Conagra Inc v McCain Foods (Aust) Pty Ltd
[1996] FCA 411
•23 APRIL 1996
CATCHWORDS
DAMAGES - measure of - interlocutory injunction granted upon applicant giving the usual undertaking as to damages - injunction subsequently dissolved - principles governing assessment of damages based on undertaking - whether damages claimed were a natural consequence of injunction
Air Express Ltd v Ansett Transport Industries (Operations) Pty Limited (1981) 146 CLR 249
Lumley Life Ltd v IOOF of Victoria Friendly Society (1991) 36 FCR 590
CONAGRA INC v McCAIN FOODS (AUST) PTY LTD
No. NG154 of 1991
Beazley J
23 April 1996
Sydney
IN THE FEDERAL COURT OF AUSTRALIA )
) No. NG154 of 1991
NEW SOUTH WALES )
DISTRICT REGISTRY )
)
GENERAL DIVISION )
BETWEEN:CONAGRA INC
Applicant
AND:McCAIN FOODS (AUST) PTY LTD
Respondent
CORAM:BEAZLEY J
PLACE: SYDNEY
DATE: 23 April 1996
MINUTES OF ORDERS
The Court declares:
The respondent is entitled to damages in respect of lost sales of Healthy Choice and Traditional dinners from April 1991 to December 1992 in an amount of $1,770,000.
The respondent is entitled to damages in the sum of $55,000 for expenses incurred as a result of ceasing production, distribution, promotion and advertising of Healthy Choice when the injunction was granted, including interest.
The respondent is entitled to interest in accordance with the Federal Court Act 1976 (Cth) except for the period 1 November 1992 to 27 May 1993.
The Court
grants liberty to the parties to bring in short minutes of
Order to be filed in the Registry in respect of any agreed calculation of the monies payable to the respondent pursuant to my judgment.
grants liberty to the parties to apply, including in respect of any issue between the parties in relation to the question of the assessmemt which ought to be made in respect of the effect of taxation upon the monies which are payable to the respondent pursuant to my judgment
.
Note:Settlement and entry of orders is dealt with in Order 36 of the Federal Court Rules.
IN THE FEDERAL COURT OF AUSTRALIA )
) No. NG154 of 1991
NEW SOUTH WALES )
DISTRICT REGISTRY )
)
GENERAL DIVISION )
BETWEEN:CONAGRA INC
Applicant
AND:McCAIN FOODS (AUST) PTY LTD
Respondent
CORAM:BEAZLEY J
PLACE: SYDNEY
DATE: 23 April 1996
REASONS FOR JUDGMENT
Introduction
On 11 April 1991, the applicant (Conagra), upon giving the usual undertaking as to damages, obtained an interlocutory injunction restraining the respondent (McCain) from distributing and advertising its "McCain Healthy Choice" frozen dinner products. The interlocutory injunction was dissolved on 4 June 1991. An appeal against that order was dismissed by the Full Court of the Federal Court on 14 April 1992. McCain claims damages from Conagra pursuant to its undertaking.
The principles governing an award of damages flowing from an undertaking as to damages were not in dispute. In Air Express Ltd v Ansett Transport Industries (Operations) Pty Limited (1981) 146 CLR 249, Gibbs J said at 312:
"...[T]he only damages to which a defendant is
entitled are those which he has sustained by reason of the grant of the injunction. The generally accepted view is that the damages must be confined to loss which is the natural consequence of the injunction under the circumstances of which the party obtaining the injunction has notice."
Mason J at 324-325 stated:
"The object of the undertaking is to protect a party, normally the defendant, in respect of such damage as he may sustain by reason of the grant of the interim injunction in the event that it emerges that the plaintiff is not entitled to relief. It is no part of the purpose of the undertaking to protect the defendant against loss or damage which he would have sustained otherwise, as for example, detriment which flows from the commencement of the litigation itself. That is loss or damage which the defendant must bear himself, as he does when no interim injunction is sought or granted. Consequently, it is for the party seeking to enforce the undertaking to show that the damage he has sustained would not have been sustained but for the injunction."
See also Lumley Life Ltd v IOOF of Victoria Friendly Society (1991) 36 FCR 590. The parties disagreed with the application of these principles to the present case. The dispute centred upon the lost sales arising from the grant of the injunction. I will deal with the differing approaches adopted by each later in these reasons. However, there was agreement on certain aspects of the claim. It is convenient to record those now.
First, the parties agreed that McCain's profit margin was 25% of gross sales. Secondly, it was agreed that McCain incurred a number of expenses as a result of ceasing, production, distribution, promotion and advertising of Healthy Choice when the injunction was granted, in the sum of $55,000, including interest.
Background
McCain manufactures and markets various frozen food products, including frozen dinners. McCain has been in the frozen dinner market since about 1985, selling a range of frozen dinner products including a range sold under the label "Traditional" frozen dinners (Traditional dinners). By 1990/1991, McCain held approximately a 21% share of the total frozen food market in Australia, second only to the Petersville group of companies and ahead of competing brands such as Nestle and Sara Lee.
In late 1989, McCain believed it had identified a gap in the Australian frozen dinner market for a healthy style frozen dinner product, such as Conagra marketed in the United States under the label "Healthy Choice". It decided to introduce such a product in Australia, also under the label "Healthy Choice". It decided to produce the product in three varieties, fillet of lamb, fillet of veal and glazed chicken.
McCain launched the product in late March 1991 having commenced production in the previous month. The timing of the launch was based upon McCain's perception of seasonality of the frozen dinner market. Seasonality is the market term used to describe the seasonal demand for a particular product. Mr David Boyle (Boyle), McCain's marketing manager, gave evidence that the frozen dinner market was subject to marked seasonality. He said that from his experience with Traditional dinners and his tracking of competitive activity in the frozen dinner market, sales for frozen dinners showed an increase in March/April each year, with a "sales plateau" in July/August/September. The market thereafter declined and flattened out until the following March. The sales history of Traditional dinners revealed a more specific trend. There was a sharper lift in sales in March/April, a peak in May/June with a slow decline until December/January. Sales would commence to increase in about February of the following year.
In Boyle's opinion, seasonality was highly relevant to the timing of the introduction of a new product into the market place. He said that the success of a new product was largely determined by the ability to attract trial purchasers and to convert those purchasers to repeat purchasers and ultimately to brand loyal purchasers. To capture trial purchasers, it was important to introduce the product during the "up-swing" in the season. That meant introducing Healthy Choice in February/March, to obtain the optimum benefit of the seasonal demand.
The timing of the promotion of a new product was also an important factor in the launch of a new product. Boyle considered that the optimum time to commence television advertising was shortly after the product was launched. Accordingly, McCain planned television advertising for Healthy Choice to commence on 14 April 1991.
Boyle also gave evidence that the first entrant to the new product market section gained a significant advantage as all potential trial purchasers were available to that entrant for an initial purchase and for potential repeat purchases and development of brand loyalty. It followed that later entrants to the market had reduced numbers of potential trial purchasers as, by that time, brand loyalty will have been generated for the first entrant. This view was supported by all experts.
Another factor critical to the successful launch of a new product was the reaction of trade buyers. In 1991, McCain had approximately 52 frozen food buyers employed by major supermarket chains and independent distribution warehouses who, between them purchased nearly 100% of McCain's manufactured frozen food products sold at retail level. McCain's practice in respect of the launch of a new product was for its senior sales representatives to approach the buyers, inform them of the proposed product and product launch and to negotiate "listing" of the product. "Listing" is of particular importance for frozen food products as there is only a fixed amount of freezer space available, both for storage of the manufactured product in the "distribution pipeline" and for display at retail level. The introduction of a new product to the market requires the displacement of existing frozen food products from available freezer space in both situations. Unless buyers are attracted by the product, the manufacturer is not able to secure any commitment to purchase. However, if a commitment is obtained, warehouse space is allocated to the new product and a "listing" in the buyer's organisation is achieved.
Once a product has been listed, shelf space, or facings, as it is referred to in the industry, must be allocated in each store in which the product is to be sold. In Boyle's experience, buyers assess the retail level of products over a period of about 3 months. If the product does not meet the buyers' expectations, it will either be delisted by the organisation concerned or the buyer will advise the supplier to improve the profitability of the product to the organisation. If, after that, the product does not reach buyers' expectations, the product will be delisted. Boyle said that for marketing purposes, McCain considered that a product had been delisted by an organisation when no orders had been placed for a product for a period of three months. He said that relisting can be difficult to achieve once a buyer considers that a product has failed.
Boyle's evidence was supported by David Colwill, formerly National Category Manager - Meals and Desserts with Coles New World Supermarket chain. Since September 1991, Colwill has provided consultancy services to organisations in foodstuffs. Colwill said that in his experience, the procedures used by Coles New World were similar to those used by other supermarket chains.
Colwill confirmed that it is the function of a supermarket buyer to both add new lines and delist products from a supermarket's inventory. Manufacturers of intended new products met with buyers to introduce the new product to them, including information relating to the trading terms the manufacturer was prepared to offer, the expected product profitability, pallet stacking configurations, use by codes, packaging, price, advertising and marketing plans, competitive activity and the manufacturer's strategy. A buyer negotiated with a manufacturer to determine the level of financial support which the manufacturer was prepared to devote in support of the new line. These monies were used to purchase promotional activity of varying types agreed to by both parties. Sometimes, the buyer required a manufacturer to devote part of these monies to a slotting allowance or fee. The retailer used such money to offset any costs incurred in adding a new item.
The buyer then generally evaluated the proposed new product, for example, by investigating products of a similar nature and assessing the performance of that product over a 13 week period. The buyer then made a decision as to whether the supermarket ought to buy the product. The buyer's decision in this regard was, to a large extent, judgmental. However, Collwill said that in his experience, the buyer's recommendations tended to be fairly accurate.
Once a new item had been added to the supermarket's inventory, it underwent a stringent review process to ensure that its place on the supermarkets shelves was being maintained and that consumers were accepting the product entry. The review process was conducted each four weeks, usually over a total period of 13 weeks, by way of either scan data or warehouse withdrawal data. The review process provided, on average, two opportunities to enhance the market plan if the product was not performing to expectations. If, in the buyer's estimation, the product was not performing adequately after the 13 week period, the product was likely to be delisted.
Launch of Healthy Choice
Boyle made certain decisions relating to the packaging and advertising of Healthy Choice which also affected McCain's marketing of Traditional dinners. First, he decided that the packaging for Traditional dinners should be redesigned and that it and Healthy Choice be marketed in complementary packages. Secondly, he decided to devote McCain's advertising budget to Healthy Choice and to use the same advertising theme as had been used for Traditional dinners. That theme involved a vicar eager to eat a Traditional dinner. The advertisement was changed to have the vicar eating a Healthy Choice product. Boyle had decided upon the promotion of Healthy Choice only as, at the time of the launch, McCain had insufficient funds to advertise both products. Boyle had deliberately decided upon the complementary packaging and a continuation of the vicar theme so as to attract "pull-through" support for Traditional dinners.
