that destroys 1,000 acres, its buffer pool contribution could cover the loss. Similar to a clearinghouse, the pool may be able to cover losses greater than the contribution of a single member. If the fire destroyed 2,000 acres then other projects’ pool credits could be used to cover the additional 1,000 acres. 

However, if an entire project is reversed (or deemed illegitimate) it may consume a substantial portion of the buffer pool if something similar to the “cover 2” risk management is not in place. Additionally, because the pool contributions are made in credits, not e.g. US dollars, if a project is deemed illegitimate, all of its contributed credits may become worthless. The moral hazard is clear: all other pool contributors may be left covering the loss of a high-risk project that loses only the (now worthless) credits it contributed. There may not be any skin in the game for a project of questionable quality. Unfortunately, even after covering the loss of another project, pool contributors may yet be left even worse off. 

In the case of a substantial loss to the buffer pool, the depletion of the pool credits may have the residual effect of increasing the perceived overall risk of all projects in the registry operating the pool. If the pool’s credibility is damaged by a loss from a project deemed low quality, buyers may hesitate to purchase credits from any projects that participate in a registry because the pool reserves may not be sufficient to cover future losses and the likelihood of such losses may be perceived to be higher if other registry projects also turn out to be of questionable quality; an example of the classic Lemon Problem. In that case, high-quality pool contributors may have an incentive to withdraw their pool contributions and seek another registry where buyers have greater confidence. This may create the potential for something akin to a bank run. 

In a clearinghouse, members’ guarantee fund contributions are segregated. They are able to withdraw only their contributions, or what remains of them if they are allocated a pro-rata portion of a loss in a default. If carbon credit projects are able to withdraw from a registry that operates a buffer pool with commingled credits, the rules governing withdrawal are important. In a post-loss scenario where a substantial portion of the pool credits have been depleted, if projects aren’t limited to their pro-rata share of the remaining pool they may have an incentive to rush to withdraw and pull all of their original credits contributed to the pool. This might look like the Silicon Valley Bank failure where depositors who pulled their deposits out at the first sign of trouble got their funds, but others who waited faced the potential risk of loss of their deposits. In that case, the Federal Reserve stepped in and guaranteed all the deposits. Registries may have no such guarantor. The result could, in theory, be a registry with fewer projects, and a reduced or non-existent buffer pool. 

Lessons from financial markets for carbon markets 

In the aftermath of events such as Herstatt Bank post-trade mechanisms were designed to manage settlement risk. Clearinghouses are one, very effective, mechanism. There are others including a system called Continuous Linked Settlement (CLS) in foreign exchange markets that functions more like an escrow agent and does not require a guarantee fund. CLS manages settlement risk but not market/replacement risk. That’s a topic for another article but the upshot is in an escrow-type model counterparties at least get their money (or assets) back in the event of a default. 

A clearinghouse isn’t the solution for many of the carbon credit products on the market now, but it might be in the future for exchange-traded products. Buffer pools may need to be reimagined instead. Like with subprime mortgage portfolios, a wrapper that offers perceived risk diversification may obfuscate the actual risks. 

The takeaway is that financial markets have several centuries worth of hard lessons to draw upon. The voluntary carbon market can be improved by incorporating lessons and designs from financial markets into VCM products and infrastructure to improve risk management and increase overall confidence in the market. Keeping products transparent and simple, being clear about the risks, and offering insurance products to manage them may present a better near-term approach.

illuminem Voices is a democratic space presenting the thoughts and opinions of leading Sustainability & Energy writers, their opinions do not necessarily represent those of illuminem.