Singh v Bains & Anor: Appellate Deference on Findings of Fact and the “Compelling Evidence” Threshold Despite Misplaced Reliance on Burden-of-Proof Last-Resort Principles
1. Introduction
Singh v Bains & Anor ([2026] EWCA Civ 408) concerns an attempt by the claimant/appellant to obtain
specific performance of an alleged oral agreement that he would hold 50% of the shares in
GB Retail Limited (GBRL). The first defendant/respondent was the alleged promisor and the controlling director; GBRL was the
second defendant.
The High Court (Chancery Division) found as a fact that no such oral agreement existed and dismissed the claim. The claimant appealed
primarily on the basis that the trial judge (i) approached the fact-finding exercise incorrectly (including resorting to “burden of proof” ideas),
and (ii) failed to engage with supposedly compelling corroborative evidence (notably Companies House filings, a £39,000 payment,
later property acquisition in joint names, and the structure of shareholdings in related ventures).
The Court of Appeal dismissed the appeal. While it accepted there was “some force” in criticisms of the trial judge’s route through the evidence
(including his reference to exceptional “last resort” burden-of-proof material), it held that the judge’s conclusion was rationally available,
and the appellant failed to demonstrate any overlooked compelling evidence or other identifiable error meeting the stringent appellate threshold.
Key issues
- Whether the trial judge’s finding that no oral contract for a 50% shareholding existed was “plainly wrong”.
- Whether the judge overlooked or failed to address apparently compelling evidence contrary to his conclusion.
- Whether the judge’s use of “last resort” burden-of-proof guidance (from Phipson and case law) infected the result.
- What level of reasoning is required where a party alleges a failure to grapple with key evidence.
2. Summary of the Judgment
The Court of Appeal reaffirmed orthodox principles on appeals against findings of fact. It held:
-
The appeal was a pure question of fact; the appellant had to show the trial judge was plainly wrong, not merely that
the appellate court might have decided differently.
-
Although the judge was wrong to treat the matter as if it might require a “last resort” burden-of-proof analysis (and should ideally have
invited submissions before relying on un-cited authorities), he also made an alternative, orthodox finding preferring the respondent’s case
on the balance of probabilities.
-
The appellant did not establish that the judge overlooked “compelling” evidence. The four highlighted topics (the £39,000 payment, the joint property purchase,
the rationale for incorporating GBRL, and the 67:33 share split in related companies) were either not compelling, not pleaded with forensic clarity, consistent
with the respondent’s profit-sharing case, or otherwise adequately encompassed by the judge’s overall credibility findings.
-
The judge’s adverse view of the appellant’s reliability—especially shifting accounts (including about BCL)—was open to him and central to the result.
Accordingly, the appeal was dismissed and the High Court’s finding that there was no oral agreement for an equal shareholding in GBRL stood.
3. Analysis
3.1 Precedents Cited
The Court of Appeal’s reasoning is anchored in appellate restraint and adequacy-of-reasons jurisprudence, drawing a line between (i) imperfect expression or
structure in a judgment and (ii) a reviewable failure to consider material evidence or an irrational conclusion.
Volpi v Volpi [2022] EWCA Civ 464, [2022] 4 WLR 48
This was the Court of Appeal’s principal statement of the standard of review on factual appeals. The court relied on the “well-trodden path”
summarised by Lewison LJ, including:
- Non-interference unless the judge was plainly wrong.
- “Plainly” is not about appellate disagreement; it asks whether no reasonable judge could have reached the decision.
- An appellate court assumes the trial judge considered the evidence unless there is compelling reason to think otherwise.
- Judgments should not be “picked over” with narrow textual analysis.
In Singh, these principles set the tone: the appellant’s criticisms about missing discussion of certain facts had to meet a high bar to show
true oversight rather than selectivity in reasons.
Lord Reed’s formulation supplied the “identifiable error” framework: absent material error of law, baseless critical findings, demonstrable misunderstanding,
or demonstrable failure to consider relevant evidence, appellate interference is unjustified unless the decision “cannot reasonably be explained or justified”.
The appellant’s route to success therefore required demonstrating one of these recognised categories, most plausibly a failure to consider relevant,
compelling evidence. The Court of Appeal held he did not do so.
ACLBDD Holdings Ltd v Staechelin [2019] EWCA Civ 817, [2019] 3 All ER 429
This authority was used to frame the judicial duty to give reasons without demanding exhaustive treatment of every point. The court endorsed:
reasons must show the basis of decision, but need not be elaborate and need not address every submission.
In Singh, it supported the respondent’s answer that the judge’s failure to address each “island” of evidence did not itself prove error.
