Product development part 6

All data in our NPV analysis is expressed by probability distributions. How different it is when compared with standard Excel analysis. Very much, because we also took care of RISK. Now we can run simulation.

With the help of @Risk for Excel from Palisade Inc we are able to consider all possible outcomes of our project. But not just that. We will also know the probabilities associated with all outcomes. That is very powerful message.
Note standard ‘what if scenarios’ do not produce any conclusive outcome. Because, this not just one variable that changes, all variable do change within their ranges and frequencies. Also some variables may be correlated what was omitted here.

A simulation with @Risk is represented by number of interactions. An interaction is one outcome based o given set of variables. We can set the number of interactions and the number of simulations. Each set of variables is randomly selected. The whole process; selection variables, computing each outcome and summarizing all outcomes is referred to as the Monte Carlo simulation. The end result produced is in form of probability distribution and associated with it statistics.
That provides us with comprehensive information on what can happen with our project, enabling informed decision. For example, assume the NPV analysis can produce outcome as on this picture.



The outcome of our NPV analysis shows us all possible scenarios our project can deliver. Should we go for it depends on our attitude towards risk.
Most people would not mind to bet on heads and flip a coin when $1 is at stake. The chance to win is 50%. But how many people would play this game with $1 million? Not many. In general more money we plan to invest then greater security we want to have. But this is still individual decision. On the picture above we see, about 99% of all outcomes are above our minimal expectations, what is when NPV is equal zero. Possibility for negative NPV exists but is associated with very low probability. It appears to be attractive project.

Product Development – Part 5

After estimating project cash outflows we need to do the same with cash inflows. Here the risk factor plays bigger role because we look further ahead. Imagine our project covers 7 years. How can we know how much we will sell 7 years from now or what our costs will be? We cannot be certain about anything. What numbers then to use in our NPV analysis? We need to build a model to determine the future free cash inflows.

Estimated product costs in period zero need to be escalated into future periods by expected inflation. On the other hand estimated sales we need to escalate by some kind of indicator reflecting the economic growth. The fact is we cannot be sure about any of these numbers.

For example inflation, what number to use in years 1, 2, 3 till 7. In most countries inflation data is commonly available. We can learn about its history. But still what numbers to use in the future? Can we forecast inflation? My opinion is we should not. Inflation is a unique time series, it is not independent. I mean inflation is controlled by actions taken by authorities. As a result inflation will be maintained within some acceptable range. Authorities will not allow it to go too high neither to assume negative value (deflation). My approach is to describe inflation history statistically as probability distribution. For example; ten years history of inflation (CPI) in the Republic of Botswana looks like that;



It is described in Excel by function; RiskExtvalue(0.078496,0.012537), what represent a probability distribution with parameters in brackets. Remember, we use add in to Excel called @Risk from Palisade Inc.

Now we need to escalate the estimated product costs to future periods by this probability distribution. Just by multiplying cells in Excel as in standard modeling. We may have the estimated product costs in year zero also expressed by probability distribution. We therefore escalate probability distribution product costs by probability distribution representing inflation.

Estimating sales
Again, the best what we can obtain are educated guesses. We can look for expert advice, run surveys among them or we can use another similar product historical sales data as a proxy for our product and forecast future sales. We can also include the economic growth factor like forecasted GDP growth. In the end with the help of @ Risk we obtain some numbers that will indicate sales ranges and associated with them frequencies. Each year sales will be represented by probability distribution. Sales less costs give us the free cash flow from the project. (there will be other items like interest, fees, taxes,..., ignored here)

We end up having all data in our NPV analysis expressed by probability distributions. How different it is when compared with standard Excel analysis. Very much, because we took care of risk. Now we can run simulation. (in the next post)

Product Development – Part 4

The fact is we cannot be absolutely sure about any particular outcome of our project. Yet we need to make a decision before spending money. That money may be not ours, it may belong to other people who invested in us, in our business. How can we make a decision?
Standard NPV and IRR analysis are not enough. We need to find a way of assessing the risk associated with our project. How to do it?
We need to reexamine all data, starting from the project outflows. This part is easy. Our outflows may be based on quotations from contractors. In such cases the responsibility for numbers accuracy falls on the contractor. But there are also the unexpected events. For example; natural disasters. There may also be additional cost forced on the contactors due to our fault. For example we did not get the necessary permissions from local authorities on time. That may be the authorities fault but our responsibility and cost to the contractor. The best is to examine database with cost history related to similar projects and compute by what percentage past project costs differed from original. That we can call contingencies. If we come to a conclusion that usually cost of similar projects were overrun by some percent that is what we need to include in our NPV analysis. The question is how? Use Excel with @Risk from Palisade. Inc ref www.palisade.com Using @Risk we can model these outflows as a probability distribution. For example; assume we predicted the project outlays are $10000 and history tells us we may overrun it by up to 30%, we will model it as follows; Triang (10000,10000,13000). 'Triang' is a probability distribution function added in to Excel by @Risk.


On horizontal axis we see varies amounts we can spend on this project. On vertical axis associated with them frequencies. 90% of all possible outflows will be between $10 076 and $12 329. As seen we provide for varies possibilities and their frequencies. One could say, this is the same as what if scenario analysis. Not the case. Each what if scenario assumes one only value of the variable. By using probability distribution, each scenario assume a value and associated with it frequency what makes this process much more accurate. We will see it in the next part.

