Commercial Arbitration: When Business Partners Fall Out and the Small Print Takes Over

Most business relationships begin with optimism. The contract, if there is one, is signed with handshakes and goodwill. The arbitration clause, if anyone reads it at all, is treated as boilerplate — something the lawyers insisted on, buried in the governing law section, probably copied from a previous deal. Nobody expects to use it.

Then something goes wrong. A payment is missed. A delivery falls short. A joint venture partner starts making decisions that the other partner did not agree to. The relationship sours. Lawyers are instructed. And suddenly that boilerplate clause becomes the most important paragraph in the document.

Arbitration clauses appear in commercial contracts for a variety of reasons. Sometimes they are there because one party insisted on confidentiality — court proceedings are public; arbitration is not. Sometimes they reflect a genuine preference for a specialist tribunal rather than a generalist judge. Sometimes they are there because the parties are from different countries and neither wants to litigate in the other's courts. And sometimes, frankly, they are there because the lawyers on both sides used a standard template and nobody thought to question it.

Whatever the reason, once the clause is in the contract and a dispute has arisen, it is usually binding. The parties are committed to resolving their dispute through arbitration rather than through the courts. Understanding what that means in practice — and what it costs — is something that many businesses only discover when it is too late to choose differently.

Arbitration is a private dispute resolution process in which the parties agree to submit their dispute to one or more arbitrators, whose decision — called an award — is binding on both sides. It is not mediation, which is non-binding and aimed at settlement. It is not expert determination, which involves a technical expert making a decision on a specific question. Arbitration is a full adjudicative process, with pleadings, evidence, submissions and a final determination.

In England and Wales, arbitration is governed primarily by the Arbitration Act 1996, which gives the process a solid statutory foundation. The Act gives arbitrators broad powers to manage proceedings, award remedies and deal with costs. It also limits the grounds on which an award can be challenged in court — which is one of the reasons why arbitration clauses are taken seriously. Once an arbitrator has decided, it is very difficult to unpick.

International arbitration adds another layer. Where the parties are from different jurisdictions, the arbitration may be conducted under institutional rules — the ICC, LCIA, SIAC, UNCITRAL or others — and the seat of the arbitration (the legal home of the proceedings) will determine which national law governs procedural questions. The New York Convention, to which over 170 countries are party, means that an arbitral award made in one signatory state can generally be enforced in another. This is a significant practical advantage over court judgments, which face a patchwork of bilateral enforcement treaties.

For businesses operating across borders — whether in trade, investment, technology licensing or joint ventures — arbitration is often the only realistic mechanism for resolving disputes. The alternative is litigating in a foreign court, in a foreign language, under a foreign procedural system, with uncertain prospects of enforcement. Arbitration, for all its costs and complexity, at least offers a degree of predictability.

The arbitrator is the decision-maker. In a sole arbitrator appointment, one person hears the case and makes the award. In a three-member tribunal, each party appoints one arbitrator and the two party-appointed arbitrators then agree on a presiding arbitrator (or the institution appoints one if they cannot agree). The composition of the tribunal matters enormously — not just for the outcome, but for the conduct of the proceedings.

Arbitrators are typically senior lawyers, retired judges or technical experts with relevant industry knowledge. In a construction dispute, you might want a tribunal with engineering expertise. In a financial services dispute, someone who understands derivatives or structured products. In a general commercial dispute, a barrister or solicitor with a strong litigation background. The parties have a degree of choice that they do not have in court, where the judge is assigned by the listing office.

That choice comes with responsibility. Selecting an arbitrator is a strategic decision. You want someone who will understand your case, manage the proceedings efficiently and produce a reasoned award that can withstand scrutiny. You also want someone who is genuinely independent — arbitrators are required to disclose any connections to the parties or their lawyers, and challenges to arbitrator appointments, while not common, do happen.

The fees of the arbitrators are paid by the parties. In a sole arbitrator case, this might be a few thousand pounds for a straightforward dispute. In a three-member international tribunal with senior QCs or retired judges, the fees can run to hundreds of thousands of pounds before a hearing has even taken place. This is not a hypothetical concern. It is one of the most significant practical differences between arbitration and litigation, and it is something that parties often underestimate when they first encounter an arbitration clause.

Arbitration has a reputation for being expensive. That reputation is, in many cases, deserved. The costs fall into several categories: arbitrator fees, institutional fees (if the arbitration is administered by an institution such as the ICC or LCIA), legal costs, and the costs of any experts or witnesses.

