This judgment text has undergone conversion so that it is mobile and web-friendly. This may have created formatting or alignment issues. Please refer to the PDF copy for a print-friendly version.

Breezeway Overseas Ltd v UBS AG
[2012] SGHC 41

Case Number:Suit No 114 of 2010 (consolidating Suit No 112 of 2010), Summons No 2443 of 2011
Decision Date:28 February 2012
Tribunal/Court:High Court
Coram: Yeong Zee Kin SAR
Counsel Name(s): Mr Julian Tay with Mr Freddy Lim for the plaintiffs; Mr Tan Shou Min for the first, third, fourth and fifth defendants; Ms Charmaine Chan for the second defendant.
Parties: Breezeway Overseas Ltd — UBS AG

28 February 2012

Yeong Zee Kin SAR:

1       The first plaintiff, Breezeway Overseas Ltd (“Breezeway”), is a customer of the first defendant, UBS AG (“UBS”). Breezeway is a family investment vehicle controlled by the second plaintiff, Vasanmal Murli, who is one of its directors; the other directors are his wife and two daughters. Breezeway held accounts with the Hong Kong and Singapore branches of UBS, ie the fourth and fifth defendants respectively. At the material time, Breezeway was initially serviced by the second and then the third defendants, who were his client advisers and employees of the Singapore branch, ie fifth defendant.

History of summons application and order for discovery in stages

2       The present summons was taken out by the plaintiffs for discovery to be carried out in accordance with a proposed electronic discovery protocol. The second and third defendant had, by the time the summons came up for hearing, left the employment of UBS. They had also deposed to the fact that they did not have in their possession, custody or power any discoverable documents – all discoverable documents had been turned over to UBS when their employment with UBS came to an end. As UBS was not familiar with how these documents had been organised, the use of search terms to identify discoverable documents would be conducive to the efficient management of discovery.

3       After hearing submissions from counsel, I had ordered that parties adopt an electronic discovery protocol that provided for discovery to be carried out in stages and for the use of search terms. Parties were directed to discuss and agree on the search terms and liberty to apply was granted for parties to apply to court for determination of any disputed search term. Eventually, parties came back before me for determination of disputed search terms. UBS has now appealed against my decision in relation to the search terms that had been ordered. I set out the reasons for my decision in these grounds.

The raison d’être for discovery in stages

4       In olden days, there was better discipline in maintaining a more or less complete set of documents in a single file. The paper file would usually be kept by the department concerned with the transactions in question or sometimes centrally. Duplicates of certain documents were maintained by other departments: for example, by the finance or accounts department as part of their payment records. Employees usually were not expected to maintain their personal files; and their working papers, unless incorporated into the departmental file on the matter, were usually discarded.

5       In more recent times, e-mail messages have taken over the functions of printed correspondence to a large extent. This has resulted in two phenomena – first, the proliferation of copies and second, the decentralisation of records. Electronic mail is personal in the sense that each employee has his own e-mail account. The ease of inserting multiple addressees in an e-mail makes it easy to send the same message to a number of persons. Documents are also thus sent to multiple recipients as attachments to the e-mail message. As a result, the same message appears in each recipient’s e-mail account; and the same documents may be stored on each recipient’s computer.

6       Concomitantly, the practice of maintaining a central file of paper documents that is also a complete record became a practice that was no longer assiduously observed in many organisations. Although document management systems are available that enable an organisation to maintain a central electronic record of documents and e-mails, not many organisations have implemented these systems. Consequently, it is now much more difficult for an organisation like UBS to be able to “pull out the relevant file”. It has to look for documents in the e-mail accounts and computers of all employees involved in the matter. The problem is compounded where, like in the present case, the employees who were most actively involved in the matter are no longer in their employ. Without a centralised document management system, it has to potentially look for discoverable documents in the e-mail accounts and computers of all the employees who had anything to do with Breezeway’s account.

7       This poses a challenge to the application of traditional discovery principles to modern civil litigation. The old rules worked well when all one had to do was to “pull out the relevant file” and identify documents relevant to the issues in dispute. The relevant file was centralised and the problem of duplicates not so severe. In the modern office, the relevant file translates to the e-mail account, the computer hard disk and (frequently) the network storage of each employee. The oftentimes indiscriminate copying of e-mails to multiple addressees will assure that the party giving discovery will have to expend time and resources in identifying and removing duplicates.

8       Although there are computer programs that will be able to identify duplicates – ie de-duplication software – these work only on electronic documents. The effort of identifying and removing duplicates, particularly when documents from multiple custodians and repositories have been collated, should not be underestimated. There is a more efficient and cost-effective way to manage discovery.

