1. (a) Purpose. The purpose of this section is to establish the methodology for forecasting fundable performance outcome counts to calculate performance tier funding amounts covering a time period for which performance data are not yet available. Using historical data for each public junior college, the Coordinating Board shall forecast each performance tier fundable outcome based on the applicable performance tier funding rules of this chapter in effect for the fiscal year for which the Coordinating Board is calculating performance tier funding. The Coordinating Board shall use these figures to calculate each performance tier payment for the fiscal year as established in this chapter.
2. (b) Methodology for Fiscal Year 2025 and Foundation Payment for Fiscal Year 2026. To calculate payment adjustments to Fiscal Year 2025 and the foundation payment for Fiscal Year 2026, the Coordinating Board shall forecast the total annual count of a fundable performance tier outcome for a public junior college using the exponential triple smoothing method of trend analysis with additive error, trend, and seasonality parameters applied to time series data. Except as specified in subsection (d)(1) of this section, this time series data shall use fundable certified data with the counts of fundable outcomes achieved annually by the public junior college during no fewer than the six most recent years for which data are available except as otherwise provided by subsection (c) of this section.

3. (c) Methodology. Except for forecasts conducted to calculate amounts of payments under subsection (b) of this section, the Coordinating Board shall forecast the total annual count of a performance tier fundable outcome for a public junior college using the model of those provided in subsection (b)(1)(A) of this section demonstrating the lowest root mean square error (RMSE) as established by time series cross validation, according to the following:

   1. (1) The model demonstrating the lowest RMSE for each fundable performance outcome for a public junior college will be selected and used for forecasting the total annual count of that fundable performance outcome for that public junior college.
   2. (2) For the purposes of forecasting total annual count of a fundable performance outcome for a public junior college, the Coordinating Board will use fundable certified data with the counts of fundable outcomes achieved annually by the public junior college during no fewer than the six most recent years for which data are available except as otherwise provided by subsection (d) of this section.
   3. (3) For the purposes of determining the lowest RMSE model for forecasting each fundable performance outcome for each public junior college for a given fiscal year, fundable certified data with the counts of fundable outcomes achieved annually by the public junior college no more recently than two years prior to the fiscal year for which the forecast is conducted will be included in model selection.

   4. (4) In forecasting performance tier fundable outcome counts, the Coordinating Board shall select between the models listed below. Each model's parameter specifications will be the set of parameters that produces the lowest corrected Akaike information criterion (AICc).

      1. (A) Autoregressive integrated moving average (ARIMA).
      2. (B) Exponential triple smoothing (ETS).
      3. (C) Random walk with drift (RWD).

4. (d) Other time series data.

   1. (1) In calculating amounts of payments under subsection (b) of this section, the time series data for forecasting Occupational Skills Awards and Institutional Credentials Leading to Licensure or Certification shall consist of fundable certified data with the counts of each performance tier fundable outcome achieved annually by a public junior college during no fewer than the four most recent fiscal years for which data are available. For Institutional Credentials Leading to Licensure or Certification, the Coordinating Board shall use the definition and limitations for the credential in effect during the fiscal year for which the credential was reported.
   2. (2) In conducting forecasting according to subsection (c) of this section, the Coordinating Board shall forecast performance outcome counts for which four or five years of fundable certified data are available using the available data. If fewer than four years of fundable certified data are available, the Coordinating Board shall forecast the performance outcome count as the greater of the count for the most recent available year of fundable certified data and the average of the counts for all available years of fundable certified data.
5. (e) Bounded projections. The forecasted total annual count of a fundable performance outcome for a fiscal year shall not exceed 110 percent nor be less than 95 percent of the count for the prior year. If the count for the prior year is also a forecasted value, then the maximum allowable change for the current year shall be calculated against the prior year's forecasted value as adjusted pursuant to this rule. If the value for a fundable performance outcome for the most recent actual, not forecasted data is zero, the forecast shall not be bounded in the next fiscal year. In no circumstances may an estimated fundable performance outcome be negative.
6. (f) The Coordinating Board shall forecast the number of each fundable credential in a high-demand field, as defined under subchapter T of this chapter (relating to Community College Finance Program: High-Demand Fields), for a fiscal year by multiplying the average annual percentage of the credential conferred in a high-demand field in the institution's time series data by the total count of the credential forecast to be conferred in that year.
7. (g) The Coordinating Board shall forecast the number of each fundable credential receiving an additional weight based on academically disadvantaged, economically disadvantaged, and adult learners status, as provided by §13.556 and §13.646 of this chapter (relating to Performance Tier: Fundable Outcomes) establishing performance outcome definitions for the fiscal year for which it is calculating performance funding, for a fiscal year by multiplying the average percentage of the credential conferred by the institution to students in each respective subgroup in the institution's time series data by the total count of the credential forecast to be conferred by the institution in that year.

Source Note:The provisions of this §13.624 adopted to be effective August 15, 2024, 49 TexReg 5978; amended to be effective September 1, 2025, 50 TexReg 5456.