Case Details
- Citation: [2012] SGHC 170
- Title: Breezeway Overseas Ltd and another v UBS AG and others
- Court: High Court of the Republic of Singapore
- Decision Date: 16 August 2012
- Coram: Lee Seiu Kin J
- Case Number: Suit No 114 of 2010 (Registrar's Appeal No 412 of 2011)
- Procedural History: Appeal against part of the decision of the learned Senior Assistant Registrar (SAR) regarding disputed keyword search terms in an electronic discovery (e-discovery) process
- Plaintiffs/Applicants: Breezeway Overseas Ltd and another
- Defendants/Respondents: UBS AG and others
- Parties (key roles): First plaintiff: Breezeway Overseas Ltd (British Virgin Islands); Second plaintiff: Mr Vasanmal Murli (director and person exercising effective and complete control); First defendant: UBS AG (global financial services firm); Fourth and fifth defendants: Hong Kong and Singapore branches of UBS AG (collectively “the Bank”); Second and third defendants: former employees (Susan Abraham and Vikrant Kanyal)
- Legal Area: Civil Procedure – Discovery of documents; e-discovery; keyword searching; relevance/privilege/confidentiality
- Statutes Referenced: Rules of Court (Cap 322, R 5, 2006 Rev Ed), in particular O 24 (discovery)
- Cases Cited: [2010] SGHC 125; [2011] SGHC 223; [2011] SGHC 61; [2012] SGHC 41; [2012] SGHC 170
- Judgment Length: 9 pages, 4,620 words
- Counsel: Freddy Lim (Lee & Lee) for the plaintiffs; Tan Shou Min (Drew & Napier LLC) for the first, third, fourth and fifth defendants; Charmaine Chan (Legis Point LLC) for the second defendant
Summary
Breezeway Overseas Ltd and another v UBS AG and others [2012] SGHC 170 is a High Court decision addressing how keyword searches should be conducted in the context of e-discovery in Singapore civil litigation. The appeal concerned part of an earlier Registrar’s decision on disputed keyword search terms for electronic repositories (mailboxes, network profiles, and instant messaging records) of relevant bank employees. The court’s focus was not merely on whether certain keywords were “too broad” or “low relevance”, but on how keyword searching fits within the discovery framework under O 24 of the Rules of Court—particularly the requirement that discovery be “necessary either for disposing fairly of the cause or matter or for saving costs”.
Lee Seiu Kin J allowed the appeal in part. The court upheld the SAR’s approach for most of the disputed keywords, but adjusted the time period for the keyword “protest” to reflect the plaintiffs’ own case as to when the relevant protest occurred. Importantly, the court also rejected the Bank’s broader concern that, under the earlier e-discovery authority, it would not be open to review search results for relevance, privilege, and confidentiality. This decision therefore provides practical guidance on (i) tailoring keyword searches and (ii) the extent to which parties may assess the output of keyword searches to ensure discovery remains fair and appropriately bounded.
What Were the Facts of This Case?
The underlying dispute arose from alleged misrepresentations and related conduct by a bank in connection with leveraged bond transactions. The plaintiffs were Breezeway Overseas Ltd, a company incorporated in the British Virgin Islands, and Mr Vasanmal Murli, a director of the company who was said to exercise effective and complete control over it. The first defendant was UBS AG, a global financial services firm headquartered in Basel and Zurich. The fourth and fifth defendants were the Hong Kong and Singapore branches of UBS AG. For convenience, the judgment refers to UBS and its relevant branches collectively as “the Bank”.
Between February and July 2008, the plaintiffs alleged that they relied on representations made by the Bank and took loans from the Bank’s branch to purchase a series of leveraged bonds. The leveraged bonds included: (a) 6.625% ICICI Bank Bonds (13 February 2008); (b) 7.335% Bank of Moscow Bonds (14 February 2008); (c) 7.335% Bank of Moscow Bonds (16 April 2008); (d) 6.609% VTB Cap Bonds (15 April 2008); (e) 8.25% VTB Cap Bonds (July 2008); and (f) 5.75% Kaupthing Bank Bonds (6 March 2008). The leveraged bonds were later sold around 13 August 2008, and the loan proceeds were applied towards the purchase of 13.625% Venezuela Bonds around the same time.
The plaintiffs’ case was that, in a meeting at the Bank’s Suntec City office, the Bank’s employees represented that the loans were fixed loans and/or fixed to maturity, meaning they could not be recalled before maturity. The plaintiffs further alleged that they were not told the first plaintiff had to provide “collateral” to secure the loans, nor were they informed of which assets were allegedly collateralised. These allegations formed the basis for claims including misrepresentation, breach of fiduciary duties, gross negligence, wilful misconduct, and breach of contract.
