I just finished reading the judgment of the Court of Appeal in the GN Savings & Loans licence revocation saga.
(Quite a few people seem confused. The ruling wasn’t about “GN Bank.” There was a prior downgrade to the Savings & Loans category done with the consent of GN Bank’s directors before the licence revocation.)
Quick overview of the judgment
Reading the judgment, I kept smiling because it reinforces my long-held view that policy literacy requires multidisciplinary fluency. No way that judgment can be made sense of in anyway without some basic familiarity and willingness to engage with law, banking, public finance, financial regulation, accounting, auditing, risk management, etc. etc. And, of course, political economy. But this reality seems lost on quite a few people who believe that somehow the complex policy issues facing society can be addressed through siloed commentary. Anyways.
The judgment is well argued, at least judicially. But I think their Lordships retreated to their comfort zone and made the whole issue about “fairness and reasonableness” constitutional tests when quite a big part of the dispute relates to PRUDENTIAL REGULATION.
The judgment, if it becomes precedent, will thus seriously clip the wings of the Bank of Ghana. I doubt that this was the intention of their Lordships. Let’s go over the core findings and issues.
The Appeal Court held that the Bank of Ghana’s (BoG’s) 2019 decision to revoke GN Savings & Loans (S&L) licence on the grounds of insolvency was unfair and unreasonable because, in the court’s view, BoG did not properly account for Government of Ghana / MDA indebtedness to GN S&L. This “indebtedness”, represented by Interim Payment Certificates, should, in their Lordships’ view, have been regarded as receivables assigned to GN S&L, and therefore part of their “asset base”. Properly accounted for, GN S&L’s assets would have exceeded their liabilities. Hence, even if they were cashflow insolvent, they may have still been balance sheet solvent.
The Judges immediately recognised that this holding needed fortification. Because according to the Bank of Ghana Law (Act 930), one can revoke a financial institution’s licence EITHER because they are balance sheet insolvent OR because they are cashflow insolvent. There is no dispute that GN S&L were struggling to pay customers and thus was cashflow insolvent.
The Judges dealt with the issue by introducing a caveat into the law. Even if a financial institution is cashflow insolvent but its balance sheet strength could generate liquidity down the line, then the central bank would be unreasonable to declare it insolvent and revoke the licence. In short, temporary liquidity problems do not amount to insolvency.
Those of us in public policy would immediately recognise the parallel with IMF debt sustainability analysis. Ghana was asked to restructure its debts because the IMF ruled that the country’s problems were insolvency-related rather than due to mere temporary liquidity challenges.
The issue here is that there are very elaborate technical tests for determining what is temporary illiquidity versus structural insolvency. The Court naturally shied away from that analysis simply by arguing that the Bank of Ghana failed to place any analysis about asset valuation, impairment policies, and long-term liquidity measures before it. Effectively, they were blaming lawyers of the central bank for doing a shoddy job in their representations before the Court.
Without a judicial test to guide the regulator in future, it would be interesting to see how it determines when seemingly impaired balance sheet items can still be regarded as good enough to address liquidity constraints, and over what period or horizon. What is “short-term” in this regard exactly?
Another pillar of the court’s reasoning relates to the essence of the “burden of proof.” The High Court, whose decision GN S&L was appealing, had effectively required the Group Nduom to prove GN was solvent. The Court of Appeal said BoG was the party claiming it was insolvent, and so they should be the ones required to justify their claim.
As far as the Appeal Court was concerned, the High Court should have given real weight to the claimed Government/MDA debts that GN says its affiliate Gold Coast Advisors (GCA) had assigned to it. The ultimate debtors – COCOBOD, Road Fund, GETFund, and ministries, etc. – were all treated as government entities. At this point, the Court makes the argument in passing that lending to the government should be regarded as safe conduct.
Next, the Appeal Court relied heavily on the BoG-appointed supervisor/advisor’s report. Before the licence revocation, BoG had appointed an expert to supervise GN’s operations. The final report of the expert was before the Court. In the Judges’ view, the report demonstrated that the supervisor did not recommend immediate revocation and instead recommended remedial measures: removing or restricting Dr Nduom, restricting affiliate transactions, improving ICT/legal systems, branch rationalisation, training, software changes, and assisting GN to recover IPC proceeds.
On the back of the analysis, the Court made the following sweeping orders: the revocation decision is quashed; BoG is ordered to restore the licence; the receiver is ordered to hand over possession, management, control of assets and activities to the pre-revocation shareholders; assets are to be handed over “in their current state”; and third-party interests created during receivership are to be handled case by case with hearings.