According to Boyle, McCain obtained an initial high level of buyer acceptance for Healthy Choice. By 11 April 1991, the date of the grant of the injunction, McCain had received firm orders from its customers representing 33% of the total available grocery trade. Within one month of the lifting of the injunction on 4 June 1991, McCain received orders for Health Choice representing 91% of the total trade. Ordinarily, it took at least 3 months to achieve orders to that level. This evidence was not challenged.
Grant of Injunction on 11 April 1991.
On 8 April 1991, Conagra commenced proceedings against McCain seeking, amongst other orders, an order to restrain it advertising, distributing and selling Healthy Choice. On 11 April 1991 Gummow J granted an interlocutory injunction in the following terms:
"Upon the applicant by its counsel giving to the Court the usual undertaking as to damages, the Court orders:
That the respondent by itself, its servants and agents be restrained from, in trade or commerce, selling, offering for sale, promoting or advertising any article of food in the packaging, examples of which are depicted in Exhibit 1 and Exhibit A."
Exhibits 1 and A referred to in the injunction were examples of the Healthy Choice product.
In accordance with the practice note of the court which was in force at 1991 (see Practice Note No 3 issued 7 May 1990), the usual undertaking as to damages was in the following terms:
"Where, in relation to an interlocutory injunction or undertaking to the Court, an undertaking as to damages is to be given, a "usual undertaking as to damages" will be an undertaking to pay to any party adversely affected by the interlocutory injunction or undertaking such compensation (if any) as the Court thinks just, in such manner as the Court directs. "
Immediately the injunction was granted, McCain ceased production and distribution of Healthy Choice. McCain also ceased selling, and offering for sale, promoting and advertising Healthy Choice. Boyle gave instructions that deliveries out of McCain warehouses were to cease and that no further orders could be taken until further notice.
The injunction was dissolved by order of the Court on 4 June 1991: see Conagra Inc v McCain Foods (Aust) Pty Ltd (1991) 101 ALR 461; (1991) 22 IPR 175.
On 4 June 1991 McCain recommenced production of Healthy Choice. Distribution, selling and promotion commenced with full effect on about 12 June 1991. Television advertising commenced in the first week of July.
Alleged effect of injunction
McCain alleges that the injunction had a number of serious consequences for it. First, it lost the advantages of being the first entrant in the market, as, in April/May 1991 a competitor, Pasta House, launched a new "healthy style frozen dinner product range" under the brand name "la Del'Lite". At the time, Pasta House had approximately 3.0% share of the frozen dinner market. However, its la Del'Lite brand achieved a 4.2% share of the total frozen dinner market, without television advertising support. Boyle considered that the success of la Del'Lite was due to the advantages it obtained as the first entrant in the market for a healthy style frozen dinner product. He believed that McCain was disadvantaged in being a later entrant into the market.
Secondly, even though the injunction was dissolved on 4 June 1991, it was, by then, too late in the season to establish an acceptable level of repurchase of the product towards the end of the upswing season. Subsequently, and, in part in consequence of this, Boyle believed that Healthy Choice suffered some delistings as there were inadequate sales levels and profits to retailing organisations.
Thirdly, even after the injunction was dissolved, there was a delay in getting product back onto the market. Production had to be restarted, warehouse picking slots and shelf facings for retail display had to be regained, and orders had to be obtained. Because of this, Boyle did not arrange for television advertising to recommence immediately, as there would be no product available for consumer sales. However, he was of the view that by time advertising commenced in July, it was too late to establish an acceptable level of repurchase.
Fourthly, Boyle considered that the sale of Traditional dinners also suffered as they did not have the benefit of the pull-through effect of the association with Healthy Choice which had been part of Boyle's marketing strategy for 1991. Boyle said that it was not feasible to re-run the 1990 advertisement for Traditional dinners as the packaging had been changed for the 1991 season. Nor did he arrange any new advertising for Traditional dinners as he was confident the injunction would be dissolved. Had he advertised Traditional dinners, he would have had no funds available to promote Healthy Choice later in the season. He considered that it was essential to the success of the new product to give Healthy Choice the promotional support.
Boyle said that these factors led to sales of both Traditional dinners and Healthy Choice being lower than anticipated in the 1991 season. He said that this had a flow on effect on both products in 1992, and in the case of Healthy Choice, until February 1993, sales being lower than he otherwise expected them to be during these periods.
It would appear that Conagra accepted that McCain lost sales of Healthy Choice during the period of the injunction. It denies that it otherwise lost sales from the effect of the injunction. Both parties called expert evidence seeking to demonstrate or deny any other loss. Before dealing with the expert evidence, it is convenient to consider the claims made by McCain.
McCain's launch of Healthy Choice was based upon an expected annual volume of sales of 96,000 dozen cases. This projection was based on the following considerations. McCain's Healthy Choice concept was derived from Conagra's US product of the same name. In the first year, Conagra had obtained a 4.4% share of the total dinner market for its product. McCain uses A.C. Nielsen (Australia) Pty Limited's SAMI data for its market analysis and projections. There was no dispute amongst the experts that A.C. Neilsen was a reputable source for this information. SAMI data records the movement of goods from participating warehouses to retail outlets. It includes data from independent distributors such as Davids Holdings as well as the major retailers who handle their own distribution. According to the SAMI data, the Australian frozen dinner market had a value of $85.3 million and a volume of 2.1 million dozen cases at the end of September 1990. At that time, the market was exhibiting an annual volume growth of 3.7%. Boyle estimated that if growth of roughly 4% was sustained, 1991 sales would be in the order of 2.184 million dozen cases. McCain assumed that that growth would be sustained and that McCain would be able to obtain the same market share for its product as Conagra obtained for its Healthy Choice product in the United States. On those assumptions, its expected volume of sales was 96,000 dozen cases per annum. McCain also assumed that its sales across its individual products fillet of lamb, fillet of veal and glazed chicken would be approximately equal.
Prior to the launch of Healthy Choice, Boyle considered that the projected sale figures of 96,000 dozen cases was both achievable and conservative. In practical terms, it required only a small number of sales in each accepting supermarket each week. Boyle's view was based upon McCain's market share, which at the time was approximately 25% of the total Australian market for frozen dinners; the annual growth rate of the frozen dinner market; the expected effect of television advertising, focussed as it was on Healthy Choice; and the benefits McCain expected from launching the product at the beginning of the upswing in seasonal demand and from being the first entrant onto the market. In addition, based on historical sales of Traditional dinners, Boyle considered the projected sales of 32,000 dozen cases per individual Healthy Choice product to be conservative.
Boyle said that it was "fundamental marketing sense", particularly in the case of the launch of a new product, to carefully monitor and evaluate both the product launch and the on-going performance of the marketing and product concept. McCain undertook that monitoring and evaluation through Sutherland Smith, its market research advisers.
In December 1991, Sutherland Smith identified six "key limiting factors in 1991" as responsible for the failure of Healthy Choice to register "acceptable" market share results. One was the late introduction of Healthy Choice onto the market. The others were: a potential price barrier for new purchasers of Healthy Choice, particularly for those who were not already purchasers of Traditional dinners; lack of branding of the Healthy Choice name in advertising; possible confusion between Healthy Choice and Traditional dinners, e.g. due to the similar packaging; a lack of activity aimed at stimulating trial of Healthy Choice; and product experience not matching expectations. Boyle's own view was that the critical factor causing losses over that period was McCain's initial inability to adequately promote the product in the important up swing period in April 1991. In addition, he said it was not possible, at the end of 1991, to adequately assess the impact of the various factors identified by Sutherland
Smith. According to Boyle:
"Even at the end of 1991, the then marketing and sales results available to McCain were unclear due to the market abnormality of the injunction, the absence of McCain frozen dinner advertising at the commencement of the seasonal "up swing" and a marked reduction in consumer trial of McCain frozen dinners in the critical seasonal "up swing" with consequent reduction in the number of consumers engaged in repeat purchasing of McCain products in the second half of the 1991 season.
That is, even though McCain was aware of difficulties in the market place and of some confusion in the market place between Traditional Dinners and Healthy Choice, it was not possible to discern whether this was a fault with the McCain marketing strategy for the Healthy Choice/Traditional Dinner products or was the result of interference with the McCain marketing strategy caused by the injunction during the first half of the 1991 season."
McCain alleges that, as a result of not being able to properly assess the impact of the factors identified by Sutherland Smith at the end of 1991, appropriate adjustments could not be made to the marketing of Healthy Choice or Traditional dinners for the 1992 season. Due to the uncertainty, only minimal changes were made to the product range, packaging and advertising concept. Boyle said it was not until some time into the 1992 season that it became evident to him that there were difficulties with the overall marketing strategy over and above the effects of the injunction. McCain alleges that as it was not able to clearly identify the cause of the problems with the marketing of Healthy Choice before the start of the 1992 season, it continued to suffer losses until early 1993 in the case Healthy Choice and until late 1992 in the case of Traditional dinners.
McCain commissioned a new Sutherland Smith report in the second half of 1992. Based on its findings and recommendations, a new strategy was implemented for the 1993 season and the product effectively "relaunched". This involved a new advertising theme, new packaging and a substantial expansion of the varieties of dinners in the range. McCain abandoned its Traditional dinner range as well as it Budget Gourmet range. The relaunch was directed to attracting a group of new consumers to Healthy Choice distinct from those loyal to Traditional dinners.
The expert evidence
As I have already stated, each party retained experts to provide opinions as to the loss which McCain sustained as a result of the injunction. McCain retained Kenneth James Le Hunt, John Michael Legge and Garry Alexander Edser. In addition, Boyle gave evidence of McCain's loss based upon a regression analysis. Conagra retained Professor Roger Layton. The experts commented upon the evidence of the other and each expert was extensively cross examined. Each party also retained accounting experts. The expert accountants were not cross examined. I propose to deal with Boyle's evidence and the evidence of each of the experts in turn.
Boyle's evidence of McCain's loss
Boyle made an assessment of losses both for Healthy Choice and Traditional dinners. Boyle estimated a total loss in the sum of $1,431,800 being a loss of $793,800 in respect of Healthy Choice and $638,000 in respect of Traditional dinners.
