Abstract

Performing production cost and profitability analysis of the new technology

The assessment of production costs of newly developed technologies starts with the establishment of a process scheme. Based on these schemes, basic input–output mass and energy flow analyses can be performed. The input–output balance is broken down into raw material energy and other required materials (e.g., chemicals) and for different received outputs. The costs and revenues are then defined using this balance and price (the wholesale price paid by a manufacturer) data, leading to an operating profit estimate. Furthermore, considering costs of equipment, infrastructure, personnel, maintenance, and other fixed costs (e.g., sales costs and plant overheads) in addition to these variable production costs, the overall profitability of the technology can be assessed when making assumptions on production revenues. Usually, intensive sensitivity analyses are performed in combination with such calculations to consider uncertainties in the process assumptions–based business environment (Hytönen and Stuart 2012).

For the production cost and profitability analysis of the case study, a traditional early stage process design method was used. First, the technical assessment was conducted. The overall production concept was designed together with researchers and equipment manufacturers. The process parameters were gathered from experimental research results, literature, and mass and energy balances of the process in selected scenarios where models were made using a process simulator (Balas; balas.vtt.fi). Then, for the economic assessment, expected prices and price ranges for the feedstock, energy, chemicals, and other consumables were gathered. Then variable costs based on the prices and the mass and energy balance were defined in all of the scenarios and fixed costs were defined. Fixed costs can include capital charges, labor, maintenance, and other cost considerations. Maintenance and other fixed costs are defined as using factors commonly used in the chemical process design. For labor costs, the expected personnel needs are estimated and capital cost estimates converted into capital charges using a capital recovery factor.

Prices and a fixed cost basis used in the economic analysis are listed in Table 2. The bark is assumed to be obtained without any costs. Similarly, the extracted bark is assumed to be priced to zero. Even though a part of the bark is extracted and, as a result, the total energy flow to the boiler is decreased, the cosolvent losses to the bark and the drying will increase the heat content of the extracted bark. These changes in the energy flow to the boiler are included in the cost model. Incremental energy production (steam and electricity) is modeled and is seen as extra cost or added revenue in the cost figures.

Table 2. Prices and fixed cost calculation basis.

View Fullscreen

Performing a B2B survey

B2B surveys are used as a tool to generate market information, especially when the developed products are aimed at new markets and applications beyond the involved companies experience and in cases where the available secondary data are not sufficient (Stern and Schwarzbauer 2008, Stern 2009). Depending on the target product, three major issues can be addressed by these kinds of surveys:

• Potential product price

• Potential sales volumes

• Customer requirements (technical, legal, economic)

It is very important to realize how these three variables are frequently interrelated. Therefore, these three variables should not be addressed separately. A price needs to be related to certain product properties and sales volumes and vice versa. Consequently, an initial choice between basic options has to be made:

• Product properties are fixed at minimum or principal requirements of the processing companies. Therefore, price and volume can be measured at defined levels (e.g., maximum volume–mass strategy, maximum price–niche strategy, and current or average market).

• Price is fixed at the substitute's price level (e.g., commodity markets). In such cases, sales volumes can be usefully defined in relation to varying product properties.

• Sales volumes are limited (e.g., specialty markets). In such cases, potential product prices can be usefully defined in relation to varying product properties.

• If volume and product properties are more or less fixed contingent valuation methods (willingness to pay, auction experiments, and Price Sensitivity Meter [PSM]), they can easily be applied.

• In other cases, all three factors need to be restricted in a predefined range to allow the application of choice experiments (analytical hierarchy process, conjoint, and discrete choice).

For investigating the potential price of a product that does not exist yet, several techniques based on contingent valuation approaches (Mitchell and Carson 1989) are available, such as direct approach or conjoint analysis (Weiner and Zacharias 2004).

In all cases, bias and a lack of knowledge about their directions and dimensions need to be taken into account. According to Mitchell and Carson (1989), the existence of strategic and compliance bias must be considered. Furthermore, a lack of motivation due to the hypothetical character of the question may lead to an increase of biased effects.

