What If CVC Had Taken Its CIETAC Award Straight to Singapore in 2019?

First, the timeline

Everything below comes from published decisions of the Hong Kong, Singapore, Chinese and US courts:

• 26 Feb 2015 — Hong Kong: worldwide freezing order and asset disclosure orders against Zhang Lan and related entities
• 2 Mar 2015 — Singapore: freezing orders; Credit Suisse and Deutsche Bank accounts frozen
• 5 Mar 2015 — Beijing: CIETAC arbitration formally commenced
• 28 Apr 2019 — Beijing: two partial awards issued (US$50,434,209 and US$92,029,457.28 plus interest)
• 12 Jul 2019 — Beijing: Zhang Lan's side applies to the Second China International Commercial Court (CICC) to set aside the awards
• 29 Jul 2019 — Hong Kong: leave granted to enforce the awards
• 15 Nov 2019 — Hong Kong: Zhang Lan's side applies to set aside the enforcement orders, or to stay pending the Mainland set-aside application
• 23 Apr 2020 — Hong Kong ([2020] HKCFI 622): the court imposes a Hadkinson order on Zhang Lan and conditions the corporate respondents' adjournment on security
• 20 May 2020 — Hong Kong: judgments entered recognising and enforcing the awards ("the HK Judgments")
• 11 Nov 2020 — Singapore: the HK Judgments are registered as Singapore judgments
• 29/31 Dec 2020 — Beijing: the CICC dismisses the set-aside application; the ruling is final and unappealable
• 2 Nov 2022 — Singapore ([2022] SGHC 278): the court finds the account funds beneficially owned by Zhang Lan and appoints receivers
• 27 Jun 2023 — Singapore ([2023] SGHC(A) 22): affirmed on appeal

One comparison in that timeline is the key to everything:

The HK Judgments were registered in Singapore on 11 November 2020 — six weeks before the Beijing court ruled on the set-aside application.

Question 1: What would the direct route have looked like?

To be clear from the outset: a direct, statutory recognition-and-enforcement pathway to Singapore existed. Hong Kong was never a mandatory stop. China and Singapore are both New York Convention states. A CIETAC award is a "foreign award" under Part 3 of Singapore's International Arbitration Act, and Singapore courts are firmly pro-enforcement. The question was never whether the direct route existed. It was what would happen on it — which is exactly what the rest of this section maps out.

Obstacle 1: A pending set-aside application at the seat. By July 2019, Zhang Lan's side had already applied in Beijing to set the awards aside, arguing the tribunal's composition breached the arbitration agreement. A distinction matters here. Article V(1)(e) of the New York Convention deals with awards that have already been set aside or suspended at the seat — a ground for refusing enforcement. A set-aside application that is still pending, as Zhang Lan's was, engages Article VI instead, mirrored in s 31(5) of Singapore's IAA: the enforcing court has a discretion to adjourn enforcement pending the seat court's decision, and may order the debtor to provide security as the price of the adjournment. Singapore case law treats this squarely as a discretionary balancing exercise — weighing whether the set-aside application is genuinely arguable against the risk that it is a pure delay tactic. The outcome cannot be known in advance, but the moment that application is filed and argued, the creditor no longer controls the clock. The CICC did not rule until the very end of December 2020. In the worst case, the direct route could have left final execution in Singapore in limbo for a year or more. To be clear, the creditor would not have been helpless in the meantime — security could be sought, and the existing freezing orders would have kept the assets in place. What is genuinely lost is certainty over the timetable: when the money actually moves becomes a function of the Beijing set-aside proceedings.