For the purposes of assessing the loss of Healthy Choice, Boyle commenced with the projected sales figures for Healthy Choice of 96,000 dozen cases. He adjusted this figure for seasonality over 1991 and subsequent years. He calculated loss of profits by applying a figure known as the "Gross Margin (average)" to the loss of gross sales revenue of approximately 25%. As I have stated earlier the gross margin figure was eventually agreed between the parties.
Boyle's initial analysis had been based on certain projections using a statistical tool known as a regression analysis. By the time he gave his evidence, historical figures were available. He said that a comparison of the projected figures with historical figures revealed that by March 1993, Healthy Choice sales reached projected sales. He believed that was likely to continue. The end of February 1993 thus marked the end of the claim of loss in respect of Healthy Choice.
A comparison of projected and historical figures for Traditional dinners revealed that in November 1992, actual figures reached projected figures and since that time had remained consistent with or above the projected sales. November 1992 thus marked the end of the claim of loss in respect of Traditional dinners.
In cross examination, Boyle agreed that his original projection of 96,000 dozen cases was a guesstimate, made to secure approval for the launch of the new product. He also accepted that the figure was originally calculated by a Mr Yung, McCain's product manager who reported to Boyle.
Notwithstanding that Boyle agreed that the original projection was a guesstimate, he considered that the market history of Healthy Choice demonstrated that McCain's initial projections were well based. Healthy Choice initially achieved a 4.4% market share by August 1991. However, its share thereafter declined, so that its average market share in 1991 was 2.5-3.0% of the frozen dinner market. This share had been achieved even though McCain had not been the first entrant into the market, had lost the benefit of the initial seasonal up-swing for frozen dinners and did not have the benefit of advertising support. Market research carried out on behalf of McCain in late 1992, indicated Healthy Choice could achieve a 8% market share. By April 1993, according to McCain's estimates, Healthy Choice held an approximate 10% market share. La Del'Lite had achieved a 4.2% share of the frozen dinner market without television advertising. Further, the United States experience demonstrated that the market for such products was increasing rapidly. By the end of 1990, the low fat, low cholesterol share of the frozen dinner market in the United States had grown to approximately 27.8% and Conagra's Healthy Choice products had achieved 7.7% share of the total frozen dinner and entree market.
Boyle agreed in cross examination that the projection of 96,000 dozen cases was based upon an annual growth rate of 4%. It was suggested to him that, in fact, as at December 1990 when this projection was made, the market was in serious decline. Boyle's
response to this was that whilst that was so, he did not recollect that the decline had exhibited itself in the MAT figures. In fact, the evidence revealed that there had been a slight increase in the market of 0.3% in the preceding 12 months to December 1990. Thereafter, there was a decline of 5.9% in 1991 and 2.8% in 1992. In relative terms, it can be said that that was a small decrease, especially given that Boyle considered his estimate to be conservative. He pointed out that, in any event, any recessionary effect in the market would be caught up in the projections. Further, the figure of 96,000 dozen was based upon two factors, first a growth rate of 4% and secondly a market share of 4.4%. If the market share proved to be greater than 4.4%, there would be a corresponding increase in the number of cases sold.
Legge's assessment of Boyle's evidence
Legge agreed with Boyle's view that the frozen dinner market was significantly seasonal. His own inquiries revealed that the market grew about 52% between January and May in each year. He said that the majority of new buyers and buyers whose previous brand loyalty may have weakened with the passing of time, joined the market during this period. He considered that such buyers are likely to remain with their first choice of variety and brand for the rest of the season.
According to Legge, revenue per square metre of stores in the supermarket chain, rather than the profit earned on any single transaction was the primary control measure used for a supermarket's inventory of products. Thus, products which were perceived to be slow moving would be deleted unless there was some exceptional reason for the supermarket chain to retain it, such as, an exceptionally high profit per item of product or a large proportion of supermarket patrons who would change their store preference if the line was deleted.
Legge said that Boyle's regression analysis was an appropriate and well known method for estimating trends. He considered that Boyle's projection, based upon this model, was "cautious and carefully done, albeit, the methodology was a bit rough and ready". Subject to that qualification, he confirmed that Boyle's approach was acceptable and appropriate to produce a reasonably satisfactory projection. He recognised, however, the inadequacy of statistical techniques. As he said:
"there is no way a computer model is going to say what people actually do".
He expressed the same point later when describing his intention in his own statistical approach:
...I'm hoping to get an average figure; a feel for what [consumers} probably might have done under the circumstances as disclosed to me".
Legge was cross examined on his endorsement of Boyle's approach. Boyle had based his analysis on a moving annualised total (MAT). A MAT figure uses historical sales data. In this case, Boyle had used a MAT figure for the 20 month period prior to the grant of
the injunction. Legge agreed that a preferable approach would have been to take a central moving average as the basis for the regression analysis. That would have involved using the historical sales data for the 6 month period before and after the grant of the injunction and averaging that data so as to ascertain a central moving average. It was suggested to Legge in cross examination that, if a central moving average was used, there would have been no loss of profit. Legge said that he would be surprised if that was the case. There was no evidence to support the proposition put in cross examination.
Layton's criticism of Boyle
Layton criticised Boyle's approach. He said it was an inappropriate method to estimate McCain's losses in relation to Healthy Choice as it assumed that the initial assumptions underlying the original forecasts were correct and that the failure of the sales to match the original expectation must be attributable to the injunction. It ignored available data as to other irregular elements, seasonal factors and trends.
Layton plotted Boyle's seasonality factor against that derived by Census II and that derived by utilizing the Winter method. Layton asserted that Boyle had been led into error by using an analysis which made assumptions masking the true seasonality of frozen dinners. Boyle had identified March/April as the commencement of the selling season and May as the peak sales period. The statistical methods used by Layton identified a later start in the selling season and identified two peak periods
- August and June respectively. Layton had concluded from these factors that "the injunction did not fall in the critical selling period." Layton withdrew his criticism of Boyle and amended his statement that the injunction did not fall in the peak selling period. He conceded that the peak selling period ran from April to October. He also conceded that Boyle was correct in his assessment that the injunction did affect sales at the beginning of the peak season for the product. However, Layton maintained his view that the pattern of seasonality was changing over time. He considered that by relying on only 2 years data, the underlying drift in seasonality that was taking place in the frozen dinner market was masked.
Layton considered that other aspects of Boyle's approach were questionable. In particular, he said that the Australian frozen dinner market was less than that estimated by Boyle. Boyle conceded an error in this regard - but noted by his error he had overestimated the market by 1.55%. Layton considered that Conagra's 4.4% market share in the United States, upon which Boyle had based his projection, was not necessarily transferable to the Australian market. He said much depended upon other factors such as the product, its presentation, its pricing and positioning, and the effectiveness, timing and strength of promotional campaigns. Most importantly, he considered that McCain's marketing approach would have to be highly effective for it to attain a 4.4% share within 3 months, as projected. He conceded however that the advertising proposed for April 1991 was appropriate timing.
Layton did a comparison of sales and advertising expenditure over 1991, 1992 and 1993. He concluded that the vicar advertising theme was flawed in relation to Healthy Choice. He said his view was reinforced by Sutherland Smith's December 1991 report, which recorded that whilst the vicar advertising was effective in making an impression on consumers, it was not effective in making them aware of the Healthy Choice product. However, in his cross examination, Layton conceded that his conclusion as to the effectiveness of the advertising was largely influenced by what he read in the Sutherland Smith report rather than by his comparison of sales and advertising figures.
Layton identified a number of other difficulties which McCain had or may have had in relation to Healthy Choice which might have affected sales. These were: a potential price barrier for new buyers, particularly those who were not already Traditional dinner buyers; possible confusion in the market between the two products, which may have been attributable to the similarity in packaging; the fact that the December 1991 Sutherland Smith report intimated that consumer product experience did not match consumer expectation; the fact that other material from McCain indicated that there were difficulties with product formulation for Healthy Choice up until late June 1991 and that recipe reformulations were being undertaken until the second half of 1992; the limited range of Healthy Choice - he considered the decision to increase the range to 5 products for the 1992 season and its further enlargement in 1993 was a recognition of this; and the fact that McCain was having distribution difficulties
with its products generally which would have impacted on supply to the market, although he was not able to assess the significance of that impact. It will be seen that Layton obtained this information from Sutherland Smith's December 1991 report.
Layton was also critical of Boyle's suggestion that McCain lost the value of obtaining trial purchases during the period of the injunction, thus losing the pattern of trial and repeat purchase necessary for a successful launch. In the first place, according to Layton, the opening months of the winter season did not comprise the only or most significant trial period so that the injunction would not have had the impact suggested. However, in cross examination, Layton conceded that this assessment would have to be modified as he had previously wrongly analysed the McCain peak periods. He also conceded that McCain had missed the first peak period due to the injunction. However, he maintained his argument that it did not miss the second peak period in August. However, Layton's view in this regard fails to come to grips with Boyle's assertion that, by August it was too late to gain brand loyal purchasers. Secondly, Sutherland Smith advised that, according to their research, there was no clear pattern of trial and repeat purchases and that the trend was for two trial periods, one in March and a second in July/August. Layton had no independent data to support this, but accepted it. On that basis, he considered that the relaunching of the Healthy Choice advertising in June 1991 after the injunction was lifted was well timed to promote the Healthy Choice product during the second trial period. This view also overlooks the inability to obtain brand loyal purchasers. It should also be noted that advertising did not recommence until July.
MCCAIN'S EXPERT EVIDENCE
Kenneth James Le Hunt
McCain retained Kenneth James Le Hunt (Le Hunt) to provide a projection of future sales of Traditional dinners. Le Hunt is a senior consultant with America Software (Australia) Pty Limited (American Software). American Software is the Australian distributor of the forecasting and distribution software designed by American Software Inc., and which is sold under the name "American Software Forecaster 2000". This product is capable of producing forward projections for the sale of a manufactured product based on historical sales. It uses various statistical and arithmetic algorithms to provide a best fit to historical data. It uses the formula so obtained to project forward, taking into account such factors as historical trends and seasonality.
Le Hunt used actual sales history for Traditional dinners for the 3 year period April 1988 to February 1991 in his assessment of lost sales. He considered this period provided an "excellent" base for a short term projection. The computer programme selected a "best fit" statistical model from those provided by the programme. Two models were selected. Le Hunt said:
"The technique is to use all the above models to retrofit the demand history. The programme will perform a linear regression analysis to compute a starting point and will then "smooth" in each historical period of demand to complete an historical forecast. Having completed that it will ascertain a starting point (permanent component) and then using identified trend and seasonality patterns, if appropriate, project out a forecast for two years.
In selecting the most appropriate historical forecast to use, the system will record the amount of "error" (estimated forecast variance to the actual demand) and choose the model (and the appropriate weighting factors) that give the "least error"."
