The frequency of observed transactions incorporated into the privateMetrics database is approximately 25 transactions per month. Between 2013 and 2022, over 300 transactions were observed annually. Scientific Infra & Private Assets (SIPA) uses these transactions to calibrate its asset pricing model monthly, allowing the indices to reflect newly obtained factor prices from these updates.

Here’s a more detailed explanation:

• Monthly Calibration: The privateMetrics asset pricing model is recalibrated monthly using the latest available transaction data, ensuring the model reflects the current state of the market.
• Transaction Volume: SIPA observes over 300 transactions per year, which translates to more than 25 transactions each month. The baseline factor model is constructed using 5,438 private market transactions that occurred globally between 1999 and 2022.
• Data Sources: These transactions are sourced from PitchBook, Capital IQ, fund manager annual reports (entry and exit values only), and processed via artificial intelligence.
• Model Updates: Each transaction is used to update the model through a dynamic estimation process using a Kalman filter, which retains the memory of past factor prices while allowing the new transaction to influence the relationship.
• Factor Prices: As recent private market transactions are updated, newer factor prices are produced. Factor prices refer to the premium (or discount) that an investor is willing to pay to seek exposure to a specific factor of return in private companies.
• Financial Data Updates: Updates to the annual financials of the private company produce new factor loadings.

In the event of very few or no reported transactions in a given month, the factor prices estimated in the prior period will be carried over. Even with zero PE deals, updated factor loadings will still be obtained due to updated company financials and changes in the public markets, which will inform an updated valuation estimate for the entire universe.