[2010] SGHC 200

Festina Lotus SA v Romanson Co Ltd
[2010] SGHC 200

Case Number	:	Originating Summons No 240/2010/H
Decision Date	:	16 July 2010
Tribunal/Court	:	High Court
Coram	:	Tay Yong Kwang J
Counsel Name(s)	:	G. Radakrishnan (Infinitus Law Corporation) for the Appellant; Jonathan Yuen and Joana Teo (Harry Elias Partnership LLP) for the Respondent.
Parties	:	Festina Lotus SA — Romanson Co Ltd

16 July 2010

Judgment reserved

Tay Yong Kwang J:

Background Facts

1 The Appellant, Festina Lotus S.A, is the proprietor of trade marks which are variations of “FESTINA and device” (“Appellant’s various marks”) in Singapore in Classes 3, 9, 14, 16, 18 and 25 ICGS. The Respondent, Romanson Co. Ltd, filed a trade mark application for “J.ESTINA and device” (“Respondent’s mark”) by way of trade mark application number T05/25465Z with respect to Class 14 in relation to “wrist watches, electric watches, desk clocks, alarm clocks, necklaces, rings, bracelets, earrings, medals and brooches” (“Respondent’s application”). The Appellant filed a notice of opposition on 22 September 2006 to oppose the Respondent’s application.

2 The Principal Assistant Registrar of Trade Marks (“PAR”) dismissed the Appellant’s opposition. The Appellant appealed against her decision in Festina Lotus SA v Romanson Co Ltd [2010] SGIPOS 2 (“GD”) in respect of her findings under ss 7(6) and s 8(2)(b) Trade Marks Act (Cap 332, 2005 Rev Ed) (“TMA”).

Facts pertaining to the Appellant

3 The Appellant’s trade mark in respect of Class 14 was registered in Singapore under trade mark number T9004265I (“Appellant’s mark 1”). The Appellant’s mark 1 is a composite mark made up of a crest with the following elements:

(a)

A crown device positioned on top of the crest;

(b)

Two prancing horses positioned on the sides of the crest;

(c)

A curved banner positioned at the bottom of the crest with the word “FESTINA” emblazoned across the curved banner.

4 The Appellant’s marks in respect of Classes 9, 18 and 25 (“Appellant’s other marks”) consist of the same device as described above with the addition of the word “FESTINA” in larger, bold font as a plain word mark positioned at the bottom of the device. The Appellant’s various marks and their corresponding Classes are as follow:

Trade Mark Number	Class	Registration Date	Mark
T9004265I	14 “Watches, clocks and other chronometric instruments”	12 June 1990
T0612082G	09 “Spectacles”	4 July 2005
T0612083E	18 “Leather and imitation leather; goods made of these materials not included in other classes; animal skins and hides; trunks and suitcases; umbrellas; parasols and walking sticks; whips, harness and saddlery”	4 July 2005
T0612084C	25 “Clothing and footwear”	4 July 2005

5 The Appellant, by way of Statutory Declarations made by its Managing Director, Mr Javier Ribas, stated that it was founded in 1902 in Switzerland. The brand “FESTINA” was said to be one of the leaders in the mid-range watch sector in markets in Europe and currently distributes and sells watches and time pieces in more than 50 countries worldwide. It also sponsors professional cycling events and is the official time keeper of the Tour de France, amongst others. The Appellant also tendered evidence of its annual turnover and copies of selected invoices attesting to sales where it has a presence in East Asia, ie, in Hong Kong and Japan. There are no specific sales figures available for Singapore as the Appellant’s various marks have not been used here.

Facts pertaining to the Respondent

6 The Respondent’s mark, on the other hand, consists of the word “J.ESTINA” with a crown device positioned on the top of it as seen below:

7 In a Statutory Declaration made by the General Manager of “J.ESTINA” division of the Respondent, Mr. Myoung-chul Shin, it was deposed that the Respondent is a South Korean manufacturer of watches established in April 1998. The watches manufactured include luxury jewellery watches, prestige-grade watches, business watches, fashionable watches and casual watches.

8 The Respondent ventured into the jewellery market in 2002. On 16 December 2005, it filed for registration of the Respondent’s mark by way of application number T05/25465Z for Class 14 goods with respect to “wrist watches, electric watches, desk clocks, alarm clocks, necklaces, ring bracelets, earrings, medals and brooches” in Singapore. As the Respondent’s mark has not been used here, the Respondent tendered evidence of approximate sales value of jewellery, watches and horological articles under the Respondent’s mark in countries such as China and Vietnam.

The decision of the PAR

PAR’s findings with respect to s 8(2)(b) TMA

9 Section 8(2)(b) TMA reads as follow:

Relative grounds for refusal of registration

8. — (1)

(2) A trade mark shall not be registered if because —

…

(b) it is similar to an earlier trade mark and is to be registered for goods or services identical with or similar to those for which the earlier trade mark is protected,

there exists a likelihood of confusion on the part of the public.

[emphasis added]

10 The PAR found that:

(a)

The Appellant’s mark 1 was the relevant “earlier trade mark” for the purposes of s 8(2)(b) TMA. The Appellant’s other marks were therefore not considered, at [53] of GD;

(b)

The Appellant’s mark 1 and the Respondent’s mark were visually dissimilar, at [59] of GD;

(c)

The Appellant’s mark 1 and the Respondent’s mark were aurally similar, at [61] of GD;

(d)

The Appellant’s mark 1 and the Respondent’s mark were conceptually dissimilar, at [64] of GD;

(e)

The goods in the Respondent’s application in respect of “wrist watches, electric watches, desk clocks, alarm clocks” were identical to the Appellant’s registered goods in Class 14, ie, “watches, clocks and other chronometric instruments”, at [71] of GD;

(f)

The goods in the Respondent’s application in respect of “necklaces, rings, bracelets, earrings, medals and brooches” were similar to those of the Appellant’s registered goods in Class 14, ie, “watches, clocks and other chronometric instruments”, at [73] of GD;

(g)

There was no likelihood of confusion between the Appellant’s mark 1 and the Respondent’s mark, at [84] of GD.

Accordingly, the appellant failed on this ground of opposition.

PAR’s findings with respect to s 7(6) TMA

11 Section 7(6) TMA stipulates:

Absolute grounds for refusal of registration

7. — (1) ...

(6) A trade mark shall not be registered if or to the extent that the application is made in bad faith.

The PAR found that the evidence tendered before her did not show that the Respondent’s actions fell short of the standards of acceptable commercial behaviour observed by reasonable and experienced men in the relevant area of trade, at [113] of GD. Thus, the Appellant’s opposition under bad faith failed.

The Appellant’s submissions on appeal

12 The Appellant submitted that there are visual, aural and conceptual similarities between the Appellant’s mark 1 and the Respondent’s mark.

13 The Appellant’s approach during the course of submissions was a strategic one. It first sought to demonstrate that the goods in Classes 9, 14, 18 and 25, ie, those which the Appellant’s various trade marks were registered under, could be broadly and rationally categorised as “fashion accessories”. Flowing from this logic, the Appellant argued that the Respondent’s goods in Class 14 were at least similar to the Appellant’s goods in respect of Classes 9, 14, 18 and 25 as they were all “fashion accessories”. The implication of such, as submitted by the Appellant, was that the relevant “earlier mark” for the purposes of s 8(2)(b) TMA included not only the Appellant’s mark 1 (as held by the PAR) but also the Appellant’s other marks. This was crucial to strengthening the Appellant’s case as the Appellant’s other marks (as opposed to the Appellant’s mark 1) bore greater visual similarity to the Respondent’s mark, a point conceded by the Respondent.

14 Applying the global appreciation test, the Appellant took the position that a likelihood of confusion existed on the part of the public between the Respondent’s mark and the Appellant’s mark 1 (and the Appellant’s other marks if the Court accepts the Appellant’s submission above) within the meaning of s 8(2)(b) TMA.

15 For the claim of bad faith under s 7(6) TMA, the Appellant submitted that the PAR erred in finding that there was no bad faith on the Respondent’s part. The Appellant submitted that Respondent was very likely to have been aware of the Appellant’s mark and its reputation. This, compounded by the Respondent’s failure to respond by way of counterstatement to the allegation of bad faith and the unsatisfactory explanation given by the Respondent regarding the derivation of the brand name “J.ESTINA” suggested that the Respondent must have copied the Appellant’s mark. Thus, the application to register the Respondent’s mark in such circumstances must have been made in bad faith.

The Respondent’s submissions on appeal

16 The Respondent took the view that the Appellant’s appeal against the PAR’s findings in respect of s 8(2)(b) TMA was misconceived. Faced with the Appellant’s attempt to categorise the goods in Classes 9, 14, 18 and 25 as being similar in nature which would entitle the court to consider the Appellant’s other marks, the Respondent submitted that there was no overlap between the specification of goods for the Appellant’s marks in Classes 9, 18 and 25 with that of the Respondent’s Class 14 goods. Hence, the Respondent maintained that the relevant “earlier mark” that the court was entitled to consider was confined to the Appellant’s mark 1 for the purposes of s 8(2)(b) TMA.

17 The Respondent also submitted that there were no visual, aural or conceptual similarities between the Appellant’s and Respondent’s marks and that there was no likelihood of confusion between the two marks.