Obstacle 2: The Article V defence window stays open. One common misconception first: Singapore courts do not re-try the arbitration on its merits. They would not re-decide whether Zhang Lan made false representations, or recalculate the damages. But direct award enforcement does mean the debtor can raise, afresh in Singapore, every ground listed in Article V and IAA s 31 — validity of the arbitration agreement, tribunal composition, due process, scope, public policy. The court doesn't retry the case, but it does still have to work through whatever of those statutory grounds the debtor actually raises. And this was not a passive debtor. The public record shows breaches of Hong Kong disclosure orders, a contempt finding carrying a 12-month sentence, and continued resistance in the US proceedings as well: a 2023 appeal to the Second Circuit ended with the appeal dismissed and the underlying 2023 SDNY opinion vacated; after the SDNY confirmed the attachment order and entered a confirmation-and-enforcement judgment in March 2026, a third party asserting its own interest filed a fresh notice of appeal in April 2026, and disputes over turnover of assets and appointment of a receiver were still being briefed as of May 2026. Worth noting too: Article V's language is that recognition "may be refused," not "shall be refused" — the enforcing court keeps residual discretion even where a ground is technically made out. So an open Article V window doesn't mean the debtor is likely to win; it means every ground she raises has to be substantively worked through, win or lose, and that takes time. Opening a brand-new front with that window fully open would have handed her exactly the battlefield she wanted.

Obstacle 3: Hong Kong's specific leverage doesn't transplant to Singapore. This point needs precision, because it's easy to overstate. Singapore was not a blank slate. The CVC side had a substantial track record there too: a freezing order on 2 March 2015, followed by CS and DB freezing the SETL accounts, and then two pre-action discovery actions (OS 305 and OS 307) that pulled account-opening documents and a complete transaction record back to December 2013 out of both banks — upheld on appeal ([2016] SGHCR 3; [2016] 4 SLR 1392). That evidence later became central to proving Zhang Lan's beneficial ownership in 2022. So the real distinction isn't "Hong Kong accumulated leverage, Singapore accumulated nothing." It's that the two accumulated different kinds of record: Singapore built an asset-side case file — freezing, bank disclosure, transaction evidence. Hong Kong built that too, but on top of it accumulated a debtor-conduct file — five years of Zhang Lan personally breaching court orders, culminating in a contempt finding and imprisonment. That second layer is what mattered in April 2020, when Hong Kong ([2020] HKCFI 622) both imposed a Hadkinson order on Zhang Lan — refusing to hear her application unless she first complied with the outstanding disclosure order — and conditioned any adjournment for the corporate respondents on security of 40% of the total award, plus indemnity costs, with the whole application dismissed if security wasn't posted by 4pm on 14 May 2020. A tool like the Hadkinson order runs on that court's own record of the debtor's conduct before it. It doesn't travel with the case. What genuinely could not have been moved to Singapore was not the asset-tracing work — that existed on both sides — but the judicial pressure Hong Kong had specifically built against Zhang Lan's conduct.

The verdict of the war-game: the direct route was not a dead end — but it would have imported the entire timing uncertainty of the pending set-aside, an open Article V defence window, and a forum missing the debtor-conduct leverage Hong Kong had specifically built up, all at once.

One clarification matters here, to avoid a common misreading: the Hong Kong route did not dodge the Beijing set-aside proceedings. Hong Kong faced the exact same request from Zhang Lan's side to wait for Beijing. What the Hong Kong court did was manage that risk — using the accumulated Hadkinson order, disclosure record, and the 40% security condition to make delay costly, so enforcement could keep moving regardless. The outcome: Hong Kong judgments were entered on 20 May 2020, and Singapore registration completed on 11 November 2020 — while the two Beijing set-aside applications weren't dismissed until 29 and 31 December 2020.

Same variable — a pending set-aside — faced by both routes. The difference isn't who avoided it. It's who had the tools to manage it.

Question 2: What exactly does the "judgment bridge" buy you?

"Judgment bridging" means not taking the arbitral award itself into the target jurisdiction, but first converting it into a court judgment in an intermediate jurisdiction, then moving the judgment. A note on terminology: this is a label of convenience used in this article, not a formal doctrine found in Hong Kong or Singapore statute or case law. In this case, the structure delivered value on five distinct levels.