How discovery may be conducted in stages

9       Conducting discovery in stages requires that parties identify at the close of pleadings both the issues in dispute and the witnesses that are key to these disputed issues. For the purpose of electronic discovery, these witnesses are referred to as custodians: an emphasis on their role as custodians of both knowledge of the relevant facts as well as custodians of the relevant documents.

10     Once the key custodians are identified, attention is turned to the repositories of electronic documents in their possession, custody or power. The repositories will typically include their e-mail accounts, hard disks on their desktop and notebook computers, removable storage media, online storage locations on the network, etc. Again, it is not necessarily the case that every repository under the possession, custody or power of each custodian has to be within the scope of the initial stage of discovery. Much depends on the amounts at stake and the significance of the issues in dispute: ie proportionality.

11     Having identified the custodian and the repositories, it is open to parties to either identify the relevant storage media or folders within storage devices (eg hard disks) to be disclosed or parties can agree that a reasonable search be conducted on the identified repositories. Where the custodian has been assiduous in his electronic filing, he may have organised his e-mails and documents into structured folders, much like how paper files used to be kept. In such felicitous circumstances, it will in fact be possible to “pull out the relevant file” by producing a copy of the relevant electronic folders and their content. This will be almost like traditional discovery. The party giving discovery may review the identified electronic folders for irrelevant e-mails and documents before making a copy of the relevant electronic documents for his adversary.

12     Oftentimes, e-mails and electronic documents are not so neatly organised and a keyword search may need to be conducted to identify documents for discovery. In the parlance of our practice directions on electronic discovery, this is a “reasonable search” – an electronic search conducted on the repositories using a set of agreed search terms with reasonable limits: Robin Duane Littau v Astrata (Asia Pacific) Pte Ltd [2011] SGHC 61, at [22]. The search is delimited in two respects.

(a)     First, parties have to decide whether the entire storage device (eg hard disk) or storage medium is to be searched, or only certain folders and sub-folders. For hard disks, parties should usually identify the relevant folders and sub-folders as modern operating systems usually establish a set of folders where documents are stored. This will obviate the necessity of searching a multitude of folders where operating system, application software, library and configuration files are stored.

(b)     Second, parties have to agree on the time period during which the relevant documents were created or received. This will serve to exclude documents responsive to the search term but which are likely to be irrelevant. The time period may differ for different key words, repositories or custodians.

13     The search results are presumed to be relevant without the need for a further review of each document for relevance: Sanae Achar v Sci-Gen Ltd [2011] 3 SLR 967; [2011] SGHC 87, at [23].

14     Conducting discovery in stages will ameliorate the effects of the proliferation of copies and decentralisation of records. By focusing on the key custodians, the party giving discovery need not (at least initially) go beyond those persons who are likely also to be the key witnesses of fact at trial. As has been observed in Goodale & Ors v The Ministry of Justice & Ors [2009] EWHC B41 (QB), at [22]:

In terms of a search one should always start with the most important people at the top of the pyramid, that is, adopt a staged or incremental approach. Very often an opposing party will get everything they want from that without having to go down the pyramid any further, often into duplicate material.

15     It must be stressed that in conducting discovery in stages, it is open to parties to proceed with a subsequent stage involving other custodians and repositories after the conclusion of the initial stage. Subsequent stages will be conducted in the same manner as the initial stage: identification of additional custodians and repositories, drawing up search terms and agreeing on the limits for reasonable searches, etc. In the limited experience of this court, a subsequent stage has not been necessary in any case thus far. However, that is very different from saying that subsequent stages will never be necessary. In any event, parties may always make an application for specific discovery under O 24, r 5 based on documents disclosed in the initial stage of discovery.

The order for the initial stage of discovery

16     An order for discovery in stages has to be tailored to the facts of each case, the issues in dispute and the custodians involved. Crucially, the extent of the order must be proportionate to the amounts at stake and the significance of the issues in dispute. Ultimately, the order is calculated to enable the cost-effective management of the discovery stage of the proceedings.

17     In the present case, the following custodians were identified:

(a)
Susan Abraham, the second defendant who was Vasanmal Murli’s client adviser;

(b)
Anandraj Jain, the client adviser who took over from Susan Abraham;

(c)
Vikrant Kanyal, the third defendant who was Head India Desk 2 and supervisor of Susan Abraham and Anandraj Jain;

(d)
Vikram Malhotra, Head India Desk 1 and who took over Vikrant Kanyal’s supervisory responsibilities when the latter left UBS’s employ;

(e)
Andreas Reber, the Regional Market Manager of the South Asia Division which included the India desks; and

(f)
Kurt Kumschick, a senior and long-time employee of UBS to whom Andreas Reber reported and who was the one that acquired the Breezeway account for UBS.

