- Question 1: At any of these dates, did Intel have a contingent liability as defined by SFAS #5?
June 30: Intel has discovered the flaw No contingent liability, no disclosure. According to Intel, a series of tests has showed that an error would occur only once every nine billion random calculations, or every 27,000 years for most users. Therefore, the chance that customers would encounter errors in calculations on their Pentium-driven PCs is slight and the event that customers would request chip replacement is remote.
This means that the implementation of a replacement program is not probable, nor is the cost of such a program reasonably estimable because the theoretical participation rate is close to or equal to 0%.
October 31: Dr. Nicely has posted information about the flaw on the Internet and started an active discussion group No contingent liability, maybe disclose. The discussion started by Nicely continued to flame accusations and threats to Intel, so the replacement of Pentium chips might become an event that is more likely than remote but less likely than probable.
However, since Nicely was involved in intense and continuous number-crunching far beyond that of a typical user, he was more likely to encounter errors and could not represent a typical user. Without more information on the frequency of the flaw encountered by a typical user, Intel might not be able to decide on a replacement plan, so that replacement costs were not reasonably estimable. Therefore, if anything, Intel might be on the hook to make a disclosure of a potential future expense.
November 25: Article in Electrical Engineering Times has appeared, a story has been broadcast on CNN, and articles have appeared in the New York Times and Boston Globe. No contingent liability should disclose. The Internet discussion caught the attention of reporters and has caused a negative impact on Intel’s stock price, which dropped two percentage points. As the media continued to report the Pentium flaw to consumers, the event that consumers would request a replacement of Pentium chips became at the very least reasonably possible, if not probable.
However, Intel might still not be able to estimate reasonable replacement costs for its Pentium chip because it has yet to see how its major customers react to the event and start a conversation with them on deciding a replacement plan. Therefore, the size of the program is still very uncertain. Intel should, therefore, make a disclosure.
December 17: IBM announces that it has stopped shipments of its computers with the flawed Pentium chip Have contingent liability. Given that IBM – Intel’s major customer – would probably return flawed Pentium chips and require Intel to account for their loss, the replacement event was likely to occur.
Intel’s management should definitely have started negotiation with major computer sellers to decide a return policy for its customers, so the replacement costs are estimable, as it has gathered cost information already. According to SFAS#5, when the future event is likely to occur and reasonably estimable, the company should record a contingent liability.
- Question 2: Assume that in December 1994, Intel’s management decided to expand its Pentium chip replacement program by offering to supply a replacement chip to all purchasers of a defective Pentium chip, regardless of how they use it.
Intel will provide a new chip-free of charge, but will not pay for any other costs. What expense/liability should Intel reflect on its 1994 financial statements? GAAP, through SFAS #5, advises that a contingent liability must be disclosed if “it is probable that… liability had been incurred at the date of the financial statements. ” From the knowledge gained in Accounting 202, the rule of thumb under GAAP as to the definition of “probable” is that an outcome is considered probable if its likelihood of occurring is greater than 70%.
We assume here that the probability of there being at least some economic benefit being sacrificed in the future (resulting in the creation of a liability now) due to customers demanding replacement chips is greater than 70%. This first condition is given by SFAS #5, however, is only a necessary but not sufficient condition for recognizing a contingent liability in the financial statements. SFAS #5 states further that in addition to a loss being probable, its amount must also be reasonably estimable.
Obviously, this second condition is vague and leaves much up to the discretion of management. However, we feel (and we assume throughout this case that we have the same information as Intel’s management) that the amount is reasonably estimable, given the information we have. This being said, the expected value of the future loss that will actually be incurred is much harder to estimate, but we can, with the given information, assign a range of values to the loss, each associated with an uncertain probability of occurring.
GAAP states that (and this is again coming from Accounting 202), should this be the case, companies should recognize as a loss contingency the lower bound of their estimated losses. We will, therefore, do this. First, we must detail our assumptions and calculations that will lead to what we ultimately recommend recording as an expense. Given that we know that the cost to replace each chip (and its accompanying heat sinks) ranges from $50-$100, we will choose the lower bound of this in accordance with GAAP.
Considering that the replacement program will extend to all customers (we assume that the firm will be liable for replacing all 6,000,000 chips already shipped and that it will not ship the remaining 2,000,000 chips it still has in inventory), we estimate that the participation rate in the program will be 25%. This is a very conservative estimate, considering that the costs associated with having a service provider actually replace the chip are estimated to average $400, reflecting the cost to the service provider of an average of $289 plus a profit margin for the service provider.