18 Given its stance that the two marks were completely different, the Respondent submitted that there was no issue of bad faith that needed to be addressed. As the Appellant had provided no distinct proof of unacceptable commercial behaviour, its claim against the Respondent in respect of bad faith under s 7(6) TMA must necessarily fail.

Decision of the court

Standard of Review of Present Proceedings

19 Order 87 rule 4 Rules of Court (Cap 322, R 5, 2006, Rev Ed) (“ROC”) applies to appeals against the decision of the Trade Marks PAR pursuant to s 75(2)(a) TMA. O 87 r 4 ROC reads as follows:

Order 87

Trade Marks Act

…

Appeals (O.87, r.4)

4. — (1) …

(2) An appeal shall be by way of rehearing and the evidence used on appeal shall be the same as that used before the Registrar and, except with the leave of the Court, no further evidence shall be given.

[emphasis added]

20 Parties were not in dispute on the standard of review of the present appeal and were in agreement that “rehearing” in relation to an appellate review of a Trade Marks PAR’s decision under O 87 ROC takes reference from “rehearing” in an appeal from the High Court to the Court of Appeal.

21 The standard of review of the High Court in its appellate capacity against the decision of a Trade Marks PAR can be surmised from the following passage quoted from Laddie J’s judgment in SC Prodal 94 SRL v Spirits International NV [2003] EWHC 2756 (Ch) which was endorsed in Future Enterprises Pte Ltd McDonald’s Corp [2007] 2 SLR(R) 845 at [5]:

It is not the duty of this court to overturn a decision of the Trade Mark Registry simply because it comes to the conclusion that it might have decided the case differently had it, that is to say the High Court, been the court of first instance. It has to be demonstrated that the decision at first instance was wrong in a material way; that is to say there must be some significant departure from a proper assessment of the law or the facts.

[emphasis added]

Thus, given the highly subjective nature of assessing similarity and the likelihood of confusion under s 8(2)(b) TMA, the appellate court should not disturb the findings of fact of PAR unless there is a material error of principle.

Appeal against the PAR’s decision under s 8(2)(b) TMA

22 The text of s 8(2)(b) TMA has been set out at [9] above.

23 As enumerated by Woo Bih Li J (“Woo J”) in Ozone Community Corp v Advance Magazine Publishers Inc [2010] SGHC 16 (“Ozone Community Corp”) at [36], there are three conditions under s 8(2)(b) TMA which the Appellant has to satisfy in order to succeed in its opposition to the Respondent’s application:

(a)

The Appellant’s mark 1 “FESTINA and device” and the Respondent’s mark “J.ESTINA and device” are similar (“Condition 1”);

(b)

If Condition 1 is satisfied, then that the Respondent’s mark “J.ESTINA and device” is to be registered for goods identical with or similar to those for which the Appellant’s mark 1 “FESTINA and device” is protected (“Condition 2”);

(c)

If Conditions 1 and 2 are both satisfied, then there exists a likelihood of confusion on the part of the public by virtue of the similarity of the marks and of the goods (“Condition 3”).

24 Notably, these conditions are cumulative in the sense that all three conditions must be satisfied before s 8(2)(b) TMA can be successfully relied on. The need for a conjunctive reading of these three conditions was also emphasised by the Court of Appeal in The Polo/Lauren Co, LP v Shop In Department Store Pte Ltd [2006] 2 SLR(R) 690 (“The Polo/Lauren Co”) at [8] albeit in the context of the trade marks infringement provision, ie, s 27(2)(b) TMA, the wording of which mirrors that of s 8(2)(b) TMA:

While we can see some link and overlap between the first and the third conditions, it is clear that they are different. The fact that a sign is similar to a registered mark does not automatically mean that there will be a likelihood of confusion on the part of the public. That is a question of fact to be determined by the court, taking into account all the relevant circumstances. Moreover, the question of similarity is really a matter of degree. In a broad sort of sense, the greater the similarity between a mark and a sign, the greater will be the likelihood of confusion. However, if either of the first two conditions is not satisfied there will not be any need to go into the third question of determining whether there exists a likelihood of confusion. Neither does it mean that if the mark and the sign are similar, and they are used on similar goods, that there will ipso facto be confusion in the minds of the public. If that was intended, s 27(2)(b) would have been phrased differently. As presently worded, there is no presumption of confusion once the two aspects of similarity are present. …

[emphasis added]

Condition 1: Similarity of marks

25 The Respondent’s mark and the Appellant’s mark 1 are both composite marks which consist of a pictorial element and a word element. The two marks are as follow:

Respondent’s mark	Appellant’s mark 1

The decided cases suggest that there are three aspects to similarity: visual, aural and conceptual. Whether there is a similarity of marks in these three aspects is essentially a question of fact, see Johnson & Johnson v Uni-Charm Kabushiki Kaisha (Uni-Charm Corp) [2007] 1 SLR(R) 1082 at [9].

26 As noted in [10] above, the PAR found that these two marks were aurally similar but were not visually and conceptually similar.

(1) Visual Similarity

27 The Appellant submitted that the PAR had erred in concluding that the two marks were visually dissimilar. It was accepted that the Appellant’s mark 1 was a composite mark consisting of a crest device with the word “FESTINA” emblazoned across the curved banner below the pictorial element.

28 The PAR at [57] and [58] of the GD made a comparison between the two marks and found:

(a)

In relation to the Respondent’s mark, “J.ESTINA” was a dominant component as the word “J.ESTINA” constituted half or more of the Respondent’s mark. As for the crown device, it was much smaller in size in relation to the word “J.ESTINA”.

(b)

As for the Appellant’s mark 1, the crest formed the dominant component while the word “FESTINA” was subsumed within the curved banner of the crest. It was opined that crests were normally intricately designed with complicated elements such that, generally, consumers would not be able to recall their exact details save for the impression that the crest was part of the mark.

29 It was the Appellant’s submission that crests and coats-of-arms are nowadays commonly used on a wide variety of goods ranging from watches, foodstuffs and beverages. It was said that the consumers’ eyes are usually drawn to the words in such a mark in order to ascertain the origin of the goods. In other words, even where a crest forms a dominant part of the trade mark, it will have a “negligible effect” when the average consumer recalls the trade mark. Thus, the Appellant’s mark 1 would be remembered and accordingly be described as “FESTINA” (ie, word element) since the crest element (ie, pictorial element) cannot be verbalized. This, combined with the imperfect recollection of an average consumer, in terms of visual recognition, is likely to militate against distinguishing between the two marks “FESTINA” and “J.ESTINA”.

30 The European Court of Justice (“ECJ”) has emphasised that the assessment of similarity between two marks must be based on the overall impression created by them, in particular, of their distinctive and dominant components, see Sabel BV v Puma AG, Rudolf Dassler Sport, Case C-251/95 [1997] ECR I- 6191 and Lloyd Schuhfabrik Meyer v Klijsen Handel BV, Case C-342/97 [1999] 2 CMLR 1343. Thus, in deciding whether there are visual similarities between the Appellant’s mark 1 and the Respondent’s mark, I agree with the PAR that with respect to the Respondent’s mark, the word “J.ESTINA” formed the dominant part of the mark whereas for the Appellant’s mark 1, the crest was the dominant aspect f. To this extent, there is no visual similarity between the two marks.

31 The Appellant submitted that there are instances whereby the overall impression conveyed to the relevant public by a composite mark may be dominated by one or more of its components, see Matratzen Concord GmbH v OHIM, Case T- 6/01 [2002] ECR II-4335 at [32]. Medion AG v Thomson Multimedia Sales Germany & Austria GmbH, Case C- 120/04 [2005] ECR I-8551 (“Medion”) was also cited by the Appellant as authority that it is still possible that a non-dominant component of a composite mark might have an independent distinctive role in the composite mark, see [30] of Medion. Thus, the Appellant’s position was that even if “FESTINA” was not the dominant element of the Appellant’s mark 1, its distinctive role would support its contention that the two marks were in fact visually similar.

32 The Appellant’s reliance on Medion should be read in context in order to fully appreciate the reasoning behind the decision of the ECJ. The ECJ envisaged that notwithstanding that the overall impression of the average consumer may be dominated by one or more components of a composite mark, there might be instances whereby a non-dominant element of a composite mark would nevertheless be so distinctive (viz, by virtue of its widely-known commercial name) such that the public may be led to believe that the goods or services at issue were derived from economically linked companies, in which case the “likelihood of confusion” must be held to be established, see [31] and [32] of Medion.

33 As seen above, the driving force behind the ECJ’s reasoning that there existed a “likelihood of confusion” between the two marks was because the non-dominant but distinctive element of the composite mark was one that was in itself already widely-known, see [34] of Medion. Thus, it is pertinent to note that Medion does not purport to confer upon otherwise dissimilar marks a greater degree of visual similarity by virtue of some distinctiveness in the earlier mark.

34 Thus, even if one considers the word “FESTINA” to be non-dominant but distinctive, it is not sufficient in and by itself to establish that the Appellant’s mark 1 and the Respondent’s mark are thereby visually similar. I therefore accept that there is no visual similarity between the Appellant’s mark 1 and the Respondent’s mark.

