Online Postgraduate Courses for the Electronics Industry

Provided By The University of Bolton

• Home
• About Us
• Questions & Answers
• Courses
• Students
• Links

Useful Tools

Business Issues of Microelectronics

Unit 1: Business Issues

This unit covers investment issues such as profit, factory cost, risk and return on investment in dealing with new technology.

Contents

• 1.1 Introduction
• 1.2 Determining Return
  • 1.2.1 The effect of production release delays
  • 1.2.2 Profit
• 1.3 Returns - What Controls them?
• 1.4 Unit Costs
• 1.5 Software Costs
• 1.6 Investment Issues
• 1.7 Choosing Projects
• 1.8 Determining Return
• 1.9 Final Review
• 1.10 Further Study
• 1.11 Self Assessment Questions

Notional workload: 5 hours

1.1 Introduction

Business is an activity in which money is invested in product development in the expectation of a future return. Business success is gauged by the level of return achieved on the investment made in its development. Long-term business success depends upon sustaining a timely sequence of high yielding successful new products. If you don't, you go out of business.

The catch with business lies in the return being recovered in the future. The future can be neither predetermined nor fully controlled. So the investment inevitably invokes uncertainty and therefore risk that the expected project return will not be achieved. A succession of poor returns can break you.

Many engineers feel uncomfortable in the presence of uncertainty. They are accustomed to an environment in which their efforts will generate a product that meets its statement of requirements and tend to gloss over its (sometimes) late release to the market. This illustrates a dichotomy. Engineers tend to focus on performance-related issues rather than purely business ones and management do the converse. This is not a criticism but a condition that should be borne in mind.

You should complete this unit with a more rounded vision and clearer understanding of the issues facing management.

The rate of return required of a project is related to its level of risk. The riskier the project, the longer the odds on development success and the higher the return required. Clearly, the level of risk will be affected by your choice of technology - how the product is made. So will other non-technical issues such as its appearance, how it is priced, marketed, manufactured, costed, funded, resourced, managed and so on.

This module deals with how business and technology issues interact and how this interaction can be harnessed to the benefit of the project.

In many ways top management is like gambling at the races. It has the same basic activities - assessing the risks and the qualities of several alternatives (runners) in the light of recent performance, the characteristics of the particular racecourse, the state of the economic "course", the chosen jockey, the weather and so on. Then it involves assessing the odds before betting (investment) money in products that in your expectation will, over a period and with ups and downs, make money. Overall you must make more money than you spend. The bookies do.

[Back to Top]

1.2 Determining Return

Accounts are usually made up monthly and the data is quantified at these intervals. Performance over time can be presented in tables or graphically in a number of ways.

Graph 1.   Various methods of presenting financial data

The first graph shows the data in steps with heights equal to the month's total.  The second shows a  smooth curve through the points defined at mid-month with the month's total. A third way would be to connect these mid points directly with lines of the required slope. Each has a particular attraction and value. But remember that the source data is quantified and any conclusions drawn must recognise the limitations of sampled data in extracting conclusions from them. For instance, the data is unavoidably delayed in its collection, analysis and publication. The publication may also introduce delays of its own. It is clear that the results will not be available until after the data is published and will be subject to the bandwidth limitations imposed by the data collection and handling procedures.

As an example, after a particularly successful year as a General Manager, I tried to find out what was the secret of this success. I believe that I found it, but only by going down to weekly results because the data fluctuated violently. When looked at in weekly detail, the profits all came from short periods of high activity. High activity seemed to generate high returns provided that it was not sustained. It appeared also to be highly non-linear. This is an example of the limitations that can accompany monthly data sampling.  This text will use the smooth curve version simply because it most clearly illustrates the points to be conveyed to you.

Return is normally gauged in terms of interest received. There are two types of interest in general use. The first and simpler of the two is "simple interest". It provides an annual return that is a constant percentage of the value of the investment, the "principal". It pays out a fixed sum each year that it is in force. The second, compound interest, adds this year's interest to the next year's principal so that the return increases year on year. It is appropriate to longer-term situations whereas the simple interest is applied without penalty to short periods less than a year or possibly two. In most cases the current year's excess of income over costs will be largely spent on new product development or the acquisition of new tools and equipment. It is unusual for large sums to be "banked" over long periods.