1. It narrows the defence window from Article V to registration grounds. The HK Judgments entered Singapore under the Reciprocal Enforcement of Foreign Judgments Act 1959 (REFJA), which covers final Hong Kong money judgments. To set aside a REFJA registration, a debtor is limited to a short statutory list: no jurisdiction in the original court, fraud, breach of natural justice, public policy, lack of finality. Compare that with the Article V menu. This needs one qualification: the point isn't that arbitration-specific defences vanish outright — it's that they no longer arrive as an independent ground at the registration stage. If the underlying facts happen to also fit one of REFJA's own grounds (jurisdiction, natural justice, fraud, public policy), they could in principle still be argued that way. But as a freestanding "the tribunal was improperly constituted" objection — the point Zhang Lan's side ran everywhere — it has nowhere to land as such at registration. The accurate conclusion isn't that every arbitration-related dispute disappears; it's that the debtor faces a materially narrower statutory defence framework carrying a Hong Kong judgment than carrying the award itself.

2. The room for repeated defences shrinks over time — but not automatically. This point needs care, because it's easy to backdate a legal development that hadn't happened yet. Looking back from today's Singapore law, transnational issue estoppel does provide meaningful institutional backing for multi-jurisdiction enforcement. The Singapore Court of Appeal recognised in Merck (2021) that a competent foreign court's final decision on a point can bind that point in Singapore; India v Deutsche Telekom (2023) went further, confirming that in the right circumstances a seat court's ruling on an award's validity can carry issue-estoppel effect too. But two dates matter here: Deutsche Telekom is a 2023 decision, while the HK Judgments were registered in Singapore in November 2020. And more importantly, the CICC's final ruling on the tribunal-composition point wasn't handed down until 29/31 December 2020 — after Singapore registration. So it would be wrong to read the later Singapore case law back into the 2020 registration as if the CVC side already enjoyed that procedural advantage at the time — and in particular, "the defence was still being decided in Beijing" does not by itself create any estoppel; a proceeding that is merely pending has no res judicata effect. Issue estoppel is also never mechanical: it requires a recognisable final judgment, substantially the same issue, and the same parties (or equivalent), assessed case by case — not a rule that "lose once abroad, lose the right to raise it anywhere, forever." The more careful and accurate reading of the Zhang Lan case is this: as the Hong Kong enforcement proceedings and, later, the Beijing set-aside proceedings each reached their conclusion by the end of 2020, the space for disputing the award itself narrowed; by the time the fight reached the substantive Singapore stage — the receivership proceedings of 2022–2023 — the centre of gravity had clearly shifted from "is the award valid" to "does Zhang Lan beneficially own the SETL accounts." The "one-way valve" image can stay as a description of the general pattern in multi-jurisdiction enforcement — it just shouldn't be presented as the specific legal mechanism that explains the 2020 registration.

3. A registered judgment plugs straight into Singapore's domestic toolkit. Under REFJA, a registered judgment has the same force and effect as a Singapore judgment. Garnishee proceedings, charging orders, and — decisively here — receivers by way of equitable execution are all available in principle, with no award-to-judgment seam left for the debtor to attack. One timing caveat: registration itself doesn't mean execution can start immediately. The debtor has a window to apply to set the registration aside, and execution has to wait until that window lapses, or, if a set-aside application is actually filed, until it's resolved. So "the interface is ready" describes the legal tools — it doesn't mean the clock starts the day of registration; the real execution timetable still depends on how much of that window the debtor uses. The Singapore courts appointed receivers over the SETL accounts on exactly that registered-judgment foundation in 2022, affirmed in 2023.

4. Different disputes ended up in the forum best equipped to handle them. Hong Kong enforces Mainland awards under the 1999 Mainland–Hong Kong Arrangement, has two decades of experience doing so, and — in this case — held five years of procedural record against the debtor. The arbitration-procedure defences were absorbed there, where the Hadkinson leverage existed. The beneficial-ownership defence was then resolved in Singapore, where the assets sat, on the strength of an already-judgmentised debt. This describes the outcome that emerged — not a claim about what the CVC side's lawyers deliberately intended, which the public record simply doesn't reveal.

