Frequently asked questions
Q: Where does the data come from?
A: The basic research data is produced by a single analyst on demand. In other words, once a client subscribes to our service, we commission an analyst at our subcontractor to produce a full model of the client company's data, including realized historical figures and projections for future performance. Since we provide a full DCF valuation engine, we need to project the company's performance "until infinity" (i.e., until a steady state is reached in terms of growth, and a terminal value can be calculated) to determine a fair value.
Q: What are the future projections based on?
A: The analyst building the base model first makes projections until terminal value (10 years into the future) based on their best judgement. To this base model, we then add consensus estimates (averages of the estimates made by several analysts) of the most important variables: net sales, operating profit, pre-tax profit, EPS and dividend per share. These figures are typically available for a couple of years into the future, with more detailed forecasts often available for the next few quarters. These consensus estimates overwrite the projections by the individual analyst for the years where a consensus is available. The system then automatically reconciles the model to remain consistent even after the consensus figures have been introduced.
Q: How do you collect the consensus estimates?
A: There are several options for how this can be done. We can have the analysts send their reports directly to one of our subcontractors (e.g., Vara Research) and have them compile the consensus. The client's IR department could also compile the consensus themselves, and simply input the figures into our system. Or, we can use an existing consensus estimate service (such as SME Direkt) where one is available for the client company.
Q: How is consensus data shown in the system?
A: Consensus estimates are shown with a highlighted background. A mouse-over of a consensus figure shows how many estimates were used to compile the consensus. Clicking the consensus figure shows a more detailed breakdown in a pop-up window, including minimum, maximum, mean and median figures.
Q: How are key figures calculated, after consensus estimates and basic research data have been combined?
A: The key figures are mostly based on the consensus estimates, although the base model (i.e., the individual analyst's research) also has influence on the key figures. Here are a couple of practical examples:
The price to earnings (P/E) ratio is calculated using the earnings per share (EPS) obtained directly from consensus estimates, and the previous day's closing share price, retrieved from an external data provider. The analyst's research figures are therefore not used in calculating P/E ratio.
The share price in the price to book value (P/BV) ratio is still the previous day's closing price, but the book value (BV) is taken from the balance sheet, which is based on the data entered by the analyst. In practice, the consensus estimates also play a major role in determining the BV. E.g., the current year's BV is calculated from the previous year's BV, paid dividend, and the current year's earnings. The first two are historical figures, and therefore not subject to debate. The earnings are obtained from the consensus estimate.
The same is true for most key financial ratios (e.g., ROI, ROE) and valuation figures (e.g., EV/EBIT, EB/EBITDA, dividend yield, PEG). While they draw on several items from the income statement and balance sheet, discretionary elements (earnings forecasts by an individual analyst) are rarely included in the calculation. The ratios are based on realized historical data, objective figures (e.g. number of shares) and consensus estimates.
In some cases the ratios do include estimates from the analyst, but often the effect of small variations in these values is minimal. For instance, in calculating ROI, operating profit (a consensus estimate) is divided by invested capital. While invested capital is primarily derived from historical data and consensus estimates, the (discretionary) judgement of how much non-interest bearing debt is included in the balance sheet does have some effect on the ratio. In these cases the estimates seldom differ much from the realized historical figures (historical proportion of non-interest bearing debt) and even with a 30% forecasted change from current levels, the effect on the ratio itself (e.g., ROI) is marginal.
Q: How do you determine divisional, quarterly, and other detailed figures if these are not available in consensus estimates?
A: Outside consensus estimates seldom include divisional estimates, and sometimes even quarterly figures from some companies are not available. In this case, the figures are generated based on higher level consensus estimates. For instance, divisional figures are calculated based on group level consensus estimates, using historical divisional figures to determine the allocation to each division. Similarly, if quarterly estimates are not available, then historical quarterly performance is used to allocate the annual estimates to quarters - this preserves any historical cyclicality within the year.
|
