Infrastructure Forecast Methodologies
Industry Forecasts
BMI’s industry forecasts are generated using the best-practice techniques of time-series modelling and causal/econometric modelling. The precise form of model we use varies from industry to industry, in each case being determined, as per standard practice, by the prevailing features of the industry data being examined. BMI mainly uses ordinary least squares (OLS) estimators and in order to avoid relying on subjective views and encourage the use of objective views, uses a ‘general-to-specific’ method. BMI mainly uses a linear model, but simple non-linear models, such as the log-linear model, are used when necessary. During periods of ‘industry shock’, for example a deep industry recession, dummy variables are used to determine the level of impact. Effective forecasting depends on appropriately selected regression models. BMI selects the best model according to various different criteria and tests, including, but not exclusive to:
- R2 tests explanatory power; Adjusted R2 takes degree of freedom into account;
- Testing the directional movement and magnitude of coefficients;
- Hypothesis testing to ensure coefficients are significant (normally t-test and/or P-value);
- All results are assessed to alleviate issues related to auto-correlation and multi-collinearity.
BMI uses the selected best model to perform forecasting.
It must be remembered that human intervention plays a necessary and desirable role in all of BMI’s industry forecasting. Experience, expertise and knowledge of industry data and trends ensures that analysts spot structural breaks, anomalous data, turning points and seasonal features where a purely mechanical forecasting process would not. Within the infrastructure industry, this intervention might include, but is not exclusive to, new investments across sectors or cancelled projects; general investment climate and business environment changes; changing domestic or regional trends; macroeconomic indicators; and regulatory changes.
Example Of Construction Value Model(Construction value)t = β0 + β1*(GDP)t + β2*(inflation)t + β3*(lending rate)t + β4* (population)t + β5*(government expenditure)t + β6*(construction value)t-1 + εt
Construction Industry
A number of principal criteria drive our forecasts for each construction and engineering variable:
Construction GDP And Infrastructure Spending
Figures for construction GDP and infrastructure spending are based, where possible, on national accounts as published by relevant central banks, as well as primary government/ministry sources and official data. Where these are unavailable, construction GDP forecasts are based on a range of variables including:
- Stated infrastructure and development programmes;
- Likely increases owing to related urban or industrial sector developments;
- Political factors (such as an electorally motivated public works programmes).
Construction as a percentage of GDP is calculated using BMI’s own macroeconomic and demographic forecasts.
Employment Within The Construction Industry
These figures are forecast based on:
- The growth or otherwise of the construction industry;
- Company results and expansion plans.
Data Methodology
Construction
Construction Value
Our data is derived from GDP by output figures from each country’s national statistics office (or equivalent). Specifically, it measures the output of the construction industry over the reported 12 month period in nominal values (i.e. domestic currency terms). As it is derived from GDP data, it is a measure of value added within the industry (i.e. the additional contribution of the construction industry over other industries, such as cement production). Consequently, it does not measure the nominal value of all inputs used in the construction industry, which, for most states would increase the overall figure by 50-60%. Furthermore, it is important to note that the data does not provide an indication of the total value of a country’s buildings, only the construction sector’s output in a given year.
This data is used because it is reported by virtually all countries and can therefore be used for comparative purposes. However, it is important to note that, where we are able to locate them, data released by national statistical offices or industry groups or associations for the overall value of the construction sector also taken into account and published by us.
Growth
Our data and forecasts for real construction measures the real increase in output (rather than nominal growth, which would also incorporate inflationary increases). In short, it is an inflation adjusted value of the output of the construction industry year-on-year. Consequently, real growth will – in virtually all instances – be lower than the nominal growth of our ‘construction value’ indicator.
Construction Industry, % Of GDP/Construction Value (US$)
These are derived indicators. We use BMI’s Country Risk team’s GDP and exchange rate forecasts to calculate these indicators.
Capital Investment
Total Capital Investment
Our data is derived from GDP by expenditure data from each country’s national statistics office (or equivalent). It is a measure of total capital formation (excluding stock build) over the reported 12 month period. Total capital formation is a measure of the net additions to a country’s capital stock, so takes into account depreciation as well as new capital. In this context, capital refers to structures, equipment, vehicles etc. As such, it is a broader definition than construction or infrastructure, but is used by BMI as a proxy for a country’s commitment to development.