In or about March 2009, the Bank issued margin calls requiring the plaintiffs to raise substantial sums within a short period, failing which the Bank threatened to liquidate the plaintiffs’ assets. The plaintiffs protested against the Bank’s decisions to reduce the “loanable value” of the leveraged bonds and to make the margin calls. The e-discovery application in this case was therefore aimed at locating electronic communications and documents that would shed light on the Bank’s representations, the loan terms, the collateral arrangements, and the events leading to the margin calls and protests.
What Were the Key Legal Issues?
The immediate legal issue was procedural and discovery-focused: whether the SAR was correct in allowing certain keyword searches and disallowing others for the purpose of e-discovery. The appeal concerned nine of the ten keywords that the SAR allowed. The only keyword not in issue was a proximity search for “fixed” within ten words of “maturity”, which was excluded from the appeal because the Bank’s search engine was said not to be capable of performing proximity searches.
Beyond the keyword list itself, the case raised a more substantive concern about the discovery process in e-discovery. The Bank argued that the disputed keywords were of low relevance and that conducting searches using them was not necessary for the fair disposal of the action or for saving costs. A further “sub-text” was the Bank’s interpretation of earlier e-discovery guidance—particularly the decision in Robin Duane Littau v Astrata (Asia Pacific) Pte Ltd [2011] SGHC 61—suggesting that the parties might not be entitled to review the search results for relevance, privilege, and confidentiality. The Bank’s position was that, if review were not permitted, keyword searches could be over-inclusive, producing irrelevant material that would not be properly filtered.
Accordingly, the court had to decide (i) how to apply the O 24 discovery necessity requirement to keyword searches in e-discovery, (ii) how to assess relevance and proportionality when keywords are “generic” or potentially broad, and (iii) the extent to which a party receiving keyword search results may review them to manage relevance, privilege, and confidentiality concerns.
How Did the Court Analyse the Issues?
Lee Seiu Kin J began by situating the dispute within the discovery framework under O 24 of the Rules of Court. Under O 24, a party must disclose documents that are or have been in its possession, custody or power and that are relevant to the issues in dispute, subject to the requirement in O 24 r 7 that discovery must be “necessary either for disposing fairly of the cause or matter or for saving costs”. The court emphasised that e-discovery is not a departure from discovery principles; rather, it is a technological method of implementing those principles. The question therefore remained whether the proposed keyword searches were necessary in the O 24 sense.
The court then addressed the SAR’s approach to keyword selection. The SAR had ordered that the Bank conduct reasonable keyword searches on defined electronic repositories for specified time periods. The parties had attempted to agree on a set of keywords; the Bank agreed to seven keywords and objected to 23. The Bank’s objection was that many proposed keywords were too generic and broad given that the employees whose documents would be searched were banking professionals in wealth management. The SAR, after preliminary searches and review of the number of hits, allowed ten keywords and disallowed the remaining 13. In the appeal, the court was asked to revisit nine of those allowed keywords.
In evaluating the disputed keywords, the court accepted that relevance and necessity in e-discovery cannot be assessed in the abstract. The SAR had considered both (i) the issues in dispute in the underlying action and (ii) the context in which the keywords would appear in the communications. For example, the SAR permitted “collateral” because collateralisation was directly in issue, including whether the loans were secured against collateral and whether the plaintiffs were informed of which assets were collateralised. Similarly, “Fixed Loan” and a proximity search relating to “fixed” and “maturity” were allowed because the plaintiffs’ case turned on whether the loans were fixed to maturity. The SAR also allowed “Protest” because the plaintiffs alleged that they protested against the Bank’s reduction in loanable value and the margin calls, and those protests were relevant to the Bank’s conduct and decision-making.
However, the High Court modified the SAR’s order in one important respect: the time period for the keyword “protest”. The plaintiffs had agreed that the alleged “protest” only arose on or after February 2009. The court therefore allowed the appeal relating to the time period and ordered that the keyword search for “protest” be restricted to February 2009 to June 2009. This adjustment reflects a key principle in e-discovery: even where a keyword is relevant, its search scope must be calibrated to the temporal relevance of the underlying issues. Over-broad time ranges can increase noise and cost without improving the fairness or efficiency of discovery.
