Skip to main content
Skip table of contents

2. Private Equity Shadow Pricing Approach

  • With the factor model calibrated prices, we shadow price the privateMetrics Universe of eligible private companies.

  • Universe excludes publicly listed companies, entirely government-owned companies, infrastructure companies, inactive companies, and those that are unlikely to be traded by private equity funds based on their size.

  • When data is missing for non-key factors, we impute such values based on its historical financials or country-activity median values.

  • Outliers in the data are winsorised or trimmed, to obtain a meaningful universe.

In this page, we detail our methodology to shadow price the companies in the privateMetrics Universe based on the factor model which is calibrated to private market transactions. Specifically, this page covers:

  • Filtering the universe for eligible companies

  • Treatment of missing data

  • Treatment of outliers

Universe Filters

The privateMetrics universe is defined to accommodate a broad global coverage, and hence from that perspective, we implement the below filters:

  • Publicly listed companies are excluded from the time of being publicly listed.

  • Companies that are owned entirely by governments are excluded.

  • In case, both subsidiaries and parent companies are observed, preference is given for the subsidiary as we want the eligible private companies to be the ones that are closest to their business activity.

  • Companies that are part of the infraMetrics Universe or are eligible for that are excluded.

  • Private companies that declare dissolution, bankruptcy, or in liquidation are excluded from that point in time when such corporate change happens.

  • Finally, a minimum revenue filter of U.S. $ 1 million. To accommodate company valuation histories when they start small and eventually grow big, we implement this filter as an average revenue of over U.S. $ 1 million at the time of being eligible.

Missing Factors

In our data collection process, despite the due diligence exercised, there arise scenarios where we are unable to obtain data for a specific factor. In such cases, after ensuring these are only a small proportion of our sample and not the majority, we impute the missing values. For performing imputations, we follow a specific hierarchy of methods.

First, when available past standardised historical values are used. For example, if EBITDA for 2023 is required, but EBITDA/Revenue for 2022 is available, we apply the past year EBITDA/Revenue on 2023 Revenue to generate EBITDA.

If past historical financials are not available then cross-sectional sample medians of standardised values (e.g., ratio or percentages) are used focusing on the country & activity sector of the private company with missing information.

Note that key factors such as size, leverage, etc., and PECCS dummies are not imputed, and private companies without data on these factors are not priced.

Outlier Treatments

Extreme outliers can affect the model and also skew the indices and analytics. To meaningfully tackle outliers and not allow them to affect our model, indices, and analytics, we perform the below:

  1. Some factors that are prone to outliers such as EBITDA margin are winsorised, i.e., extreme values are replaced by fixed upper and lower percentiles from the distribution.

  2. Transactions included in the model calibration are trimmed to exclude outliers.

  3. Valuations that are estimated to be extreme both in terms of levels ($ value) or ratios (Price-to-sales) are excluded, and such companies are eliminated from the entire privateMetrics Universe.

JavaScript errors detected

Please note, these errors can depend on your browser setup.

If this problem persists, please contact our support.