There is one perennial, overlooked, and devastating irony in the vast body of bank capital and liquidity regulation: the better a bank’s liquidity score, the less regulatory capital it has. Although liquidity and capital are inexorably linked when it comes to preserving bank solvency, the need to comply with two contradictory standards forces banks to change their business model to meet both ends in the middle at considerable cost to profitability and long-term franchise value. This is of course a major threat to solvency of which bank regulators are either blissfully unaware or, worse, heedless. Federal banking agencies have stoutly refused to undertake the essential cumulative-impact analysis we’ve fruitlessly urged on them most recently in Congressional testimony. A new Federal Reserve Bank of New York study shows not just why they should judge rules by sum-total impact, but also how they could do so and thereby have a much better sense of which banks might actually go broke before they do.
I refer you to the full FRB-NY paper for details. It crafts an economic-capital construct calculated by netting the net present value of financeable assets versus par liabilities as a baseline measure which can then be tested under various stress scenarios that start with illiquidity and end in insolvency and vice versa. This leads to a robust measure of survivability that combines the impact of credit risk, liquidity, and the real-world market conditions current rules ignore. In essence, economic capital is derived from the hard-nosed, real-time factors that wise investors and counterparties use to determine market capitalization and the cost at which banks gather funds. This is clearly preferable to the backward-looking, book-value way each bank regulatory requirement comes up with a different measure of resilience essentially irrelevant in the real world of shifting market values that can quickly pull a bank under no matter how resilient supervisors may think it.
To be sure, the economic-capital methodology isn’t perfect, as its authors readily admit. However, problems are principally due to data limitations, not logical flaws. And, even with partial data, the methodology still does better predicting troubled banks than bank supervisors who have missed every recent crisis and thus rescued banks that might have been saved or, if that proved impossible, closed before bailout, ensuring market discipline, not ever greater moral hazard.
The economic-capital approach has immediate and particular value as a way to make sense of the stress capital buffer (SCB) also required of all large banking organizations. The SCB depends not only on risk-based capital rules of sometimes-dubious value, but also on Fed-set stress tests that, as the central bank finally conceded, have proved a burdensome, problematic, capital standard. And, as with the basic capital rules, the SCB takes no account of liquidity risk even though stress often starts there, not in the “fortress” capital the Fed also admits banks have even without resorting to the SCB.
As FRB Gov. Bowman has recognized, the SCB is a motley mixture of the Fed’s stress-test scenarios (flawed on their own) and regulatory risk-based standards that not only contradict the liquidity ratios, but also yet another rule at odds with all the others: the leverage-capital standards. Using economic-capital ratios and market-based stress tests would permit full transparency without the gaming the Fed says justifies its opaque, and often mistaken, scenarios. As noted, economic-capital measures rely on market valuations with only a few, transparent scenarios needed to spot key vulnerabilities such as troubling drops in current asset market valuations, deposit runs, concentrated holdings, and novel-activities. Even as is, economic-capital measures are better than the SCB as is. The Fed could make this measure still better thanks to all its supervisory data supplementing transparent market valuations, not opaque models. The SCB also could and should have operational- and market-risk add-ons for each bank where supervisors determine these risks are material.
Thousands of big-bank staff dedicated solely to stress testing, millions of consultancy dollars, and hundreds of Fed staff might be troubled by a simple solvency exercise based on economic capital. The rest of us should think this through carefully and, if objective analyses validate and even improve the economic-capital methodology, then take it quickly from theoretical to the rulebook so that banks actually have a stress buffer reflecting their actual true solvency risk. This is, of course, the fundamental point of bank regulation even though it’s now clouded in so much complexity and perverse incentives that no one can truly tell which bank will make it through the next crisis.
In scientific paradigm shifts, complex theorems that get more elaborate to explain ever more failures are replaced by a thunder-clap of a simple new conclusion that explains far more and is validated far better. The current bank regulatory paradigm creaks of complexity and too many requirements built to address too many anachronistic markets and often-inexplicable models. It’s time for something simple, clear, credible, and accountable.