Graph 2.   Return on Investment

Graph 2 has four quadrants. The left-hand ones relate to time before the release of a product to market, while those to the right relate to time after it. The lower left-hand quadrant covers development costs incurred in the time up to product release while the one to its right covers any costs incurred after it.

The total return to be obtained is the total investment  over both lower quadrants plus "X"% per annum.   Problems can sometimes emerge after product release and need field support and possibly new development to clear them. This expenditure may be charged back to the project but is conventionally accounted for as an operating expense and recovered from profits.

The right-hand upper quadrant shows three curves:

• The lowest one is incremental profit and is presented as a continuous variable. It shows a rapid increase in sales at product release, a plateau of mature sales and a tail-off into obsolescence. It has the nickname of the Moby Dick curve.
  
  
• The second shows cumulative profits, again as a continuous variable. It is the running integral of the area under the first curve.
  
  
• The third shows a line of constant returns. Its slope is the rate of return from simple interest and is such that it is tangential to the cumulative profits curve. The rate of return is maximised at this point. It then reduces. At or before this point you need a new product to come on stream. Ideally, the overlap should eliminate any sag in sales as the declining income from the first is compensated by the contribution from the second.

Had we been using compound interest instead of simple interest we would have a curve of constant percentage compound return as in Graph 1 but the principle would be the same. (It's more difficult to locate the maximum return though).

The decay in profits following the point of maximum rate of return is caused by the stagnation of sales and their later decay. It is wise not to linger at this point but bring out that new product that you wisely have ready to release.

In general you should be looking to reach a total return of around 40% per annum as an absolute minimum and up to 100% or more (whatever you can get away with) for something ahead of the game. Niche market operators may survive with top end pricing and long product lives but it has become a shrinking market and the exception rather than the rule.

An alternative way of judging a project is in terms of its "payback time" - how long it takes to recover the development costs. 40% compound equates roughly to a two year payback time (1.4²).

In practice many proposals predict payback times under 12 months and a few as low as six. (We will see how this can be achieved later). In many fields time-to-market has become a key performance parameter. It gets increasingly important with shorter lifetime products. Of course development times must also be reduced in proportion.  Rapid times to market are essential, and launch delays disastrous.

One way of meeting this requirement is by successive incremental enhancement of a replicated basic common product. The increments adapt it successively to more numerous or more complex functions or to different fields of application. It can generate the succession of product releases that meets the need for short development times and low investment. To be fully exploited, this succession path and the functional interfaces must be identified before starting any design work. All of the hooks and interfaces required to accept the new functions should be incorporated at this stage.

Software development can be treated not as an investment but as an expense to be paid out of profit. It may then be advantageous to retain the same hardware core and embody new functions through software. This has the effect of transferring the charges from development investment to factory costs, trading profit for return. Combining the use of replicated hardware with well-structured software is a powerful way of reconciling the demand for reduced development times and costs combined with high returns.

1.2.1 The effect of product release delays

Graph 3 The Effect of Product Release Delays

Graph 3 shows the detail of events when and after a delay occurs within the planned development programme. It may be caused in any one of many ways. Usually there will be complex choices to be made to recover the delay. They have drastic, sometimes dramatic, effects and can be in unexpected situations. The whole project activity is frequently involved and sound decisions have to be taken and implemented as quickly as possible. A good programme allows for these effects as far as it is possible / reasonable.

Graph 3 shows the financial tracks through this process. The top curve shows the planned sales and delivery performance and below it shows the actual overrun of the planned investment. The problem is shown to appear in the incremental investment curve after the process equipment is largely installed and ready, as are the procurements. However, near the end of these activities, disaster strikes and spend rate increases rather than decaying to the planned zero.

It continues as an extension to the incremental investment programme and a continuing increase in the cumulative investment. As the recovery phase advances, the investment slows and the first unit sales follow a recovery track through which investments develop and a build-up of deliverable hardware takes place.

This phase is illustrated in graph 3 as the recovery track between late development and early sales. The recovery track will build up to high delivery levels but will never reach the originally planned volume, but it is not just the delay that strikes. The early sales peak has been missed, market prices have decayed. It is likely that the programme will not only start late - it may well end early.