5. The bridge is addition, not substitution. Getting the HK Judgments didn't leave the CVC side dependent on Hong Kong judgments alone — the record shows this directly. In the United States, the CVC side enforced the two CIETAC awards directly under the New York Convention throughout, rather than routing through the HK Judgments. Two SDNY tracks carried this: an attachment proceeding opened in 2019 (No. 1:19-mc-00536), where the court confirmed the attachment order on 24 March 2026 and closed the motion; and a confirmation-and-enforcement action in which the district court worked through the Article V grounds and entered a judgment on 31 March 2026 confirming the awards enforceable against Zhang Lan directly.

Worth flagging: the US track kept moving afterward, and moved into a new asset class. In April 2026, the CVC side sought turnover of paintings held in Zhang Lan's name and appointment of a receiver over them; a third party asserting its own interest (Apex Lead Investment Holdings Limited) then filed a fresh notice of appeal, and the turnover-and-receivership dispute was still being briefed as of May 2026. In other words, the New York track never once borrowed the Hong Kong judgments — and it has since advanced from the original Manhattan apartment to fine art, on the strength of the same arbitral awards, in the same court. So the real structure isn't "judgment replaces award." It's:

In Singapore, the right entered as a judgment, because the registration channel carried the narrower defence window. In New York, the same right entered as an award, because the New York Convention was itself already a mature channel there. One debt, two legal forms, deployed jurisdiction by jurisdiction according to which form faced the narrower defence window.

That's what a full picture of "judgment bridging" looks like — not a substitute route, but one extra tool added to the creditor's toolkit.

(Note: the US proceedings were still ongoing as of this writing; readers should check the latest SDNY and Second Circuit filings for current status.)

The honest caveats

The bridge is not a universal template. It requires a judgment-circulation channel between the intermediate and target jurisdictions — REFJA happens to cover final Hong Kong money judgments, but swap either jurisdiction and the channel may not exist. REFJA also excludes registering "judgments on judgments" of other foreign courts; a judgment recognising an arbitral award passed in this case, but that analysis is case-specific. The bridge costs money and time, which was economical here only because the Hong Kong front had existed since 2015. And the leverage — Hadkinson orders, security conditions, the estoppel dynamics discussed above — was generated by this particular debtor's conduct. A compliant debtor produces none of it.

Even a successful structure doesn't guarantee full recovery. As of the 2022 Singapore judgment, the CS and DB accounts held approximately US$22 million and US$33.4 million respectively — roughly US$55 million combined, against a debt of about US$142.46 million. Even a fully successful three-jurisdiction structure, with receivers duly appointed, only reaches money that is actually there; the balance still depends on tracing other assets, such as the New York apartment. The most sophisticated enforcement architecture solves how to reach an asset — not how much of the debt that asset actually covers.

Three takeaways for creditors

The shortest route is not the fastest route. Skipping Hong Kong saves a stop — and risks a year hostage to a pending set-aside. The detour finished Singapore registration before Beijing even ruled. Cross-border enforcement is not about minimising jurisdictions; it is about controlling the clock.

The legal form of your right is a design variable. The same US$142 million existed as an enforcement order in Hong Kong, a registered judgment in Singapore, and an award awaiting confirmation in New York. A mature enforcement plan asks: in what form should this right arrive in the next jurisdiction, so that the defence window is narrowest?

Procedural record is an asset. Alongside the money in the accounts, the most valuable thing in this case was five years of freezing orders, disclosure orders and contempt findings sitting in the Hong Kong file. At the decisive moment they became a Hadkinson order and security conditions — and made judgment conversion possible while the set-aside was still alive. Where you accumulate procedural record is where you stockpile enforcement leverage.