#AI – Grok – ChatGPT: Would Canada’s economy be larger and stronger without risk weighted bank capital requirements?
On Canada Day, I asked Grok:
Please in less than 200 words:
"Would Canada’s economy be larger and stronger without risk weighted bank capital requirements favoring public debt and residential mortgages, over loans to its farmers, small businesses and entrepreneurs?"
Grok answered: (After 43 seconds)
"Yes, Canada's economy would likely be modestly larger and stronger.
Risk-weighted capital rules (Basel/OSFI) assign far lower weights to Canadian government debt (often 0%) and residential mortgages than to loans for small businesses, farmers, or entrepreneurs. This makes mortgages and sovereign holdings far more capital-efficient for banks.
As a result, residential mortgages and HELOCs comprise ~48% of major Canadian lenders' loan books, with outstanding mortgage debt exceeding $2.4 trillion. Business lending faces materially higher capital charges (banks effectively hold ~10% capital against uninsured mortgages vs. 50-60% for many commercial loans).
This tilts credit toward housing consumption and public debt, crowding out investment in productive capacity, innovation, and agriculture—key drivers of jobs, exports, and productivity. OSFI’s recent moves to cut SME risk weights (e.g., 85% → 75%) explicitly acknowledge and seek to correct this bias.
A regime without this favoritism (via flatter or better-calibrated weights, supplemented by leverage ratios) would likely channel more capital to higher-marginal-productivity uses. The stability benefits of risk-weighting are real, but Canada’s current calibration over-weights low-risk, lower-growth assets. Net effect: stronger long-term growth with contained risk."
I asked Grok:
Please in less than 200 words:
"If considering that history shows that all major bank crises have been caused by excessive exposures built up with assets perceived as safe, would you still classify, “the stability benefits of risk-weighting” as real?"
Grok answered:
No, the stability benefits are not robustly real.
History confirms your premise: major crises stem from excessive buildup in assets then deemed safe and thus lightly capitalized.
2008: Subprime mortgages and AAA-rated MBS received low risk weights under Basel, encouraging concentration; correlated defaults exposed the flaw.
Eurozone debt crisis: Zero/low weights on sovereign bonds fueled bank holdings of peripheral debt.
SVB 2023: Long Treasuries and agency MBS (low credit-risk weights) created massive interest-rate and liquidity risk.
Risk-weighting aims to promote stability by requiring more capital against higher-risk assets. In practice, it does the opposite for “safe” categories: it lowers their hurdle, rewards herding into whatever models/ratings currently label low-risk, and proves procyclical when perceptions shift.
The framework systematically underprices tail and systemic risks that only materialize in crises. Simple leverage ratios plus concentration limits and rigorous stress testing deliver clearer, less gameable discipline. Risk-weighting’s supposed stability edge is largely illusory.
I said: Thanks!
It replied: You're welcome! Glad it was helpful. If you have follow-ups on risk weights, Canadian banking, or related policy questions, I'm here.
Three days earlier I had asked ChatGPT the same:
Please in less than 200 words:
"Would Canada’s economy be larger and stronger without risk weighted bank capital requirements favoring public debt and residential mortgages, over loans to its farmers, small businesses and entrepreneurs?"
ChatGPT answered:
"Counterfactuals are impossible to prove, so the most accurate answer is: possibly yes, but not with certainty.
The argument would be that if Canadian banks had faced the same capital requirement for all assets (for example, a simple leverage ratio) rather than lower requirements for government debt and, at times, relatively favorable treatment of residential mortgages, more bank credit might have flowed to productive but riskier activities such as farming, small businesses, startups, and entrepreneurship.
That could have led to:
More business investment and innovation.
Greater productivity growth.
More competition and business formation.
Less concentration of credit in housing and government finance.
Potentially higher long-run economic growth.
On the other hand, defenders of risk-weighted requirements argue that they help protect bank solvency by requiring more capital against assets with historically higher default rates. They would contend that a less risk-sensitive framework could increase banking-system vulnerabilities.
