(e) Is evidence of institutional investors relevant to the first stage trial?
79 There can be little doubt as to the relevance (in the sense required by s 55(1) of the Evidence Act 1995 (Cth)) of the role and behaviour of institutional investors and the significance of such evidence to the common issues.
80 Let me begin with the question of materiality of the relevant information. Materiality arises in a number of contexts.
81 First, it is embedded in ss 674(1) and 674(2)(b) itself, which requires reference to listing rule 3.1 of the ASX Listing Rules. Listing rule 3.1 in turn refers to information that "a reasonable person would expect to have a material effect on the price or value" of the relevant shares.
82 Second, and relatedly, it is expressly referred to in s 674(2)(c)(ii) in similar terms. This is elaborated on in s 677, which refers to information that would, or would be likely to, "influence persons who commonly invest in securities"; the issue as to whether s 677 only applies to s 674(2)(c)(ii) or also s 674(2)(b) and listing rule 3.1 can be put to one side for the moment. Note also that s 677 provides a sufficient condition for establishing the s 674 materiality, but it is not expressed to be a necessary condition.
83 Third, it arises separately in dealing with causation. The first two contexts refer to a reasonable person's expectation on materiality. But in the general causation context (as opposed to individual investor's actions or failures to act on the basis of information or the absence of information) where one is seeking to ascertain the effect of information or its non-disclosure on share price, what I would describe as objective materiality arises. In other words, did the non-disclosure of adverse information produce share price inflation? Did the non-disclosure of favourable information produce share price deflation?
84 Now in terms of the first two contexts, the headline test is objective and not focused on individual subjective states of mind. But how is this objective test established forensically? Usually evidence will be sought to be adduced in the following manner:
(a) One will have the direct evidence of the applicant to establish its individual causation case.
(b) One may have evidence from large institutional or wholesale investors (including fund managers) as to their expectation of materiality; this is relevant given the genus of the class referred to in s 677 of the Corporations Act (see National Australia Bank Ltd v Pathway Investments Pty Ltd (2012) 265 FLR 247; [2012] VSCA 168 at [89] to [90]). Now to say that the evidence of institutional investors may be relevant to materiality and its statutory elaboration in s 677 is not to limit materiality and the operation of s 677 to any particular genus narrower than the breadth of the statutory language used, being "persons who commonly invest in securities". And nor is it to introduce any fluctuating standard (see the discussion in Grant-Taylor v Babcock & Brown Ltd (in liq) [2015] FCA 149 at [69]). Rather, it is simply to recognise that such evidence is relevant, whatever the breadth the language of s 677 can accommodate. To say that materiality and s 677 is not a subjective test is not to deny that evidence of institutional investors' behaviour and perceptions can inform the application of the objective test.
(c) One may have expert evidence from brokers, analysts, investment bankers and capital market researchers opining on the question of materiality and expectation; this is more indirect than category (b) evidence.
(d) One will have evidence, at least on the causation question, on the purely objective materiality question ie whether the information and its non-disclosure actually produced share price inflation or deflation. Alternatively expressed, if on the day the information is disclosed to the market there is a statistically significant movement in the share price that can be attributed to the disclosed information, then this provides some evidence of materiality. But the absence of such evidence does not mean that one cannot establish the relevant s 674(2)(c)(ii) expectation as illuminated by s 677. As noted in Australian Securities and Investments Commission v Fortescue Metals Group Ltd (2011) 190 FCR 364 at [188], s 677 is not a "high threshold". The terms of s 677 "do not invite any inquiry as to whether any change in the price of securities has occurred… caused by an announcement". Nevertheless "[w]hat happened in the market, in terms of movements in share price, may assist the Court in applying the 'likely influence' test"; such an analysis is an ex post analysis, albeit that the expectation issue is an ex ante one (see also at first instance (2009) 264 ALR 201; [2009] FCA 1586 at [474] to [629] and James Hardie Industries NV v ASIC (2010) 274 ALR 85; [2010] NSWCA 332 at [531] to [540]). Now all of this is to be understood in the context of s 677.
(e) Finally, the company's own views on materiality are not irrelevant, but may carry little weight.
85 Further, on the non-disclosure case, the evidence of institutional investors may be relevant to the issue dealt with in s 676 as to what information was generally available, particularly in relation to s 676(2)(b)(i) and s 676(3). Now even though such matters deal with objective tests, the understanding and perceptions of institutional investors on matters relating to what information was in the market place, and what could be deduced, concluded or inferred therefrom, whilst not definitive is not irrelevant.
