Triumph Controls - UK Ltd & Anor v Primus International Holding Company & Ors

These proceedings concern claims for breach of warranty in respect of a share purchase agreement entered into by the claimants (“Triumph”) and the defendants (“Primus”) on 27 March 2013 (“the SPA”).

On 11 March 2019 the Court handed down judgment for Triumph against Primus (“the Judgment”).

At paragraph 498 of the Judgment, the Court’s conclusions were set out as follows:

Further directions were given for the parties to adduce additional expert evidence and this hearing was fixed to determine the outstanding quantum issues, together with any consequential matters arising out of the Judgment.

The exercise the Court required the experts to carry out was to determine the discounted cash flow (“DCF”) value of the Primus companies at the date of the SPA based on the properly adjusted long range plan (“LRP”).

At paragraph 470 of the Judgment, the Court found that a carefully prepared LRP should have included the following adjustments:

On 29 March 2019, Mr Fisher, forensic accounting expert for Triumph, and Mr Dearman, forensic accounting expert for Primus, produced their second joint memorandum on the quantum issues.

In their second joint memorandum, the experts agreed the appropriate valuation methodology to determine the DCF value of the Primus companies at the date of the SPA, namely:

Mr Fisher and Mr Dearman disagree on the amendments that should be made to the LRP, using the above methodology, so as to determine the DCF value of the Primus companies. They have produced the following reports:

Mr Pillai, counsel for Triumph, relies on the opinion of Mr Fisher that the Amended LRP reduces the DCF value of the Primus companies by US$ 13,359,535. Triumph’s position is that such change in the DCF value would have reduced the purchase price it would have paid by an equivalent amount and thus represents the quantum of its loss. Allowing for the contractual deductible of US$ 1.5 million, Triumph claims an award of damages in the sum of US$ 11,859,535. Triumph also claims interest on the damages at the rate of 4.5% per annum.

Mr Morgan QC, counsel for Primus, relies on the opinion of Mr Dearman that the Amended LRP reduces the DCF value of the Primus companies by US$ 428,830. Primus’ position is that neither the reduction in value assessed by Mr Dearman nor that assessed by Mr Fisher would have had any impact on the purchase price that Triumph would have been willing to pay because such values were in excess of the total price paid by Triumph of US$ 76.5 million. Therefore, Triumph has not suffered any loss as a result of Primus’ breach of warranty and is not entitled to any damages.

It is common ground that any damages are subject to the contractual deductible of US$

The matters for consideration by the Court are:

The first adjustment required to the LRP is an allowance of three months accelerated production to build buffer stock on each platform that had not been transferred at the date of the LRP.

Although the LRP summarises planned and actual transfers on a platform level, the underlying spreadsheets break down each platform to show production at a product level. In their second joint memorandum, the experts agree that adjustments to the LRP should be made on a product basis.

The required adjustments to the LRP are:

The experts agree that at the date of the LRP there were twelve outstanding product transfers from Farnborough to Thailand, of which:

The experts agree that the adjustments to the LRP for buffer stock require amendments to the timing and location of production hours. The experts agree that the quarter prior to transfer would be the period when buffer stock would be built. The LRP should include additional production hours required to build buffer stock in these quarters in Farnborough. Following the period in which such buffer stock was built, while production capability was being established in Thailand, there would be an equivalent reduction in production hours in Thailand, during the period when the buffer stock would be used.

The experts disagree on the number of products requiring buffer stock, and the timing and sequencing of the revised transfers.

As to the number of products requiring buffer stock production, the issue between the experts is whether the three-month period to build buffer stock should be applied to all twelve outstanding products or limited to the five products planned for transfer in 2013.

Mr Fisher has adjusted the LRP to allow for the time and cost of building buffer for each of the twelve outstanding products not transferred at the date of the LRP. Mr Dearman has adjusted the LRP in respect of the five products that were planned for transfer in 2013 but not the outstanding transfers planned for 2012 or after 2013.

Mr Pillai submits that Mr Fisher’s approach reflects paragraphs 468 and 470 of the Judgment. Mr Morgan submits that the Court should accept Mr Dearman’s approach because towards the end of 2012 total production transfers were ahead of schedule and therefore no additional buffer was required. Further, Primus had ample time to obtain duplicate tooling to avoid the need for buffer stock for transfers after 2013.