For the case example, an online survey among relevant processing companies in Germany, Switzerland, and Austria was used to perform the B2B survey. As described by Stern (2009), the survey was limited to these three countries in order to obtain a manageable population in terms of language, available time, and performance results. As the population included people who worked for many international companies operating all over the world, the results were not restricted to this base area. Companies were identified using the information found at Datamonitor's Research Store (http://www.datamonitor.com/store). Hence, only companies with a Web page were included in the population. In total, 164 companies were identified, of which 36 were producers of functional additives and plant extracts, 58 companies operated in the food industry, and 70 companies operated in the dietary supplement industry (Table 3).

Table 3. Companies participating in the survey.

View Fullscreen

First, a contact e-mail was sent to the selected companies explaining the aim and procedure of the study. A few days later, the companies were called by phone, if necessary, to identify the right contact person for the survey and to make an appointment regarding survey participation. Target persons were employees involved in the application of active agents, such as R&D managers and product or production managers. As an incentive for participating in the study, the companies were promised a short report of the survey results. Compared with traditional business interviews via telephone, the online survey guarantees more anonymity, especially for sensitive questions regarding prices or potential amounts sold. Additionally, visualization within an online survey was necessary with respect to the complexity of some of the questions.

In order to investigate potential product prices, the PSM developed by van Westendorp (1976) was used. It delivers an acceptable price range as a result and can be nicely utilized for a sensitivity analysis of the profitability of a new technology in terms of production cost or investment. The PSM approach asks four price-related questions that are then evaluated as a series of four cumulative distributions (Reinecke et al. 2009). The four price questions are related to the price expectations of the product:

• Price at which the product is too expensive to consider

• Price at which the product would start to get expensive

• Price at which the product would be too cheap to consider as a high-quality product

• Price at which the product would be a bargain (cheap)

With these four questions, the optimal price range can be defined. The answers to the previously mentioned questions are grouped into frequency distributions, and then the cumulative frequencies are plotted on a graph. The percentage of respondents is shown on a vertical axis with the price points on a horizontal axis. The cumulative distributions “cheap” and “expensive” are presented in the graph in such a way that they will cross each other, showing an indifference price point. The “indifference price point” represents the average price of an existing product or the market price of a dominant market leader. The indifference price is based on the experience of people with recent price levels in the market; it has been shown that the indifference price changes when conditions in the market change (van Westendorp 1976). The indifference price can also vary because of various submarkets where people might buy only cheap or expensive products. The distributions “too cheap” and “too expensive” show the “optimal pricing point.” At this price point, an equal number of people believe that a product is too cheap or too expensive. This price point represents graded resistance from the market participants and the highest market penetration.

The graphs “expensive” and “cheap” can be reversed, yielding the “not expensive” and “not cheap” distributions. If they were combined with the original “too expensive” and “too cheap” distributions, we get two new intersections. These points were called “point of marginal cheapness” and “point of marginal expensiveness” (van Westendorp 1976). The point of marginal cheapness is the price where the number of people who experience a product as too cheap is larger than the number who experience it merely as cheap. The same thing happens at the point of marginal expensiveness, where the number of people experiencing a product as too expensive is larger than the number of those experiencing the product as expensive. The range between those two points is called the range of acceptable prices.

Assessing the market impact of new capacities

When preforming a market impact assessment, usually econometric methods (see Wooldridge 2009) are used to calculate supply elasticities in order to evaluate price–quantity relationships. Basic time-series data about the value of the product and the quantities produced or traded are needed. However, these data are often not available for the product in question. In such cases, a detour (by means of using an indirect way for approaching the same goal) based on the idea of analogue markets and best available estimates is necessary. The aim is to find a similar product (in the broadest sense) for which the data of prices and quantities are freely available. Once a similar product is found, an analysis of the price–quantity relationship can be carried out in order to draw conclusions about the price–quantity behavior of the product in question.

After the selection of one or more similar products, the prices and quantities of that proxy product are analyzed. The concept of price elasticity helps quantify the extent of the expected price change, which is then adapted to the market of the product in question. Hence, it is very important to be aware that, due to rather strong assumptions, the validity of the conclusion is subject to those assumptions. Unless price-fixing agreements of the suppliers or other market disruptions, such as trade tariffs, trade restrictions, or legal price controls, distort a market, it is the price mechanism that brings supply and demand into equilibrium.