(2) Conceptual Similarity

35 The PAR at [63] of the GD noted that the addition of crest-like designs (eg, that of the Appellant’s mark 1) had over time been considered to be a design element which lends a “high-class” look to a mark. However, she opined that the same could not be said of a crown device (eg, that of Respondent’s mark) since it continues to this day to connote and evoke the idea of royal patronage. Thus, it was held that the two marks were conceptually dissimilar.

36 The Appellant submitted that the proper comparison between the Appellant’s mark 1 and the Respondent’s mark in the assessment of conceptual similarity should be that the two marks were of a combination of a “foreign sort of name” and a device which hints at royalty. On this premise, the Appellant submitted that there was conceptual similarity between the two marks. Alternatively, the Appellant submitted that the crest and crown devices were conceptually neutral.

37 The Respondent, on the other hand, submitted that the Appellant’s mark 1 and the Respondent’s mark were conceptually dissimilar as the crest device and the crown device conveyed different ideas. For instance, a crown device would convey an impression of royalty whereas a crest device, eg, crest of Singapore, would not.

38 Bently and Sherman in Intellectual Property Law (Oxford University Press, 3^rd Ed, 2009) (“Bently & Sherman”) noted at p 866 that in relation to “conceptual similarity”:

Because similarity is assessed conceptually as well as visually and phonetically, in thinking about whether marks are similar, it is necessary to take account of the ideas that lie behind or inform the earlier mark. [emphasis added]

39 A clear case of conceptual dissimilarity can be seen in Phillips – Van Heusen, T– 292/01 [2003] ECR II – 4335 where “BASS” was found to be conceptually dissimilar to “PASH” as the former was understood as a musical reference whereas the latter either had no meaning or referred to German dice.

40 Itochu Corporation v Worldwide Brands, Inc [2007] SGIPOS 9 (“Itochu”) offers a more nuanced illustration of conceptual similarity. In Itochu, the PAR (and subsequently the High Court) held that the marks “CAMEL” and “SWEETCAMEL” were conceptually dissimilar: the “CAMEL” mark evoked a sense of “ruggedness, adventure and masculinity” whereas the applicant’s mark “SWEETCAMEL” which was preceded by the adjective “SWEET” projected a totally opposite impression.

41 In the case of Ozone Community Corp, the two marks in issue were “GLAMOUR” and “HYSTERIC GLAMOUR”. The PAR found that both marks were in some way conceptually similar as both marks used the “glamour” concept, despite one appealing to the orthodox meaning while the other giving the concept a spin without completely altering its texture. This finding was upheld by the High Court at [64].

42 In the case of Imperial Tobacco Co (of Great Britain and Ireland) Ltd’s Trade Marks [1915] 2 Ch 27 (“Imperial Tobacco”), albeit decided in the context of whether the use of Prince of Wales plumes (part of the Prince of Wales heraldic achievements) which formed part of the proprietor’s trade mark amounted to an act of “calculative deception”, Lord Cozens- Hardy MR noted that the Prince of Wales feathers had for many years to a very great extent been used by traders of every kind in such a way that the mere use of them did not by itself indicate to anybody with a rational mind that the person so using it was a warrant holder. Concurring with Lord Cozens-Hardy MR, Pickford LJ further observed that:

I have no doubt whatever that many kinds of articles are at the present moment described by the names of persons distinguished in the war, but no one imagines that that practice means anything more than giving those articles an attractive name, and I cannot imagine there is anything more than that in using the name of the Prince of Wales.

[emphasis added]

Notably, this passage buttresses the PAR’s finding at [63] of the GD that the use of a crest, in the modern context, is used to lend a “high-class” look to a mark design as an attraction to the average consumer as opposed to signifying royal patronage.

43 The PAR further opined that the crown device connotes and evokes the idea of royal patronage. It is probable that the crown device would evoke a sense of royalty, which is in line with the Respondent’s brand story of an Italian princess who later became a Bulgarian queen. However, a crown device in modern usage does not signify only royal patronage. Borrowing the example of Eve J in Imperial Tobacco, I do not think that an average consumer would be disappointed in finding out that the smoking mixture she purchased was not supplied for the personal use of the Prince of Wales. In the same vein, nor would the consumer be disappointed if she discovers that the watch she purchased was not used by an Italian princess and Bulgarian queen. The crown device, in the modern day, is more likely to evoke a sense of royalty or class and status rather than royal patronage.

44 The PAR at [63] of the GD appeared to have emphasized solely on the conceptual (dis)similarity between the crown and crest devices as the invented words were conceptually neutral. I accept the Appellant’s submission that as the Appellant’s mark 1 and the Respondent’s mark are composite marks, they should be viewed as a whole. Thus, the pictorial element as well as the word element should be taken into account while assessing the conceptual similarities between the marks. The Appellant’s alternative submission that the crown and crest concepts are neutral has to be rejected as both do evoke some qualities.

45 In my view, insofar as the two marks comprise an invented word, in that “FESTINA” and “J.ESTINA” are not English terms found in dictionaries, and that both employ a pictorial element that hints at “class” and “status”, there are some conceptual similarities between the two marks, as opposed to complete or significant similarities, bearing in mind that the crest and crown devices evoke a sense of class but only the crown device evokes a sense of royalty.

(3) Aural Similarity

46 The PAR at [61] of the GD found that there were aural similarities between the Appellant’s mark “FESTINA” and the Respondent’s mark “J.ESTINA” given that there is little difference between the sounds which the letters “F” and “J” make at the beginning of the words such that the pronunciation would largely focus on “ESTINA”. The PAR also found that while the Respondent’s mark was in the form of “J.ESTINA”, it was likely that the average consumer would pronounce it as “jes-tina”.

47 At the PAR hearing below and on appeal, the Respondent sought to differentiate between “FESTINA” (pronounced as “fes-tina”) and “J.ESTINA” (pronounced as “jay-estina”) because the former comprises three syllables as opposed to the latter’s four syllables. The Respondent also submitted that Sargant LJ’s observation in London Lubricants (1920) Limited’s Application to Register a Trade Mark [1925] 42 RPC 264 at 279 was instructive:

[That] the tendency of persons using the English language to slur the termination of words also has the effect necessarily that the beginning of words is accentuated in comparison, and… the first syllable of a word is, as a rule, far [sic] the most important for the purpose of distinction.

This observation, useful as it is, should not be taken as a hard and fast rule of pronunciation in the English language. The Court of Appeal in The Polo/Lauren Co at [23] rightly held that there is no rigid principle that the first syllable of a word mark is more important than other syllables.

48 In Mystery Drinks Gmbh v OHIM, T 99/01 [2004] ETMR (18) 217 (“Mystery Drinks”), the applicant argued that the two words “MIXERY” and “MYSTERY” were aurally different as the “i” in “MIXERY” was shorter and clearer than the “y” in “MYSTERY” and the “x” in “MIXERY” was pronounced differently from the “st” of “MYSTERY”. The European Court of First Instance (“CFI”) rejected this argument as such aural differences (as submitted by the applicant) could only be detected if one carried out a fragmented analysis, syllable by syllable, and completely omitted the comparison of the pronunciation of the words as a whole. In a similar fashion, the Respondent’s distinction of the number of syllables of the “FESTINA” and “J.ESTINA” by dissecting the respective words and then analysing them at a micro-level overlooks the bigger picture that words are pronounced as a whole.

49 In any event, it is more likely than not that the average consumer when pronouncing the words “FESTINA” and “J.ESTINA” (whether pronounced as “Jes-ti-na” or “Jay-es-ti-na”) would stress “-estina”. The beginning letters of the words, ie, “F” and “J”, are consonants which would probably be given less emphasis. Thus, I agree with the PAR’s finding that there are aural similarities between “FESTINA” and “J.ESTINA”.

(4) Overall similarity between the two marks

50 As seen above, the Appellant’s mark 1 and the Respondent’s mark are visually dissimilar but conceptually and aurally similar.

51 The law does not require all three similarities to be established before the finding that the two marks are similar can be made (see Mediacorp News Pte Ltd v Astro All Asia Networks plc [2009] 4 SLR(R) 496 (“Mediacorp”) at [32]). As noted in Bently & Sherman at pp 863-864:

Typically, tribunals consider [visual, aural and conceptual similarities] in turn, before reaching an overall conclusion. In so doing, they have observed that a trade-off can occur: visual and conceptual differences, for example, can offset aural similarities.

[emphasis added]

52 Hence, the relative importance of each type of similarity would vary with the circumstances of each case, in particular the type of goods and the type of mark. The case of Mystery Drinks serves as a useful illustration on this point. There, the CFI at [225] held that:

Since the goods in question are also consumed after being ordered orally, the aural similarity of the signs in question is in itself sufficient to give rise to likelihood of confusion. [emphasis added]

53 There is much wisdom in the “trade-off” approach as it accords a degree of flexibility and latitude to the courts in coming to a conclusion (with regard to the overall similarity of marks) against a contextual background by adapting to the ever evolving commercial reality. For instance, it recognises that in this modern age, one is able to purchase a timepiece over the Internet, as opposed to the more conventional trade channels such as departmental stores or specialised boutiques. It is not denied that an average consumer would rely on his own visual recognition of a trade mark to differentiate among the various brands for timepieces. However, this visual recognition is often accompanied by the “verbalisation” of the mark, ie the name of the mark. The significance of the verbalisation will, in turn, depend on the particular mode of purchase of the consumer.