Using this technique, Le Hunt produced two projections. The first was an historical forecast based on the best fit identified by the prescribed technique. He incorporated seasonality factors based on the equivalent seasonality of the historical data (the first projection). This projection produced a loss of $1,406,300 for the period April 1991 to March 1993. Le Hunt considered that there was a larger "error" in this projection than he would have liked. He said this was due to the irregularity of some of the historical data, particularly in 1988 and 1989.
He then modified selected parameters from that forecast to filter out some of the demand irregularities. This was produced a projection of $1.580,700 million for the same period (the second projection). Le Hunt believed that this forecast provided a better historical fit, and that therefore, as a matter of theory, future projections should be more accurate. However, he observed that a statistical system such as he used for the second projection will not readily know if an item is "new" or stable and so will be slow in predicting a turning point in the demand. He considered that people were better positioned to interpret such turning points and expected at least another 3 or 4 months of demand to be applied to the system before it projected a forecast for the item that indicated a decline in the product life. He thus preferred the historical forecast notwithstanding its greater marginal error.
Le Hunt agreed in cross examination that products generally exhibit a pattern of growth, stability and decline. He considered that his forecast indicated a product decline, which was already evident in the historical data provided. Le Hunt also conceded that there were a number of factors which might operate in the market place which were not taken into account in his model. For example, the model did not take into account the impact of an associated product coming onto the market which might take sales away from the original product. Nor did the model contain a technique to remove any abnormal demands, such as might be produced, for example, by "spectacularly successful advertising". Likewise, it did not take into account the effect of recession. However, Le Hunt pointed out that if the historical data upon which the projection was based had recessionary components in it, this would be reflected in the projection. Le Hunt contrasted his estimated losses with Boyle's estimated loss of $1,431,800. He considered the variation to be inconsequential.
Le Hunt conceded that statistical forecasts were wrong "all the time" as there are always unexpected factors at work. He accepted that a fair way of summarising his projection was "the best attempt in advance to model what may occur in the realisation that it probably will not occur".
Legge's assessment of Le Hunt
Legge was not familiar with the software used by Le Hunt. However, he considered that Le Hunt's historical projection provided a good approximation to the actual sales during the historical period. He believed, therefore, that the programme was mathematically soundly based. He also considered that Le Hunt's statistical projection was based on a normal statistical method. For these reasons, he thought that Le Hunt's model was likely to produce a sound projection.
Layton's criticism of Le Hunt
Layton considered that there were several difficulties with the methodology used by Le Hunt. First, he noted it was a variant of the Winter's method. Typically, that method required 4 years of data to form estimates of the seasonal effects evident in data. He accepted that whilst shorter periods could be used, the results had to be used cautiously. Indeed, he suspected that the difficulties described by Le Hunt which caused him to make adjustments were as a consequence of the short time series used. Secondly, he was uncertain as to what procedure Le Hunt used for the estimation of the initial seasonal effects. For that reason, he could not repeat the analysis carried out.
Thirdly, Layton doubted Le Hunt's data replacement. Whilst he considered it reasonable to adjust for known once-off shocks, he said he would not have adjusted the data in the absence of good external reasons. Le Hunt had not identified any. Next, Layton made the same criticism of Le Hunt's analysis as he did of Boyle, namely, it was based on pre-injunction data. He observed that a Winter's analysis of the full data series indicated that the injunction had little more than a passing impact on Traditional dinner sales.
John Legge
McCain retained John Legge (Legge), a lecturer in the School of Innovation and Enterprise of Swinburne University of Technology. Legge is a Master of Business, a Fellow of the Australian Institute of Management and a member of other professional bodies. Legge commenced in the computer industry in 1968 as a programmer. He later transferred to marketing and has been in the marketing industry for 15 years. He has a sound knowledge of the grocery industry. He is the author of 4 books on marketing and business. In 1991 - 1993 Legge lectured in Marketing and Business Planning in the Innovation and Management programs of the School of Innovation and Enterprise. Swinburne University as part of the Master of Enterprises Innovation and the various Diploma programs. He has also written two texts on management.
Legge stated that in the case of a product with a sales history, mathematical models may be used to obtain reasonably accurate projections of future sales. He stated that these estimates were particularly accurate where products were sold in large numbers at a relatively low price. In such cases, a sudden departure
from an expected pattern of sale would almost invariably be found to be due to some external event. In such circumstances, it was a relatively straightforward matter to estimate the degree to which such an event had impacted upon the expected pattern of sales. Legge considered that McCain Traditional dinners fell into this category.
Initially, Legge was requested to give an expert opinion as to whether Boyle's and Le Hunt's projections were "within the bounds of possibility". For this purpose, Legge carried out a basic form of statistical modelling against which he assessed their projections. On the basis of this modelling, Legge gave a general endorsement of the methods used by Boyle and Le Hunt. I have referred above to his evidence and cross examination relating to their projections.
Conagra's expert, Professor Roger Layton, criticised Legge's assessment of Boyle's and Le Hunt's analysis. He said:
"[i]f the market turns down for an external reason, the long term trend line will over-estimate the true level of sales."
Legge responded to this criticism by pointing out that his initial brief was a limited one and at the time he reviewed Boyle's and Le Hunt's projections, he had not constructed a detailed model of McCain sales incorporating media and distribution data, nor had he been able to review the McCain sales data for 1993.
McCain subsequently requested Legge to compute the revenue loss to McCain arising from the injunction using a distinct methodology from that employed by Boyle and le Hunt.
Legge's methodology: general
In accordance with his further brief, Legge produced an assessment of McCain's loss in respect of Healthy Choice and Traditional dinners using a Bass-type computer model. The Bass model is named after Professor Bass (Bass) of Texas USA. Bass proposed that two distinct phenomena were present during the growth of a new market. Certain people join the growing market in response to the sales and marketing efforts made by the suppliers and their distributors. These people are generally referred to in the literature as "externally influenced adopters". Other people join the growing market in response to information received from others. These people are called "internally influenced adopters".
According to Legge, Bass and other researchers have been able to demonstrate that mathematical models incorporating the concept of internal and external influences are able to reproduce the pattern of growth observed in numerous real markets. He said that statistical tests have shown that Bass models have been able to explain a very high proportion of the change in purchasing between the launch of a product and the on-set of maturity in its market. Other research, including research carried out by Bass, indicated that the rate of externally influenced adoption depends, at least in part, on marketing effort, including
advertising and point of sale display. According to Legge, Bass had validated this research. Legge had also carried out research which he said validated the statistical model. Legge was not challenged in respect of this evidence. He conceded, however, that although his model was a sophisticated one, it was "fraught with problems in terms of accuracy, because of its necessary dependence on data which is incomplete". He explained that that was a problem endemic to all statistical models. As he said:
"The whole process of making a model is to abstract certain data and the only form of reality is what was actually happening in the supermarkets and there is no way we are ever going to go back and find that".
Legge relied upon the following matters in his assessment of McCain's losses:
(1)McCain were restrained from distributing or advertising Healthy Choice during the period of the injunction.
(2)McCain were damaged by being prevented from selling part of their completed production and being forced to dispose of certain raw materials.
(3)Sales which McCain might reasonably expect to have gained during the period the injunction was in force and the further period whilst Healthy Choice was being put back into production and subsequent distribution, were not made, with a consequent loss of profits on these sales.
(4)Advertising, prepared for Healthy Choice, (and which was also intended to support Traditional dinners) could not be shown until July 1991, after the peak selling period for the products.
(5)Arrangements made by McCain with supermarket chains to display Healthy Choice was disrupted and McCain's credibility as a reliable supplier was compromised.
Overall these matters were not is dispute.
Legge's model
Legge constructed two Bass-type models of the frozen dinner market, (separate models for Healthy Choice and McCain's complete frozen dinner range) which "assume[d] that entry into the market may follow either external or internal stimulus and demonstrated their adoption by purchasing one unit of product". Legge introduced a "consumption extension" to the models which involved the following assumptions.
(1)That having purchased the product, consumers would form an opinion of it and, if they found it satisfactory, would make a further purchase in the following month.
(2)That having entered the market, such a consumer would remain in the market and purchase another seasonally adjusted unit of the product in each month.
(3) That certain consumers would leave the market each month for exogenous reasons - this proportion of consumers was known as the "churn".
(4)That some consumers who had left the market may re-enter it - however, no distinction was made between such consumers and those who entered the market for the first time.
Healthy Choice model
Legge incorporated the following parameters into the Healthy Choice model:
(1)The size of the potential market including current users. This parameter acted as a limit on the impact of advertising expenditure on the model by computing the proportion of influence that was "wasted" by being exerted on committed users. In this way, the model did not permit sales beyond a specified volume of product.
(2)The churn rate before April 1993.
(3)The churn rate after April 1993. Legge said that the churn rate for Healthy Choice was quite high as compared to the rate for Traditional dinners.
(4)The advertising effectiveness co-efficient. To derive this co-efficient, Legge used actual historical data from 1991 onwards, 1991 data for the 1990 year and the same data, scaled back, for the years 1989 and 1988. This co-efficient was used to convert advertising dollars into monthly sales figures, so that partial external influence due to advertising was determined.
(5)The display effectiveness co-efficient. This co - efficient, which converted the number of facings per store into a dollar value for sales of product, demonstrated the apparent sensitivity of sales to the number of facings as reported by McCain's merchandisers. By using this parameter, the partial external influence due to display could be determined. Legge said that the data available to him indicated that Healthy Choice purchasers tended to be impulse buyers, much more than was the case with purchases of Traditional dinners.
(6)The satisfaction parameter. This parameter measured the average customer response to the use of the product before the re-launch and the average customer response to the use of the product after the re-launch. Thus the model was able to take into account the fact that, before April 1993, the product was not as attractive as it was after that date.
Legge calibrated the model with known data to find the parameter set that provided the best fit to the actual data, so as to estimate the probable market response under different volumes of advertising, distribution and product acceptance, using the Microsoft Excel "Solver" function. The information against which the Healthy Choice model was calibrated was: monthly warehouse withdrawals from March 1991 to September 1993, using seasonally adjusted SAMI data supplied by Layton; monthly media expenditure from March 1991 to October 1993, provided by McCain, seasonally adjusted using Layton's co-efficients; the monthly average number of facings, provided by McCain, seasonally adjusted using Layton's parameters.
Legge's model for McCain's complete frozen dinner range
Legge used 7 parameters in the model for the complete McCain frozen dinner range, defined as for the Healthy Choice introduction model but split over different periods. The effective dates used to describe the churn were: from January 1987 to June 1989; and from July 1989 to the end of the simulation. Two parameters were used to describe overall product satisfaction: one from January 1987 to March 1993, just before the Healthy Choice relaunch; and the other from April 1993 to the end of the simulation. The second satisfaction parameter was constrained to the same value as the corresponding parameter in the Healthy Choice model.