The sort of performance shown in graph 3 would break most companies that do not have sufficient reserves, either in finance or credibility, so keep up a tolerable performance.

1.2.2 Profit

Graph 4. Optimal Operating Conditions

Broadly speaking, profit is the difference between costs and selling price. You may be able to control some of your costs but selling prices and release timing will normally be strongly influenced by your market and as such, near impossible to control. (Price insensitive products are a special case and are usually associated with niche markets).  Timing is also important in relation to competing products in the same market sector. So is good market intelligence. In general the first company to market will capture a major share of the potential sales, the second will get a large slice of the rest but the selling price, and hence profits, will have been eroded. The others may be lucky to see profits at all.

The timing of product releases is significant and influences risk. Delays in development are at least as serious as a cost over-run. Worse still, the two usually come together. Minimising the risk of cost or timescale overruns is as essential to business survival as are low costs or high selling prices. Reduced time to market will usually result in higher profits.

Breaking a market up into specialist segments (product differentiation) is a way of improving your chances of being first to market in a particular field. Using an established product as the basis for a new one with enhanced or additional functions is, as we have just seen, a good way of reducing project risks - and costs and timescale.

An analysis of the level of success achieved by different types of project has shown that success is related to project class in the following 'high to low' sequence. a) An existing product enhanced
b) An existing product cost reduced
c) An existing product with new features
d) An all new product The correct choice of technology is important in enabling the best business performance to be achieved. At this point you should also see more clearly when and why expediency can be more desirable than elegance in the way in which a development programme is put together.

[Back to Top]

1.3 Returns - What Controls Them?

The return from a product is its accumulated profits.  It is the rate of accumulation that is significant.

Profit is the difference between the price paid by the buyer and the costs already incurred by the seller. So first we will look at costs and cost structures.

Costs are usually divided into fixed costs and variable costs.  Fixed costs are those that are independent of the number of units made. Examples are:  interest on loans, local taxes, rent, depreciation, and so on.  Variable costs are those that vary with the number of production units made - for example, bought-out component costs, materials (e.g. solder, flux, paints, sheet metal, printed circuit boards, components), manufacturing labour costs and so on.  The total cost of a given production volume is the fixed costs plus the variable costs for that volume, as shown in Graph 5.

Graph 5.   Cost versus Production Volume

Of course the reality is more complicated. For instance, if sales volume becomes higher than expected it may be necessary to extend the production facility, purchase additional equipment and so on. That will increase the fixed costs but create compensating savings in the variable costs.  Graph 6 shows the effect of such a discontinuity in both fixed and variable costs.

Graph 6.   Discontinuity in Costs Leading to Reduced Total Cost at Higher Volumes

In practice the variable costs are rarely constant with volume either and may be subject to discontinuities of their own - for example at quantity/price breaks in materials purchasing. But these remarks are cautionary. The simple picture is sufficient to illustrate the necessary points. If real life were that simple then good accountants would not be so sought after!

[Back to Top]

1.4 Unit Costs

The cost per unit made is the first differential (the incremental slope) of the Cost/Volume graph (Graph 4). In general, unit cost falls as volume increases.

Graph 7.   Cost/unit versus Production Volume

The unit cost is initially high because the fixed costs are spread over few units. It falls rapidly as sales build up. The total cost/unit  is a non-linear function of production volume.

[Back to Top]

1.5 Software Costs

It is not usual to include a variable cost attributed to software and the actual cost incurred in manufacture is very small.  Software development is conventionally regarded as an investment and a charge is included in the operating costs. This also allows for any fixes required in service.

1.6 Investment Issues

Direct Investments

There may be "direct" investment in the equipment required in assembly and test of products involving microelectronics. Such a cost will normally be recovered through an "overhead" charge.

For project cost comparison it is often more convenient to attribute a share of the investment to a particular project. In this connection it is usual to write off an asset such as a  design workstation in three years but to exclude any software charges. Software is usually regarded as having no resale value and not therefore as an amortisable investment, but as an expense that is written off in the year in which it is incurred. The share of the investment attributed to a given project is a fraction of the total hardware expense. The fraction is taken as the ratio of the time it is used on the project to its three year "life". A twelve month project is charged with 12/36 times the investment value. The software investment is most likely to be charged to the specific project unless multiple use is certain within the first year after its acquisition.