54 Shaker de L. Laudato & C. Sas v Limiñana y Botella, SL, Case C-334/05P (“Shaker”) highlights the significance of “verbalization” of a mark. In Shaker, the applicant sought to register a composite mark for alcoholic beverages which included the word “LIMONCELLO” above an image of a plate decorated with six lemons. This was opposed by the owner of a Spanish word mark “LIMONCHELO” with respect to goods in Class 33. The CFI held that the signs were dissimilar because of the dominance of the dish feature in the applicant’s mark which was absent in the opponent’s mark. The authors of Bently & Sherman commented at p 864 that “this finding was itself surprising, given the general tendency to emphasize verbal components”. The ECJ overturned the CFI’s finding and reiterated that the assessment was a global one and should not be conducted by comparing only one element.

(A) Applying the significance of verbalizing a mark to the context of purchase of a watch

55 By way of an example, if a consumer chooses to purchase a “FESTINA” timepiece online, he could either buy it from a “FESTINA” e-store (assuming there is one) or from a website which hosts a variety of brands. In the former situation, the name or word element “FESTINA” has to be typed into the address bar or search engine for the consumer to be brought to the desired website. In the latter situation, the name or word element “FESTINA” will be found in a list of brand names on the relevant website. In both instances, one needs to recall the word element, “FESTINA”, as opposed to the crest element of the trade mark. Hence, the visual impact of the mark, ie, the crest, will be less significant as compared to the aural and visual impact of the word “FESTINA” despite the crest element being the dominant component of the mark.

56 In a situation where a consumer chooses to buy a “FESTINA” timepiece from a shop, the visual recognition of the crest device in the trade mark may be of greater significance. Nevertheless, there is still a need to verbalize the word “FESTINA”. An average consumer usually identifies a trade mark by virtue of its pictorial element (if any) and its name/ word element. Even where the pictorial device forms the dominant element of a trade mark, it would be extremely rare for a consumer to describe the product by its pictorial element because the description of, say, the crest, is a much more difficult task, given its intricacy, as compared to the relatively easy articulation of the word element/ brand name. Thus, the aural similarity between the Appellant’s mark 1 and the Respondent mark would militate against the visual dissimilarity in this instance.

(B) Distinctiveness in the earlier mark

57 The issue of distinctiveness of the earlier mark “FESTINA and device” was not discussed in the GD in the assessment of similarity of marks. Likewise, parties on appeal did not submit on the point of distinctiveness. Woo J has observed in Ozone Community Corp at [67] that the distinctiveness of the earlier mark, either inherent or acquired, is an important factor in the inquiry into the similarity between the marks under s 8(2)(b) TMA.

58 Registration per se is inconclusive as to the distinctiveness of the earlier mark (The Polo/Lauren Co at [72]). Chao Hick Tin JA in The Polo/Lauren Co at [23] also said:

We recognise that some marks are inherently distinctive because they consist of inventive words without any notional or allusive quality. An example would be "Volvo".

[emphasis added]

Given that “FESTINA” is an invented word, it possesses inherent distinctiveness and such distinctive quality would be taken into account in the overall assessment of similarity between the marks.

59 Bently & Sherman offers some useful observation on the interplay between the distinctiveness of an earlier mark and the similarity of marks at pp 866-867 (footnotes omitted):

Distinctiveness. The question of whether marks are similar will often be dependent on the inherent or acquired distinctiveness of the mark (for the goods or services for which it is registered). This has a number of effects.

First, the less distinctive the earlier trade mark, the less literal or visual alteration is necessary to ensure that the later mark is not similar. … In Reed Executive plc v Reed Business Information Ltd [[2004] RPC (40) 767], Jacob LJ stated that ‘where a mark is descriptive [as opposed to distinctive] small differences may suffice to avoid confusion’… Where the distinctiveness of the earlier trade mark is very low, the later mark will have to be in close proximity for it to be similar. …

Second, if the earlier mark is highly distinctive, then a mark that has been substantially modified might nonetheless be similar. …

[emphasis added]

60 As “FESTINA” is an invented word with inherent distinctiveness, a modification of at least a substantial degree has to be done for the Respondent’s mark to be found to be dissimilar to it. A comparison of the two composite marks shows that:

(a)

A crown device can be seen in both marks;

(b)

The Appellant’s crest device is not present in the Respondent’s mark;

(c)

Six out of seven letters in the Appellant’s word element of its composite mark, ie, “ESTINA”, are found in the Respondent’s mark;

(d)

The first letter “F” in the Appellant’s word element in its composite mark is replaced by a “J” and “.” in the Respondent’s mark.

The word element of the Respondent’s mark “J.ESTINA” does not seem to be very dissimilar from the Appellant’s mark “FESTINA”. It is likely that the imperfect recollection of a consumer who rarely has the opportunity to juxtapose the two marks and make a comparison may think the marks are similar, see Lloyd Schuhfabrik Meyer v Klijsen Handel BV, Case C-342/97 [1999] 2 CMLR 1343, 1358-1359.

61 Given that there is significant aural similarity between “FESTINA” and “J.ESTINA” and taking into account the inherent distinctiveness of the earlier mark “FESTINA”, I conclude that there is similarity between the two marks within the meaning of s 8(2)(b) TMA.

Condition 2: Similarity of goods

62 As noted in Bently & Sherman at p 869, “the question of whether goods or services are similar depends on the facts of the case”. The goods which the Respondent’s application relates to are “wrist watches, electric watches, desk clocks, alarm clocks, necklaces, rings, bracelets, earrings, medals and brooches”. On the other hand, the goods for which the Appellant’s mark 1 (ie, Class 14 goods) was registered include “watches, clocks and other chronometric instruments”.

63 The PAR found that there was identity between “wrist watches, electric watches, desk clocks, alarm clocks” and “watches, clocks and other chronometric instruments”, see [71] of GD. As for the remaining goods of the Respondent’s application, ie, “necklaces, rings, bracelets, earrings, medals and brooches”, the PAR found that there was some similarity between them and the Appellant’s Class 14 goods as watches and jewellery have similar use as being articles of adornment although watches have the additional function of being able to tell the time, see [73] of GD. I agree with these findings made by the PAR.

64 The PAR held at [53] of the GD that the goods in Classes 9, 18 and 25 (ie, goods for which the Appellant’s other marks were registered) were not identical or similar to the goods of the Respondent’s mark. On appeal, the Appellant argued that the PAR erred in this respect. For ease of reference, the goods in Classes 9, 18 and 25 are as below:

Classes of goods which the Appellant’s other marks were registered for	Specification of goods in the Respondent’s application
Class 9 Spectacles	Class 14 Wrist watches, electric watches, desk clocks, alarm clocks, necklaces, rings, bracelets, earrings, medals and brooches.
Class 18 Leather and imitation leather; goods made of these materials and not included in other classes; animal skins and hides; trunks and suitcases; umbrellas; parasols and walking sticks; whips, harness and saddlery
Class 25 Clothing and footwear

65 The Respondent submitted that the goods falling within Classes 9, 18 and 25 were evidently dissimilar from those of Class 14 as there was no overlap of goods. The Respondent objected to the Appellant’s attempt to categorise the goods in Classes 9, 14, 18 and 25 as “fashion accessories” as this would entail an expansion of the categories to include goods which otherwise would not be so classified. Such an unwarranted expansion was said to have the effect of defeating the whole purpose of having a system of classification. However, this submission may not be entirely correct. The fact that goods are classified into various classes under the ICGS does not preclude a finding that those goods are nevertheless similar in nature. This is because goods are so classified for administrative convenience whereas the question of whether goods are similar is an issue of substantive law, see El Corte Ingles SA v OHIM, Case T-443/05 [2007] ETMR (81) 1340 (“El Corte Ingles”) at [38].

66 The Appellant, on the other hand, submitted that the relevant goods were “fashion accessories” and cited British Sugar Plc v James Robertson & Sons Ltd [1996] RPC 281 (“British Sugar”) which was endorsed by Ozone Community Corp at [75] as the proper test to be used in determining whether the goods were similar. The factors that were to be taken into account were (British Sugar at 296-297):

(a)

The respective uses of the respective goods;

(b)

The respective end users of the respective goods;

(c)

The physical nature of the goods;

(d)

The respective trade channels through which the goods reach the market;

(e)

In the case of self-serve consumer items, whether in practice they are respectively found or likely to be found on the same or different shelves; and

(f)

Whether the respective goods are competitive or complementary. This inquiry may take into account how those in the trade classify goods, for instance whether market research companies who act for the industry put the goods or services in the same or different sectors.

67 The Appellant further submitted that QS by S. Oliver Trade Mark [1999] RPC 520 (“QS”) was a constructive indicator of “similarity of goods”. In QS, it was held that leather goods, leather bags, purses, etc would be regarded as “clothing accessories”. This was because these items were found to be complementary to clothing, had the same users and trade channels, and both would be found adjacent to each other in retail outlets.