This participation rate also takes into account the fact that only an estimated 5% of users will actually be inconvenienced by their chips’ flaws. Therefore, the participation rate, under this program that only pays for chip replacement, is likely to be even lower than 25%, but, due to the lack of forecasting data available, we feel that this is appropriate given the conservatism principle. Calculation of Intel’s contingent liability: For $50/chip: 6,000,000 chips total*25% participation rate*$50 per chip = $75,000,000 contingent liability
In the financial statements, this would be recorded as follows on the 1994 statements: DR Pentium chip replacement expense 75,000,000 CR Provision for chip replacement program 75,000,000 In 1994’s financial statements, this new expense would show up on the income statements as a decrease of EBT by $75 million, leading to a decline in net income. We would probably include this inside of 1994’s cost of sales as a non-recurring item (with an appropriate note disclosure alerting investors to this treatment).
This is because this replacement program, while unusual, will probably not prove all that infrequent in the future, as Intel’s future products could very well suffer from defects that may lead to a similar replacement contingency expense. Overall, this would lower net income by $75,000,000*(1-Intel’s tax rate), while decreasing Intel’s provision for taxes by $75,000,000*(Intel’s tax rate). Moving on to balance sheet effects, this would decrease 1994’s retained earnings by $75,000,000*(1-Intel’s tax rate) and decrease Intel’s current liabilities by $75,000,000*(Intel’s tax rate).
However, the creation of the provision itself, which would probably be classified as a current liability, would then increase current liabilities by the full $75,000,000, leading to a balanced Assets = Liabilities + Shareholders’ Equity equation. As users request chip replacements (which would occur in subsequent years), Intel would record the following for each period DR Provision for chip replacement program (# chips shipped * $60) CR Inventory (# chips shipped * $60) We focus here only on the expense/liability that would be recorded in association with the replacement program.
If the likelihood of lawsuits being successful against the firm is deemed probable (with a likelihood greater than 70%), and if Intel can reasonably estimate the amount that it may have to pay out, it would also record an expense and a contingent liability related to these legal issues. If this amount is not estimable, then Intel would only disclose the potential legal expenses in its notes. While this does not relate to the replacement program, Intel would also have to write-down the value of the defective chips that it currently has in its inventory. This would be done as follows: DR Loss on inventory (due to product defects)
CR Inventory The amount of this write-off would fluctuate based upon the value at which these chips are currently held. The loss would flow through the income statement exactly like the contingent liability expense, as detailed above. This inventory write-off would also be recorded in the scenario given in question 3, so we will not talk about it in that question again.
- Question 3: How would your answer to question 2 change if Intel also offered to pay for the labor and direct incidental costs in addition to offering to supply a new chip to all individuals?
Again, we must first detail our assumptions and calculations. We will again use the lower bound for hardware replacement costs ($50/chip). It is uncertain from the question of whether or not Intel has secured a favorable service provider agreement, which could decrease its labor and direct incidental costs. Therefore, we will generate expense/liability pairs for two scenarios. Under the first scenario, we assume that Intel has no service provider agreement and will, therefore, pay, on average, $400/chip for labor and incidental costs.
In the second scenario, given that the average cost to service providers to replace the chips is $289, we assume that Intel could negotiate a more favorable rate of $340/chip. We again assume that the replacement program will extend to the 6,000,000 chips shipped. We will adjust our participation rate upwards here since customers will no longer need to incur the approximately $400 in labor and incidental costs associated with replacing their chips. We will, therefore, work with a participation rate of 40%, since many individual and non-intensive users will probably not go through the effort of securing replacements.
This 40% takes into account the 33. 33% of business clients (who are more likely to seek replacements) who presently own defective chips and also provides for an uncertainty premium of 6. 6666% above this. We will also assume that the estimates given for the labor and direct incidental costs include the $10/chip in labor and overhead that we factored into our calculations for the last problem, hence, our price per chip will stay at $50. Calculation of Intel’s contingent liability: For $50/chip in hardware replacement costs and $400/chip in labor and direct incidentals: 6,000,000 chips total*40% participation rate*($50+$400 per chip) = $1,080,000,000 contingent liability •For $50/chip in hardware replacement costs and $340/chip in labor and direct incidentals: 6,000,000 chips total*40% participation rate*($50+$340 per chip) = $936,000,000 contingent liability We are conflicted as to which of these two estimates to use. On the one hand, the conservatism principle would advise us to record the highest expense/liability.