The information against which the complete product line was calibrated was as follows:
(1)McCain's record of factory shipments over the periods January 1988 to December 1990 and January 1993 to March 1994, seasonally adjusted using the method described by Layton;
(2) McCain's record of monthly media expenditure from January 1991 to March 1994 seasonally adjusted;
(3)Monthly media expenditure in previous years was taken to follow the same pattern and phasing as McCain's 1991 plan;
(4)McCain's records of monthly average facings for McCain's Traditional Dinner lines for the period October 1992 to March 1994 seasonally adjusted;
(5)Monthly average facings in previous years were to follow the same (almost invariant) pattern.
Legge constructed five figures or graphs (two for Healthy Choice, and three for the full product line) using certain spreadsheet tools to adjust the parameters to give the best possible fit to the data. The figures were attached to his report.
Figure 1 was a calibrated model for the introduction of Healthy Choice. It depicted actual deseasonalised data (for this purpose Legge used Professor Layton's coefficients). The model then found the best fit for those deseasonalised figures. This produced the coefficient of determination R2 of 98.8%. This meant that only 1.21% of the variation from an average level of sales could not be explained by the model using the selected parameters. According to Legge, a statistical test with such a high coefficient of determination suggested that the result could not be due to chance.
Legge next sought to predict the achievable outcomes for Healthy Choice in the absence of the injunction. Figure 2 graphically depicted the result and contained 3 predictions. Under the first prediction, the media expenditure and supermarket exposure planned by McCain before the injunction were restored and maintained at the same level through 1992. The simulation measured the direct effect of the injunction on McCain's Healthy Choice launch, but took no account of the disruption caused to McCain's marketing organisation by the injunction. Thus the effect on McCain's Healthy Choice sales was simulated on the basis that the only change was to reverse delistings which Legge believed occurred 1992. Legge undertook this task as he believed
that there were delistings and that they were a result of the disruption to distribution as distinct from any particular problem with the product. Legge later explained that by delisting he meant that, according to the merchandising data with which he was supplied, the average level of facings was down in 1992 over the level in 1991. He said that in his view, this reduction in facings was strong evidence of delisting. This produced the most conservative estimate of the loss sustained by McCain due to the effects of the injunction, with there being lost sales apparent in the period between about December 1991 and say, May 1992. On this basis, Legge predicted a revenue loss of $849,000 between April 1991 and December 1992. Legge revised this estimate in line with the co-efficient of determination as the model only explained 95% of the sales variance. The adjusted estimate under this prediction was $838,000 (prediction 1).
The second prediction assumed that McCain had made some improvements to its product because whatever deficiencies were in the product had been corrected by the start of selling season 1992, leading to a growth in sales over 1992. At this stage, Legge did not allow for a complete re-launch of Healthy Choice with the success and impact of the April 1993 re-launch. The same media and exposure parameters were used as in the first prediction. Legge then assumed that the packaging and product changes introduced in 1993, would have been introduced in 1992 if the injunction had not disrupted McCain's marketing organisation, but not the new varieties launched in 1993, nor the higher media expenditure incurred in that year. The loss thus
predicted was $1,561,000 between April 1991 and December 1992. He adjusted this figure in line with the coefficient of determination to arrive at a figure of $1,542,000 (prediction 2).
The third prediction reproduced in 1992 what in fact occurred in 1993 after the relaunch. It used the media and exposure figures from the first prediction until March 1992. Actual figures from a year later were used for the period to March 1993. Finally, actual figures from April 1993 were used. This prediction assumed that McCain's successful relaunch of Healthy Choice in April 1993 could have taken place a year earlier if the injunction had not been granted. In making this assumption, Legge expressed the opinion that the strong seasonal nature of the market was such that it was quite possible that a relatively small delay in planning an event such as the Healthy Choice relaunch could set the execution back by a full year. The loss predicted by this test was $14,119,000. As readjusted in line with the coefficient of determination, the loss of revenue was $13,946,000 (prediction 3).
Legge stated that modelling Traditional dinner sales presented a different set of problems to those solved while developing the Healthy Choice model. In part, this was due to the fact that Healthy Choice itself may have captured some Traditional dinner customers. Thus, Legge's approach was to fit a model to all the available data for uncontested periods. That is, from January 1988 to December 1990, after the initial product launch and before the Healthy Choice launch; and from January 1993 to March
1994 when the effects of the Healthy Choice launch and the injunction had largely warn off.
Legge then made various predictions in respect of the full range of McCain dinners. These were graphically depicted in figures 3, 4 and 5.
In figure 3 Legge attempted to find a best fit to the data that was unaffected by the injunction so as to isolate the effects of the injunction. For that purpose, the two years from January 1991 to January 1993 were taken out of the calculation, representing the period over which McCain considered it continued to suffer damage from the effects of the injunction. By this process, Legge sought to obtain a view of what might have happened to McCain in the absence of any event (in this case, the injunction) which might have affected sales. Figure 3 produced a coefficient of determination for the full product line of 86% using the Corel function in Microsoft Excel. Legge said that he used the Microsoft Corel function as he considered it to be more reliable than his own hand done calculations.
In figure 4, Legge undertook the same task using quadratic smoothing to minimise the uneven distribution of the product to warehouses. Legge explained quadratic smoothing as follows:
It means that you take each point and you replace it with half its own value plus a quarter of each of the outriders.
And by a point do you mean ‑ ‑ ‑ ?‑‑‑Each monthly sales report. So I'm not deseasonalising, I'm just trying to get out some of the month to month ups and downs. Again I'm not really in a position to comment whether figure three or figure four is more authoritative. In either case, the correlation in the 80 per cents is statistically very significant. Suggests that the line is not a bad fit to the data."
The coefficient of determination rose to 95% when a single level of quadratic smoothing was applied to the data.
Figures 3 and 4 revealed that actual sales were consistently below the trend line in 1991 and 1992.
Figure 5 depicted, on a seasonal basis, the magnitude of lost sales of Traditional dinners and Healthy Choice between April 1991 and December 1992. Legge described figure 5 as representing a minimally shocked situation taking into account only the delisting of Healthy Choice at the start of 1992. He said that if there were further effects due to the disruption in market planning, preparation and new advertising and new varieties, figure 5 would cease to be an adequate demonstration of the loss.
The model estimated a loss of $700,000 in 1991 and $1.4 million in 1992, totalling $2.14 million for Healthy Choice and Traditional dinners. Adjusted to reflect the co-efficient of determination, the estimated loss thus predicted was $1,770,000.
Legge concluded that the real damage to McCain in the period was in respect of Healthy Choice, although the injunction had some impact on Traditional dinners. Legge's approach therefore was to identify the Healthy Choice loss and by subtraction, ascertain the loss sustained in respect of Traditional dinners. By this process, Legge estimated the loss for Traditional dinners between April 1991 and December 1992 in the sum of $932,000.
Legge's estimate of McCain's total loss of sales
From figure 5, Legge determined that as a result of disruption
to McCain's advertising and distribution in 1991, McCain lost sales of at least $1,770,000 between April 1991 and December 1992. McCain also suffered loss due to disruption in their marketing department in 1991 caused by the injunction and its aftermath, and the further disruption in 1992 as they were diverted from product and category management by the need to repair their relationships with the major retail chains. Legge considered this in turn delayed the relaunch of the Healthy Choice range until April 1993. Legge acknowledged that there were problems with the Healthy Choice product as originally launched, in particular in relation to the vicar advertising, lack of product variety and a shortfall in customer expectation in respect of the product. Legge considered that the effect of the injunction caused these problems to persist longer than they otherwise would have. He said that a competent marketing executive (as he would describe Boyle) would have addressed the problems promptly, once they had become apparent. However, there was confusion at the end of 1991 as to the cause of the problems McCain was having with Healthy Choice. Legge considered that Boyle's response in not seeking to rectify these matters in time for the 1992 season was reasonable.
He considered that but for the injunction, there was a high probability that McCain would have remedied the packaging and advertising in 1992 so that the actual outcome would have resembled or exceeded prediction 2 in Figure 2. He was also of the opinion that there was a 50% probability that Mr Boyle would have achieved a complete relaunch of the Healthy Choice range by April 1992, so as to resemble prediction 3 in Figure 2. On this basis, Legge estimated that likely loss on the combined range of Healthy Choice and Traditional dinners to be $6,745,000. He arrived at this calculation as follows:
50% probability of $6,973,000
Prediction (3)
Additional 40% $610,000
probability of
prediction (2)
Deduct Prediction (1) ($838,000)
as counted with the
Full Line estimate
Total $6,745,000
In cross examination, Legge agreed that in an ideal world, the parameters which he used in his model would be "let free" for the purposes of constructing the figures produced by the model. He said that he had started with all the parameters free, but had progressively frozen the parameters, as he observed that the model tended to become unstable on certain values. Legge denied that he fixed the parameters so as to achieve the best fit. Rather, he said the best fit was obtained by the co-efficient of determination. He explained that the coefficient of determination cannot exceed 1 and he took no steps which would reduce it. He observed that if the model achieved a co-efficient of determination of 1, the data had been statistically explained in a perfect way. Provided the parameters used pushed the answer closer to 1, the model provided a statistically better explanation of the data than any other numbers chosen for the parameters. Legge considered that as he had obtained a co-efficient of determination of approximately 98%, he had produced a substantially accurate model. He conceded that as there were a number of parameters, there may be a number of solutions to the combination of parameters. However, he said that he had spent many hours trying various combinations and, as a matter of experience and judgment, considered that the figures at which he had arrived provided the best answer.
Garry Edser
Garry Edser is a senior consultant with SPSS Australasia Pty Limited (SPSS). SPSS is a wholly owned subsidiary of SPSS Inc., an American company which has developed software products designed to carry out statistical analysis and related functions. As Edser's brief was, essentially, to validate Professor Layton's work, it is convenient to deal with his evidence after having considered Professor's Layton's methodology.
CONAGRA'S EXPERT EVIDENCE
Professor Roger Layton
Conagra engaged Roger Layton (Layton), the Dean of the Faculty of Commerce and Economics at the University of New South Wales as its expert. Layton was the foundation President of the Australia and New Zealand Marketing Association, and, amongst other memberships, is a Fellow of the Marketing Association of Australia and New Zealand. He is a co-author of a number of books on marketing and has published numerous articles and papers. Layton does not have any practical experience in marketing. He has some experience in relation to monitoring the performance of products in the course of consultancy work with market research companies. However, he is familiar with the literature dealing with the various aspects of a new product launch.