68 Building on this line of reasoning, the Appellant submitted that goods falling within Classes 9, 18 and 25 were frequently designed and manufactured by a single entity as they were all considered as fashion accessories and were complementary to the consumer’s style or get up. The Appellant pointed to various stores in Singapore such as “Calvin Klein” and “DKNY” where a diverse range of clothing and accessories, such as clothing, bags, shoes, jewellery and watches, could be found in the same store as they were intended to be put together as a particular look or style.

69 Furthermore, there is now a trend among fashion designers to either license out their name or engage in brand extension. For instance, a brand that was previously well-known for handbags may now be selling sunglasses, clothes or even phones with its trade mark affixed on such goods. Thus, it was the Appellant’s submission that a consumer would expect that goods falling within one of the “fashion-related classes”, eg, Class 14 (watches) and Class 25 (clothing and footwear), to have originated from the same source or from economically-linked undertakings.

70 This idea of “aesthetic harmony” or “aesthetic complementarity” was discussed in Bently & Sherman where two conflicting lines of cases decided by differently constituted CFI were given some consideration. In the case of Sergio Rossi SpA v OHIM, Sissi Rossi, Case T-169/03 [2005] ECR II-685, the CFI held that women’s bags and women’s footwear were not “aesthetically complementary” as the applicant had failed to demonstrate that this aesthetic complementary nature had reached a stage of “true aesthetic necessity” in the sense that the consumers would think it unusual or shocking to carry a bag which does not perfectly match their shoes. Subsequently, a CFI of five judges held that bags and clothes were sufficiently complementary to be regarded as similar, see El-Corte Ingles at [38]. In the absence of an ECJ decision, the authors of Bently & Sherman at p 871 opined that the second view, which was more open to the idea that consumers might draw connections between goods because they are searching for “aesthetic harmony”, to be more consistent with the general idea of a fact-sensitive “global appreciation” by the Court.

71 As the Appellant rightly observed, there is a growing phenomenon of brands expanding into goods of various kinds in recent years. Sometimes, the crossing-over takes place between goods that one does not commonly associate with each other, for instance, cigarettes and clothing. The idea of licensing out one’s trade mark or engaging in a sister brand or diffusion line has inevitably led to the same trade mark being found on a multitude of goods available in the market. In this regard, there is force in the Appellant’s submissions that there is a sense of “relatedness” between goods such as necklaces and clothing and a broad classification of these items as “fashion accessories” or “lifestyle goods” may be justified.

72 Applying the British Sugar test to the present facts, it would be reasonable in the modern context to regard the goods in Classes 9, 14, 18 (trunks and suitcases and, in some instances, even umbrellas and parasols) and 25 as complementary in nature and are likely to be of similar uses, targeting almost identical end users and employing similar if not identical trade channels by which the goods reach the market. Thus, there is some similarity between “necklaces, rings, bracelets, earrings, medals and brooches” in the Respondent’s application and the goods for which the Appellant’s other marks are registered in Classes 9, 18 (the items specified above) and 25.

73 However, the Appellant’s inclusion of the other goods in Class 18, such as walking sticks, whips, harness and saddlery, as “fashion accessories” would appear to be too over-reaching. To say that these items are “fashion accessories” which are similar to “necklaces, rings, bracelets, earrings, medal and brooches” would be to put an unnatural strain on the average consumer’s perception of what fashion accessories are. Thus, these items are not similar to the Class 14 goods in the Respondent’s application.

74 The implication of the above finding that there is similarity between the goods in the Respondent’s application and the Appellant’s registered goods in Classes 9, 18 and 25 is that the Appellant’s other marks can be considered against the Respondent’s mark in determining whether there is a similarity of marks. For this purpose, the two competing marks are:

Respondent’s mark	Appellant’s other marks

Taking into account the Appellant’s other marks in the assessment lends even more force to the finding that there is similarity of marks as there is much greater visual similarity between the Appellant’s other marks (as opposed to the Appellant’s mark 1) and the Respondent’s mark.

75 To summarize the findings thus far:

(a)

There is identity between the goods in the Respondent’s application for “wrist watches, electric watches, desk clocks, alarm clocks” and the Appellant’s goods in Class 14 for “watches, clocks and other chronometric instruments”;

(b)

There is similarity between “necklaces, rings, bracelets, earrings, medals and brooches” in the Respondent’s application and the Appellant’s Class 14 goods;

(c)

There is similarity between “necklaces, rings, bracelets, earrings, medals and brooches” in the Respondent’s application and the Appellant’s goods in Classes 9, 18 (trunks and suitcases and, in some instances, even umbrellas and parasols) and 25.

Indeed, with today’s creative proliferation of goods morphing and merging in functionalities and forms, one begins to wonder whether the classification of goods for trade mark purposes is becoming dated and whether it reflects present-day commerce.

Condition 3: Likelihood of Confusion

76 Having found that there is similarity of marks and identity in some goods and some similarity in others, we now turn to the assessment of whether there is a likelihood of confusion on the part of the public. As stated earlier, there is no presumption of confusion even where similarity of marks and goods are made out, see Ozone Community Corp at [81] and the Court of Appeal’s decision in The Polo/Lauren Co at [8].

77 Lee Seiu Kin J at [43] of Mediacorp dealt with the issue of “likelihood of confusion”:

The following principles as to what constitutes "likelihood of confusion" can be distilled from the cases:

(a) The court is entitled to look outside the mark and the sign, as well as the articles, to assess whether there exists a likelihood of confusion (Polo/Lauren ([29] supra) at [8]).

(b) The question of likelihood of confusion has to be looked at globally taking into account all the circumstances including (Polo/Lauren at [28]):

(i) the closeness of the goods;

(ii) the impression given by the marks;

(iii) the possibility of imperfect recollection;

(iv) the risk that the public might believe that the goods come from the same source or economically-linked sources;

(v) the steps taken by the defendant to differentiate the goods from those of the registered proprietor; and

(vi) the kind of customer who would be likely to buy the goods of the applicant and opponent.

(c) One should not determine the likelihood of confusion based on the man in a hurry, and the test should be the "ordinary, sensible members of the public" (Polo/Lauren at [31]). The average consumer is someone who would exercise some care and good sense in making his purchases (Polo/Lauren at [34]). However, the "average consumer" need not necessarily mean the general public, as more specialised products might be purchased by a more specific cross-section of the public (Nation Fittings (M) Sdn Bhd v Oystertec plc [2006] 1 SLR(R) 712 at [97]).

(d) The ambit of the protection offered by the Act to a registered proprietor of a trade mark should be guided by the underlying aim of the trade marks regime, which is to ensure that consumers do not confuse the trade source of one product with another (Polo/Lauren at [32]). The policy of s 8 of the Act is to protect the public against deception. In Weir Warman Ltd v Research & Development Pty Ltd [2007] 2 SLR(R) 1073 , V K Rajah JA observed that (at [41]):

It is plain that the trade mark register is intended to be a comprehensive and accurate record of trade marks currently in use. As such, it is crucial to maintain the accuracy of the register and to ensure that undeserving and invalid trade marks are removed from the register without undue delay and complication. In addition, the register is meant to assure and endorse the function of registered trade marks as badges of origin ... so as to protect the public against deception. Hence marks that are misleading or likely to cause confusion or no longer serve their function should not be accepted or allowed to remain on the register, as the case may be. [emphasis added]

To this end, mere association which the public may make between two trade marks because of their semantic content is not in itself a sufficient basis for concluding that there is a likelihood of confusion, in the absence of any possibility of a misapprehension as to the origin of the goods and services (Richemont International SA v Da Vinci Collections Pte Ltd ("Richemont") [2006] 4 SLR(R) 369 at [20]).

(e) The greater the exposure and use of a particular registered mark, the greater its reputation is likely to be, and therefore the greater the protection likely to be afforded to it (The Polo/Lauren Co at [34]).

78 In relation to (a) above, ie, where “a court is entitled to look outside the mark and the sign”, Woo J in Ozone Community Corp at [84] to [87] conducted a thorough review of the conflicting line of cases which differed on whether a court is entitled to consider extraneous factors in determining whether there is a likelihood of confusion and concluded in the affirmative. In addition to that, the ECJ in Canon Kabushiki Kaisha v Pathé Communications [1998] ETMR 366 and Lloyd Schuhfabrik cases noted that the global assessment of likelihood of confusion implies some interdependence between the relevant factors, and in particular, similarity of marks and of goods.

79 As pointed out earlier, both the Appellant’s and the Respondent’s marks have not been used in Singapore. Thus, an analysis of the issue of “likelihood of confusion” has to be made on the basis of a notional, normal and fair use of the marks, see React and Device Trade Mark [1999] RPC 529, 532.

80 It is fair to say that the average consumer in Singapore is one who is literate, educated, exposed to the world and unlikely to be easily hoodwinked, see McDonald’s Corp v Future Enterprises Pte Ltd [2005] 1 SLR(R) 177 at [64].

