Why This Work Exists
The mathematics of resilience was built for advanced economies. Basel III stress tests, supply chain risk models from Western consulting frameworks, infrastructure resilience standards designed for OECD telecoms — all of them assume conditions that do not hold across most of the world.
Emerging markets face a different problem set. Financial systems span banks, non-bank financial companies, microfinance institutions, and informal credit networks. Supply chains run through fragmented logistics with concentrated chokepoints. Digital infrastructure is built faster and adopted at greater scale than anywhere in the developed world, with regulatory frameworks racing to catch up. And the shocks — currency crises, climate disruption, geopolitical realignment — arrive harder and recover slower. Cristo Labs develops the computational methods this gap requires.
Research Agenda
Our current research is organised around five questions:
- Resilience of digital infrastructure. Real-time payment systems, identity infrastructure, and open digital networks are now systemically important across emerging markets. How do you measure their fragility before disruption forces the question?
- Supplier network resilience under uncertainty. Multi-criteria decision frameworks combined with Monte Carlo methods to rank supplier networks where data is sparse and risks are correlated.
- Cross-country resilience comparison. Building defensible resilience rankings across emerging market economies despite divergent data definitions, regulatory perimeters, and reporting standards.
- Stability scoring for non-bank financial institutions. How do you produce auditable, comparable resilience scores for non-bank lenders, microfinance institutions, and cooperative banks operating outside Basel frameworks?
- Regime detection in emerging market credit cycles. What can hidden Markov models and statistical mechanics reveal about fragility windows that conventional early-warning indicators miss?
Methodological Approach
Our work draws on three intellectual traditions that rarely converge in applied research.
- Machine learning. Hidden Markov models for regime detection, fuzzy clustering for institutional grouping, kernel methods for resilience classification under sparse and heterogeneous data.
- Multi-criteria decision analysis. VIKOR enriched with Monte Carlo uncertainty, used for ranking and comparison where conventional weighting schemes fail to capture institutional heterogeneity.
- Statistics. Statistical distributions applied to institutional state spaces, producing stability metrics that are physically meaningful and mathematically auditable.
Every method we publish is reproducible. Every result is auditable. Every framework is built to be extended by other researchers. We publish open computational supplements alongside our papers wherever possible.