The finding made in the Judgment was that the LRP failed to make allowance for building a buffer in respect of each outstanding platform, save where duplicate tools were available or a direct transfer was made. Each of the twelve products identified by the experts was outstanding at the date of the LRP, none was a direct transfer and none had duplicate tools available. The argument that no buffer was needed for the outstanding products because transfers of other products were ahead of schedule was considered and rejected by the Court at paragraphs 445 to 452 of the Judgment. No evidence was produced by Primus to demonstrate that duplicate tooling would be made available for the subsequent transfers or that any allowance for the costs of acquiring the required tooling was included in the LRP. On the contrary, the contemporaneous documents produced by Primus in 2012 simply indicated that the relevant products did not have duplicate tooling and buffer was required. Therefore, allowance should have been made in the LRP for the time and cost of building a buffer in respect of each of the twelve outstanding products prior to transfer.

As to the timing and sequencing of the transfers, the issue between the experts is whether the LRP transfer schedule should be amended before incorporating the threemonth buffer period and, if so, whether any changes should be made to the sequencing of the transfers.

Mr Fisher has adjusted the LRP transfer schedule to reflect the delayed start to the four planned transfers outstanding at the end of 2012, keeping the original sequencing logic for all subsequent transfers. Firstly, he has made an adjustment to the transfer schedule to show the true status of the four transfers that remained outstanding at the date of the LRP, by moving the starting point for transfer of each of those products from 2012 to 2013. In making that adjustment, he has kept the original logic links in the transfer schedule, thereby delaying all subsequent transfers by an equivalent period. Then Mr Fisher has made a second adjustment, by adding a further delay of three months to the start of each of those transfers, to reflect the need to build buffer stock.

Mr Dearman has moved the four outstanding products from 2012 to the first quarter of 2013, without any buffer or transfer period. He has adjusted the transfer schedule by adding a three-month buffer period to each of the five product transfers planned for 2013. He has made no adjustments for the other products.

Mr Pillai submits that Mr Fisher’s approach reflects the Court’s finding at paragraph 460 of the Judgment that although the financial output of the LRP for the first three quarters of 2012 was manually overwritten with actual financial results, the underlying assumptions were not updated for future transfers. As a result, the LRP wrongly assumed that for 2013 onwards, all transfers planned for 2012 had in fact occurred and no adjustment was required for the 2013 transfer figures. The original transfer sequence for all products must be retained to avoid compression of the programme. There is no evidence that the Farnborough facility had sufficient capacity to accommodate the increased production hours calculated by Mr Fisher within a compressed programme. Further, Mr Pillai submits that Mr Fisher’s approach is consistent with the supplemental joint memorandum in which the experts agreed that they should retain the order of any transfers when making revisions to the schedule.

Mr Morgan submits that Mr Fisher’s approach is flawed. It is contrary to the Judgment, which rejected Triumph’s pleaded case that a properly adjusted LRP would include a longer period of six to eighteen months’ delay for the transfer of production to Thailand. He illustrates the extreme impact of Mr Fisher’s method by reference to the A330-200 trailing edge panels. The schedule indicated transfer of this product starting in the third quarter of 2012 but Mr Fisher has shifted this first, to the first quarter of 2013 to reflect initial delay in transfer and then, to the second quarter of 2013 to allow for buffer to be built. This nine-month delay has a consequential impact on all subsequent transfers if the original sequence is maintained. Mr Morgan submits that Mr Fisher’s current stance

I accept Mr Fisher’s opinion that the transfer schedule needs to be adjusted to reflect the accrued delay to the outstanding 2012 transfers. Time must be allowed for manufacture of a buffer and transfer of the delayed products to Thailand. However, I reject his assumption that this requires a delay of up to six months to be built in for every subsequent transfer, prior to adjustment to accommodate the three-month buffer period. Although the original sequencing and duration of each subsequent product transfer should be maintained, there was scope for recovery of the accrued delay to the 2012 transfers at the beginning of 2013. At the time of the LRP, transfers of other products were ahead of schedule, as considered in paragraphs 445 to 452 of the Judgment. This would release capacity to manufacture those products that were behind schedule. In any event, as set out in the supplemental joint memorandum, there was spare capacity to accommodate the consequential increased production required by compression of the programme.