81 Having considered the various factors, the PAR held (at [84] of the GD) that there was no likelihood of confusion on the part of the public as:

(a)

The goods in question, ie, watches and fashion accessories, had some areas of overlap but were not substitutes and were normally bought after much scrutiny and not at a whim, at [76] of GD;

(b)

An average price of $400.00 was assigned to the Appellant’s and Respondent’s watches respectively. The PAR opined that at such a price, the consumers were not likely to be careless with their purchase, at [77] of GD;

(c)

Consumers of fashion accessories would be a “conscious lot” who would exercise “some form of brand loyalty” thus taking pains to ensure that the purchased goods were of a particular brand, at [78] of GD;

(d)

Against the backdrop of a proliferation of different brands and products, consumers would be more deliberate and circumspect in their selection and purchase, at [79] of GD;

(e)

The mode of sales via specialist boutiques and departmental stores would mean that the visual appearances of the marks would be the main differentiating factor, at [80] of GD;

(f)

The Respondent had taken steps to differentiate their goods, viz, jewellery and watches, from the Appellant’s by way of development of a brand story which portrays a romantic, fairytale-like image of an Italian princess and Bulgarian queen adorned with her tiara and accompanied by her pet cat. In contrast, the Appellant’s goods tend to portray a more mature and sensual image, at [81] to [83] of GD.

82 The average consumer is exposed to a variety of brands. There is also an increasing trend of brands venturing into diffusion lines so as to target different sectors of consumers. For instance, a brand that used to be associated only with female users may now branch into products for male users as well. Similarly, a brand known for its upmarket products may want to make inroads into a more affordable product range. Consumers generally regard the main and diffusion lines to be of the same source. Bently & Sherman at p 872 discusses what “likelihood of confusion” under the UK equivalent of s 8(2)(b) TMA connotes (footnotes omitted):

In its ‘classic’ form, consumers must be confused about the source or origin of the goods or services – that is, they must be confused as to the designer, manufacturer, selector or supplier of the goods or services. In other words, classic confusion is concerned with the situation where consumers believe that the goods or services emanate from one organization, but they, in fact come from a different independent organization. This classic form of confusion has been extended to accommodate broader understandings of the source of goods or services. Consequently, a person will be confused for the purposes of [UK equivalent of s 8(2)(b) TMA] if they incorrectly assume that there is some broader kind of economic connection between the users of marks, for example, that the goods are being provided by a subsidiary or licensee of the trade mark owner. …

In considering whether there is a likelihood of confusion, the tribunals must consider whether there is a genuine and properly substantiated likelihood of confusion: it is not enough that confusion is hypothetical and remote. The term ‘likelihood’ indicates probability rather than possibility. …

[italics in original]

83 Similar to City Chain Stores (S) Pte Ltd v Louis Vuitton Malletier [2010] 1 SLR 382 (“City Chain Stores”), the Appellant’s goods in Class 14 include watches. The Court of Appeal in City Chain Stores at [56] held that since a watch is a product which is commonly available and purchased by the general public, the average consumer would be the general public. Thus, it follows that the nature of the members of public is a relevant consideration. The Court of Appeal went on to say as follows:

The average consumer is not an unthinking person in a hurry but someone who would exercise some care and good sense in making his purchases …

Thus, it was held when luxury goods are purchased, they are bought after careful inspection and deliberation.

(1) Is there a possibility that the consumers would be confused that “FESTINA” and “J.ESTINA” are from the same or economically-linked undertakings?

84 In considering this question, the function of the respective marks as badges of origin of the goods becomes especially important. The marks “FESTINA” and “J.ESTINA” both end with “-ESTINA” which is part of an invented word with no meaning in itself. Apart from their similarity discussed earlier, one must also consider the possibility that the consumer would see the letters “ESTINA” as a common suffix and therefore associate goods with the same (invented) suffix “ESTINA” to be from economically-linked sources.

85 In the case of Durferrit GmbH v OHIM, T-244/01 [2003] ECR II-1589, “NUTRIDE” and “TUFFTRIDE” were held by the CFI to be dissimilar despite sharing the same suffix “-TRIDE. However, there are counter examples such as Julian James’ Application: Opposition of Smart GmbH [2005] ETMR (93) 1096 where “CARSMART” was held to be similar to “SMART” (albeit in part due to “CAR” being descriptive). Notably, these counter-examples serve to illustrate that rulings are “fact-specific” and as Bently & Sherman notes at p 866, such examples have virtually no value as precedents.

86 It was held in City Chain Stores at [58] (approving the approach in Richemont International SA v Goldlion Enterprise (Singapore) Pte Ltd [2006] 1 SLR(R) 401 at [20]) that in the absence of any possibility of a misapprehension as to the origin of the goods or services, the mere association of the public between Solvil’s “Solvil Flower” and Louis Vuitton’s “Flower Quatrefoil mark” based on their similar use was not a sufficient basis for concluding that there is a likelihood of confusion on the part of the members of public.

87 Notably, unlike City Chain Stores where there was a stark disparity in price between the parties’ watches, ie, $200 and (at least) $4,000 respectively, there is no such disparity between the Appellant’s and Respondent’s watches, both of which are marked at around $400. It is important to recognise that “a 50 pence purchase in the station kiosk will involve different considerations from a once-in-a-lifetime expenditure of £50,000” (as per Pumfrey J in Picasso v OHIM, T-104/01 [2002] ECR II-4359). It may be argued that it is unlikely that the average consumer would freely and whimsically spend $400 on a watch without much consideration prior to the purchase and that although the average consumer might do a “double-take” when looking at “FESTINA” and “J.ESTINA” by virtue of similarity of marks, at a price of $400, the consumer will exercise some degree of prudence and pause for a moment to ensure that he is purchasing a watch from the desired brand.

88 Even if there is an association between the two marks owing to the common “-ESTINA”, this association by itself, short of a misapprehension of origin of the goods, may not warrant a finding of likelihood of confusion. Some evidence would need to be adduced to indicate such an association on account of the advertisements, methods of sale and/or packaging of the Appellant’s and Respondent’s items or for any other reason. Otherwise, the risk of confusion is merely hypothetical and speculative because there is simply no evidence of any such confusion arising on the part of the consumers; see City Chain Stores at [57].

89 In this regard, the steps taken by the Respondent to differentiate its goods from those of the Appellant’s become pertinent. As Lai Kew Chai J in The Polo/Lauren Co, LP v Shop In Department Store Pte Ltd [2005] 4 SLR(R) 816 at [21] opined, where the user of a potentially infringing sign had taken pains to distinguish his products from the registered proprietor’s goods, the effect might be that the likelihood of confusion, if any, was merely hypothetical or speculative. The Respondent here submitted that it had taken steps in differentiating its goods from the Appellant’s given that its target consumer was wholly different from the Appellant’s. The Respondent referred to its brand story which projected a dainty image of an Italian princess and Bulgarian queen adorned with her tiara and accompanied by her pet cat. Accordingly, it was argued that this brand story was likely to draw the attention and patronage of younger female consumers. In stark contrast, the Appellant’s goods were advertised in more mature materials thus conveying an image of maturity and sensuality aimed at older and more established female consumers.

90 Unlike the situation in City Chain Stores where Solvil watches averaging $200 in price were pitted against Louis Vuitton watches costing at least thousands of dollars, we have here goods of similar pricing which may be described as low to medium-range, bearing in mind the standard of living in Singapore. There was no evidence that the Appellant’s or the Respondent’s watches were sold in specialty stores (unlike those of Louis Vuitton). We are comparing marks which bear words and pictures resembling each other and not merely four-petal stylised flower designs (Louis Vuitton’s flower design in City Chain Stores was considered not to be that well-known standing on its own anyway). We also do not know how the parties will market or advertise their goods in Singapore as both do not have a presence here at the moment. With the similarity in the marks and in the goods as discussed earlier, there is a real likelihood that the consumer may mistake one brand for the other or believe erroneously that the two are related and have the same origin or come from economically-linked sources.

91 The Respondent’s brand story does not serve as a differentiating measure as there is no brand story on the Appellant’s part. In any event, merely having the brand story on its website is far removed from having it in routine advertisements. Indeed, by giving a European setting to its brand name, the Respondent (with its origin in the Far East) increases the likelihood of confusion between FESTINA and J.ESTINA. I will have more to comment on the brand story later when I deal with the issue of bad faith.

92 In all the circumstances of the case, I am of the view that the likelihood of confusion is a probability and not a mere possibility. With respect, the PAR’s opinion on this issue was premised on her findings from which I have departed materially. Accordingly, the Appellant is entitled to succeed in its opposition based on s 8(2)(b) TMA.

Appeal against PAR’s decision under s 7(6) TMA

93 s 7(6) TMA stipulates as follows:

Absolute grounds for refusal of registration

7. —(1) …

(6) A trade mark shall not be registered if or to the extent that the application is made in bad faith.

94 The PAR held at [113] of the GD that the Respondent’s act of registering “J.ESTINA” was not sufficient to show that its act fell short of the standard of acceptable commercial behaviour observed by reasonable and experienced men in the relevant trade. Thus, the PAR opined that bad faith was not established.

95 The Appellant submitted that given the Appellant’s establishment since 1902 and the successful sales it has in Hong Kong and Japan, the Respondent was very likely to have been aware of the Appellant’s mark “FESTINA and device”. The Appellant also took issue with the Respondent’s failure to tender any written submissions on the issue of bad faith. In particular, the Appellant argued that the Respondent failed to provide evidence of any “Jovanna Estina” or even any “Estina family” with ties to royalty (in its brand story).