Capital Investment (US$), % Of GDP, Per Capita
These are derived indicators. We use our Country Risk team’s population, GDP and exchange rate forecasts to calculate them. As a rule of thumb, we believe an appropriate level of capital expenditure is 20% of GDP, although in rapidly developing emerging markets it may, and arguably should, account for up to 30%.
Government Capital Expenditure
This is obtained from government budgetary data and covers all non-current spending (i.e. spending on transfers, salaries to government employees, etc.). Due to the absence of global standards for reporting budgetary expenditure, this measure is not as comparable as construction/capital investment.
Government Capital Expenditure, US$bn, % Of Total Spending
These are derived indicators.
Construction Sector Employment
Total Construction Employment
This data is sourced from either the national statistics office or the International Labour Organization (ILO). It includes all those employed within the sector.
Construction Employment, % y-o-y; % Of Total Labour Force
These are derived indicators.
Average Wage In Construction Sector
This data is sourced from either the national statistics office or the ILO.
Infrastructure Business Environment Ratings
BMI’s Infrastructure Business Environment Ratings (IBER) provide a numerically based evaluation of prospects for the infrastructure sector in each state that we cover. Our approach is threefold. Firstly, the risks rated capture the operational dangers to companies operating in this industry globally. Secondly, we attempt, where possible, to identify objective indicators that may serve as proxies for indicators that were previously evaluated on a subjective basis. Finally, we use BMI’s proprietary Country Risk Ratings (CRR). Overall, the ratings system – which is integrated with those of all the industries covered by BMI – offers an industry-leading insight into the prospects and risks for companies across the globe.
Ratings Overview
Conceptually, the ratings system is divided into two distinct areas:
Limits Of Potential Returns
An evaluation of sector’s size and growth potential in each state and also broader industry/state characteristics that may inhibit its development.
Risks To Realisation Of Returns
An evaluation of industry-specific dangers and those emanating from the state’s political/economic profile that call into question the likelihood of anticipated returns being realised over the assessed time period.
For each category and sub-category, each state is scored out of 100 (100 being the best), with the overall IBER a weighted average of the total score. Importantly, as most of the countries and territories evaluated are considered by BMI to be ‘emerging markets’, our IBER is revised on a quarterly basis. This will ensure that the IBER draws on the latest information and data from across our broad range of sources, and the expertise of our analysts.
Table: Infrastructure Business Environment Indicators |
|
Indicator |
Rationale |
Limits to potential returns |
|
Market structure |
|
Construction expenditure, US$bn |
Objective measure of size of sector – the larger the sector, the greater the opportunities available |
Sector growth, % y-o-y |
Objective measure of growth potential – rapid growth results in increased opportunities |
Capital investment, % of GDP |
Proxy for the extent the economy is already oriented towards the sector |
Government spending, % of GDP |
Proxy for extent to which structure of economy is favourable to infrastructure/ |
Country structure |
|
Labour market infrastructure |
From BMI’s Country Risk Ratings (CRR). Denotes availability/cost of labour. High costs/low quality will hinder company operations |
Financial infrastructure |
From BMI’s CRR. Denotes ease of obtaining investment finance. Poor availability of finance will hinder company operations across the economy |
Access to electricity |
From BMI’s CRR. Low electricity coverage is proxy for pre-existing limits to infrastructure coverage |
Risks to potential returns |
|
Market risk |
|
No. of companies |
Subjective evaluation against BMI-defined criteria. This indicator evaluates barriers to entry |
Transparency of tendering process |
Subjective evaluation against BMI-defined criteria. This indicator evaluates predictability of operating environment |
Country risk |
|
Structure of economy |
From BMI’s CRR. Denotes health of underlying economic structure, including seven indicators such as volatility of growth; reliance on commodity imports, reliance on single sector for exports |
External risk |
From BMI’s CRR. Denotes vulnerability to external shock – principal cause of economic crises |
Policy continuity |
Subjective rating from BMI’s CRR. Denote predictability of policy over successive governments |
Legal framework |
From BMI’s CRR. Denotes strength of legal institutions in each state – security of investment can be a key risk in some emerging markets |
Corruption |
From BMI’s CRR. Denotes risk of additional illegal costs/possibility of opacity in tendering/business operations affecting companies’ ability to compete |
Source: BMI |
|
Project Finance Ratings
BMI’s Project Finance Ratings (PFR) provide a globally comparative, numerically based assessment of the risks facing major infrastructure projects in the power, transport, and commercial construction sectors. It evaluates the degree of uncertainty facing projects that are generally characterised by the following: long construction period; high construction costs; difficulty in redeploying project assets (e.g. power stations) to other uses; earnings generated only after construction completed. The PFR draws on BMI’s broad analytical expertise. The methodology incorporates our industry-leading Country Risk Ratings (CRR), drawing on our 25-year expertise in assessing political, economic and business operational risk, as well as our in-depth knowledge of the global infrastructure industry.