For the remaining keywords, the court agreed with the SAR’s order. The judgment indicates that the SAR’s reasoning was grounded in the uniqueness or contextual relevance of the terms to the leveraged bonds and the banker-customer relationship. Keywords such as “Bank of Moscow”, “Kaupthing”, “Kuznetski”, and “Republic of Venezuela” were permitted because the leveraged bonds were relatively unique to the plaintiffs’ relationship with the Bank, making it less likely that the searches would capture large amounts of unrelated material. Likewise, keywords “ICICI” and “VTB” were permitted but restricted to particular repositories (mailboxes and P-drive of specified employees) in view of the preliminary hit counts. This demonstrates that the court endorsed a proportionality approach: where a keyword is potentially broader (e.g., common bond issuers or desk references), the search can be narrowed by repository and time period to manage over-inclusiveness.
The court also addressed the Bank’s concern about over-inclusiveness and the ability to review search results for relevance, privilege, and confidentiality. The Bank’s argument relied on an interpretation of Robin Duane Littau [2011] SGHC 61. In that earlier decision, the court had discussed preliminary searches and the purpose of keyword searches in identifying potentially relevant documents. The Bank contended that, under that authority, it would not be permitted to review search results for relevance, and therefore the output might be over-inclusive. Lee Seiu Kin J disagreed with the Bank’s interpretation. The court’s reasoning, as reflected in the excerpt, was that the Bank’s concerns were “substantially assuaged” because the High Court did not accept the Bank’s reading of Robin Duane Littau. While the truncated extract does not reproduce the full doctrinal discussion, the outcome is clear: parties are not locked into a mechanical “hit list” approach. Instead, the discovery process must allow for appropriate filtering and handling of relevance, privilege, and confidentiality so that discovery remains aligned with the substantive requirements of O 24.
In practical terms, this means that keyword searches are a starting point for identifying potentially relevant documents, but they do not replace the legal assessment of what must be disclosed. The court’s approach supports a workable e-discovery model: use keyword searches to locate candidate documents, then apply legal review to determine relevance and to protect privileged or confidential material. This balances efficiency with fairness, ensuring that e-discovery does not become an unbounded exercise in collecting irrelevant information.
What Was the Outcome?
The High Court allowed the appeal in part. The principal modification was to restrict the keyword search for “protest” to the period of February 2009 to June 2009, consistent with the plaintiffs’ agreement that the relevant protest occurred only from February 2009 onwards. This ensured that the search scope matched the temporal relevance of the underlying allegations.
For the remaining disputed keywords, the court agreed with the SAR’s orders. The effect of the decision is that the Bank was required to conduct the ordered keyword searches on the specified electronic repositories and time periods, subject to the adjusted timeframe for “protest”. The decision also clarified that the Bank’s concerns about an inability to review search results for relevance, privilege, and confidentiality were not accepted, thereby supporting a more legally grounded and controlled e-discovery process.
Why Does This Case Matter?
Breezeway Overseas Ltd v UBS AG [2012] SGHC 170 is significant for practitioners because it provides concrete guidance on how Singapore courts will supervise e-discovery keyword searches under O 24. The decision reinforces that keyword selection is not purely technical; it is a legal exercise tied to the issues in dispute and the necessity requirement for discovery. Lawyers advising clients on e-discovery protocols should therefore expect courts to scrutinise both the relevance of keywords and the proportionality of search parameters such as time periods and repositories.
The case also matters because it addresses the practical problem of over-inclusiveness. Banks and other document-heavy parties often worry that broad keyword searches will generate large volumes of irrelevant material, increasing review costs and potentially undermining confidentiality. The High Court’s rejection of the Bank’s restrictive interpretation of Robin Duane Littau supports the view that parties can and should review search outputs to manage relevance and to handle privilege and confidentiality. This is essential for ensuring that e-discovery remains fair and efficient rather than turning into a purely mechanical data extraction exercise.
From a precedent perspective, the decision contributes to the developing Singapore jurisprudence on e-discovery. It aligns with earlier authorities that emphasise preliminary searches, iterative refinement, and court supervision of keyword protocols. For litigators, the case is a useful reference point when negotiating e-discovery terms, especially where parties disagree on whether certain keywords are “too generic” or whether the search process should be limited to avoid unnecessary costs.
Legislation Referenced
- Rules of Court (Cap 322, R 5, 2006 Rev Ed), O 24 (Discovery of documents), including O 24 r 7 (necessity for disposing fairly or saving costs)
Cases Cited
- [2010] SGHC 125
- [2011] SGHC 223
- [2011] SGHC 61 (Robin Duane Littau v Astrata (Asia Pacific) Pte Ltd)
- [2012] SGHC 41 (Breezeway Overseas Ltd v UBS AG)
- [2012] SGHC 170 (Breezeway Overseas Ltd and another v UBS AG and others)
Source Documents
This article analyses [2012] SGHC 170 for legal research and educational purposes. It does not constitute legal advice. Readers should consult the full judgment for the Court's complete reasoning.