Issues Arising
Because the Court stayed firmly in the “administrative decision-making reasonableness” domain (what lawyers sometimes call “Wednesbury unreasonableness”), the judgment is understandable as a reaction to a perceived administrative defect: BoG allegedly had evidence from its own supervisor that pointed to remedial action rather than immediate revocation, and yet revoked without sufficiently engaging the IPC/receivables issue. BoG had information about assets that could be used to cure the liquidity constraint, but still chose not to take them into account. Etc.
Still, I am not too sure that the Court’s decision not to wade into the bank-resolution analysis on prudential grounds led to a full and safe disposition of the issues. The judgment feels thin.
BoG’s 2019 revocation notice alleged much more than a disputed receivable problem. It said GN had a Capital Adequacy Ratio of about negative 61%, a severe liquidity crisis, numerous depositor complaints, failure to meet the 10% cash reserve requirement since Q1 2019, failure by shareholders to restore capital/liquidity, and only GH¢30.33m of IPCs confirmed by the Ministry of Finance, which BoG said still did not address a claimed GH¢683.66m capital deficit. The Court very hurriedly skimmed the regulator’s findings, pointed to claims of indebtedness made by GN exceeding the amount of the deficit and then pronounced balance sheet solvency.
BoG also alleged that the insolvency was largely caused by related-party placements/facilities to Groupe Nduom entities, including GH¢761.55m placed with Ghana Growth Fund / Gold Coast Advisors and Gold Coast Fund Management / Blackshield; that some of those funds were used to repay investment-company customers or finance contractors; and that the IPCs were not direct GN-to-Government transactions.
The regulator’s assertions raise serious prudential issues that the Court should have treated with more seriousness because they go to the heart of whether the assets GN claims it has are, enforceable, owned by GN, timely collectible, unencumbered, and valued after impairment.
Under Act 930, banks must maintain policies for non-accrual, provisioning and bad-debt treatment for loans and other exposures. Thus, the IPCs the Court chose to recognise would need to go through “risk adjustment” to ascertain their real value.
The problem, as the Court itself acknowledged, is that these IPCs are the subject of ongoing dispute, seven years on. It is still not clear how much the government of Ghana admits to owing; whether, in light of GN’s other conflicts with third parties, they could have been converted to cash in GN’s favour; and whether the timeline of any conversion would have been realistic under the conditions of commercial deposit-taking banking.
The global accounting standards raise similar issues. IFRS 9 requires expected-credit-loss analysis based on probability-weighted cash-flow expectations, and not hope of eventual collection. IFRS 13 fair-value logic also requires market-participant assumptions, credit risk, timing, and valuation uncertainty. Fair value goes beyond mere invoice face value.
On BoG’s version, the IPCs did not provide assurance since only GH¢30.33m was confirmed by the Ministry of Finance, while the alleged capital deficit was far larger. The Court rejected BoG’s treatment of the evidence, but it did not itself perform a valuation in the prudential sense before deciding to give full weight to GN’s claims.
The Court’s reasoning may throw a wrench into prudential regulation
Where the Court’s judgment introduces some rather hairy concerns is the treatment of liquidity realisation timelines in a banking safety and soundness context.
Banks are confidence institutions. Their core liability – deposits – is often payable on demand or short notice. If a bank cannot pay depositors, the fact that it may hold long-dated, disputed, illiquid receivables is rarely enough assurance. In bank resolution, liquidity failure can itself become solvency failure because asset fire sales, depositor runs, reputational collapse, and regulatory sanctions destroy franchise value, necessitating early intervention. This is the central equation in prudential regulation that seemed to have completely escaped Their Lordships.
The Financial Stability Board’s Key Attributes say resolution should be initiated when a firm is no longer viable or likely no longer viable, with no reasonable prospect of recovery, and that entry should occur before balance-sheet insolvency and before equity is fully wiped out. The BIS summary says something similar.
One cannot fault the Court for privileging administrative fairness, however. They understood that a regulator cannot ignore relevant evidence. They made it clear that BoG should have treated its own advisor/supervisor’s report with more seriousness. I am a long-term critic of the Ghanaian central bank. As I read the judgment, I found myself wistfully remembering many of my clashes with the BoG. A central bank should be much more clinical than the Ghanaian version is wont to be.
But the Judges did not, in my view, demonstrate an adequate grasp of modern prudential supervision.
Modern safety-and-soundness supervision ask questions such as:
Is capital real, loss-absorbing, and immediately available?
Are related-party exposures within limits and independently underwritten (BoG flagged GN’s persistent breach of single obligor limits, which itself can be a source of serious asset impairment)?
Are assets properly classified and provisioned?
Are liabilities payable on demand being met?
Is governance fit and proper?
Can the institution produce reliable accounts?
Are liquidity ratios being met?
Is there a credible capital restoration plan?
Are depositors and financial stability at risk?