While we believe the resulting scores are a reliable guide to project finance risks, it should be emphasised that the PFR assesses broad industry risks, not individual projects. This has several implications. First, there will be instances where the risk profile, for example the supply of inputs, of particular projects is markedly different from general risks in the industry. Second, the PFR will not take into account measures by private sector project participants to mitigate risk when structuring finance – for example, by securing a substantial equity involvement from the sponsoring agency or government. The PFR is best used to evaluate the breadth and depth of risks facing major infrastructure projects, which will in turn affect the source, availability and cost of finance. Thus, in an environment of limited global finance for such projects, it provides a leading indicator for the cost of financing major projects and the pace at which infrastructure development will occur in each state.
Ratings Overview
To reflect the life-cycle of infrastructure projects, the PFR is divided into two distinct sub-ratings:
Design And Construction Risks
This evaluates risks in the broad assumptions underpinning construction cost projections. Specifically, it assesses uncertainty in the political, economic and regulatory environment, and input cost volatility.
Sector Operational Risk
This evaluates risks in revenue projections during the operational period of a project. It assesses uncertainty regarding supply and cost of inputs, and sale of outputs, including the regulatory, market and political environment.
Ratings Components
The rating uses 10 subjectively measured indicators and around 40 separate indicators/datasets. The weighting of each indicator and group of indicators (inputs, regulatory, market risks, etc) reflects their relative importance, and the relative risk level that they pose for sponsors and equity holders. The relative influence of each indicator and group of ratings on the final score can be found below. The score next to the segment sub-heading indicates the weighting of the segment in the final rating, and the score next to each indicator shows each indicator’s influence within the segment.
Table: Design And Construction Phase |
|||
Indicator |
Definition |
Rationale |
Weighting |
Inputs |
|
20% (segment) |
|
Domestic: inflation |
Average consumer inflation 2002-2009, adjusted for standard deviation |
High and uncertain inflation raises risks to input cost projections |
60% |
International: long-term currency volatility |
Standard deviation of monthly average of past 12 months of data, plus standard deviation of moving average |
Currency volatility increases risks to cost projections of imported goods |
40% |
Political environment |
25% (segment) |
||
Market orientation |
Measure of government intervention in economy, using data for government expenditure and revenue from state-owned enterprises; average trade tariff rates; tax levels; trade bureaucracy; and history of FDI inflows |
Governments with strong commitment to free markets are unlikely to make sudden changes to the investment/trade regime |
20% |
Security risk |
Measure of level of security threat (external and internal) facing a country |
The higher the security risk, the higher the risks to infrastructure assets in terms of physical security (as they tend to become targets) and insurance premiums, which rise with security risks, increasing costs to sponsors |
20% |
Long-term policy continuity |
BMI’s evaluation of level of broad governmental policy consistency over past decade |
Strong policy continuity between elites (within or across parties) minimises risks that new legislation will alter the business environment |
20% |
Characteristics of polity |
BMI’s evaluation of system of government and constitutional framework against ideal type |
Democratic governments with strong, independent institutions are less prone to sudden policy shifts |
20% |
Rule of law |
Evaluation of breadth and depth of government’s ability to protect individuals and property |
Strong rule of law reduces direct threats to assets during construction |
20% |
Legal/regulatory risks |
20% (segment) |
||
Corruption |
Subjective measure of level of corruption |
Transparency is essential to planning of predictable input delivery and cost and the predictability of officials’ decisions |
50% |
Contract enforceability |
World Bank Index of cost, procedures and time taken to recover a bad debt |
Confidence in the legally binding nature of contracts is essential to minimise domestic counterparty risk |
50% |
Economic/financial risks |
35% (segment) |
||
Domestic: economic stability |
BMI’s long-term economic rating, which incorporates 20 indicators to assess risk of an economic crisis |
Economic stability reduces risks to project activity (e.g. due to financial problems at suppliers) |
30% |
International: availability of finance |