Arbitrator fees are typically charged at an hourly or daily rate, or calculated as a percentage of the amount in dispute under institutional scales. A three-member tribunal in a significant commercial case might cost £50,000 to £150,000 in arbitrator fees alone, before any hearing. The LCIA's 2023 cost data showed average arbitrator fees of around £30,000 per case, but that average conceals enormous variation — straightforward cases at one end, multi-year mega-disputes at the other.

Legal costs follow a similar pattern. A party in a well-resourced international arbitration might spend £500,000 or more on lawyers, experts and preparation. Even in a domestic commercial arbitration of moderate complexity, legal costs of £50,000 to £150,000 per side are not unusual. The question of who pays those costs — and on what basis — is determined by the arbitrator in the award, applying principles broadly similar to those in English litigation: costs follow the event, subject to the arbitrator's discretion.

The practical implication is that arbitration is not always the right mechanism for smaller disputes. A claim worth £50,000 can easily generate costs that exceed the value of the claim. Parties who find themselves in arbitration over a modest sum should think carefully about whether a negotiated settlement — even an imperfect one — makes more economic sense than pursuing the process to an award. The arbitration clause does not require you to fight to the end; it simply determines the forum if you do.

Once a dispute has been referred to arbitration, the process broadly follows a pattern that will be familiar to anyone who has been involved in commercial litigation, though with significant procedural flexibility.

The claimant serves a notice of arbitration (or request for arbitration under institutional rules), setting out the nature of the dispute and the relief sought. The tribunal is constituted. There is usually a preliminary procedural hearing at which the timetable is set — pleadings, document production, witness statements, expert reports, and the hearing itself.

Document production in international arbitration is a significant exercise. Unlike English litigation, which has broad disclosure obligations, international arbitration typically follows the IBA Rules on the Taking of Evidence or a bespoke document production protocol. Parties request specific categories of documents from each other; the tribunal rules on disputed requests. The process is less extensive than English disclosure but can still generate substantial volumes of material.

Witness evidence is given by written statement, with oral cross-examination at the hearing. Expert evidence follows a similar pattern. The hearing itself — where witnesses are examined, submissions are made and the tribunal asks questions — can last anywhere from a day to several weeks, depending on the complexity of the case.

After the hearing, the parties typically exchange post-hearing briefs, and the tribunal deliberates. The award — which must be in writing and signed by the arbitrators — follows. In a complex case, it may take six months or more from the end of the hearing to the award. The award is final and binding, subject to the very limited grounds for challenge under the Arbitration Act 1996 or the applicable institutional rules.

One of the features that makes arbitration attractive — and occasionally frustrating — is the limited scope for challenging an award. In English law, the Arbitration Act 1996 provides three grounds: lack of substantive jurisdiction (section 67), serious irregularity (section 68), and appeal on a point of law (section 69).

Section 67 challenges — jurisdiction — arise where a party contends that the arbitrator had no power to decide the dispute. This might be because the arbitration clause was invalid, or because the dispute fell outside its scope. These challenges are relatively rare but can be significant where the clause is ambiguous or where one party was not properly party to the agreement.

Section 68 — serious irregularity — is the most commonly attempted ground of challenge, and the most rarely successful. The threshold is high: the irregularity must have caused or be likely to cause substantial injustice. Courts have consistently held that errors of fact or law, even significant ones, do not constitute serious irregularity. The arbitrator is entitled to be wrong. What the court will not tolerate is a fundamental failure of due process — a party not being given the opportunity to present its case, or an award dealing with issues that were never argued.

Section 69 — appeal on a point of law — is only available if the parties have not excluded it (many institutional rules do exclude it) and requires the court's permission. The court will only grant permission if the question of law is one of general importance or the decision is obviously wrong. It is not a mechanism for re-arguing the merits.

The practical message is that once an arbitral award has been made, it is very likely to stand. Parties who are unhappy with the outcome should take advice quickly — time limits for challenges are short — but should also be realistic about the prospects. The courts are reluctant to interfere with the arbitral process, and for good reason: the finality of the award is part of what makes arbitration valuable.

Arbitration is not always a slow, deliberate process. Sometimes a party needs urgent relief — to freeze assets, to preserve evidence, to prevent a counterparty from taking steps that would render the eventual award worthless. Arbitral tribunals have power to grant interim measures, and most institutional rules include emergency arbitrator procedures for situations where the tribunal has not yet been constituted.