18     For each custodian, the following repositories were identified: (a) e-mail accounts (or mailboxes) and instant messaging platforms, and (b) documents in their personal network profiles stored online in UBS’s servers. This being general discovery, I did not think that any of the issues between parties warranted the restoration and searching of backup tapes. Hence, I ordered that the reasonable search be conducted on the current personal network profiles of employees still under the employ of UBS; and for ex-employees, UBS need only restore the personal network profile as at the last date of his employment or the last date of the relevant period, whichever was earlier.

19     Based on the issues that were raised during submissions, I also ordered that the reasonable search be further limited to the following periods for the respective custodians:

(a)      February and August 2008. This was the period when Breezeway alleged that misrepresentations were made resulting in its decision to purchase the leveraged bonds. A reasonable search limited to this period was ordered to be conducted on the repositories of Susan Abraham, Anandraj Jain, Kurt Kumschick and Vikrant Kanyal.

(b)      February and June 2009. This was the period when Breezeway alleged that inaccurate information was provided leading to its decision to liquidate the leveraged bonds, which in turn led to UBS reassessing the value of Breezeway’s collateral and making a margin call. A reasonable search limited to this period was ordered to be conducted on the repositories of all identified custodians.

20     As parties had not discussed the search terms to be used in the reasonable search, an adjournment was granted for parties to discuss and agree on a set of search terms to be used, with liberty to apply for a determination in the event of disputes over the search terms to be used. I further directed that parties perform a preliminary search using the disputed keywords before applying for a determination on the disputed search terms.

21     Apart from the order for discovery to be conducted in stages and the use of reasonable searches set out above, I made several other orders for specific discovery and interrogatories. As these are not relevant to the subject matter of the Registrar’s Appeal, I need not set out the reasons for my decision.

Disputes over search terms

22     Parties were, unfortunately, unable to agree on the list of search terms to be used for the reasonable search. Good faith attempts led to an agreement of 7 search terms, but parties were unable to agree on the remaining 23. Having conducted a preliminary search, parties came back before me for determination of the search terms to be used. After hearing submissions, I allowed a further 10 search terms. I set out hereunder the approach that I had adopted in selecting the search terms to be used and my reasons for doing so.

Approach in selecting search terms

23     Search terms function as a proxy for ocular review of electronic documents by identifying documents based on keywords and search operators. The use of keywords and search operators has become almost second nature when we look for material on the Internet. With desktop search engines becoming standard features on modern desktop operating systems, this is fast becoming the natural way for many individuals to look for documents and e-mails stored on their desktop or notebook computers. Modern search engine technology is relied upon to manage the increasingly overwhelming volume of information that is amassed on personal computing devices and to ameliorate the effects brought about by lackadaisical habits in managing and categorising electronically stored documents and e-mails.

24     As has been observed earlier, electronic documents, if assiduously categorised and stored, can be retrieved efficiently. The decentralisation of records without the concomitant discipline of records management leads to the common situation where electronic documents are not properly categorised in the storage medium, thereby making it difficult for orderly retrieval. The problem is compounded where, like the present case, the employees are no longer with UBS and UBS has to discharge its discovery obligations.

25     As a proxy for ocular review, selection of keywords and search operators has to be made with reference to the pleadings and the issues in dispute; in other words, they must be relevant: Robin Duane Littau v Astrata (Asia Pacific) Pte Ltd. The selection must be made having in mind the ultimate goal: the identification of relevant documents. In selecting keywords and formulating search terms, we ought to guard against the tendency to cast as wide a net as we can in order to obtain the highest number of documents in the search results. This has emerged as a pattern of behaviour in certain types of cases, where the party seeking discovery tries to use searches to draw out as many documents as possible from his adversary, with the hope that he may find the proverbial smoking gun. That would, to my mind, be tantamount to trawling or the emptying out of the adversary’s filing cabinets while hiding behind the mask of search terms.

26     As an aside, if a party makes use of search terms to identify documents which he then puts through a process of review in order to identify discoverable document as a means of discharging his discovery obligations, the situation is a little different. As he is not imposing this on his adversary, he can choose to manage the discovery of his own documents in any manner he deems necessary. However, there may potentially be the question whether, if he had adopted a patently inefficient and resource-intensive method of managing discovery, the entire costs of the discovery effort ought to be recoverable as part of party-and-party costs. This is an issue which may need to be addressed at some point in the future when the appropriate case comes up for consideration.

