(Photo by Christophe Pallot/Agence Zoom/Getty Images)

A Slalom in Songdo


Hot on the heels – or perhaps the skates, skis and snowboards – of the winter Olympics, the 19th meeting of the GCF Board convened at its HQ in Songdo, Korea, at the end of February.

To briefly summarise the headlines, the meeting saw several new Board members take their seats, and elected two new co-Chairs, Lennart Bage of Sweden, for the developed country constituency, and Paul Oquist of Nicaragua for the developing countries[1]. The chemistry between the two seemed good, especially given that Bage is a newcomer to the Board, and the meeting got through a fair bit of business.  It approved 23 projects, with a value of USD 1.09 billion, bringing the total GCF portfolio to 76 projects with a value of USD 3.73 billion.  Few projects in Africa, LDCs or SIDs had been presented, but the project approvals included the first under the so-called “Simplified Approval Process” (a process, though, as the Civil Society observer noted wryly, still entailed application paperwork running to 70 pages).  The meeting also approved long-awaited policies on Indigenous Peoples and on Environmental and Social procedures, and a further USD 60 million in grants to help countries get ready to make applications to the GCF.  A private sector outreach plan was also OK’d, though not before the ever-present suspicion of the private sector among some Board members had surfaced in more than an hour of discussion and some tortuous in-meeting redrafting of the decision-wording.

Much else, however, had either never made it onto the agenda or fell by the wayside during the meeting.  HR policies requested more than two years ago were still not before the Board, one member complained. The Fund still lacked a communications policy, said another. Proposals on a two-stage approvals process (designed to weed out sub-standard applications before they reach Board level) were also delayed.  A paper on the knotty issue of how to calculate incremental cost [2] had not been provided to Board members in time to be considered. A Gender and Social Inclusion policy was deferred for further consideration, as was a proposal to trial a project-based accreditation process, replacing the GCF’s glacially slow full accreditation process for certain applications. Investment committee proposals on establishing criteria for investment eligibility were also put back to the next Board meeting. Meanwhile, a report on the continuing astonishing levels of unfulfilled conditions attached to projects, which are preventing disbursement of cash to them, was not even discussed by the Board, for the second meeting in a row [3].

The continuing absence of so many critical policy and practice guidelines – which Board members themselves regularly deplore – was reflected in the more than two hours spent on general discussion of project related issues before any projects were actually presented.

An emerging GCF ‘sweet spot’ for projects?

When discussion of the projects did begin, it was clear that there is an emerging GCF ‘sweet spot’ into which certain types of project fall.  These get the Board’s nod with little or no comment.  The IADB, which presented two such projects – in Paraguay and Argentina –  seems to have become adept at finding this sweet spot, which consists of reasonable levels of co-financing by non-GCF entities and straightforward investment sectors such as renewables and energy efficiency.

Smaller projects, for obvious reasons, also tend to attract less attention, but this is not always the case, and two such projects were among the four proposals that exposed tensions within the Board and required reference back to their sponsors before they were finally approved.

In both of these smaller projects, Board rumblings were along one of the main fault lines between developed and developing country members, namely the ‘devolution’ of disbursement of funds to local levels.  This is an issue over which, as we have reported elsewhere, development finance institutions are increasingly being brought to task as well.  In this context, developed country Board members talk about “operational” risks – for example, that funds won’t be taken up by local communities. Their real fear, however, is probably more what we might call ‘Ethiopian Spice Girl Syndrome[4] – that money over which they don’t have total control might ultimately find its way to recipients that an aid-hostile media or lobby could make mischief with.  Speaking to support one of the projects, the Civil Society observer noted that it presented a “strong learning opportunity, piloting devolved decision-making processes” and that the GCF regularly takes ‘operational’ risks on very expensive investments like funds-of-funds, “so our priorities are not right if we can’t take high risk in low cost proposals.”

Sessions of the Board are broadcast live on the GCF website, the most recent meeting being available here

Another frequent bone of contention at the Board is the issue of GCF risk being substituted for risk that could be taken by other public finance institutions.  A case this time around was a facility for cities in Brazil to install LED street lighting and finance energy efficiency initiatives of one kind or another.  In this $1.3bn project, whose promoter (or ‘Accredited Entity’ in GCF parlance) was the World Bank, GCF was to contribute a $186 million concessional loan (at 0.75%), while the WB contribution was a $200 million ‘contingent’ loan (that is, a loan drawable only in the case of defaults).  One Board member complained that “with the right policies to guide the Secretariat, we could have obtained the same [climate] impact [from this project] with far less GCF funding” – and indeed it is hard to see why the GCF and not the World Bank itself is supplying the concessional loan in a project of this kind.