Product Development – Part 3

In order to conduct net present value NPV and internal rate of return IRR we need;
• Determine the time scope of this analysis
Of course that will depend on the project life cycle. Remember, the product life cycle; concept, introduction, growth, maturity and decline. We need to try to earn desired return on new product at maturity stage. For example for new car model that will be about 4 years
• Determine the free cash flows from future periods
Forecast future revenue and expenses, build pro forma financial statements in order to determine free cash flow from product. This may be some task, with product revenue there are no problems but very often it is difficult to separate ‘this product costs’ from other costs or expenses within company. But possible, methodology of activity based costing may help.
• Determine the discount rate
Weighted average cost of capital WACC is commonly used as the discount rate. The discount rate is used to bring future free cash flows to present, to account for time value of money. WACC will depend on how the project is financed, in specific on cost of debt and cost of equity. Cost of equity is sometimes difficult to determine. Capital Assets Pricing Model is commonly used to estimate the cost of equity. (more on that later) Here I will only highlight, the discount rate should not be used subjectively it hould be based on market realities.
• Build the model
Of course Excel is appropriate toll for such model. Excel also provides variety of standard functions including NPV and IRR.

The model is simple comparison between future cash outflows (negative values) and cash inflows (positive values) from project but brought (discounted) to present. If the resulting value is positive the project will exceed our minimum expectations, if it is zero the project meets our minimum expectations, it is negative we should stop here, the project will not meet our expectations. The rate at which NPV equals zero is the internal rate of return. From the same model we can determine the payback period, in fact this is discounted payback period.

So assume we got our NPV as a positive number, let’s say $1 million. But how sure are we to achieve that. The fact is the probability of getting that million is 0%. More on that in part 4.

Product Development – Part 2

It is obvious, we will engage in new product development only when convinced that our expectations are met. But how can we know that? Before we spend most of the money on new product development we need to conduct project evaluation. In coming to this stage we had to spend some of the money already developing the product concept or initial design and market research. That allowed us to obtain estimates and asses the market. Now what to do with this data?
We need to start with assessing what do we want to gain from this product? The minimum would be to at least recover our investment. But that will leave us indifferent even if successful. Normally we would try to recover the investment and earn something on top of that. For example 10%, but will that satisfy us knowing that we can earn the same on the saving account while facing practically no risk. No, we want more than that because we need to compensate ourselves for the risk we take. Well, it also depends on the product nature and the market we face. I would say it is necessary to aim at no less than the best alternative return. We need to remember, the future is uncertain. Any deterministic analysis has 0% probability to occur. We need therefore to look for a range of returns with our minimum return better than our best alternative.
There are three prime analyses we need to conduct. The Net Present Value, Internal Rate of Return and the Pay Back Period. All of them have some short comings but there is nothing more universal to replace them with. Analysis in Part 3

Product Development - Part 1

The fact is if your business wants to gain above average earnings it must be able to offer attractive, unique or differentiated products to its market and do it consistently. In order to achieve that, your business needs product development strategy or at least a program. Where to start?

At first it is necessary to determine where is your market, what do they need and what can you do within your competencies. In the next step you need to develop the product design or at least specification. This is perhaps the most important. Consideration needs to be given to; who will use it, how will they use, in what conditions, what safety standards are required, are there any government regulations, what raw materials and components will be required, from whom at what prices and lead times, what processes will be used, how much production will cost, …… You need to create a check list.

There are also the marketing issues; how will you tell the market about your product, how to distribute it, what guaranties to give, what will be price, packaging,… And why would anyone buy for us? What are the competitive products or close substitutes? How does the product rate against them?

That is a lot to think about. Of course for businesses already operating who upgrade its offering this process is easier and done with more confidence.

Once the spec and design exist you will be able to determine the production process its requirements, and costing. What about price? You need to start at price the end user will pay. Then debit the expected retail and distribution margins to arrive ‘free on board’ price (FOB). How does it compare with costs? Are we in business? Not yet, there are still the capital costs to consider. How much do we need to invest in tooling, machinery, space,…
Having production costs and capital estimates still does not tell us whether we have a business case. We still need to do additional analysis in order to determine whether this project will be good for us. Net present value, internal rate of return and a business plan – in the next post.

Product Strategy – Part 3

Products are goods that have physical shapes characterized by dimensions and performing useful functions. Products are tangible, they consume raw materials and are properties. The last is very important. As a property or intellectual property products give its owners market advantage. However it depends on their nature to what extend such advantage can be realized.
Product types decide what we can do with them or what market we can service. There are four major market characteristics; perfect competition, monopolistic competition, oligopoly and monopoly.

Perfect competition, in theory perfection is assumed to exist where we have plenty of buyers and sellers in each product category and neither group is able to set prices. Forces of supply and demand set price equilibrium for each product. Products in this category are identical and common, produced by many and consumed by millions. No one owns the product rights as intellectual properties. For example agriculture products like grain or potatoes, or a paving brick. Because market sets the prices the only strategy available to such producers is low cost. OK there may be few extra things like product quality, availability, appearance, and even discounting by the producers. But why to discount if you can sell all at market prices at least in environment of perfect competition.
Monopolistic competition, here, there are still many producers (sellers) and buyers but products are differentiated. The functions they perform or the way they look,.., are specific or unique. They represent intellectual property. For example photo camera or CD player. However close substitutes exist what along with the demand schedule sets a ceiling on price. Differentiation strategy is most common here.
Oligopoly is like monopolistic competition but with fewer producers, the products tend to be more complex and more technologically advanced. For example cars. Strategies employed in this market are based on differentiation and focus.
Monopoly is very rare and even if exist most likely it will be regulated. Utility companies are close to monopolies but not because of uniqueness of their products (or services). That is mainly due to acquiring exclusive rights to operate.

Products do have life cycles. That requires regular product development efforts and payment of associated with them costs. That calls for product development strategy. It is amazing that only small number of companies do something about that. Can anyone imagine a motor company offering the same model for more than 20 years? Companies that have innovation culture prosper. More on that in the next post.