27     The proper approach in crafting search terms is to have firmly in mind the effects of decentralisation of records without the concomitant discipline of proper records management. Search terms are, on this view, a means of re-establishing a system of categorisation of electronic documents and e-mails. Of course, this is not the only purpose. Search terms can legitimately be used to identify documents based on new parameters that would not have been possible with traditional filing or paper record-keeping practices. It would be wrong to eschew these new options offered by search technology to the discovery process. These may be adopted if they facilitate the ultimate goal of efficient management of the discovery process in a manner proportionate to the amounts at stake and the significance of the issues in dispute.

Decision relating to the disputed search terms

28     This is the approach which I think ought to be adopted in determining the relevance of keywords and the formulation of search terms. First, commence with the specific before expansion to broader search terms. Specific search terms would include the following:

(a)      Unique reference numbers. For example, bank account numbers or client account numbers where the context is in a banking relationship. In the context of other commercial transactions, if one party has in place a file reference number or account identification number, these may be used as well. This very closely approximates the traditional paper filing system.

(b)      Names of specific projects. This can be an important keyword particularly where the dispute arises from a developmental project or commercial transaction which has been assigned a project name, eg Robin Duane Littau v Astrata (Asia Pacific) Pte Ltd.

(c)      Keywords which identify the key witnesses (or custodians). For example, e-mail addresses, contact numbers and names or initials. Search terms may be formulated based on such keywords. For example, e-mail addresses of two key witnesses appearing in the same e-mail in order to identify e-mail conversations between them, eg Robin Duane Littau v Astrata (Asia Pacific) Pte Ltd.

(d)      Significant events and locations. Depending on the facts, there may have been a significant meeting which took place. It has proven useful in some cases to make use of the meeting location or a short-hand reference to a key meeting as a search term. This may identify correspondence and documents that have been generated surrounding the event. However, care must be taken in selection of locations. In the present case, there were some meetings which took place at the offices of UBS. The search term “Suntec meeting” did not turn up any hits in the preliminary search. Using “Suntec” alone would have turned up too many, as the standard e-mail footer from employees of UBS contains their office address, which is located in Suntec. However, locations have been useful as a keyword for some other cases, eg meeting at the lobby of Furama hotel.

29      Next, we can consider search terms incorporating keywords which are unique to the facts of the case or the context of the dispute. The capabilities of the search engine – viz the search operators available – to be used are particularly important for this stage.

(a)      Product names. Where the dispute centres around the purchase of certain product, the product may be used. However, one should bear in mind the facts of the case. Product names alone would, for example, be unsuitable if the product is one commonly sold by one of the parties.

(b)      Unique phrases. For some industries, there may be unique terms or phrases that can be used to identify significant correspondence.

30     As has been observed elsewhere, we should avoid words which are commonly used, either in daily usage or in the context of the industry: see Robin Duane Littau v Astrata (Asia Pacific) Pte Ltd. Additionally, words which are legal concepts that may be part of the lawyers’ vocabulary and thought process may not always be helpful as keywords, eg breach of contract, confidential, etc. If such keywords are to be used, the search term will have to be carefully crafted.

31     The choice of keywords and formulation of search terms depends on the facts of each case and the issues in dispute. The guidelines and approach set out above provides a means to navigate these waters, but cannot be in any way considered definitive or exhaustive. Further, familiarity with the capabilities of the search engine to be used will in almost all cases be particularly helpful. This will assist the court in formulating search terms which could include keywords which are common words, as in this case (see below).

32     Applying this approach to the keywords which were placed before me, I had permitted an additional 10 search terms. The reasons for permitting them are set out below. First, it ought to be noted that parties had agreed to the use of 7 keywords which fall within the categories of unique reference numbers (ie bank account numbers) and keywords which identify key witnesses.

33     I had permitted keywords based on financial products, ie leveraged bonds, which Breezeway had purchased. These may be further categorised into two baskets.

(a)     Products which were fairly unique to the banker-customer relationship between UBS and Breezeway. These were leveraged bonds which had been recommended to Breezeway by the client advisers:

(i)
“Kuznetski” and “Bank of Moscow” – these refer to 2 leveraged bonds which are in dispute, the “7.335% Kuznetski Loan Participation Lot 1 Notes 2006 (Bank of Moscow Reg S)” and “7.335% Kuznetski Loan Participation Lot 2 Notes 2006 (Bank of Moscow Reg S)”;

(ii)
“Kaupthing” – refers to one of the leveraged bonds in dispute; and

(iii)
“Republic of Venezuela” – similarly refers to one of the leveraged bonds in dispute.

(a)     These keywords were permitted to be run on the repositories and for the periods previously ordered.