Indirect Investments

There are three primary items of indirect expense:  maintenance, training and consultancy advice.

Maintenance will be allowed for in overheads, whether it be software or hardware maintenance.

Training may be demanded in order to complete a specific project. In this case the whole cost is attributed to it, otherwise it may be apportioned to several projects.

Consultancy is almost invariably project-specific. As such it is charged wholly to the project.

As with costs, there are "direct" and "indirect" elements of investment in microelectronics.

[Back to Top]

1.7 Choosing Projects

You have £200,000 to invest.  Which of the following would you choose?

(a) a current account with one of the major banks
(b) a building society savings account
(c) an endowment policy
(d) a unit trust
(e) a high earning investment trust
(f) penny shares What do you need to know to help you choose?

Returns over five years come in different guises (share value or Interest) and with differing risk levels. They can be summarised as in the table below:
 

The correct answer to the opening question is "a selection from the various types of investment." The objective is to have a "guaranteed" worst possible consequence which you can support, yet keep open the possibility (remote as it may be) of making a big killing with money you can afford to lose.

Choosing your project portfolio is a process of a similar kind and, in business, conservatism has a good track record. Each option has a particular characteristic:

(a)  for the services which accompany it
(b)  security
(c)  special purposes such as tax efficient pension funds
(d), (e) & (f) a spread of risk/return combinations for balancing the portfolio.

The same applies to choosing a project "portfolio" to pursue. In practice you will select more projects than you can handle - some of them will drop out on the way. One way of handling this situation is to assign a "probability" of proceeding to each project and to factor the predicted return in the same way. You plan your future developments to include projects with total factorial returns that are consistent with your company target and projected costs that are consistent with what you believe you can fund. Your portfolio might look like this:

Note that Project C gives the second highest returns even when factored by 1/10. A full assessment table would include an assessment of scaled costs. Cost and return do not always scale by the same factor.

[Back to Top]

1.8 Determining Return

There are several versions of "return". The simplest is the accumulated profits generated by the project. It implies that there is no interest contribution. Profit is put to use as soon as it is generated.

The second is to include annual interest payments with the profits being invested over the accounting year and accruing interest. The interest is put to use at the end of the year. All the interest therefore appears as simple interest.

The third version is compund interest. Here the profits are invested and the annual interest payment are re-invested until the project is complete when the funds become available for re-use.

The choice of procedure is critical at the strategic level of management but not in comparing different projects - providing that the same procedure is used for each project being compared.

Simple interest is easy to use. Compound interest will give higher returns - so it looks better. Immediate use of profits gives conservative performance estimates and is less speculative. It is up to you which you choose. You can a expect a new product to be loss making initially. It should only last for a short period. It results from things like last minute problems debugging the assembly and test training curve and from incentive pricing for launch customers. It can either be treated as a negative profit or an increase in investment. Graph 1 shows it as negative profit.

The graphs we have seen show idealised continuous curves. In practice they will be more often presented as steps of a given period.

Example

Graph 8: Profits as a Percentage of Investment

From the graph you can deduce the maximum return on investment (120% pa at 2½ years of sales). You can deduce the payback time (10 months) and see that the return will remain high (better than 100%) even though profit has fallen to zero after 4.3 years.Later on you will be able to track project performance and take early corrective action if it is required.

[Back to Top]

1.9 Final Review

You should now have a clearer understanding and a greater sympathy with business, management and its enforced responsibility for the partly uncontrollable.

An essential characteristic of business is that outstanding success is generated rarely and often unexpectedly. So is disastrous failure. In both instances it is frequently due to the uncontrollable but only rarely due to the incapable.

An outstanding feature of microelectronics is its rate of evolution. The majority of the business issues involved are the direct consequence of coming to terms with the technology and its capabilities.

Twenty  years ago there were no design systems. Chip layout was all done by hand using coloured sticky tapes and plotted out at 400 times full-size. Chip layouts were as large as 2m 50cm high and 5m long. The feature size was 8 microns and an essential tool was a step-ladder. Compare looking back twenty years with today and try to imagine where the technology will be in twenty years from now.

The only comfort is that the issues will probably remain the same. Only the numbers and the processes may change.

[Back to Top]

1.10 Further Study

[Back to Top]

Site Search

Powered by Google
Site Map