Layton was asked to provide his expert opinion on:
whether the available data showed any loss of sales arising from the grant of the injunction;
if so, the likely range of the quantum of that loss; and
whether the projections used by McCain's witnesses to estimate lost sales were sustainable.
Layton considered that the most appropriate way of estimating lost sales of Healthy Choice and Traditional dinners, both during and after the period the injunction was in force, was to consider the whole of the data available from 1987 and try to discover, in that data, the impact of the injunction. This involved a "decomposition" of the data. It did not involve making projections.
This approach differed significantly from that used by Boyle, Le Hunt and Legge. For the purposes of estimating lost sales of Healthy Choice, Boyle compared actual sales to forecasts of sales prepared by him in 1990, based on sales of Conagra's Healthy Choice product in the US and concluded that the difference between the two was attributable to the injunction. For the purposes of estimating lost sales of Traditional dinners, Boyle, Le Hunt and Legge had taken available data up to the date of the injunction, prepared forecasts from that data and then compared the forecast with actual sales. Layton's approach involved an assumption that there was no impact from the injunction. Legge, in his detailed modelling, assumed that the injunction did have an impact on sales.
Layton pointed out that Boyle, Le Hunt and Legge's first analysis focussed on data for a period up to the commencement of the injunction and on the basis of that data, attempted to forecast the level of sales likely to have been achieved if the injunction had not occurred. He did not consider this to be a desirable approach in an ex-post facto analysis, either generally or in respect of the frozen dinner market. He said that the market was one in which competitive pressures as well as external events played an important role. In such circumstances, he considered that the imposition of an assumed long term trend in the face of actual data could produce misleading results.
Layton identified 3 factors which might cause a variation in the
volume of product sales over time:
irregular or random events such as disruptions in supply, price variations and advertising campaigns;
the seasonality of the product; and
underlying trends in the market.
He characterized the injunction as being a factor of the first type. He considered that, in order to identify and quantify its effect, it was necessary to isolate the impact of the injunction from other random or irregular events and from the seasonality and trend factors.
Layton employed a variant of the procedure called the Census II Decomposition Method (included as X11ARIMA in the Statistical Package for Social Science) to analyse the relevant data. That data included not only the SAMI data but also TART data in relation to advertising expenditure. The data revealed that there had been a rapid growth in the market from 1987 to 1990; that the market peaked in 1990, fell slightly through 1991 and 1992 and began to recover in 1993. Prices during that time had tended upwards until 1990, followed by a relatively stable price structure through to 1993.
Layton said that the method he employed was widely used for the decomposition of data series such as those under consideration here. He said it was the preferred tool in such exercises, because it was more flexible in its assumptions than procedures such as simple or exponential smoothing. He also said it was more sophisticated than the procedures used by Le Hunt and by Legge. This criticism of Legge however, was made prior to Legge constructing his Bass type models.
Layton explained that the Census II method involved the following steps to arrive at a multiplicative decomposition of a data series:
(1)An initial estimate of seasonal factors, based on a 12 month centred moving average.
(2)The elimination of unusual or extreme values for seasonal factors, such as strikes or once only promotions, to obtain an average seasonal factor for any month so affected.
(3)The calculation of a first stage seasonally adjusted series, which is then tested for residual seasonality.
(4)This series is then smoothed, using a weighted moving average so as to eliminate randomness.
(5)The irregular or random component is estimated by dividing the seasonally adjusted series (step 3) by the smoothed series (step 4).
(6)Final estimates of the seasonally adjusted series and the irregular component are obtained using the smoothed series in step 4 in place of the moving average in step 1 and repeating the earlier steps.
(7)Tests were then carried out to determine if the decomposition of the original series had been successful.
Using this procedure, Layton obtained an estimate of trend and seasonal factors. He observed that there was a long term decline in the total frozen dinner market, following a peak in September 1990. The decline continued through March 1991 to October 1991. This was followed by a slight rise in the seasonally adjusted data in late 1991 and a further decline in the first part of 1992. In 1993, the seasonally adjusted data showed an upward trend.
Layton identified August as the peak selling month for the frozen dinner market, followed by May and then by July. He said there was no clear single peak in the winter season. He also observed that the seasonal pattern was changing slowly. In particular, the early months of the calendar year were growing in relative importance. The traditional winter months of June, July and August were declining. He concluded:
"[in] simple terms...frozen dinners appeared to be losing their seasonal character".
Layton had concluded that McCain did not suffer any loss beyond the period of the injunction and that the only loss suffered was in respect of Healthy Choice. However, he subsequently amended this and accepted that McCain would have suffered a loss for a period of about a month after the injunction.
For the purpose of coming to his conclusion, Layton said that he assumed that Healthy Choice had achieved a high level of trade
acceptance; that planned advertising would be at an adequate level, which he assumed would be high; and that there would be a negative reaction from consumers when the product was suddenly withdrawn. However, he had not used, and said that it was not necessary to use, any precise data as to these factors because he was not constructing a model using numbers of that kind.
Layton agreed that there was an advantage to a manufacturer in being the first supplier of a new product to the market, with the product being supported by advertising leading to the peak season and in having the product remain in the market. However, he did not consider that the absence of these factors necessarily meant there would be an impact on sales beyond the period of the injunction.
He also agreed that the continued availability of the product was important in obtaining repeat purchases and that repeat purchasers were critical to a product's performance. However, he said that in this case, he was driven by the market data which showed that the product initially achieved its expected level of market share. Accordingly, he concluded that the only effect of the injunction was that McCain would have received the expected level of sales in April\May rather than in July\August.
Layton considered that, as the sales data after Healthy Choice was reintroduced onto the market in June 1991 showed that the product reached very close to the market share McCain had estimated, the injunction had only a very slight effect on sales. He said that he interpreted the subsequent fall in market share as being due to consumer resistance, related to the product or concept, rather than delay caused by the injunction.
Layton had assumed that as the product was a frozen food product, there would have been a very short time lag between the time the injunction was dissolved and the time that the product was back on the supermarket shelves. However, he conceded that if it was a three to four week lag, the effect of the injunction would continue for that period.
Layton also undertook an analysis of other major brands in the frozen food market to identify the role of seasonality and trend in their sales performance. Both Griffs and Nestle exhibited a rise in sales, peaking in the 1990 winter season, followed by a decline to a lower level of sales in 1991 with a stable trend until 1993. Layton said the decline was sharp in the case of Griffs and that Nestle had had a slower decline. The same overall trend was exhibited in Nestle's Lean Cuisine sub-brand. He considered this product was of interest, as it was directed to the "diet/low calorie" portion of the market, which he considered was comparable to Healthy Choice. He noted, however, that Lean Cuisine had a different seasonal pattern to that for the frozen dinner market as a whole. In particular, Lean Cuisine commenced its upward trend in February rather than March, as was the case with frozen dinners. He considered that this difference may have reflected consistent promotional activity by Nestle early in each calendar year. It might also have meant that
seasonality was not as important a factor in products directed to the "health conscious".
Layton considered that the absence of a clear trend of trial and repurchase and the diminished seasonality observed in relation to Lean Cuisine suggested that frozen dinners directed to the "health conscious" were less likely to have a trial period. Layton drew this conclusion on the assumption that Lean Cuisine and Healthy Choice were likely to behave in a similar manner.
In cross examination, Layton agreed that the point he sought to make in relation to Lean Cuisine was that there was a difference in seasonality between Lean Cuisine and frozen dinners generally. He also agreed that he had treated Lean Cuisine and Healthy Choice as comparable products and the Lean Cuisine experience might be applicable to Healthy Choice. The point he sought to make therefore was that the seasonality of a product could be changed by promotion and positioning of the product in the market. Layton also agreed that when he stated in his report that advertising of Lean Cuisine between February and May:
"...may also provide evidence that the traditional perception of the seasons is less applicable to products directed to the health conscious than it is to other products".
He regarded Lean Cuisine and Healthy Choice as similar products. However, he was not prepared to concede that if his assumption was wrong, that would affect his conclusions as both products were directed at specialist market segments. However, there was
no evidence to establish that Lean Cuisine and Healthy Choice were in fact comparable products.
Layton agreed that there were advantages in having as many consumers as possible trialling the product in the initial peak period. He accepted therefore that to this extent the injunction had an impact on sales. However on the basis of the historical data, he doubted whether consumers who would have tried the product early in the season would have tried it again in the second peak period.
Layton observed that McCain had reached its historical peak sales in 1989, a year earlier than the total market and the other major brands. Its sales then tended downwards throughout 1990, the year in which the other brands were peaking. Its sales had a secondary peak in 1991, a further downward trend in 1992 and then a substantial recovery in 1993. He considered that the 1989 peak may have been due to a temporary reduction in the average price for McCain. He considered that the reduction in price may have reflected McCain's recognition that there was one of the factors affecting the performance of Healthy Choice was its price.
Layton's assessment of lost Healthy Choice sales
Layton accepted that McCain lost sales during the period the injunction was in force, that is during April, May and June 1991. Layton estimated that if it was assumed that sales would have grown from their $60,000 March level by $50,000 per month to a peak level of $262,000 in July 1991, the loss was no more than
$317,000. Layton concluded that the ineffectiveness of the marketing concept together with the difficulties identified by Sutherland Smith in its December 1991 report, was the cause of the losses sustained by McCain in the period outside the injunction.
Layton's assessment of lost sales for Traditional dinners
Layton's analysis using the Census II method revealed that there had been a significant shift in the seasonal pattern of sales of Traditional dinners away from the middle months of the year and towards the first four months. He considered that this was not due to any promotional strategy, as McCain's advertising had, traditionally, been directed to the period May to September in each year. In addition, he said there was a long term downward trend in 1991 in the sales of McCain Traditional dinners. Layton considered that the injunction may have affected the sales of Traditional dinners in one of two ways. First, the delay in advertising may have delayed the onset of the slowing effect. Secondly, there may have been long or short term adverse effects on sales of Traditional dinners. However, he said that the data showed little evidence of any long term shift in the sale of frozen dinner products commencing around the time of the injunction. Nor did the data indicate any long term shift commencing during and after the injunction. Layton accepted however that although there was no evidence of long term effect, it was possible that there may have been a short term effect.
Using both the Census II and the Winter models, Layton
ascertained that there was a drop in sales in the order of $101,800 on the Census II methodology or $152,000 on the Winter methodology. However, Layton was of the opinion that the low result in June was as consistent with the operation of random factors characteristic of the behavior of the market as with the effect of the injunction. He concluded therefore that the range of possible lost sales was between nil and $101,800.