(b)     There were 2 other leveraged bonds – “ICICI” and “VTB” – which were heavily traded by the India desk and which had been recommended to Breezeway. For these, I allowed them but with restrictions. Based on the number of preliminary hits, these were restricted to the e-mail mailboxes of Susan Abraham and Vikrant Kanyal (including the e-mail archive), but not the documents stored on their personal network profile stored online in UBS’s servers.

34     One of the issues in dispute is whether the loans which Breezeway had taken were secured against collateral, and if so, whether there was any identification of which banking assets had been marked as collateral against the loans. There are also disputes of fact in relation to requests by UBS for additional collateral or margin call. Hence, I had also permitted the keyword “collateral” but restricted to their e-mail mailboxes (including the e-mail archive), having in mind the number of hits in the preliminary search.

35     Another of the issues in dispute is whether the terms and interest rates of the loans taken by Breezeway was “fixed to maturity”. This is a major issue in dispute as Breezeway’s case is that they were advised to take these loans in order to purchase the bonds on a leveraged basis. In so doing, Vasanmal Murli had insisted that the terms and interest rates of the loans were fixed to maturity. Hence, I allowed the search term using a proximity search of “fixed” within 10 words of “maturity”. For this same reason, I also allowed the search term “fixed loan”.

36     I further allowed the keyword “protest”. Although “protest” is a common word, its use in the context of the correspondence between Breezeway and UBS is particularly relevant to one key issue: UBS’s alleged unilateral decision to reduce the “loanable value” of the leveraged bonds which was in turn used by UBS as a basis for the margin call. Breezeway had protested to UBS against this reduction and margin call. In this context, the keyword “protest” was relevant; and the preliminary search did not reveal a surfeit of hits. Hence, the keyword was allowed.

Review for privilege and provision of copies in lieu of inspection

37     Although the search results are deemed to be relevant and no further review for relevance is necessary, time was allocated for UBS to conduct a review for documents and e-mails in the search results which may potentially be privileged or for which there may be good reasons to prevent disclosure, eg documents subject to banking secrecy and which clearly relate only to accounts of the bank’s other customers. In this privilege review process, search terms can also be used effectively in order to identify privileged or confidential information. For example, using the e-mail address of external counsel to identify privileged documents; or account numbers of other customers of the bank handled by the client advisers in order to identify documents which relate to these other customers. Once identified, these documents may either be excluded from the search result or a redacted copy disclosed instead.

38     The list of documents serves the role of identifying documents which are disclosed and (for the party entitled to inspection) a means of identifying documents for inspection (and the supply of copies). Any subsequent dispute over disclosure may be settled by reference to the list of documents. When soft copies of electronically stored documents are provided in their native format on finalised optical discs, the list of documents is effectively supplanted by the optical disc. The optical disc functions both as the carrier of copies and also a (directory) listing of documents so supplied. Furthermore, since the optical disc is finalised and no new files may be added (and similarly, no existing file removed), the risk of tampering is significantly reduced.

39     Hence, I ordered that after UBS has completed its privilege review, discovery was to take place by the provision of softcopies of documents and e-mails in their native formats on a finalised optical disc. E-mails from the Outlook e-mail system can be provided in native PST files. I further directed that parties be dispensed from enumerating each electronic document in the list of documents as individual items. Instead, parties were to categorise and organise the documents and e-mails into folders and sub-folders in the optical disc. The optical disc is to be identified as a single item in the list of documents and its contents given brief descriptions, preferably in accordance to the categorisation in its folders. This conflates the traditionally separate steps of disclosure and supply of copies, and should result in significant savings in time and costs that would otherwise have to be expended for the preparation of the list of documents.

40     Discovery of electronically stored documents would therefore take place by the provision of the optical disc together with the abbreviated list of documents. For paper documents, the traditional practice of individual enumeration was to be observed.

41     Finally, electronic copies are near perfect duplicates of the original electronic document, particularly where the copy is in native format with metadata intact. The need for inspection of the digital original would usually be unnecessary. Hence, it is appropriate that a different regime be adopted: one based on a request for inspection as opposed to the traditional regime in general discovery where the notice in the list of document makes an offer for inspection (see O 24, r 9). For soft copy documents, inspection was therefore ordered to be deferred and O 24, r 10 was ordered to apply to the list of documents as it applies to pleadings and affidavits. Further, the time for objections under O 27 was ordered to run from the date of exchange of the optical disc or the date of actual inspection, whichever is the later.

Copyright © Government of Singapore.

Back to Top

This judgment text has undergone conversion so that it is mobile and web-friendly. This may have created formatting or alignment issues. Please refer to the PDF copy for a print-friendly version.

Version No 0: 28 Feb 2012 (00:00 hrs)