86 Further, their evidence may be relevant to establishing whether the relevant information was of a kind that a reasonable person would not expect to be disclosed (see the listing rule 3.1A carve outs to listing rule 3.1).
87 In terms of the third context, namely, whether the non-disclosure of information had a price inflationary effect, a statistical tool for analysing this involves linear regression analysis known as event studies. One variable involved in the correlation analysis looks at the movement in market price over a time frame. Now of course the market prices used are a composite of numerous investors' expectations and transactions and cancel out different classes of investors and their idiosyncratic expectations, choices and transactions.
88 The essential elements of such studies are the following:
(a) First, one takes a market and/or industry share index and obtains data for the movement in that index over a relevant period. What "relevant period" does one use? Some methods would use an index for a period before the announcement and look at the relationship of the movement in that index against the movement in the company's share price over that time. So, for example, and perhaps usually, one might use an estimation window of one year. Other methods would actually also include data after the corrective disclosure to anchor the index with the company's "true" share price; at that time the share price will have the inflation or deflation removed. Another aspect to consider is whether one uses any data at all for the period of the contravention. Some models use only data for the index and data for the company's share price before the contravening non-disclosure period. Other models would also include within period data.
(b) Second, one looks at the movement of the share price for the particular company over that same relevant period.
(c) Third, what is modelled is the relationship of the share price movement for the particular company against the movement of the index that has been chosen over the relevant period. What is hoped to be produced is a statistically significant linear trend using regression analysis, so that one can then make predictions as to the company's share price based upon the movement in the index.
(d) Fourth, one takes the day when (and immediately after) the information is disclosed to the market. At that time one has the actual share price for the particular company immediately after disclosure.
(e) Fifth, one compares this actual share price with what one can predict for the share price based upon the relationship with the movement of the index as referred to in subparagraph (c). In essence, one is looking at what one would have predicted for the change in share price on the day by reference solely to normal background or market conditions (economic, general investor market confidence etc). One then compares this predicted share price with the actual share price immediately after disclosure. The difference between the predicted and actual price is then taken to be attributable to the information that was disclosed. If the actual price falls from what one would have predicted (based on normal market movements), then one has established that the share price was inflated by reason of the non-disclosure (but then deflated when full disclosure was made). If the actual price rises from what one would have predicted, then one has established that the share price was deflated by reason of the non-disclosure (but then inflated when full disclosure was made).
(f) Sixth, share price impacts of an event can usually only be revealed if the following conditions are present:
(i) The event is a well-defined news item.
(ii) The time that the news reaches the market is known.
(iii) There is no reason to believe that the market anticipated the news.
(iv) It is possible to isolate the effect of the news from market, industry, and other firm-specific factors which also simultaneously affect the company's share price.
89 Now this is the simple case. But there is an important feature to note which is relevant to the present discussion. Event studies rely on the "semi-strong" version of the efficient capital market hypothesis, which states that share prices in an actively traded security reflect all publicly available information and respond quickly to new information. The "strong" form is that prices incorporate both public and private (including inside) relevant information. The "weak" form excludes contemporary information of any sort; present price is taken to reflect only historic price changes.
90 If one is not able to proceed on the efficient capital market hypothesis and to satisfy these conditions, then one cannot assume that the difference between the actual share price and the predicted share price (based on the relationship of the movement with the index) will produce the quantum of the inflation/deflation.
91 Now in order to use event studies at the first stage trial in the proceedings before me, one matter that will need to be established on the evidence is the "semi-strong" version of the efficient capital market hypothesis. I accept, of course, that an event study may itself show one aspect that in and of itself partially supports demonstrating market efficiency, that is, the speed of the apparent cause and effect relationship between an unexpected event or financial release and a response in the share price.
92 Unsurprisingly, reference may have to be made to the well-known proxy factors referred to in Cammer v Bloom 711 F. Supp. 1264 (D.N.J. 1989) at 1286 to 1287 for determining market efficiency. These have been supplemented (see for example Krogman v Sterritt 202 F.R.D. 467 (N.D. Tex. 2001) at 477 to 478) including looking at the volume and patterns of trading by institutional investors (O'Neil v Appel 165 F.R.D. 479 (W.D. Mich. 1996) at 503).
93 Evidence concerning the behaviour of institutional investors may well be relevant in analysing or establishing any assumption used as to the "semi-strong" version of the efficient capital market hypothesis. Moreover, one may also need to consider how market price is affected by factors other than informational inputs. For example, some large investors may engage in portfolio rebalancing exercises so that their spread of investments matches weightings of stock in an index. Further, they may vary their investments to meet their own liquidity or taxation requirements. Further, market price may be affected not just by human assimilation of information but by computerised trading programs where complex algorithms implement trades on metrics that are pre-determined.