Layton considered that there was only a possibility that the earlier commencement of the advertising might have slowed the decline. This view was based upon the fact that there was a slight recovery in the overall market in this period. In addition, whilst there was an improvement in sales in July and August 1991 at the time when the advertising was again in place, that improvement was not sustained in September and October. Likewise, in 1992, sales in May were lower than the projected level but June and July were up. Layton considered that if the trend adjustment had begun earlier with earlier advertising, the estimated loss was $185,100.
During the course of cross examination, Layton made a number of concessions. For example, he had stated in his report that one of the difficulties Healthy Choice had experienced was a potential price barrier for new purchasers. He said that the:
"- gradual reduction in case prices by McCain from 1991 through to 1993 may reflect its recognition of this problem and it's attempt to overcome it."
He agreed that he meant to convey by this that McCain had lowered its prices. He agreed this was speculative and that if his premise was incorrect his inference was untenable. He ultimately conceded that this was a false premise. After some further urging, Layton withdrew the statement in the report.
Layton had stated that the seasonality of McCain products had shown a greater variation than other products in the market. However, in coming to this conclusion, Layton had used ex-factory data and had not made the appropriate adjustment for this when identifying the seasonality of the product. Layton was thus forced to concede, as I have commented upon earlier that, contrary to what he had said in his report, Boyle had correctly analysed the seasonality of McCain traditional dinners and that seasonality for Healthy Choice began in April and ended in October. Layton had also stated that over time, McCain products were becoming less seasonal. He was forced to recognise that in coming to this conclusion, he had used certain incorrect data.
Layton had concluded from the data that the pattern of McCain advertising was generally to advertise from May through to September in each year. This general conclusion failed to specify that in each year, McCain had commenced advertising in March, and that at least by April, the advertising was substantial. Layton was not prepared to make any concession on this matter. However, I consider that Layton's statement, overlooking as it did the substantial advertising by April in each year, was an unacceptable gloss on the true position.
Layton's approach in relation to Healthy Choice involved constructing a graph on the basis of actual sales for March. He then drew a straight line graph between sales in March and July, assuming that July represented the peak in sales. The straight line assumed equal increments of sales in between those months. Layton agreed that this approach was based on his view that he was entitled to extrapolate the results from 1992 actual data (as he had graphically depicted them) to graphs of results of 1991. He denied this approach was circular. He said:
"...My understanding of the whole pattern of the data, and I think it is important to look at the whole data, all of the data available to us, was that in 1991 and in 1992 the sales figures were very similar. If anything 1992 was slightly down on 1991, but that there had been a radical change in the nature of the advertising, in the product range and in the packaging in 1993. It was the fact that there had been a radical change that suggested that that was what was needed in order to bring success to this particular product line, that the 1993 figures showed that quite clearly. 1991 and 1992, the reason for believing that they were the same was that very little had happened in terms of the underlying market offer, to suggest that there had been any variation between the two years. Fundamentally the two years were offering - were made up of fundamentally the same basic product line advertising with some variation in packaging.
...
...I looked at the 1992 graph and noted that it was similar to 1991 and that would have been one factor in believing there was no loss other than in the first few months of 1991, then looked at the 1993 figures and noted the substantial change that had taken place there, that in fact the whole market offer had been changed, which suggested the underlying problem in the data, in that McCains was experiencing, was not the injunction, but the nature of the market offer that was surrounding Healthy Choice in 1991 and 1992. As they addressed that the problem went away."
However, this process of extrapolation was flawed in a number of respects. In the first place, Layton used different periods. In the case of 1991, he used actual sales data from June to September. However, in 1992, he used actual sales data from March till June 1992. Secondly, Layton did not take into consideration the possibility that the effect of the injunction extended beyond June 1991. If the effect of the injunction did extend beyond June 1991, Layton conceded that the similarity in the graphs for 1991 and 1992 may have been due to the effects of the injunction.
In drawing a straight line between March and July on the basis that he was trying to estimate what effect on ex factory sales would have been if the injunction had not been present. It was implicit in his approach that the decline in the graph after July 1991 was attributable to a decline in consumer sentiment towards the product. However, Layton was forced to concede that the peak in the graph he had drawn was substantially due to the fact that there was no product in the warehouse. He conceded that, using SAMI data, the peak would move to August and would reflect the fact that product was moved from warehouse to fill supermarket shelves. He also conceded that it would have been more accurate to use SAMI data to assess consumer purchases.
Layton had questioned whether Conagra's experience in the US was transferable to Australia. However, he agreed that he had no practical experience upon which to draw in making any assessment on that issue. He agreed, however, that promotional activity would be one factor which would influence whether the US results would be transferable. He accepted that McCain's advertising was appropriately timed.
Layton agreed that he had come to his conclusion as to the effect of the injunction in part, by reference to the views expressed by Sutherland Smith. It followed that in doing so, his conclusion was dependant upon his having correctly understood their material. I refer to this further below.
Layton assumed for the purposes of his approach that the injunction had no effect on the sale of Traditional dinners. He said having made this assumption, the X11 Arima process which he used would reveal whether there was any distortion or residual element unexplained by this assumption. However, Layton agreed that if, as was the case, actual data after the date of the injunction was used, it was not surprising that a smoothing, by means of the X11Arima process would substantially reproduce actual sales. He maintained however, that the X11 Arima process would have thrown up any discernable effect from the injunction.
Layton interpreted a decline in consumer sales after August as referable to a decline in consumer sentiment. However, in coming to this conclusion different concepts were compared. Layton sought to match a peak which was not referable to consumer sentiment, to a decline which was referable to consumer sentiment. Secondly, his interpretation failed to take into account the seasonality of the product and therefore the ability to attract repeat and subsequently brand loyal customers.
Criticisms of Layton's Report
Boyle criticised Layton's report on a number of bases. He said that Layton failed to take into account the fact that the decline in the frozen food market in 1991 was likely to have been due, at least in part, to the absence of McCain advertising in the vital periods of seasonal demand. Boyle said this was of particular significance as McCain had a 25% market share of the frozen dinner market at that time, second only to the Nestle/Findus brand dinners. Boyle expressed the view that the absence of McCain advertising had a depressing effect, not only on McCain sales but on the market generally. He considered the absence of advertising was as much a "once off shock to a system" for which it was "reasonable to adjust" as was actual advertising. Accordingly, Boyle was of the view that Layton erred in treating as appropriate market data the historical data which itself had been distorted by the injunction.
Boyle also produced a series of tables which indicated Layton had made a number of errors of transcription of the TART data supplied by McCain to Conagra. Layton did not accept this was so, but conceded that the material may have changed or the numbers may have been "misobserved". He later admitted that there were at least two errors in the data. He also said there had been gaps in the data supplied by McCain.
Next, Boyle pointed out that Layton had erred in the conclusions he drew from the SAMI data. In particular, he said that Layton
failed to recognise that SAMI data is based on retail selling price (including trade margins) whereas McCain sales data is based on manufacturers' sales price "ex factory". This error was apparent in Layton's statement that the difference between the two sets of data "relates to timing, inventory effects, transfer pricing and some problems in measurement". Layton also failed to recognise that there was a lag factor of about 1 month between the McCain ex factory sales data and the SAMI data. This led Layton to conclude, wrongly, according to Boyle, that the season commenced later than that estimated by Boyle. Layton again conceded this point.
It will also be recalled that Layton conceded his error in placing the critical selling period in August and June, rather than May. Boyle contended that both on his analysis, and by correcting Layton's approach on the Winter and Census II methods, the upswing occurred in March/April for ex factory sales which was the period during which the injunction was granted. Layton also failed to take into account the need to introduce a seasonal product at a time which is optimal to obtain the benefit of the upswing in the season.
Boyle contended that Layton had failed to read the 1992 Sutherland Smith Report as a whole when he relied upon the statement that research suggested "there is no clear pattern of trial and repeat and that the trend is for trials in both March and July/August each season". He pointed out that Layton either overlooked or, in any event, omitted that part of the report which stated "Findus always gets a head start. It is crucial that we rid ourselves of this handicap and fight hard for market share in the February/March period each year", together with other material in the report which dealt specifically with the need to have a strong presence in the market both in March and July/August of each year.
Boyle was also critical of the "aggressive smoothing" which Layton had applied which Layton said "preserved the seasonality and trend evident in the data but removes the random elements." Boyle expressed the view that although the monthly peaks had been smoothed, Layton had substantially and consistently overestimated sales in the December/January trough. The effect, according to Boyle, was to suppress the seasonality range. In particular, Boyle was critical of Layton's view that the December/January trough contained "a substantial random or irregular component". The forward smoothing technique which Layton used meant that his conclusion that the frozen dinner market was losing its seasonal character was suspect. In any event, according to Boyle it was irrelevant, because any loss of seasonality was minimal in the relevant years. Further Layton's own analysis (see his table 3) showed that during the period 1988 to 1993, March was becoming more important each year.
Boyle then identified an error relating to McCain's product pricing. On the one hand, Layton said that from 1990 onwards, McCain's prices had remained relatively stable. He also stated that prices were lowered in 1992 and 1993 and concluded that this gradual reduction might reflect McCain's recognition that its prices were too high and was an attempt to overcome it. Boyle pointed out that these statements were inconsistent. In any event, McCain did not reduce it list price ex factory for any product in its frozen dinner range during the period 1991 to 1993, except for a minor change in Traditional dinners in Western Australia to effect a national pricing policy. Boyle concluded that Layton's view that prices had dropped may have come from a misreading, or a misunderstanding of McCain's data. In particular, he may have failed to recognise that SAMI data would include McCain's lower priced products such as the Budget Gourmet and Looney Tunes range. The apparent reduction in price may also have reflected the fact that some retailers, such as Franklins, are low price operators, so that their sales would have effected a drop in the average price recorded in the SAMI data for McCain products despite the absence of change to McCain list prices. Further, Layton failed to recognise that promotional activity may cause a drop in the average price reported in the SAMI data, or that as products become more successful at a consumer level, grocery retailers become more competitive in their product selling strategy and product pricing, resulting in reduced margins and lower shelf prices.
Boyle is next critical of Layton's assessment that for a 4.4% market share to have been achieved in 3 months, the concept and its implementation would need to be highly effective. According to Boyle, this ignored the fact that within 3 months of the injunction being dissolved, Healthy Choice had achieved that share. It further overlooked the fact that many supermarket grocery buyers only allow a 13 week period for new products to perform before reviewing them for possible delisting. Boyle expressed the view that, the fact that this market share was achieved quickly but not maintained, confirmed McCain's claim that the initial trial at this time of year was not converted to an adequate level of repeat purchase, thereby limiting the development of loyal users. He said that the reason for this inadequate level of repeat purchase was not clear to him at the end of 1991 as they were not differentiated in the McCain market research report prepared by Sutherland Smith at that time. Sutherland Smith had identified several key limiting factors including the late introduction of Healthy Choice. He said that it was not until well into the 1992 season that sufficient information was available from Sutherland Smith research to enable McCain to make appropriate corrective action.