There are several reasons why demand and supply are able to shift over time. In the case of demand, it could simply be that expectations or tastes change. It could also mean that income increases or that prices of related goods decrease. For the supply, similar reasons are true. This shows an increase or a decrease of input prices, advancement in technology, or a change in expectations. Increasing prices are due to either a reduction in the supply or an increase in the demand. Finally, prices decrease because the underlying determinants of supply become stronger or the demand-shaping determinants become weaker. The fundamental question to be answered by the market impact assessment for the case example is, How could prices of bioactive substances react on an increase in the supply of such a substance?

This increase in supply is assumed to be due to the construction of a new production site that, for its size, has the potential to contribute significantly to the quantities traded worldwide and thus has an impact on prices. In order to conduct an econometric analysis to assess the impact of additional product volumes on a small market, a complete and standardized time-series data set was needed. As no such data set is available for bioactive substances in question, a proxy product for which such data are easily accessible had to be found.

The Comtrade database of the UN Statistics Division (United Nations 2012) was accessed to search for potential proxy products. A multiple step-by-step process, including data analysis and fundamental and historical product research, took place to select the most appropriate product from an initial list of 14 products. Selection criteria were the following:

• Natural product (ideally plant or even wood based)

• No obvious raw material supply constraints

• Application in food or pharmaceuticals

• High value–low volume

• A history of growing production capacities

• Comparable stable demand situation

Finally, it was decided to analyze licorice (Glycyrrhiza glabra) as a proxy for bioactive substances because of the best overall fulfillment of the above criteria. The concept of elasticities is used to quantify the relationship between a dependent and an independent variable. As such, the price elasticity of demand measures the percent change in quantity demanded in reaction to a 1 percent change in the price. Since it is impossible, without an in-depth analysis, to observe all of the necessary data to display supply-and-demand curves on a yearly and worldwide basis, no statement can be given about the underlying force of a price change. It may be that, in the case of a world price increase, supply-side shocks, such as drought, flooding, disease, or infestation, took place and destroyed bigger parts of the stock or that demand-driven shocks, such as larger or richer populations, were driving prices up. What can be done is analyze the available information (i.e., prices and quantities at certain times) in the long run. When analyzing the development of quantities and corresponding prices of licorice over time where prices are falling and quantities are increasing, it may be argued that the forces driving an increase in supply were stronger in the long run. And as it might be true for many agricultural and agricultural-based products, tremendous advances in production technology reduced the cost of inputs or substituted the input factor labor. This then increased overall supply and thereby decreased product prices.

The focus of this short analysis is the quantification of a price change, which was assumed to be induced by a supply shock, namely, the construction of the production site, which was able to significantly influence the quantities traded on the world market. Because the data about supply and demand are not available for the short term, reactions and the transition to an equilibrium cannot be studied. Yet this does not hinder the analysis of long-run effects.

In order to quantify the price–quantity relationship of licorice, basic time-series analysis techniques were applied. The Comtrade database that was used provides both value (in US dollars) and volumes of exports and imports for each country. This allows for the calculation of world prices as the weighted average of all country unit prices where weights are determined by calculating the market share of a country. In this study, prices (given in US dollars) were based on the data for exports and were then deflated by the US Consumer Price Index in order to eliminate misinterpretations due to inflation.

The Comtrade database offers data not about the production of a good but rather about the quantities traded between one country and another. For this reason, aggregated exports are used as a proxy for the worldwide production of licorice.

Before analyzing the relationship of prices and quantities in a regression, the two variables were transformed by taking a natural logarithm. This is often done in economics for interpretational reasons in order to normalize the distribution of input data, among other reasons (Baltagi 2008, Wooldridge 2009).

After regressing the log-transformed price for the log-transformed quantity, the Durbin-Watson test suggested correcting for a serial correlation. For this reason, the Prais-Winsten estimator was applied, which uses the generalized least-squares method to estimate the parameters in a linear regression model where the errors are assumed to follow a first-order autoregressive process (e.g., Andrews 1993). Furthermore, observations were systematically reduced in order to check for the sensitivity of the regression.