96 The Appellant cited Polo Spirit Pte Ltd v S. Tous S.L [2009] SGIPOS 3 for the proposition that when bad faith is alleged by either party, the party who is second in time has the onus of explaining how he came about using or registering the mark. Moreover, the Appellant sought to rely on Kundry SA’s Application: Opposition by the President and Fellows of Harvard College [1998] ETMR 178, for the proposition that where the applicant had not rebutted such serious claims of bad faith by way of counterstatement or evidence, the opposition would be allowed to proceed.

97 The Respondent, on the other hand, submitted that since the two marks were completely different, no issue of bad faith arose in the present case. In any event, the derivation of the Respondent’s mark was sufficiently explained in the brand story, ie, the initial “J” stands for “Jovanna” which is derived from the name of an Italian princess and Bulgarian queen known as “Jovanna Elizabeth Antonia Romana Marie” and revolves around princess Jovanna’s name, her tiara and her pet cat “Jenna”. The Respondent argued that it never maintained that the Respondent’s mark was indeed the princess’ name.

98 An allegation that a trade mark has been applied for in bad faith is a serious one. In “Royal Enfield” Trade Marks [2002] RPC 24 at [31] (endorsed by the Court of Appeal in Future Enterprises at [78]) it was held that:

A plea of fraud should not lightly be made ... and if made should be distinctly alleged and distinctly proved. It is not permissible to leave fraud to be inferred from the facts (see Davy v Garrett (1877-78) L.R. 7 Ch.D. 473 at 489). … It should not be made unless it can be fully and properly pleaded and should not be upheld unless it is distinctly proved and this will rarely be possible by a process of inference.

Notably, this sets a high threshold for an allegation of bad faith to be met by the Appellant.

99 The findings of Geoffrey Hobbs QC in Demon Ale Trade Mark [2000] RPC 345 with regard to the expression “bad faith” was cited with approval in Nautical Concept Pte Ltd v Jeffery Mark Richard and another [2007] 1 SLR(R) 1071 (“Nautical Concept”) at [19]:

[There are] moral overtones which appear to make it possible for an application for registration to be rendered invalid ... by behaviour which otherwise involves no breach of any duty, obligation, prohibition or requirement that is legally binding upon the applicant.

The logical consequence of this is such that once bad faith is established, the application to register a trade mark in question will not be allowed even if the mark does not cause any confusion or breach of duty.

What is “bad faith”?

100 Bad faith is to be determined as at the date of application and matters which occurred after the date of application which may assist in determining the applicant’s state of mind as at the date of application can be taken into consideration, see Ferrero SpA’s Trade Marks (Kinder), O/279/03 [2004] RPC (29) 253.

101 Lindsay J in Gromax Plasticulture Ltd v Don & Low Nonwovens Ltd [1999] RPC 367 at 379 discussed what the concept of “bad faith” connotes:

I shall not attempt to define bad faith in this context. Plainly it includes dishonesty and, as I would hold, includes also some dealings which fall short of the standards of acceptable commercial behaviour observed by reasonable and experienced men in the particular area being examined. Parliament has wisely not attempted to explain in detail what is or is not bad faith in this context; how far a dealing must so fall-short as to amount to bad faith is a matter best left to be adjudged not by some paraphrase by the courts (which leads to the danger of the courts then construing not the Act but the paraphrase) but by reference to the words of the Act and upon a regard to all material surrounding circumstances.

102 As Bently & Sherman at p 851 observes, the courts have struggled to find a satisfactory definition of “bad faith” and have ultimately aligned the definition in trade mark law with that operating in other areas of civil liability, particularly the so-called “knowing assistance” in breach of trust, see Harrison’s Trade Mark Application [2005] FSR 10 (“Harrison’s”). In Harrison’s, Sir William Aldous discussed the notion of “bad faith” at [24] and [26]:

Clearly the court, when considering bad faith, cannot apply a purely subjective test, called by Lord Hutton [in Twinsectra Ltd v Yardley [2002] 2 AC 164] the “Robin Hood test”. The dishonest person or one with low standards cannot be permitted to obtain trade mark registrations in circumstances where a person abiding by a reasonable standard would not. The registration of trade mark is designed to enable bona fide proprietors to protect their proprietary rights without having to prove unfair trading. Registration is not provided to help those with low moral standards.

…

For my part, I would accept the reasoning of Lord Hutton [in relation to “dishonesty”] as applying to considerations of bad faith. The words “bad faith” suggest a mental state. Clearly when considering the question of whether an application to register is made in bad faith all the circumstances will be relevant. However the court must decide whether the knowledge of the applicant was such that his decision to apply for registration would be regarded as bad faith by persons adopting proper standards.

103 The authors of Bently & Sherman at p 852 note that instances of “bad faith” can be placed into three categories:

(a)

No intention to use the mark;

(b)

Abuse of a relationship;

(c)

Where the applicant was aware that a third party had some sort of claim to the goodwill in the mark.

104 The following cases demonstrate the meaning of bad faith according to the above categorization:

(a) No intention to use the mark

105 In the case of PT Swakarya Indah Busana v Dhan International Exim Pte Ltd [2010] 2 SLR 109, Lai Siu Chiu J found that after registering the mark “EMPEROR MARTIN” in 2002, the defendant did not intend to use the mark. Instead, on the shirts sold by the defendant was the defendant’s tailored mark (as opposed to the registered mark) which showed the word “MARTIN” in a much more prominent font than that of “Emperor”. Lai J found that there was an obvious and conscious effort on the part of the defendant to copy the plaintiff’s mark in the usage of the word “MARTIN”. The plaintiff had a total of six marks registered in Singapore, all of which included the name “MARTIN”) and there was a deliberate preference to use the defendant’s tailored mark instead of the defendant’s registered mark. Thus, the lack of bona fide intention to use the mark in the form that was applied to be registered was found to be an act of bad faith.

(b) Abuse of a relationship

106 In Nautical Concept, Tan Lee Meng J found the appellant there to have acted in bad faith when it sought to register its “JWest” mark. It was held that in the light of the appellant’s history of dealings with the respondent and its blatant attempt to take advantage of the respondent’s trade mark, ie, “Jeffery West” and “JW” time and time again in the past, its attempt to register the “JWest” mark was regarded as an unacceptable business tactic adopted in relation to the respondent’s trade mark. The appellant’s attempt to explain the derivation of its mark was found to be too far-fetched; at [26] of Nautical Concept. Tan J agreed with the PAR’s decision there that the “sheer elaborateness of the explanations” for the selection of one letter of the mark (the letter “J”) made it difficult to believe the appellant’s reasons for choosing the mark “JWest”. This was especially so when there was a glaring omission on the appellant’s part to mention that its managing director once distributed shoes under the mark “Jeffery West” and “JW” for the respondent.

107 A finding of bad faith can also be made where the applicant was aware that someone else intends to use and/or register the mark, particularly where the applicant has a relationship, for example as employee or agent, with that other person, or where the applicant has copied a mark being used abroad with the intention of pre-empting the proprietor who intends to trade in the specific country, see Daawat Trade Mark [2003] RPC 11 at [93]. Similarly, where an applicant attempts, by means of registration, to seize a trade mark of a third party with whom he had a contractual or pre-contractual relationship, the registration would also be found to have been made in bad faith, see Lancome Parfums et Beaute & Cie’s Community Trade Mark Application [2001] ETMR 89 (“Lancome Parfums”) at [33].

108 In Rothmans of Pall Mall Ltd v Maycolson International Pte Ltd [2006] 2 SLR(R) 551 (“Rothmans”), having regard to the totality of evidence, Lai Siu Chiu J at [36] to [41] found that various instances of bad faith could be inferred in that case. For instance, the suspicious circumstances surrounding the incorporation of the respondent company suggested that it was merely a “shell company” used by the Hertlein brothers to propagate the use of the “Fairlight” mark in relation to cigarettes. Moreover, despite a Europe-wide injunction which at that time prevented the Hertlein brothers from using the “Fairlight” mark and packaging, they have repeatedly shown their improper motives to capitalise on the appellant's reputation and goodwill by their attempt to sidestep the terms of the injunction, as well as by their persistence in associating themselves with the proprietor's trade marks. Such an association with the appellants can be seen from the Hertlein brothers’ lack of candour when explaining their advertising strategies which blatantly claimed that "Fairlight" cigarettes were similar in taste and packaging to the appellant’s "Rothmans" cigarettes or that “Fairlight” cigarettes resembled the appellant’s "Rothmans" cigarettes.

109 Other illustrations of bad faith can be also be seen in Interkrenn Maschinen Vertriebs GmbH’s Community Trade Mark [2002] ETMR 27 whereby the applicant conclusively showed that there was a person who was linked to the trade mark proprietor and who had obtained knowledge of a planned trade mark of the applicant during a business meeting. This scenario was dubbed as a classic case of “trade mark theft”.

110 During the course of submissions, parties were invited to address the court on whether “bad faith” was a stand-alone, distinct ground of opposition which was independent of attending confusion or whether a “twin-approach” should be taken (in the sense that bad faith can be relied upon in order to tilt the balance in deciding whether there was confusion between the two marks).