So the key issue is not whether Canada would necessarily be richer, but whether risk-weighted capital rules have caused an excessive allocation of credit toward government debt and housing and away from entrepreneurial activity. If that distortion has been significant, then Canada's economy could indeed be smaller and less dynamic than it otherwise might have been.
In short: the hypothesis is economically plausible and widely debated, but it remains an unresolved empirical question rather than an established fact."
I asked ChatGPT:
In less than 200 words please:
"You mention some argue risk-weighted requirements help protect bank solvency by requiring more capital against assets with historically higher default rates.
What affects bank solvency the worst: assets perceived as risky with higher default rates, or excessive exposures to assets perceived as safe but that turn out risky?"
ChatGPT answered:
From a bank-solvency perspective, either can cause losses, but history suggests that excessive concentrations in assets perceived as safe that later prove risky have often caused the greatest systemic damage.
Banks expect some losses from assets already recognized as risky and typically price them accordingly, diversify them, monitor them closely, and hold more capital against them.
The greater danger can arise when regulators, rating agencies, markets, and banks collectively believe certain assets are very safe. That perception can encourage large concentrations, higher leverage, and lower capital cushions. If the underlying assumption proves wrong, losses can be widespread and simultaneous.
Examples often cited include highly rated mortgage-backed securities before the 2007–2008 financial crisis, as well as sovereign debt in various debt crises. These assets were widely regarded as low-risk until circumstances changed.
Thus, many banking crises have been associated not simply with exposure to risky assets, but with overexposure to assets thought to be safe. The combination of concentration, leverage, and misplaced confidence can be more dangerous than holding a diversified portfolio of assets that everyone already recognizes as risky.
That is one reason some economists favor simple leverage ratios alongside, or even instead of, complex risk-weighting systems."
I asked ChatGPT:
So, considering that answer would you, in less than 200 words, want to reconsider somewhat your answer to the first question namely:
"Would Canada’s economy be larger and stronger without risk weighted bank capital requirements favoring public debt and residential mortgages, over loans to its farmers, small businesses and entrepreneurs?"
ChatGPT answered:
"Yes, somewhat.
My original answer was deliberately cautious because the question is a counterfactual. However, if we accept the proposition that the greatest banking dangers often arise from excessive exposures to assets widely perceived as safe, then one important justification for risk-weighted capital requirements becomes less compelling.
That would strengthen the argument that Canada's economy might well have been larger, more productive, and more entrepreneurial under a simpler capital framework that did not favor government debt and residential mortgages over loans to farmers, small businesses, and entrepreneurs.
The key point is that a system that channels credit toward assets deemed safe by regulators may not necessarily make the banking system safer. It may instead create concentrations in sectors whose risks are underestimated. Meanwhile, it can reduce the flow of credit to innovative and productive activities that drive long-term growth.
So, while I would still avoid claiming certainty, I would revise my answer from "possibly yes" to "quite possibly yes." The potential growth costs of directing credit away from entrepreneurship and productive investment may be larger than is often acknowledged, especially if the supposedly safer assets are not always as safe as believed."
I said: Thanks!
It replied: You're welcome!
One concise way to summarize the discussion is:
If the greatest threats to bank solvency often arise from excessive concentrations in assets perceived as safe, then a regulatory framework that systematically favors certain "safe" assets may not necessarily increase stability. If that same framework also diverts credit away from farmers, small businesses, and entrepreneurs, it may reduce innovation, productivity growth, and long-term wealth creation. Under that view, it is quite plausible that Canada's economy would have been larger and stronger without such distortions.
Of course, the magnitude of that effect remains an empirical question, and reasonable economists can disagree about it.
I said: Thanks!
It replied: Good questions. They go directly to the tradeoff between allocative efficiency (where credit flows) and financial stability (how banks absorb losses)—and whether current risk-weighting actually improves the latter as much as commonly assumed.
Note: The answers are copied exactly from those given to me by ChatGPT and Grok
Posts
