Sector accounting betas for estimating the cost of equity of non-listed firms in Colombia.
CAB-Ke v1 is a reliability-tiered methodological framework for estimating the sector-level cost of equity (Ke) under the data constraints typical of emerging markets. Built on the SIREM panel over 2007–2025, excluding the 2013–2014 IFRS transition: 3,704 firms and 44,246 firm-year observations.
The research and the full replication material
CAB-Ke was born from a concrete problem: in Colombia, valuation professionals lack locally calibrated sector betas that can be defended under scrutiny. This framework closes that gap with a reliability-tiered approach and delivers the full material needed to replicate it.
The problem
Colombian practice rests on two unsatisfactory alternatives: (a) importing Damodaran's emerging-market betas without local validation — this provides international comparability but is not calibrated to Colombian data; (b) computing idiosyncratic firm- or sector-level betas without a coherent reliability framework — leaving the professional defending ad-hoc methodological choices before arbitration panels, transaction committees, or regulatory authorities.
The CAB-Ke proposal
The appropriate professional standard in data-constrained markets is not mechanical statistical purity, but transparent estimation under uncertainty with an explicit sector-level reliability classification. CAB-Ke delivers a tiered sector matrix that distinguishes three evidence standards — strict academic, strong professional, and indicative — and assigns each sector to one of five tiers.
How the sector βU sector
Three component methodologies with robust inference
Hill-Stone (1980) on a leave-one-out factor
Profitability beta: the slope of sector ΔROA on aggregate market ΔROA excluding the own sector (LOO). Tukey-biweight M-regression (c=4.685) with stationary-bootstrap inference, Politis-Romano (L=4, 500 replications).
Distress-augmented beta
Credit-cycle-augmented beta in the spirit of Campbell, Hilscher, and Szilagyi (2008). Aggregate default probability built from firm-year Z-scores, standardized with rescaled SE and censored.
Beaver-Kettler-Scholes
Cross-sectional regression of beta on sector D/E, log assets, and ROA volatility. Provides an independent cross-sectional benchmark and an input to the fallback blend for Tier 4 sectors.
Five reliability tiers
Each sector is assigned to the highest tier for which its local evidence qualifies
Orthogonality with Damodaran (+0.06)
The Spearman correlation between the CAB-Ke βU and Damodaran's emerging-market benchmark is +0.06. The two methodologies are complementary, not substitutes: Damodaran reflects the average sectoral composition of emerging markets and supports international comparability; CAB-Ke reflects local Colombian accounting evidence and supports sectoral discrimination within Colombia. Reporting both columns is recommended.
Sovereign floor binding in 22 of 31 sectors
In pesos, the sovereign floor defined as TES 10Y + 150 bp is binding in 22 of 31 sectors (~71%). USD Ke retains full sectoral discrimination and is the recommended reporting column; COP Ke under the sovereign floor is a conservative professional consistency rule, not a no-arbitrage result.
Partial reconstruction on 26 of 31 sectors
A partial-reconstruction bootstrap exercise covering 26 of the 31 sectors with 500, 1,999, and 4,999 replications is consistent with the central finding: the strict symmetric academic criterion is not met under any replicate count, and tier stability holds for 24 of 26 sectors; two boundary sectors shift by one tier as the replicate count increases — behavior consistent with the threshold-based classification design.
Download the paper and cite the work
Working paper · version 1 · 2026
CAB-Ke v1 Paper · English
Accounting-Based Sector Betas and the Cost of Equity in Data-Constrained Emerging Markets: A Tiered Framework for Colombia.
↓ Download the paper (PDF) ↓ Download PDFPanel SIREM 2007–2025
Unified Superintendencia de Sociedades panel. Decree 2649 window (2007–2012) + IFRS (2015–2025). Excludes the 2013–2014 transition.
↓ In preparationResults matrix · 31 sectors
Tables 5A/5B of the paper with component βs, βuncapped, βcapped, bootstrap intervals, USD Ke and COP Ke for the 31 sectors + 2 excluded.
↗ See in the paper (PDF)Pipeline R + Python
Audited implementation of the full pipeline with bit-identical results across languages: pre-processing, beta estimation, Vasicek combination, Fisher conversion.
↓ Under review@article{jimenez2026cabke,
title = {Accounting-Based Sector Betas and the Cost of Equity in
Data-Constrained Emerging Markets: A Tiered Framework
for Colombia},
author = {Jim{\'e}nez Trivi{\~n}o, Jhon Alexander},
year = {2026},
journal = {Documento de trabajo},
url = {docs/cab-ke/papers/CAB_Ke_v1.pdf}
}
Results for each sector
Summary sector matrix for the 31 non-financial sectors plus 2 excluded. βU is the final tier-specific unlevered beta. βL applies Hamada with τ=35% and the sector median D/E. COP Ke uses Fisher conversion with long-run inflation (π_COL=3.0%, π_US=2.3%) and the sovereign floor when it binds. Macro snapshot: Rf=4.20%, ERP=4.70%, CRP=2.85%.
| CIIU | Sector | Tier | βU | βL | Ke USD | Ke COP |
|---|
Los βU per sector match those published in Tables 5A/5B of the paper. Full values with bootstrap intervals and component breakdown (βHS, βDA, βBKS, βuncapped, βcapped, βVasicek) are available in the paper appendix.
Calculate your company's Ke
The calculator applies the CAB-Ke v1 framework using the published sector βU , your financial D/E, and the Hamada formula. Everything runs locally in your browser — nothing is uploaded to any server.