Helpfully, Mr Fisher has provided an alternative calculation of the DCF value in his fourth report that makes corrections for the above points. He has adjusted the start date for each of the four delayed 2012 transfers without allowing for any consequential impact on the remainder of the transfer schedule. He has then inserted a three-month delay to each of the twelve products to allow for building the buffer stock.

Initially, there was a dispute as to the additional production hours that should be inserted into the LRP to build three months’ buffer but that has now been resolved. Mr Fisher has added additional production hours to build three months’ buffer stock for each of the twelve outstanding products to meet customer demand in Farnborough and Thailand during the transfer period. In his second supplemental report, Mr Dearman allowed for additional production hours limited to those necessary to build buffer equal to the production which was planned to be made in Thailand in the transfer period. However, having considered Mr Fisher’s analysis on this issue, in his third supplemental report, he accepts that the balance of production could not continue in Farnborough without duplicate tools. Therefore, he agrees with Mr Fisher’s approach on this issue.

Adjustment is required to the LRP to allow for the additional cost of three months’ enhanced production to reduce arrears.

As set out in paragraph 421 of the Judgment, Mr Fisher calculated that it would require 5,693 production hours to clear outstanding arrears at the date of the LRP. The experts agreed that such arrears could be burned down over a period of three months. Mr Dearman calculated that it would take one month to eliminate the further arrears that would accrue during that three-month period.

In their second joint memorandum, the experts agree that the 5,693 production hours relate to arrears at the time of the LRP for: (i) products that were scheduled to be transferred to Thailand in the short term; and (ii) products manufactured in Farnborough that would not be transferred to Thailand.

The experts disagree on the number of production hours required to reduce arrears, the adjustments to be made to the LRP to reflect such enhanced production and the treatment of revenue from such production.

As to the production hours required to burn down the arrears, Mr Fisher has increased production at Farnborough by 5,693 hours to model the burning down of accrued arrears at the date of the LRP. He explained in his first and third reports that those additional production hours would not eliminate all arrears at the date of the LRP; he excluded from his calculation products already transferred to Thailand, products for transfer in 2015/2016 and spares. However, he has included in his calculation the additional production hours required to eliminate accrued arrears for products due for transfer in the short term and products manufactured in Farnborough that would not be transferred to Thailand. Mr Fisher has allocated 2,582 additional production hours required to reduce arrears for products to be transferred to Thailand to the buffer stock period for each product. He has allocated the remaining 3,111 additional production hours required to reduce arrears for the products to be retained at Farnborough to the beginning of 2013.

Mr Dearman has allowed 1,751 production hours to eliminate the accrued arrears in respect of products scheduled for transfer to Thailand from 2013. He has excluded 830 hours in respect of products due for transfer prior to 2013 and 3,111 hours in respect of products that continued to be manufactured at Farnborough. Mr Dearman explains in his second supplemental report that he has been instructed that the LRP included an allowance of 5,000 hours in the fourth quarter of 2012 to reduce arrears. Therefore, he has adjusted the LRP by re-allocating the cost and associated revenue of the 1,751 hours from the end of 2012 to the quarter prior to transfer of each product to Thailand.

Mr Pillai submits that Mr Dearman’s approach is flawed. Primus did not plead and has not produced any evidence to support a new case that the LRP included an allowance of 5,000 hours to reduce arrears. Mr Dearman criticises Mr Fisher’s allowance of 5,963 additional production hours on the basis that this includes arrears on products that were either planned for transfer prior to 2013, or not planned for transfer to Thailand at all. Mr Pillai submits that Mr Fisher’s figure is correct because it is representative of the volume of arrears that could be cleared within three months. Mr Dearman’s calculation fails to give proper effect to the Court’s direction that the LRP should be adjusted to include a period of three months accelerated production to reduce arrears.

Mr Morgan no longer relies on the new argument. However, he submits that the 5,963 production hours are a maximum for eliminating arrears and it was only necessary to reduce arrears on those products that were scheduled for transfer from 2013 onwards.

I accept Mr Fisher’s approach as correct. The LRP included production and other costs and revenues to meet current demand but made no allowance for additional production costs to reduce arrears. For Primus to reduce arrears it was necessary to introduce accelerated production alongside planned production. This requires adjustment to the LRP to model such accelerated production.

