What if bank regulators had not missed, or fallen asleep during the lectures on Bayesian conditional probabilities?
Opining 2013:
What's perceived as “most dangerous”, the risky, the trucks, is what currently in banking must transit at the slowest speeds, the lowest allowed bank leverages; while those perceived as the safest, like sovereigns, residential mortgages and AAA rated securities, are those allowed to go through our residential neighborhoods at the highest speeds, the highest allowed leverages.
I do understand, it is hard to internalize that, at least when it comes to banking, that which is perceived as safe is so much more dangerous to the system than that which is perceived as risky. Sadly, way too many missed their lectures on conditional probabilities.
https://subprimeregulations.blogspot.com/2013/10/worse-than-high-speed-allowed-trucks-is.html
Opining 2018:
What is the conditional probability of assets being dangerous to bank systems when conditioned to that bankers have perceived these assets as risky?
Assets perceived by bankers as risky become safer, not riskier.
Assets perceived by bankers as safe become riskier, not safer.
Regulators who base their capital requirements for banks on that what’s perceived as risky is more dangerous to the bank systems than what’s perceived as safe, is that because they have never heard about conditional probabilities?
https://subprimeregulations.blogspot.com/2018/08/risk-weighted-capital-requirements-for.html
Dialogue with ChatGPT 2023:
I asked: "If the risks for banking systems are much conditioned to how credit risks are perceived, would it not be useful to base the risk weighted bank capital requirements on the conditional probabilities?"
It answered: "Yes, by utilizing conditional probabilities, regulators could tailor risk weights more precisely to reflect the true credit risk associated with different assets. This approach could help mitigate the potential for misalignment between risk weights and actual risk, which is important for maintaining the stability and soundness of the banking system."
https://subprimeregulations.blogspot.com/2023/06/ai-chatgpt-openai-should-risk-weighted.html
Dialogue with Grok 2025:
I asked: “Small businesses and entrepreneurs are inherently riskier bets than government debt—default rates are higher, and public debt (especially from stable governments) is often seen as a safe haven.” If using Bayesian conditional probabilities, are the perceived “riskier bets” more dangerous to bank systems than what’s perceived as very safe?
It answered: “Yes: Small-business loans might ding banks regularly but rarely break them. Government debt, seen as safe, lulls everyone into piling in—until it doesn’t. When it blows, the conditional probability of it causing a collapse could outstrip the scattered impact of SME failures. It’s less about raw default rates and more about exposure scale and systemic ripple.”
https://subprimeregulations.blogspot.com/2025/04/ai-grok-chatgpt-what-about-bank.html
Posts