Cited within ACLBDD Holdings Ltd v Staechelin for the proposition that a judge need not provide “reasons for his reasons”. This helped the
Court of Appeal reject an approach that would effectively require trial judges to pre-emptively refute every potential appellate argument.
English v Emery Reimbold & Strick Ltd, DJ & C Withers (Farms) Ltd v Ambic Equipment Ltd, Verrechia (t/a Freightmaster Commercials) v Comr of Police of the Metropolis [2002] EWCA Civ 605, [2002] 3 All ER 385, [2002] 1 WLR 2409
Reiterated (via ACLBDD) as a leading statement on sufficiency of reasons: the judgment must show the basis on which the judge has acted.
The Court of Appeal treated the trial judgment as meeting that standard notwithstanding its contested structure.
Simetra Global Assets Ltd v Ikon Finance Ltd [2019] EWCA Civ 1413, [2019] 4 WLR 112
This was crucial to the appellant’s theory: a judge should deal with “apparently compelling” evidence contrary to the conclusion and explain why it is rejected.
The Court of Appeal accepted the principle but applied it strictly: the four highlighted matters were not of a kind that necessarily demanded discrete,
express treatment, particularly given how they were pleaded (or not pleaded) and how they fitted (or did not fit) with the central oral-agreement question.
Simetra thus operated as a safety valve—permitting intervention where the judgment’s silence indicates true oversight—but it did not assist the
appellant on the facts.
These authorities were cited by the trial judge (without prompting by counsel) via Phipson on Evidence for the rare scenario where a court cannot
decide which account is more likely and disposes of the issue by burden of proof.
The Court of Appeal held there was “some force” in the complaint that it was wrong to look to these “last resort” principles in a case with ample evidence.
Importantly, however, it treated this as an imperfection rather than a vitiating misdirection, because the judge also made an
alternative merits-based finding.
3.2 Legal Reasoning
(a) The appellate lens: fact-finding is primarily for the trial judge
The Court of Appeal emphasised that it was not conducting a rehearing. The appellant’s submissions were characterised as “island-hopping”: extracting discrete
facts and asking the appellate court to reassemble the case differently. Under Volpi v Volpi and Henderson v Foxworth Investments Ltd,
that is impermissible unless the overall conclusion is irrational or the judge demonstrably missed something decisive.
(b) Misplaced “last resort” burden-of-proof guidance: error acknowledged, impact denied
The Court of Appeal accepted that the trial judge was wrong to treat the case as one where the burden-of-proof “last resort” might be needed and noted it was
“unfortunate” that he did not invite submissions on that un-cited material. This is a clear cautionary message: if a judge plans to rely on an authority or line
of analysis not argued, procedural fairness may call for inviting submissions.
Nevertheless, the appellate court read the judgment as ultimately deciding the case on the balance of probabilities, reinforced by an express
alternative finding preferring the respondent’s version. The misstep therefore did not undermine the result.
(c) Structure of the judgment: witness-by-witness vs chronological/iterative evaluation
The Court of Appeal was candid that a witness-by-witness approach was “probably not an ideal way” to handle multi-year commercial dealings; a chronological
framework may better allow iterative testing against probabilities and documents. But it refused to convert a stylistic/structural critique into an inference
that the judge failed to consider the totality.
(d) “Compelling evidence” arguments: why the four headline points did not move the needle
The appellant alleged four items were compelling corroboration of an equity (shareholding) deal, not merely a profits arrangement. The Court of Appeal rejected
each as insufficiently decisive to demonstrate oversight or irrationality:
-
£39,000 payment (Nov 2015): The judge did not overlook it (he mentioned it). It was not pleaded as a capital contribution; disclosure was lacking;
ledgers treated it as a loan; and the pleaded consideration for shares was services, not capital. Against the backdrop of the appellant’s credibility problems,
it was not a “must-address” item.
-
Property purchase (No. 2292): Even if intended for GBRL initially, it was ultimately owned personally and not treated as a company asset; the point
was not pleaded as now argued; it was not compelling evidence of an equity-sharing oral contract.
-
Rationale for incorporating GBRL: At most a credibility point, not core. The court also observed the judge did address it (notably referencing it in
paragraph 112 of the High Court judgment as identified on appeal).
-
67:33 structures (BMC/OPF/BMPF): Consistent with profit participation via GBRL rather than proof of who owned GBRL. The judge’s scepticism about
Baljit Singh’s alignment with the appellant was open to him.
