Recoverability of funding costs: a necessary debate, but not a simple one

The Singapore International Commercial Court’s 2 June 2026 decision in DTH and DTI v DTF and others is a timely contribution to a question funders, claimants and tribunals have been grappling with for some time: when a funded party succeeds, should the cost of obtaining that funding be recoverable from the losing party?

The Court did not decide that third party funding costs can never be recovered in a Singapore-seated arbitration. Its task was narrower. The applicants sought to set aside, or remit, a costs award after the tribunal majority concluded it had no power to award approximately US$14.6 million in funding costs. The claimants had recovered a buyout award of about US$14.7 million, so the commercial point was stark: if the funding return was payable but irrecoverable, the claimants said their victory was close to pyrrhic.

There can be cases where a funding return, particularly one based on a multiple of deployed capital, risks consuming a very substantial part of the damages ultimately awarded. That is not good for claimants, and nor is it a good look for the funding industry when it occurs. A funding agreement should not become the tail that wags the dispute. It should not make sensible settlement impossible, or distort the claimant’s decision-making because the economics of compromise become hostage to the funder’s return.

At Erso, we take that concern seriously. We try to avoid matters where, on a realistic range of outcomes, there is a material risk that the proceeds will be swallowed by the funding structure. That assessment begins by challenging the damages assumptions, even before we have considered the merits, enforcement, budget and duration. The objective is alignment: a claimant able to pursue a meritorious claim; lawyers able to run it properly; and a funder compensated for genuine risk without changing the natural trajectory of the dispute.

But DTH also illustrates why hindsight can be a dangerous lens. The Singapore court noted that the claim had been advanced on a substantially higher valuation: the applicants had sought a buyout figure of around US$65 million, and the US$14.7 million award fell far short of that. The case was not funded as a low-value claim in which the funding return was predictably going to devour the award. It became that case only after the tribunal’s quantum decision. Litigation and arbitration remain inherently uncertain. Even careful damages analysis can be overtaken by findings on causation, valuation, expert evidence, mitigation or remedy.

Without funding, many claimants would never reach that point at all. The Court recognised that funding enabled the applicants to bring the claim. In many disputes, funding is the bridge between having rights in theory and enforcing them in practice. The policy question is whether, and in what circumstances, the cost of that bridge should be shifted to the party whose conduct made the proceedings necessary.

There are, broadly, three possible outcomes. First, the claimant prevails with a high recovery, resulting in a sizeable net return even after settling its funding cost. Secondly, the case loses on the merits. In that scenario, the funder loses what is often a substantial investment made over a prolonged period, and receives no return. That will be disappointing for the claimant, but the court or arbitral tribunal has determined that the claimant was not entitled to damages, and at least the claimant is not out of pocket for having pursued the case. Then there is the third version: there is a recovery, but one that leaves little or no net return to the claimant after settling the funder’s return.

The funder does not control the outcome in any of these scenarios. It can try to mitigate the risks by declining matters that are likely to lose, or where the damages theory appears flawed or overzealous. But ultimately litigation and arbitration are high-risk processes. A claim may be responsibly funded on one view of quantum and decided very differently by the tribunal.

The Singapore decision approached the issue through the narrow gateway of public policy and set-aside. The applicants argued that access to justice for impecunious but deserving claimants required recoverability of funding costs. The Court rejected that formulation as too narrow to constitute public policy in the Article 34 sense, and held that even an erroneous refusal to award funding costs would not “shock the conscience”. It also stressed that access to justice is not one-way: the respondent’s position matters too. A losing party may be exposed to liability shaped by a private agreement to which it was not party.

That is a real point. Recoverability should not mean automatic pass-through of whatever the claimant agreed to pay. Any principled regime must include reasonableness, proportionality, disclosure sufficient to let the respondent understand its exposure, and careful attention to causation. Was funding reasonably necessary? Were the terms market-facing? Was the amount claimed proportionate to the risk transferred? Did the respondent’s conduct increase the claimant’s need for external capital? These questions are better suited to tribunal discretion than blunt rules.

The English arbitration cases have gone further in recognising that funding costs may, in appropriate cases, fall within “other costs”. Essar v Norscot opened the door; our affiliate company, TheJudge Group gave evidence in that case as to reasonableness of the funding agreement. Tenke Fungurume showed that the issue turns not only on power but on reasonableness, both as to recourse to funding and the amount claimed. Singapore has now signalled greater caution, particularly given its statutory framework and the express prohibition on recovery of third party funding costs in SICC proceedings.

As matters stand, claimants ought to approach any negotiation with funders on the basis that there may be no recoverability of the funding fee if they prevail. Funding terms should be stress-tested against realistic outcomes, not headline damages. Waterfalls, caps, ratcheting time-based multiples and settlement provisions can all reduce misalignment. Claimants should understand from day one that a tribunal may award damages, costs, some funding costs, or no funding costs at all.

Arguably, the ideal answer is not that funding costs should always be recoverable, or never recoverable. Tribunals should have room, where the governing law and rules permit, to award them in appropriate cases, subject to rigorous scrutiny.

DTH is therefore not a defeat for funding. It is a reminder that legitimacy matters. The strongest argument for recoverability is not that funders take risk and should therefore be paid by the losing party. It is that, in an appropriate case, the cost of funding may be part of the real cost of obtaining justice. Tribunals should have the flexibility to recognise that, as Norscot did, where the funding was reasonably necessary and the terms were themselves reasonable. But that flexibility cannot mean recovery at any level, or on any terms. If a claimant seeks to shift the cost of funding to its opponent, the funding structure must be capable of withstanding scrutiny.