Contrary to Mr Dearman’s approach, the composition of the arrears used to calculate the 5,693 production hours is not material; it is simply representative of the production that would be required to significantly reduce arrears so as to obtain approval for the transfers. There was ample evidence before the Court that operational issues affected

Mr Dearman considers that credit should be given for the revenue derived from the additional production hours to clear arrears. Mr Fisher disagrees with this adjustment because it was not considered by the experts or raised in evidence at the hearing. However, he has provided an alternative calculation to assist the Court if it concludes that revenue from enhanced production should be included as a matter of principle.

I understand the logic of Mr Dearman’s argument but it would not be appropriate to allow this new adjustment at such a late stage in the proceedings. The accelerated production hours do not represent additional product that will lead to increased sales; it is accelerated production to meet historical orders. The revenue from such orders must already be included in the LRP. There might be an argument for bringing forward the receipt of such revenue in line with the accelerated production. The difficulty is that no evidence was called at the trial on the treatment of future revenue in the LRP and the experts have not considered it. Therefore, I conclude that it would be unsafe to assume that additional revenue should simply be added into the LRP without any consequential adjustments.

The variable labour rate in the LRP must be adjusted for the periods of accelerated production. As stated in paragraph 427 of the Judgment, the best evidence of the variable labour rate that should have been used in the LRP for the period required to reduce arrears and build buffer stock is the actual variable cost rate for the 2013 financial year of US$ 24.51 per hour.

The experts agree that the applicable rate is US$ 24.51 per hour but they disagree on the production hours to which that rate should be applied.

Mr Fisher has adjusted the LRP using the increased variable labour cost rate to all of Farnborough’s production during the periods of accelerated production in line with the above finding. During periods where there is no accelerated production, he has adopted the standard rate of US$ 17.94, which then falls by 0.5% for each quarter where there is no enhanced production, in line with the LRP assumptions.

Mr Dearman has applied the increased variable labour cost rate but limited to the additional production required to build buffer stock and clear arrears. The reduced rates used in the LRP are applied to all other production, including other production during periods of accelerated production.

I decline Mr Morgan’s invitation to re-write the Judgment and accept Mr Fisher’s calculations. The variable labour cost rate of US$ 24.51 per hour applies to all production during the relevant accelerated periods for the reasons set out in the Judgment.

The scrap rate in the LRP must be adjusted upwards for the periods of accelerated production. As stated in paragraph 432 of the Judgment, the LRP should have accounted for a higher scrap cost during periods of accelerated production. The best available evidence of the appropriate scrap rate is the 2012 average rate of 3.1% of revenue.

The experts have agreed that the LRP should be adjusted to increase the average scrap rate from 0.9% to 3.1% for the periods of enhanced production but they disagree on the revenue to which that rate should be applied.

Mr Fisher has adjusted the scrap cost rate used in the LRP to the revenue value of all products manufactured during the periods of accelerated production.

Mr Dearman has applied the increased scrap rate but limited its application to the revenue value of the products to be transferred to Thailand and produced during periods of accelerated production.

For the reasons set out above in respect of the increased labour cost rate, I accept Mr Fisher’s calculations. The increased scrap rate of 3.1% applies to all production revenue during the relevant accelerated periods.

The experts agree that reduced revenues, operating income and profits as a result of the transfer delays are reflected in the first three adjustments to the LRP as set out above.

Mr Dearman has proposed a further adjustment to the LRP to reverse an unintended consequence for the fixed overhead costs. For the purpose of the LRP, fixed overhead costs are calculated as a product of the fixed overhead rate and production hours at Farnborough until the end of 2013. From the first quarter of 2014 onwards, the rate is fixed using the rate at the end of 2013. Adjusting the LRP to include additional production hours for building buffer stock and reducing arrears increases the overheads rate by the end of 2013. In turn, that increases the fixed overhead rate, not only for the end of 2013 but for 2014 onwards, including 2017. That increased fixed overhead rate is used as a multiple in the terminal value.

Mr Dearman’s opinion is that fixed overheads should not alter with an increase or decrease in production. Mr Morgan submits that the unconventional treatment of overheads in the LRP has the unintended consequence of producing a very significant change in terminal value that does not reflect the fact that the increase in production hours to build buffer and reduce arrears was a temporary spike. He invites the Court to correct the Amended LRP to eliminate this anomaly.