(e) Credibility and consistency: the decisive battlefield
The Court of Appeal accepted the trial judge was entitled to treat the appellant’s shifting case as central. The appellant’s changing explanations—particularly
about whether Brand Connection Limited (BCL) was a personal venture or a “partnership asset”—were treated as going to reliability in a fundamental way.
Equally, while the respondent had inconsistencies (e.g., pleading a meeting with the accountant but later denying it), the judge did not ignore them and still found
the respondent generally credible. Appellate courts rarely disturb such evaluations, especially where the trial judge heard seven days of oral testimony and had
the advantage of observing witnesses.
3.3 Impact
(a) For future factual appeals
-
Missteps in reasoning do not automatically succeed on appeal: even where an appellate court criticises parts of the trial judge’s approach,
the appeal will fail unless the outcome is shown to be irrational or infected by a material error meeting the Henderson threshold.
-
“Compelling evidence” is a high threshold: reliance on Simetra Global Assets Ltd v Ikon Finance Ltd requires genuinely decisive
evidence that the judge’s failure to address it suggests true oversight. Evidence that is peripheral, disputed, consistent with the other side’s case, or poorly
pleaded may not qualify.
-
“Island-hopping” is disfavoured: appellants must show how the alleged omissions undermine the overall reasoning, not merely point to evidence that
could have supported a different conclusion.
(b) For shareholder/partnership-style disputes in informal business relationships
-
Companies House filings and tax treatments are not determinative of legal ownership: entries may reflect administrative practice, misunderstanding,
or one party’s instructions; they are part of the evidential mix, not a substitute for proving an agreement and proper corporate formalities.
-
Pleadings and consistency matter acutely: where the alleged contract is oral and contemporaneous documentation is thin, the court will lean heavily
on credibility and consistency across pre-action letters, pleadings, witness statements, and oral evidence.
-
Profit-sharing can mimic ownership: arrangements described in everyday language (including terms like “sanja”) may point to sharing of profits rather
than equity; careful articulation and corroboration are critical.
(c) For trial judges
-
If relying on unargued authorities or a novel analytical route, there is a practical (and sometimes fairness-based) case for inviting submissions.
-
In long-running commercial disputes, a chronological approach may better demonstrate the iterative reasoning process, though a different structure will not itself
constitute appealable error.
4. Complex Concepts Simplified
- Specific performance
-
A remedy ordering a party to perform a contract (here, to allot/transfer shares). It presupposes a binding contract with sufficiently certain terms.
- Oral agreement (oral contract)
-
A contract made by spoken words rather than writing. It can be enforceable, but proof problems are acute: the court must decide what was agreed and whether it was
intended to be legally binding.
- Balance of probabilities
-
The civil standard of proof: whether something is more likely than not.
- Burden of proof (and “last resort” use)
-
The obligation on a party (usually the claimant) to prove its case. In rare situations where a court genuinely cannot decide which account is more likely, the party
bearing the burden may lose. The Court of Appeal signalled that such “last resort” framing was inapt where the evidence permits a choice.
- Without prejudice
-
A privilege that can exclude communications genuinely aimed at settling an existing dispute from being used in court. Here, the trial judge admitted the communications
as not having the purpose of resolving a dispute.
- Unfair prejudice petition (section 994 of the Companies Act 2006)
-
A shareholder remedy where company affairs are conducted in a manner unfairly prejudicial to members’ interests. In this dispute, the appellant’s standing as a member
depended on whether he was truly a shareholder, prompting separate proceedings.
- Quasi-partnership
-
A company run like a partnership (personal relationship, mutual confidence, expectation of participation) where equitable considerations may apply. Alleging a
quasi-partnership does not itself prove legal share ownership.
- PSC (person with significant control)
-
A Companies House concept identifying persons who control the company (often through shareholding or voting rights). Being listed (or not listed) is not conclusive
of legal ownership; it is evidence to be weighed.
5. Conclusion
Singh v Bains & Anor is a strong restatement of appellate restraint on findings of fact: even where the appellate court identifies imperfections
in a trial judge’s method (including ill-judged reference to “last resort” burden-of-proof principles and a less-than-ideal witness-by-witness structure), the appeal
will fail unless the appellant demonstrates a recognised Henderson error—most realistically, that the judge overlooked apparently compelling
evidence or reached a conclusion that no reasonable judge could reach.
The decision also illustrates a practical lesson in informal business relationship litigation: where alleged equity arrangements are undocumented and accounts shift
across letters, pleadings and witness statements, courts are likely to resolve the dispute on credibility; Companies House filings, profit distributions, and later
conduct may support competing inferences and will rarely be “silver bullets” without clear, consistent, and properly pleaded foundations.