94 Let me also address some other contexts where institutional investors' evidence may be relevant.
95 On the representation case, it cannot be said that the evidence of institutional investors relating to the applicable common issues is irrelevant. What relevant message(s) is conveyed to the intended audience can partly be informed by such evidence, which may differ from the applicant's personal evidence.
96 Further, some of the pleaded representations dealing with gold production and associated forecasting rely upon inferences that it is said should be drawn. The evidence of institutional investors as to what they did infer or may have inferred may have relevance, although it would not be definitive.
97 Further, evidence of particular investors to whom the representations were made, even as part of a group, concerning whether they were led into error is relevant and accordingly admissible.
98 But to say in the present case that evidence from institutional investors is or may be relevant to such matters relevant to the common issues is a far cry from saying that the Court cannot try such matters without that evidence. First, so to contend is an exaggeration. Second, if that was to be the case, then no doubt the applicant would have to address that aspect. And if it was not addressed (and assuming the significance of such evidence as asserted by Newcrest), then that would be to the applicant's and the group's detriment.
99 Generally, it seems to me that evidence from institutional investors will be relevant if not significant evidence for determining the common issues.
100 Contrastingly, it would seem that the applicant's preferred forensic choice on the availability and materiality of information disclosed and not disclosed is to rely on expert opinion evidence and what it seductively asserts to be "market-wide" objective data. Mr Phi has deposed at [37] to [39]:
37. The expert evidence the Applicant intends to file includes evidence from, among others:
(a) a forensic economist expert, who will undertake a quantitative assessment of the impact of the Representations and non-disclosure contraventions alleged by the Applicant on the price of Newcrest securities; and
(b) a materiality expert, who will undertake a qualitative assessment of the materiality to Newcrest's market of investors of the Representations and the information the subject of the non-disclosure contraventions alleged by the Applicant (material information).
38. Based on my experience, I believe that the evidence of these two experts will:
(a) provide a proper basis to test the occurrence and material effect of each of the alleged Representations on the market for Newcrest shares, individually and cumulatively, and with sufficient regard to other factors prevailing in the market at the time; and
(b) similarly provide a proper basis to test the availability and materiality of the information the subject of Newcrest's alleged breaches of its disclosure obligations, individually and cumulatively and with sufficient regard to other factors prevailing in the market at the time.
39. In my experience in proceedings of this nature, the Court, in determining the questions set out at paragraph 38 above, will be assisted by expert evidence reflecting an assessment of the qualitative and quantitative effect of the Representations and material information on Newcrest's market of investors, considered as a whole (rather than by the necessarily partial and idiosyncratic evidence of one or a number of individual investors).
101 Whether this is a sound forensic choice is not for me to say, at least at this stage. Questions of the "best evidence", the adducing of evidence necessary to establish the factual foundations upon which experts may opine, the drawing of Jones v Dunkel and, more generally, Blatch v Archer (1774) 1 Cowp 63 at 65 type inferences are all interesting questions for trial (see more generally Australian Securities and Investments Commission v Hellicar (2012) 247 CLR 345 at [164] to [170] and [250] to [270]). Moreover, the choices to be made, and the risk assessments attending such choices, are all for the parties to make. They are not for the Court to impose in form or to conduce in substance if not in form. They are certainly not for one party to impose on the other directly or indirectly via the s 33ZF mechanism or through the likely effect of any order that might be made under s 33ZF.
102 It may be accepted that the evidence from institutions is relevant to the common issues. But that is, of course, not saying that such evidence is necessary to establishing the applicant's individual claim or one or more of the common issues. In my view, the status of such evidence cannot be so elevated. But even if it could be, that does not establish the s 33ZF test. Section 33ZF does not exist as a coercive power to compel the applicant to file evidence necessary for its case either in an individual or representative capacity. If such evidence is necessary but not adduced, its individual or representative claim will fail. But it is for the applicant and its advisers to make that forensic choice, rather than for Newcrest to invite the Court to impose it. Now Newcrest eschews putting a position that it is seeking to compel the applicant to file evidence of a particular type. True it is that the orders sought do not so directly impose in terms, but the reality is that the effect of such orders has that tendency and likely effect. And after all, the foundation of Newcrest's argument or the consequence that it seeks to achieve is that the Court should hear from and consider the conduct of two institutional investors at the first stage trial.