I decline to make the further adjustment as suggested. The LRP contains a number of anomalies and it would be inconsistent to correct this one without revising other entries in the model. Mr Fisher identified modelling errors in the LRP that he corrected as part of his report and evidence to the Court in the earlier hearing. The modelling errors included the treatment of fixed overhead costs. Primus objected to Mr Fisher’s corrections on the grounds that the issues were not pleaded and Mr Dearman considered that the use of fixed overhead costs in the LRP was appropriate. For the reasons set out in the Judgment, the Court agreed with Primus that Mr Fisher’s proposed corrections were issues that were not pleaded and the claims should not be quantified by reference to the same. For the same reasons, the Court rejects the additional adjustments Mr Dearman now seeks to make.

In summary, the Court accepts Mr Fisher’s alternative calculation. This gives an adjusted NPV of US$ 86,815,798, a reduction from the NPV used by Triumph for the SPA of US$ 5,701,570.

The Court has found that:

The experts have agreed that the difference between (a) the NPV of US$ 92,517,368 million that was shown in the TGI model using the data from the original LRP and (b) the NPV shown in the TGI model using the data from the Amended LRP represents the reduction in value of the Primus companies as a result of making the four adjustments.

The Court’s above determination as to the method for making the four adjustments produces a reduction in value of the Primus companies of US$ 5,701,570.

Mr Pillai submits that Triumph would have paid a correspondingly lower amount for the shares and, therefore, the above reduction in value represents the quantum of Triumph’s loss. He submits that a dollar-for-dollar reduction in the purchase price to reflect the reduced value of the companies is the most logical and principled outcome.

Mr Morgan submits that Triumph has suffered no loss, alternatively any loss is within the de minimis floor of US$ 1.5 million and therefore no damages are recoverable.

In my judgment Triumph would have reduced the purchase price of the companies to reflect the reduced DCF value based on the Amended LRP and Primus would have agreed to such reduction for the following reasons.

Firstly, Triumph’s pleaded case was that if the LRP had been carefully prepared, Triumph would not have proceeded with the purchase of the companies; alternatively it would have paid a lower purchase price.

Secondly, Mr Kornblatt’s written evidence was that if the adjusted forecasts had been produced by Primus, Triumph would have walked away from the deal or reduced the purchase price. In the Judgment, having considered his written and oral evidence and

Thirdly, although the purchase price agreed was not based on a mathematical formula by reference to the LRP, the evidence of Mr Kornblatt and Mr Wilkin, which I accepted, was that information in the LRP was used as part of the due diligence exercise to check the level of the purchase price. The NPV produced by the TGI model using the data from the original LRP was US$ 92,517,368 million but there is no evidence that Triumph considered offering a purchase price without building in a substantial margin for risk. Therefore, I reject Primus’ argument that there is no loss because the reduced value of the companies remains higher than the contract price. The EBIT and EBITDA figures, on which Mr Fletcher of Primus placed such weight because he recognised their significance for the purchase price, are reduced when the adjustments are made to the LRP. It is logical that a lower valuation for the companies would have reduced the price that Triumph was prepared to pay.

Fourthly, the best evidence before the Court as to the basis of calculation of a reduction in price is the reduction made following discovery of a discrepancy between the long term machining agreement and the LRP. The reduction in the bid price was equal to the amount of the discrepancy.

Fifthly, Triumph was not in competition with other bidders. The Spirit indicative bid was substantially below Triumph’s indicative bid and below any reduced price based on the adjusted value of the companies.

Finally, Primus would have accepted a reduced purchase price. In his witness statement, Mr Delaney explained that in 2012 he valued the Primus companies at US$ 40 – 45 million, with an anticipated increase in value by 2013 to US$ 47 million. His target price for the sale of the companies was US$ 50 million. In 2012 Primus had taken the decision to sell the companies. Triumph’s indicative bid must have been an attractive offer that surpassed expectation. There is no evidence to suggest that Primus would have walked away from the transaction had Triumph reduced the purchase price by US$ 5,701,570.

Having regard to the compensatory principle of damages for breach of warranty, I am satisfied that a fair and reasonable basis for assessing Triumph’s loss is the reduction in value of the companies of US$ 5,701,570.

Taking into account the contractual deductible of US$ 1.5 million, Triumph is entitled to damages in the sum of US$ 4,201,570.

The parties have agreed that an award of interest at the rate of 4.5% per annum from 3 May 2013 to the date of judgment is just and appropriate, and thereafter at a daily rate of US$ 1,181.00 until payment.

I invite the parties to make any submissions on costs in writing, if not agreed, following which I will produce a short judgment on costs.