However, SFAS #5 recommends recording the lower bound of our estimated potential losses. However, since the question does not explicitly state that Intel has an agreement with service providers to obtain lower service costs, we will record a higher expense/liability. Intel’s management may want to disclose in the notes the impact that any favorable service agreement may have on this liability going forward. In the financial statements, this would be recorded as follows on the 1994 statements: DR Pentium chip replacement expense 1,080,000,000 CR Provision for chip replacement program 1,080,000,000
In 1994’s financial statements, this would again probably appear as a non-recurring item recorded in cost of sales, which would decrease EBT by $1,080 million (with an appropriate note disclosure alerting investors to this treatment). Overall, this would lower net income by $1,080,000,000*(1-Intel’s tax rate), while decreasing Intel’s provision for taxes by $1,080,000,000*(Intel’s tax rate). Moving on to balance sheet effects, this would decrease 1994’s retained earnings by $1,080,000,000*(1-Intel’s tax rate) and decrease Intel’s current liabilities by $1080,000,000*(Intel’s tax rate).
However, the creation of the provision itself, which would probably be classified as a current liability, would then increase current liabilities by the full $1,080,000,000, leading to a balanced Assets = Liabilities + Shareholders’ Equity equation and wiping out the previous lowering of the current liabilities balance. As users request chip replacements (which would occur in subsequent years), Intel would record the following for each period DR Provision for chip replacement program (# chips shipped * $450).
CR Inventory (# chips shipped * $50) CR Labor costs, direct incidental costs (# chips shipped * $400)
- Question 4: At the end of the December 17 meeting, what should Intel’s management do? Should they expand their Pentium chip replacement program by (1) covering more individuals and/or (2) providing or paying for some or all of the (non-chip) incidental costs of replacing the defective chips? We would advise Intel’s management to expand their replacement program along both of the listed dimensions for a variety of reasons.
Regarding how management should actually expand the program, firstly, we feel that they should cover all individuals who request replacement chips. Secondly, we feel that they should not only cover all costs associated with replacing the chips themselves, but also all non-chip incidental costs. This would mean that they would have to record the $1,080 million expense/contingent liability given above in their financial statements. The first reason for our recommending this approach is that Intel must work to preserve its brand equity and future sales growth.
If management does not take sufficient action now to show its customers that the company will accept full responsibility for any flaws in its products, they may not trust the company in the future. This would lead them to purchase products other than new Pentium chips (perhaps by waiting for clones to come out). This could endanger Intel’s projected sales of 22-25 million chips in 1995, leading to far slower-than-projected sales growth. This would lead to lower future net income, thus lowering the share price because EPS will not grow as fast as analysts’ expectations demand.
Furthermore, Intel needs to win the support of IBM, one of its main customers. Losing IBM at this crucial growth stage of the microprocessor product life cycle could lead to Intel failing to achieve sufficient economies of scale to keep its margins wide, leading to the loss of major competitive advantage. This could further endanger Intel’s future growth of earnings and share price. Additionally, as stated in the case, Intel’s brand equity currently sits at approximately $6. 4 billion.
Taking an expense of approximately one billion dollars presently appears rational to us, as it will help to avoid losing this $6. 4 billion in value. Therefore, we feel that rolling out an extensive replacement program is really just making a positive NPV investment, as it will cause future net income flows to be much higher than they otherwise would be, leading to the firm, under the discounted cash flow model, to be worth more now than it otherwise would be after a market correction for mishandling of this product defect.
The second reason we recommend expanding the replacement program is that we feel Intel must set a precedent now. The company must show its customers now that whenever it ships a faulty product in the future, it will pay the full price for its errors. Failing to assure its customers now will, again, create additional uncertainty as to its future sales numbers, as customers may not have faith in Intel’s ability to deliver quality products that it stands behind 100% in the future.