(The structures for projects where more sophisticated Accredited Entities are involved are, by the way, often mind-numbingly complex for even an experienced banker to follow, making fears expressed by the Civil Society observer that, in this case, city administrations could be left with risks they don’t understand a very material consideration)

Similar issues led to the 4th project referred back to the Accredited Entity, an energy efficiency facility in Vietnam.  Here, curiously, a guarantee provided by the World Bank appears to be itself guaranteed by the GCF, but the main discussion centred around the way that the project would actually subsidise the continuation of a poor energy regulatory environment in Vietnam. The Private Sector observer noted that power is so cheap in Vietnam compared to other operating costs that “companies have no incentive to switch anyway”, and that the “high level of [historic] ODA flows into Vietnam have distorted the energy market.”

After discussions with the relevant Accredited Entities, this and the other three projects not immediately approved were passed. These discussions take place behind the scenes, so we are not able to report what assurances were given to satisfy Board concerns – a gap in the GCF’s otherwise admirable willingness to make its proceedings public.

It is difficult to discern any plan or strategy behind what gets approved, one result of the GCF’s dysfunctional accreditation system

Standing back from the details of the proposals passed, however, it is difficult to discern any plan or strategy behind what gets approved, or how these approvals might fit with any notion of portfolio management – that is, achieving targets for sectoral coverage, adaptation vs mitigation, or leverage (how much private money the GCF funding crowds in).

This hapzardness is just one result of the GCF’s completely dysfunctional accreditation process, a many-stage, many dimensional maze that can take several years to navigate.   The system has actually been stuck for the best part of a year, since the GCF had been unable until recently to staff its accreditation panel, which had become inquorate.  The shut-down has led to a queue of applicants that at current approval rates will take four or five years to process.

The effects of this accreditation requirement, which is designed to ensure that the GCF only deals with entities that meet its gold standards (even though in very many cases, as noted above, these standards don’t actually yet exist) affects every aspect of the GCF’s workings, and none of them to the good.  First, it is the main reason for the glacial pace of actual disbursements, with a large number of ‘master agreements’ with Accredited Entities remaining unsigned, so money can’t be released to them. Though the GCF had approved some $2.6 bn at the time of the Board meeting, therefore, only $170 million (6.5%) had actually been disbursed.  Second, the accreditation process deters many private sector entities from dealing with the GCF at all.  Third, it leaves the organisation prey to having before it the projects that happen to be the preferences of the Accredited Entities, rather than being able to implement any focussed strategy for the portfolio.  And lastly, even with the capacity building grants and slightly abbreviated procedures for ‘Direct Access Entities’ (GCF-speak for more local ones) the process still militates against smaller and/or community-based organisations that might be coming forward with more innovative and less ‘engineered’ projects.  As a result, large international institutions receive overwhelmingly the lion’s share of GCF funds.

For more climate finance news and comment, subscribe to our free weekly newsletter.

In persisting with this prior-accreditation system, the GCF seems intent on reinventing wheels and refusing to recognise developments all around it.  It won’t be GCF accreditation requirements, for example, that will change the behaviour of global banks over financing fossil fuels – the TCFD, government policies such as France’s Article 173, the divestment movement and shareholder activism is what will do that.  In its accreditation demands the GCF seems also not to recognise that banks are regulated, audited and rated, and thus already subject to levels of monitoring and scrutiny that the GCF couldn’t hope to emulate.  The ‘killer’ GCF argument – that accreditation helps build capacity in smaller entities – has some merit, but far greater capacity building would occur if many more of these entities were able to access money, transact regularly, and build a track record.

RFPs the solution?

In attempt to correct the scattergun effect accreditation has on the project pipeline, the GCF has attempted to create focus by issuing RFPs on particular topics.  The first of these, promoting private sector innovations, closed in August last year with 350 responses. In a process that must have taken a huge amount of scarce human resource, a shortlist of 30 has been created.  Most of the shortlisted projects seem to have been put forward by non-accredited parties, which is excellent, but it will be interesting to see how many actually decide to proceed when they realise it might take them two years just to get over that hurdle.