111 The Appellant submitted that there has to be at least some similarity between the two marks before bad faith can be established, thus alluding to a case of “copying” of the proprietor’s mark. Alternatively, the Appellant reiterated Lai J’s decision in Rothmans in which it was held at [29] that despite the fact that the marks in question are not so similar as to cause confusion, bad faith can still be established on the facts.

112 As for the Respondent, it maintained that nothing short of a “distinct proof of unacceptable commercial behaviour” should warrant a finding of bad faith. This position, however, seems to be a rather abstract postulation of what the court should do in determining whether bad faith arose on the facts.

113 From the case law discussed above, it appears that bad faith encompasses “underhand practice involving absence of good faith on the part of the [trade mark applicant]” (in the words of Lancome Parfums at [33]. As bad faith is an “absolute ground for refusal”, this suggests that it is an independent and distinct ground for refusing the registration of marks in its own right. As noted by Lai J in Rothmans, this view is consonant with Tan Tee Jim SC’s observations in Law of Trade Marks and Passing Off in Singapore (Sweet & Maxwell Asia, 2^nd Ed, 2005) at paragraph 5.71:

Once bad faith is established, the application will be refused, even though the offending conduct does not cause any confusion or breach any duty, obligation, prohibition or other legally binding requirement.

114 However, the numerous attempts in drawing the parameters of “bad faith” by various courts are vivid illustrations of the inherently abstract nature of this concept. The concept of bad faith is extremely wide in the sense that the courts can infer instances of bad faith and decide according to the justice of individual cases. Naturally, the court should also be cautious in not over-extending this concept lest it becomes a weapon of terror against competitors of trade mark proprietors.

115 The categorization by Bently & Sherman as seen above hints that despite the broad nature of the notion of bad faith, one must show some sort of nexus between the parties in dispute. Otherwise, the notion of bad faith would have to be decided in vacuum. A clear-cut example of such a nexus would be an outright copying of the proprietor’s mark such that the two competing marks are practically identical. However, the nexus may be in the guise of something more subtle. In finding a nexus between the parties, a parallel may be drawn between ss 8(2)(b) and s 7(6) TMA. For example, there may be cases where although there is some similarity of marks or of the goods or services, it falls short of confusing similarity (ie, no likelihood of confusion) within the meaning of s 8(2)(b) TMA. Nevertheless, the evidence of this similarity may be taken into account and considered against the background facts from which bad faith may be inferred. In other words, while the finding of bad faith is largely, if not invariably, based on circumstantial evidence, the party alleging bad faith needs to show some link between the parties, perhaps by way of a pre-existing relationship or some acts of association with the proprietor or some nexus between the two competing marks.

116 There may be a fine line between being inspired by another’s trade mark as opposed to blatant copying or copying with some modifications made to the mark. The former would not attract punitive measures whereas the latter would lead to the mark being rejected. Where the dividing line between “inspiration” and “infringement” should be drawn is a matter best left to the facts peculiar to any case.

Was the Respondent’s registration of “J.ESTINA” made in bad faith?

117 The Appellant argued that the lack of response from the Respondent against the claim of bad faith was far from satisfactory. Even if the Respondent’s brand story was taken as some sort of explanation of the derivation of the brand name, it was argued that the brand story was incapable of holding up against scrutiny. Thus, the only inference was that the Respondent’s mark was a product of copying of the Appellant’s mark.

118 As this was not the first time the parties have been embroiled in litigation, the Appellant sought to draw the court’s attention to previous instances of litigation between the parties:

(a)

An opposition in South Korea Opposition No. 40-2006-000245 in relation to the Respondent’s other mark “Estina Donna & Device” which was successfully opposed by the Appellant;

(b)

An opposition in relation to Korean Trade Mark Application No. 2002-0046412 for “J.ESTINA” in respect of Class 14 which was dismissed as the marks were found to be dissimilar and no likelihood of confusion was held to exist;

(c)

A decision by the Office for Harmonisation in the Internal Market (“OHIM”) in respect of Opposition No. B765356 issued on 26 April 2006 in favour of Appellant. The Respondent submitted that the OHIM had issued its decision based on evidence before it without having considered any submissions or representations by the Respondent as it had, for purely commercial reasons, not filed any observation or response to the Appellant’s opposition. The Respondent was there seeking to use the mark “JOVANNA ESTINA” instead of “J.ESTINA” in Europe;

(d)

A censure issued by the Arbitration Panel at Baselworld (Watch and Jewellery Show) on 2 April 2006 which prohibited the Respondent from using “J.ESTINA” in relation to, among other things, jewellery, precious stones, horological and chronometric instruments for the duration of the show. The Respondent submitted that the Arbitration Panel was an unofficial adjudicating body having no juridical authority over the parties and its decision should have no bearing upon the court’s decision here. Further, as the trade show was for a short duration only, it was simply not expedient to contest the Appellant’s allegations.

As the PAR rightly observed, litigation between the parties in other jurisdictions serves at best as guides to opposition proceedings here. The relevance of these decisions will hinge upon the similarity of the laws of the particular jurisdiction with those of Singapore and also upon the individual facts of the case, at [118] of GD.

119 The Appellant also pointed out that the Respondent had sought to register the mark “Jovanna Estina” in Singapore on 20 June 2008 in Trade Mark No. T 0808240Z, in respect of Class 14 goods. This date fell between the date of notice of opposition filed by the Appellant (22 September 2006) and the date the GD was rendered (8 February 2009). The fact that the Respondent thought it necessary to register an alternative mark in respect of the identical class of goods between those dates could not be coincidental. The Appellant, by way of Mr. Javier Ribas’ statutory declaration, submitted that the Respondent’s action was an indication that it was not confident that it would be able to avoid being held to be in infringement of the Appellant’s mark “FESTINA” by its use of “J.ESTINA”.

120 In Mr. Myoung-chul Shin’s statutory declaration, it was said that a commissioned group was set up in order to come up with a brand name for their jewellery/watch line. A total of 14 European-sounding names were suggested and considered and by process of elimination, “J.ESTINA” was chosen. A Report on Result of Naming Research for New Business was made on 1 October 2002 in which “J.ESTINA” was finally selected, accompanied by a note stating that “a crown device may be added”. The letter “J” in “J.ESTINA” was said to have derived from “Jovanna” who was an Italian princess and Bulgarian queen. This brand story appears on the Respondent’s website which shows a princess, a tiara, and her pet car “Jenna”.

121 A graphic presentation (in the form of a mind map) of the various names that were considered, amongst them being “J.ESTINA”, was also adduced as evidence in the Bundle of Documents, Volume II Part B. The graphic presentation showed the various factors that were taken into account in the process of naming, ie, originality, luxury, royal family, targeting consumers between 25 and 35 years of age. The graphic presentation also showed that the brand slogan was “Modern Renaissance” and the brand concept was one of “European Modern Jewellery”. Three graphs were similarly adduced to show the process of elimination of the other shortlisted13 names and also the degree of registrability of the final five names.

122 In my opinion, the Respondent’s brand story on the name “J.ESTINA” appears extremely contrived. Firstly, the use of Princess Jovanna’s name (compressed to “J.”) only explains the use of the letter “J”. It does not explain why the meaningless word “ESTINA” was chosen. The six letters in “ESTINA” are not even the initials of “Jovanna Elizabeth Antonia Romana Marie” nor do they have any nexus with the tiara or the pet cat “Jenna” which were repeatedly referred to whenever the question of derivation of the brand name was raised. The brand story certainly does not answer any of the queries regarding the derivation of “ESTINA”. As mentioned at [90] above, giving a European setting to the story merely reinforces the association between the two competing marks.

123 The graphs which supposedly illustrate the process of ultimately choosing the name “J.ESTINA” also do not explain in any way how the word “ESTINA” came into being. As the Respondent has conceded, “ESTINA” is an invented word with no meaning. The failure to furnish the court with a credible explanation of how “ESTINA” in the brand name “J.ESTINA” was derived, where the explanation is particularly pertinent since 6 out of 7 letters are in identical sequence in the two competing marks, can only lead to the irresistible conclusion that the word “ESTINA” was blatantly copied from the Appellant by a “cut and paste” job. The adding of “J” with a dot just before “ESTINA” makes little difference to the final name as both words in the competing marks (including the Appellant’s other marks) would still look alike and sound remarkably similar whether “J.ESTINA” is pronounced “jay-estina” (with the “jay” sound likely to be slurred) or “jes-tina”. Such outright copying of the Appellant’s mark is an act that falls short of the acceptable commercial behaviour observed by reasonable and experienced men in the relevant trade.

124 In these circumstances, the Respondent’s registration of T05/25465Z was undoubtedly made in bad faith. I therefore have to disagree with the PAR on this issue as well and allow the Appellant’s objection based on s 7(6) TMA. The Respondent’s application to register its mark should accordingly be refused.

Conclusion

125 For the foregoing reasons, the Appellant’s appeal is allowed on the basis of its opposition based on s 7(6) and s 8(2)(b) TMA.

126 The Respondent is to pay the Appellant the costs of this appeal which are to be taxed or agreed.

127 I thank counsel for both parties for their written and oral submissions which were of great assistance to me.

Festina Lotus SA v Romanson Co Ltd [2010] SGHC 200

Festina Lotus SA v Romanson Co Ltd
[2010] SGHC 200