The courts also retain jurisdiction to grant interim relief in support of arbitration proceedings. Under section 44 of the Arbitration Act 1996, the court can make orders for the preservation of evidence, the inspection of property, the sale of goods and — most importantly — freezing injunctions. A party who fears that its counterparty is dissipating assets or moving them beyond reach can apply to the court for a freezing order even where the underlying dispute is subject to arbitration.

The interaction between court-ordered interim relief and the arbitral process requires careful management. The court will generally be reluctant to grant relief that pre-empts the tribunal's decision on the merits, but it will act to preserve the status quo where there is a real risk of irreparable harm. Getting this right — knowing when to go to court and when to go to the tribunal — is one of the areas where experienced advice makes a material difference.

Winning an arbitration is one thing. Getting paid is another. If the losing party complies with the award voluntarily, the process ends there. If it does not, the winning party must enforce the award.

In England and Wales, an arbitral award can be enforced as a court judgment under section 66 of the Arbitration Act 1996. The process is relatively straightforward: the winning party applies to the court for permission to enforce, and once granted, the award has the same effect as a court judgment and can be enforced through the usual mechanisms — charging orders, third-party debt orders, writs of control.

International enforcement is more complex but, thanks to the New York Convention, generally achievable. A party with an award made in a Convention state can apply to enforce it in any other Convention state. The grounds for refusing enforcement are narrow — essentially limited to jurisdictional defects, procedural irregularities and public policy. Courts in most jurisdictions take a pro-enforcement approach.

The practical challenge is often not legal but factual: finding the assets. A counterparty that has moved its assets offshore, restructured its corporate group or simply has nothing worth taking presents a different kind of problem. Asset tracing, enforcement strategy and the use of ancillary court proceedings to locate and freeze assets are a significant part of post-award practice in international arbitration.

The time to think about arbitration is before the dispute arises, not after. An arbitration clause that is poorly drafted can cause significant problems — uncertainty about the seat, the rules, the number of arbitrators, the language of the proceedings, or the governing law of the arbitration agreement itself.

A few practical points worth considering. First, the seat matters. The seat determines the procedural law of the arbitration and the supervisory jurisdiction of the courts. London is a popular seat for international arbitration, with a sophisticated legal framework and courts that are experienced in arbitration matters. But other seats — Singapore, Paris, Geneva, Hong Kong — have their own advantages depending on the parties and the subject matter.

Second, the choice of institution matters. The ICC, LCIA, SIAC and others each have their own rules, fee structures and administrative cultures. The ICC's Terms of Reference process adds a layer of procedural certainty. The LCIA's rules are widely used in London-seated arbitrations. SIAC is the institution of choice for many Asia-Pacific disputes. The choice should reflect the nature of the transaction and the likely profile of any dispute.

Third, the number of arbitrators matters. A sole arbitrator is cheaper and faster; a three-member tribunal provides more deliberative depth and is harder to challenge on grounds of bias. For high-value or complex disputes, a three-member tribunal is usually worth the additional cost. For smaller disputes, a sole arbitrator makes more economic sense.

Fourth, consider whether arbitration is actually the right mechanism for your transaction. Not every commercial dispute benefits from arbitration. Where speed is critical, where the parties are both English and the dispute is straightforward, or where the amount at stake is modest, litigation may be more appropriate. The Commercial Court in London is an excellent forum for commercial disputes, and its procedures have been significantly streamlined in recent years.

Commercial arbitration is, at its core, a process for resolving disputes between people and organisations that once had a working relationship. The legal machinery — the pleadings, the document production, the expert reports, the hearing — can obscure that fact. But the human element is always present.

The business owner who built a company over twenty years and is now fighting a former partner over the terms of a shareholders' agreement is not primarily interested in the finer points of section 68 of the Arbitration Act. They want to know whether they have a case, what it will cost, how long it will take, and what the realistic outcome looks like. Those are the questions that matter.

Good arbitration advice starts with those questions. The legal analysis follows. Understanding the commercial context — what the parties actually wanted, what went wrong, what a fair outcome looks like — is as important as understanding the procedural rules. Arbitration is a tool. Like any tool, it works best when used by someone who understands both the tool and the job.

The small print, when it finally takes over, is rarely as impenetrable as it first appears. But it does require careful reading.