Act 930 gives BoG broad powers to set capital methodology, define eligible capital, prescribe capital buffers, impose additional capital for concentration risk, prescribe liquidity requirements, and impose remedial measures. If a Court intends to override that technical judgment, it has no option than to dig into the details. The judgment unfortunately skates the corners.
What now? The Devil is in the Detail
I have a lot of empathy for the Nduom family and Groupe Nduom. Their entrepreneurial travails have been legend. It is so damn hard to build any entrepreneurial edifice of consequence in Africa that one naturally feels a spiritual bond to those trying.
In the midst of GN’s issues in Ghana, GN Bank USA, based in Chicago, also started to face serious challenges leading to regulatory restrictions in 2020 that were only lifted in 2025. The already small bank has now shrunk to a third of its size since 2018. Group Nduom has suffered!
GN Bank USA/Chicago’s Trajectory Since the Group’s Issues Began in Ghana
For that reason alone, there is substantial goodwill. Quite a number of people would like to see the GN brand restored. Even the angry litigants still chasing their money from GN may see glimpses of hope in a restored GN S&L.
But I fear that the road ahead could be bumpy.
Prudential issues remain
The BoG is likely to restore the GN licence without further appeals because of the clear political position of the ruling party, which has been to dismiss the prudential issues raised by the previous BoG administration.
But operationally, BoG may worry about whether it can safely allow immediate deposit-taking without a fresh supervisory assessment. GN meanwhile would need a massive deposit mobilisation drive to have the slightest chance of operational success.
Yet, if BoG simply lets GN reopen counters and starts taking deposits, some may argue that it is taking enormous depositor-protection/insurance risk.
Assets “in their current state”
The Judges avoided massive liability risk for the government by ordering that GN’s assets should be returned in their “current state.” This phrase avoids making GN’s receiver liable for all deterioration, but it creates a swamp.
The receiver may have collected loans, sold assets, paid depositors, terminated leases, retrenched staff, transferred records, defended claims, repudiated contracts, paid taxes, and incurred professional fees. Some third parties may have bought assets from the receiver in good faith. Some loans may have been written down or settled. Some depositors may have been paid by the state or through bailout/resolution mechanisms.
Every one of those events creates a massive reconciliation problem.
Depositors and state subrogation
If customers were paid by the state, receiver, deposit-protection scheme, or bailout vehicle, those payments do not vanish. The paying entity may become subrogated to the customers’ claims. That means GN’s liability stack today may include Government or resolution-vehicle claims, not only original retail depositor claims. In patching up the balance sheet, it may emerge that liabilities now far exceeds assets, especially if the IPCs being relied upon end up heavily impaired following recent auditor general evaluations.
The Security & Exchange Commission’s (SEC’s) updates on key affiliates of GN such as Blackshield/Gold Coast are relevant here by analogy and possibly substantively because Gold Coast/Blackshield was part of the same receivable/affiliate story.
SEC has said in the recent past that Blackshield/Gold Coast clients received partial bailout payments and that GHS1.34bn had been paid to 73,541 Blackshield/Gold Coast claims, with GHS757.5m fully settling 61,734 claims. In 2024, SEC announced another GHS1.5bn release for affected investors, including Blackshield/Gold Coast and Kron Capital, on top of GHS4.46bn previously released to affected investors.
If the implication is that Gold Coast/Blackshield or Ghana Growth Fund are insolvent or subject to investor/state claims, then GN’s receivables from those entities may be severely impaired after netting off. A claim against an affiliate with massive crystallised liabilities is not equivalent to cash or, for that matter, Tier 1 capital.
Capitalisation
Which boils down to the most critical question of all: fresh capitalisation. Does GN today have:
positive net worth after impairment;
sufficient unimpaired paid-up capital;
adequate risk-based capital ratio;
adequate liquidity;
clean audited accounts;
no disqualifying related-party exposures;
reliable systems and governance;
capacity to resume deposit-taking without threatening depositors.
What all GN’s fans are hoping to hear is when the Government will pay validated receivables; if shareholders are in the position to inject fresh capital; how investors are being brought on to recapitalise; and by what process liabilities will be legally restructured. The IPCs story pales in insignificance before the sheer scale of the challenge.
The Bottomline
For independent analysts, the way forward, if GN Bank is to avoid the zombie status of being a licenced bank on paper without operational activity, appears to be the following:
Completion of the IPC validation (possibly with court cases discontinued);
A determination of what Government truly owes and to whom;
A set off or settlement of state/resolution claims;
A clear plan for shareholder recapitalisation;
A ring-fencing of old claims into a specialised resolution vehicle;
And a final reopening of the institution once a new audited balance sheet and liquidity plan has met objective BoG prudential requirements.
Fans and friends of the Nduom family deserve to bask in the court victory. They have suffered for long. But the hard work is only now beginning and the choppy waters have yet to recede.