Therefore, Intel should offer a full replacement program to give its customers additional confidence in its products, now and in the future. We also feel that Intel should offer a full replacement program now in order to reduce the impact of lawsuits filed against the company (and to preempt any lawsuits that have yet to be filed against the company). As stated in the case, some of these lawsuits may force Intel to offer a comprehensive replacement program for the defective chips, so Intel may end up needing to book this expense/contingent liability in the future anyway.
We, therefore, feel that it would be better to assume the liability now, in order to hopefully avert the need to go to court (thus saving the associated legal expenses) and to stop additional lawsuits from being filed. We realize that some lawsuits will proceed regardless of whether or not Intel implements a comprehensive replacement program, but we believe that Intel could avert at least some legal costs by taking action now. Additionally, taking action now could lead to there being less negative publicity in the future from future lawsuits.
This is important because while there has already been a fair amount of negative coverage, additional future coverage could lead more consumers to feel that their Intel chips are untrustworthy. This would lead them to want to participate in any replacement program that Intel might implement, increasing Intel’s costs. By taking action now, Intel could avoid some of this future additional negative media coverage, resulting in a lower participation rate in the replacement program it ultimately decides upon and thereby decreasing actual replacement expenses.
Our fourth reason for recommending implementing a full replacement program is that if Intel creates a large provision now, it may be able to recover some of this in later periods. If the participation rate in the program proves much lower than our estimated 40%, Intel’s provision will prove to be much too large. Therefore, Intel could recognize a gain on this provision in the future by drawing it down, which may come in useful if it needs to smooth earnings.
While relatively unethical/semi-deceitful, this potential benefit must be considered in order to get a holistic perspective of the relative benefits nd detriments associated with our recommendation. Intel should be able to draw down the liability relatively soon since its chips (like all electronics) have relatively short lifecycles. This is because customers replace their computers frequently, so if they plan on participating in the replacement program, they are likely to do so immediately upon the program’s announcement or not at all, as, should they delay replacement, they will likely just buy a new computer rather than seek a replacement.
This means that after the end of 1995, Intel should be able to conclude with a large amount of certainty that it will have reimbursed all the consumers who actually plan to participate in its program. It will then be able to draw down any remaining provision without fearing a rush of participants after doing so.
- Question 5: After the December 17 meeting, how should Intel’s management communicate its decision to the financial markets? Should they file a Form 8-K?
If Intel were to cover the incidental costs, then with the replacement program, they would be entering into a Material Definitive Agreement. According to the official document, a material definitive agreement is an agreement that provides for obligations that are material to and enforceable against the registrant, or rights that are material to the registrant and enforceable by the registrant against one or more other parties to the agreement, in each case whether or not subject to conditions.
The general instructions of filing an 8-K and the information that should be included in it for a material definitive agreement can be found at http://www. sec. gov/about/forms/form8-k. pdf (Item 1. 01) Entering into a material definitive agreement requires disclosure of certain information as listed in the document. Intel’s management needs to appropriately communicate its decision through a disclosure because they need to make their counterparties as well as potential customers who might want to sue them what the current situation might be.
They also need to calm the financial markets and let them know exactly what action they may be taking, in order to preserve brand equity. Additionally, they would want to maintain their business with IBM and start the shipments of Pentium PCs again, letting them know that they take full responsibility for the problems and are taking drastic measures to fix the issues.
- Question 6: Why do you think the FASB did not provide more “bright-line” rules for when contingent liabilities should be recognized?
Does SFAS #5 provide managers too much discretion in recognizing and estimating the amounts of contingent liabilities? Bright-line rules create incentives to follow the rule by the letter but not by the spirit. The basic characteristics according to the rule can easily be manipulated to avoid meeting the disclosure requirements, while still being reasonably possible and estimable. In addition, it would be difficult to set broadly defined rules that would work equally well across multiple industries.
Would a 70% of occurrence be appropriate for both an oil spill and a product recall? While we agree with FASB’s decision to avoid using bright-line rules, we still think that the current SFAS #5 leaves managers with too much discretion in both recognizing and estimating the amounts of contingent liabilities. Management may deny that something is probable now, in order to delay recognizing contingent liability, or they could deny that it is estimable now, even though it is estimable now, to delay recognizing a liability.
More likely, managers may underestimate the liability balance. Even though we think FASB #5 gives managers too much discretion, we were short of viable solutions, without using bright-line rules. One possibility is involving an arm’s-length professional third party to either supply estimate of liability amount or give a second opinion, but this creates many extraneous costs and extends the process.