Finally, Boyle criticises Layton's approach for failing to take into account the fact that McCain was denied the opportunity to take such corrective action in time for the 1992 season.
Legge's criticism of Layton
Legge expressed the view that whilst Layton's technique of drawing a straight line between the sales of Healthy Choice in March 1991, that is of sales before the injunction was granted, and the sale achieved by McCain in July 1991, the first complete month after the injunction was discharged, was somewhat unsophisticated, the result was little different from what he had achieved on his own model. However, Legge criticised Layton's assumption that the injunction had ceased to have any effect a month after the injunction was discharged. He noted that Layton acknowledged that the effects of the injunction could have lingered till November 1991. However, he said that, even so, Layton had not paid regard to the effect of the volume or phasing of advertising or the extent of distribution on the sales outcome, notwithstanding that Layton, in his book Fundamentals of Marketing, had stated that these factors did affect sales outcome. Legge accepted the validity of Layton's text book approach, stating it accorded with his practical research and teaching experience.
Legge next deflected the criticism Layton had made of his first report and of Boyle's and Le Hunt's assessment of lost sales. He agreed that calculations based on past data or 'extrapolations' as he referred to them as he, Boyle and Le Hunt had relied upon, are subject to a greater degree of error than interpolations, where the missing data points lie between known data points. Legge accepted that McCain's sales had been substantially unaffected by the injunction since January 1993. Legge considered therefore that data from January 1993 could be used as the base for one end of an interpolation and data before April 1991 as the other. He said he had used the data in this fashion in his second report.
However, Layton had used data from the challenged period to determine the size of the revenue shortfall in the challenged period. He considered that this necessarily led to an underestimate of the discrepancy. In addition, if the shortfall was a small percentage of the total, the overall discrepancy could be underestimated, especially late in the challenged period when the acknowledged short fall early in the period may have established a declining trend.
Next, Legge observed that Layton assumed that because there were problems with the launch itself, those problems accounted for the lost sales. Legge said Layton failed to recognise that both factors could be operating together. Layton also failed to take into account the fact that McCain failed to take corrective steps in respect of the product because of the confounding effect of the injunction.
Garry Edser
As stated above, Edser was engaged to validate Layton's results. He was not required to comment upon the validity of Layton's methodology. Edser accepted the correctness of the data used by Layton. Edser also replicated the analysis that Layton had done. Finally Edser undertook his own analysis. In doing so, he used historical data up until the date of the grant of the injunction and then used a 95% confidence limit for the purposes of forecasting sales in the period after the grant of the injunction.
Subsequently, Edser was requested to do a forecast from April 1991 to December 1992 on Traditional dinners based on actual data for the period of 4 years and 4 months up to and including March 1991. According to Edser, X11Arima was not suitable for this purpose, as it required a period of 5 years of data. Accordingly, Edser used general Arima and then applied an exponential smoothing technique.
According to his forecast, sales of Traditional dinners should have been higher from the period of the injunction.
McCain's case
I have already referred, at the commencement of these reasons, to the principles which govern the award of damages in this case.
Loss sustained in respect of Healthy Choice
There was no dispute that McCain sustained loss in respect of Healthy Choice during the period the injunction was in force. The real question in the case is whether it sustained loss beyond that period which was attributable to the injunction and whether in sustained any loss in respect of its Traditional dinner range. In large measure, the question of whether McCain sustained such further losses depends upon whether it was reasonable for Boyle not to react to the information contained in Sutherland Smith's December 1991 report. I was impressed with Boyle as a witness. He appeared to me to be straightforward in the manner he gave his evidence. I consider that he had a basic working understanding of principles which governed the making of market projections, although he was not able to use the more sophisticated techniques used by the expert witnesses. More significantly, he impressed me as having a soundly based "hands on" understanding of the market place. Conagra did not call any evidence of a marketing kind to contradict Boyle's evidence or to explain why Boyle's reaction was not a reasonable reaction given the dynamics, both of marketing and of the marketing place. Although Boyle was cross examined extensively as to his failure to take action in 1991/2, I considered that his explanation for failing to do so demonstrated that his "wait-see" response was reasonable in all the circumstances.
Having reached that conclusion, the question becomes what is the quantum of loss sustained by McCain. This answer to this question, in turn, depends upon which expert, if any, ought to be accepted or whether Boyle's own assessment ought to be accepted. I do not consider that I should accept Boyle's assessment. Even on McCain's own case, it was obtained using a methodology that was "a bit rough and ready". Likewise, I do not consider that Le Hunt's estimate for Traditional Dinners should be accepted. Although his estimate was more sophisticated than that used by Boyle, it was more limited than that utilised by Legge. Before dealing with Legge's evidence, it is convenient to deal with Layton's evidence.
Counsel for McCain submitted that Layton's evidence should not be accepted on a number of bases. It will only be necessary to deal with some of them. First, it was submitted that the evidence was logically compelling that the effect of the injunction extended beyond the period of the injunction. Factors which compelled this conclusion included the fact that McCain lost the benefit of being the first entrant in the market and the fact that it missed the important first seasonal upswing, and thus lost the opportunity of obtaining repeat and brand loyal purchasers. It was submitted that a conclusion to the contrary ignored the fact that the product when initially launched had a high level of trade acceptance and also ignored McCain's contemporaneous assessment of the impact of the injunction. I consider that these submissions are well based and reinforce the reasonableness of Boyle's decision to continue, in 1991, with the same product and market format for 1992.
It was also submitted that Layton's estimate was based on a number of errors and misconceptions. I have referred to these in the course of considering Layton's evidence, and it is not necessary to repeat them. It is also significant, in my opinion, that there was no real rebuttal of the criticisms which Boyle made of Layton's report. Boyle based those criticisms on real data or his own opinion as a person with extensive practical experience in the market. I consider that his criticisms were well based.
I am also of the opinion that Boyle's original launch plan, based on annual sales of 96,000 dozen cases, or a market share of 4.4% was achievable, as was demonstrated by the initial reaction to the launch, the fact that Healthy Choice reached its market share even before the relaunch, as well as its 1993 relaunch sales
history. The market history of la Del'Lite also supports this assessment.
Overall, I was not impressed with Layton's evidence, suffering as it did from the errors and criticisms to which I have referred. Although he said in re-examination that his various concessions did not affect his opinion, he was not requested to explain why this was so. I am of the opinion that, given the extent of error and misconception in his report, his opinion should not be accepted for the purposes of the assessment of damages in this case.
That leaves the evidence of Legge and Edser. Although Layton put forward a number of criticisms of Legge's evidence, Legge's underlying assumptions and methodology were not effectively challenged. The same may be said of Edser. However, given Legge's market experience in addition to his experience and expertise in statistics, I am of the opinion that his evidence should be accepted rather than Edser's, which was based on a purely statistical approach. The same may be said of Le Hunt's evidence.
Having said that however, it is necessary to determine which aspect of Legge's evidence should be accepted. It will be recalled that in figure 5 Legge assessed the lost sales for Healthy Choice and Traditional dinners in a total sum of $1,770,000 over the period April 1991 to December 1992. It will also be recalled that in figure 2, Legge set out 3 predictions. He did an assessment of the likelihood of each prediction being correct and expressed that likelihood in percentage terms. He made this assessment on the basis of his own experience and assessment of Boyle's marketing abilities. I agree that Boyle would, more likely than not have introduced changes to the product and marketing concept for the 1992 season so as to deal with the problems raised by Sutherland Smith. This was a reasonable inference to draw from the evidence and a reasonable assessment of what Boyle would have done but for the injunction. I do not agree however, that, but for the injunction, he would have undertaken the same complete relaunch of the product in 1992 as was undertaken in 1993. Sutherland Smith's recommendations in 1992 were more extensive than in 1991. I do not consider that there is any basis for accepting that Boyle would have reacted to more than the problems identified in the Sutherland Smith report.
It seems to me therefore that the assessment which I should accept is Legge's calculation represented in figure 5. It is true that Legge described this as a minimally shocked situation and which would not be appropriate if there were further effects of the injunction beyond delistings. Whilst that might be the case, this figure is not significantly different from the assessment of lost sales made by Boyle, which McCain's counsel submitted should be accepted. However, Boyle's assessment did not survive without criticism, even by Legge. It seems in that circumstance it is appropriate to use the more expert calculation arrived at by Legge. Accordingly, I consider that amount of lost sales which McCain suffered due to the effects of the injunction is $1,770,000.
Other losses
McCain also suffered other loss for raw ingredients, already manufactured product and the like. This loss was agreed in the sum of $55,000.
Interest
It was submitted that interest ought not to be awarded for the period 3 July 1992, when the inquiry into damages was awarded and 28 May 1993, when McCain filed its first affidavits.
The evidence obtained in the matter was technical and complex. A reasonable time would have to have been allowed for its preparation. It cannot be said therefore that the matter did not advance at all between 3 July 1992 and 28 May 1993. On the other hand, I do not consider that, in the absence of any explanation, 11 months is a reasonable time. McCain did not call any evidence to explain the failure to file its evidence earlier.
Accordingly, I can only approach the matter in a broad based fashion, taking into account the nature of the evidence which was filed and my experience in the court of how long such matters generally take. On this basis, I consider that it would have been reasonable for the affidavits to have been filed by the end of October 1992. I consider therefore that interest should not be payable for the period 1 November 1992 to 27 May 1993.
Taxation
In his written submissions, counsel for Conagra raised the issue of the adjustment (if any) which ought to be made for taxation, particularly as the company rate of taxation had changed from those which applied if McCain had earned the profits from the lost sales in 1991 and 1992.
However, in his closing address, counsel announced that the parties had agreed that the matter for my determination was the quantification of the lost sales. It was agreed that computation of the final amount payable should be left to the parties' agreement.
Accordingly, the appropriate course for me to take is to declare the amount of lost sales that I have determined McCain sustained as a result of the injunction and to make other consequential declarations reflecting my reasons.
I certify that this and the preceding 72 pages
are a true copy of the Reasons for Judgment
of the Honourable Justice Beazley.
Associate:
Dated: 23 April 1996
APPEARANCES
Counsel for the Applicant: J M Ireland QC &
R JWebb
Solicitors for the Applicant: Baker & McKenzie
Counsel for the Respondent: R C Macaw QC &
B J Hess
Solicitors for the Respondent: Alfred Tatlock
Dates of hearing: 28, 29, 30, 31 August 1995 and 1, 4 September 1995
4
4
0