A review of accreditation by a consultant was due to come to this Board meeting but has been deferred to the next.  It’s clear what the GCF is trying to achieve with this process – high standards of ESG practice and the avoidance of reputational risk – and accreditation one of the Fund’s “sacred cows,” being enshrined in its governing instruments.  But if the GCF really wants to be a channel for $10’s of billions a year, the system really has to be scrapped.  There are other ways to achieve the same ends, via due diligence and the use of covenants, and the best way of all is to establish relationships of trust with entities that you transact with regularly.  Other entities employing public funds employ these methods very successfully – the Dutch development Bank FMO, for example.  Let’s hope the GCF’s consultant is speaking to them and will present some alternatives in the summer review.

How is the GCF Measuring Up?

In our last report on the GCF Board, we proposed some metrics that a fund would normally look at to establish how efficiently it was being run.  They are listed below, along with notes on any information we have been able to glean from GCF documentation.

  1. Number and $$ of projects approved to date: 76 projects, for $3.73 bn
  2. $$ disbursed to date and projected time to full disbursal for projects approved: $170 million disbursed, representing just 6.5% of approvals.  The GCF website says it expects to have disbursed $900 million by the end of 2018.  Assuming the next two Board meetings approve $1bn each, that would still be just a 15% disbursement rate
  3. Metrics for project conditions imposed / met / unmet, and projected time to full meeting of conditions for projects approved: As at the date of the Board, 34 projects had 69 unmet conditions.  There is no indication of when these conditions might be met.
  4. Metrics for Accredited Entity pipeline and approval timescales (for example, average time to approval and signature of master agreement): In the 3 years the accreditation process has been running, 59 entities have been accredited (though 18 have still not signed the master agreement that would allow money to be disbursed to them).  100 entities are in the process, 83 at the first stage and 17 at the second.  Nearly 100 more have expressed interest but have not formally applied.  There is no forward indication of timeframes, but at the rate achieved so far (20 p.a.) the current queue of applicants would take 5 years to process.
  5. A metric for leverage achieved (in blended finance projects): As at the Board meeting, $2.6bn GCF funding had attracted $6.5bn in co-financing, a ratio of 2.5:1.  However, this includes all GCF projects, even those where GCF is the sole funder. There is no apparent reporting of the critical metric of blended finance leverage, including leverage of private sector funding.
  6. All-in costs as a % of project costs: There is no apparent reporting of this metric.  The GCF paid fees of $7.66 million on the $149 million disbursed in 2017, a rate of 5%.  In our report on the last Board meeting, we estimated that GCF costs amounted to 2.3% of funds approved.  We have no visibility on further intermediation costs, but with these two elements alone, GCF funding is costing in the region on 7.5%, which is extraordinarily high by private sector standards.
  7. Metric for person-days per $ million of projects approved: There is no reporting of such a metric, which would allow an understanding of how the GCF’s working procedures were becoming more efficient over time.
Leave a comment or join the conversation by following us on TwitterFacebook or Linkedin.

[1] The election of Mr Oquist was in no small measure ironic for followers of the minutiae of UNFCCC history.  As the Nicaraguan delegate at COP21, he it was that COP president Laurent Fabius took care not to notice waving a speaker’s flag as he brought the gavel down on the Paris agreement.  Mr Oquist had wanted to register his country’s last-minute objection to the  Paris agreement, as one that lacked ambition and was a “path to failure”. After Paris, Nicaragua joined Syria, Libya and North Korea among the handful of nations not to sign up to the agreement, a position it suddenly and somewhat mysteriously overturned in October last year.  With Mr Oquist then on the GCF Board as an alternate member, speculation at the time was that one reason for this sharp about face was the fact that it might be hard for the resident of a country that hadn’t signed up to Paris to chair one of its key implementing agencies …

[2] The additional cost of making projects, and especially infrastructure, resilient or “climate proof”

[3] The paper reports projects that have conditions attached that have a timeframe for resolution. 34 projects  (some approved as far back as 2015) have these.  Of a total of 88 conditions, 69 remain unmet.

[4] This was a famous furore in the UK in 2013, where aid agency DFID was lambasted in the anti-aid press for a grant promoting health education, which found its way in part to a girl-band in Ethiopia that was messaging on the topic in its songs

Leave a Reply

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