US and eurozone average interest rates, adjusted for Chicago BOE VIX index |
Project finance is mainly raised internationally. Price and availability depend on US and eurozone interest rates and investor risk appetite (VIX is a proxy). A global, rather than country-specific, indicator |
70% |
Source: BMI |
|||
Table: Commissioning And Operating Phase – Commercial Construction |
|||
Indicator |
Definition |
Rationale |
Weighting |
Inputs |
|
30% (segment) |
|
Domestic: inflation |
Average consumer inflation 2002-2009, adjusted for standard deviation |
High and uncertain inflation raises risks to input cost projections |
30% |
Domestic: power, imports as % of consumption |
As stated. Power is used as proxy for all utilities |
Supply of utilities is essential to functioning of asset and revenue generation |
25% |
International: long-term currency volatility |
Standard deviation of monthly moving average of past 12 months of data, plus standard deviation of moving average |
Currency volatility increases risks to cost projections of imported goods |
45% |
Sale of outputs |
|||
Regulatory |
|
20% (segment) |
|
Supply risk |
BMI’s subjective view of transparency of government planning policy |
Clarity regarding future market supply essential to forecast demand for asset |
20% |
Price risk |
BMI’s subjective view of transparency of government policy regarding price of service related to asset |
Clarity over policy/regulations covering price are essential to projecting income |
20% |
Contract enforceability |
World Bank Index of cost, procedures and time taken to recover a bad debt |
Confidence in legally binding nature of contracts is essential for minimising domestic counterparty risk |
60% |
Market risks |
|
30% (segment) |
|
Economic stability |
BMI’s long-term economic rating, which incorporates 20 indicators to assess risk of economic crisis |
An economic crisis would cut projected demand, potentially greatly |
|
Long-term currency stability |
Standard deviation of monthly moving average of the past 12 months of data, plus standard deviation of moving average |
Sharp currency movements introduces risks to value of income in international currency |
50% |
Political risks |
|
20% (segment) |
|
Market orientation |
Measure of government intervention in economy, using data for government expenditure and revenue from state-owned enterprises; average trade tariff rates; tax levels; trade bureaucracy; and history of FDI inflows |
Governments with strong commitment to free markets unlikely to make sudden changes to investment/trade regime |
10% |
Security risk |
Measure of the security threat (external and internal) facing a country |
The higher the security risk, the higher the risks to infrastructure assets in terms of physical security (as they tend to become targets) and insurance premiums, which rise with security risks, increasing costs to sponsors |
40% |
Policy continuity |
BMI’s evaluation of level of policy consistency over past decade |
Strong policy continuity between elites (within or across parties) minimises risks that new legislation will alter the business environment |
10% |
Rule of law |
Evaluation of breadth and depth of government’s ability to protect individuals and property |
Strong rule of law reduces direct threats to assets |
40% |
Source: BMI |
|||
Table: Commissioning And Operating Phase – Energy And Utilities |
|||
Indicator |
Definition |
Rationale |
Weighting |
Inputs* |
|
30% (segment) |
|
Inflation |
Average inflation, 2002-2009, and its standard deviation |
Volatile inflation risks unanticipated cost increases that may be impossible to pass on to asset users |
30% |
Crude price costs |
BMI’s Brent Crude forecasts for next five years, adjusted for standard deviation of prices over past three years |
Gas and other fuel prices correlate closely with oil. Price stability is desirable, as are anticipated future trends. This is a global, rather than country-specific, risk |
25% |
Long-term currency volatility |
Standard deviation of monthly moving average of the past 12 months of data, plus standard deviation of moving average |
Currency volatility increases risks to cost projections of imported goods or goods bought in US dollars (i.e. fuel feedstock) |
45% |
Sale of outputs |
|||
Regulatory |
|
20% (segment) |
|
Demand risk |
BMI’s subjective view of government energy policies and implications for industry demand |
Transparency regarding government energy policies is essential for evaluating demand |
20% |
Price risk |
BMI’s subjective view of transparency of government policy regarding power prices |
Clarity over policy/regulations covering price essential to projecting income |
20% |
Contract enforceability |
World Bank Index of cost, procedures and time taken to recover a bad debt |
Confidence in legally binding nature of contracts is essential to minimising domestic counterparty risk |
60% |
Market risks |
|
30% (segment) |
|
Economic stability |
BMI’s long-term economic rating, which incorporates 20 indicators to assess risk of economic crisis |
An economic crisis would cut projected demand, potentially greatly |
50% |
Long-term currency stability |
Standard deviation of monthly moving average of the past 12 months of data, plus standard deviation of moving average |
Sharp currency movements introduces risks to value of income in international currency |
50% |
Political risks |
|
20% (segment) |
|
Market orientation |
Measure of government intervention in economy, using data for government expenditure and revenue from state-owned enterprises; average trade tariff rates; tax levels; trade bureaucracy; and history of FDI inflows |
Governments with strong commitment to free markets internally and internationally unlikely to make sudden changes to investment/trade regime |
10% |
Security risk |
Measure of the security threat (external and internal) facing a country |
The higher the security risk, the higher the risks to infrastructure assets in terms of physical security (as they tend to become targets) and insurance premiums, which rise with security risks, increasing costs to sponsors |
40% |
Policy continuity |
BMI’s evaluation of level of broad governmental policy consistency over past decade |
Strong policy continuity between elites (within or across parties) minimises risks that new legislation will alter the business environment |
10% |
Rule of law |
Evaluation of breadth and depth of government’s ability to protect individuals and property |
Strong rule of law reduces direct threats to assets |
40% |
* No distinction between internal and domestic risks. This reflects BMI’s view that all projects would have fuel feedstock contracts in place prior to construction. Source: BMI |
|||
Table: Commissioning And Operating Phase – Transport |
|||
Indicator |
Definition |
Rationale |
Weighting |
Inputs |
|
30% (segment) |
|
Domestic: inflation |
Average inflation, 2002-2009, and its standard deviation |
Volatile inflation risks unanticipated cost increases that may be impossible to pass on to asset users |
40% |
International: long-term currency Volatility |
Standard deviation of monthly moving average of the past 12 months of data, plus standard deviation of moving average |
Currency volatility increases risks to cost projections of imported goods |
60% |
Sale of outputs |
|
|
|
Regulatory |
|
20% (segment) |
|
Demand risk |
BMI’s subjective view of government regulation and its record in supporting substitutes etc |
Transparency regarding government policy essential for evaluating demand |
20% |
Price risk |
BMI’s subjective view of transparency of government policy regarding price of service related to asset |
Clarity over policy/regulations covering price essential to projecting income |
20% |
Contract enforcibility |
World Bank Index of cost, procedures and time taken to recover a bad debt |
Confidence in legally binding nature of contracts essential to minimising domestic counterparty risk |
60% |
Market risks |
|
30% (segment) |
|
Economic stability |
BMI’s long-term economic rating, which incorporates 20 indicators to assess risk of economic crisis |
An economic crisis would cut projected demand, potentially greatly |
50% |
Long-term currency stability |
Standard deviation of monthly moving average of the past 12 months of data, plus standard deviation of moving average |
Sharp currency movements introduces risks to value of income in international currency |
50% |
Political risks |
|
20% (segment) |
|
Market orientation |
Measure of government intervention in economy, using data for government expenditure and revenue from state-owned enterprises; average trade tariff rates; tax levels; trade bureaucracy; and history of FDI inflows |
Governments with strong commitment to free markets are likely to refrain from sudden changes to the investment/trade regime |
10% |
Security risk |
Measure of the security threat (external and internal) facing a country |
The higher the security risk, the higher the risks to infrastructure assets in terms of physical security (as they tend to become targets) and insurance premiums, which rise with security risks, increasing costs to sponsors |
40% |
Policy continuity |
BMI’s evaluation of level of broad governmental policy consistency over past decade |
Strong policy continuity between elites (within or across parties) minimises risks that new legislation will alter the business environment |
10% |
Rule of law |
Evaluation of breadth and depth of government’s ability to protect individuals and property |
Strong rule of law reduces direct threats to assets |
40% |
Source: BMI |
|||
Sources
BMI uses publicly available information to compile the country reports and collate historical data. Sources used in Infrastructure reports include UN statistics; national accounts; infrastructure, public works, transport, energy and economy ministries; officially released company results and figures; trade bodies and associations and international and national news agencies.
