The next financial crisis will not come from the traditional banking sector. So goes conventional thinking among financial policy makers. The world’s biggest banks are now safer, according to the narrative, thanks to stricter capital requirements and frequent stress tests that have curbed the appetite for extreme risk and tightened up lax regulatory standards.
I wish I were completely reassured. But as an accountant, I know that the headline capital numbers result from a subjective calculation. Banking regulators typically spend too little time digging into how those figures are calculated. I also know that when the US financial system is healthy, as it is now, we should strive to do better at accounting for potential losses, because that might cushion the blow when the inevitable downturn arrives.
To be sure, the big banks have all passed the Federal Reserve’s stress tests with flying colors. And this reflects substantial increases in capital buffers: the 35 banks that underwent 2018’s stress test have added about $800 billion in the highest quality type of capital over the past decade, according to the Fed. The central bank has deemed that the banks would therefore be strong enough to continue lending if the economy were to plunge into another severe downturn.
But I am not the only observer who remains concerned. In a speech to Americans for Financial Reform in May, Georgetown’s Daniel Tarullo, who was a Fed governor from 2009 to 2017, questioned the robustness of the stress tests. Banks know what regulators are looking for, Tarullo observed, enabling them to “find clever ways to reshape their assets,” thereby reducing their capital levels without reducing their risk exposures. And he also cast doubt on a Fed proposal to create a “stress capital buffer” to stop banks from running down their capital cushions by using dividend payments. Such a buffer, Tarullo argued, could actually prompt banks to take on even more risk.
This raises the tricky question of how capital levels are calculated. At times, banking regulators could be accused of fixating on the level of capital requirements without adequately taking into account how loan losses are provisioned for in a bank’s financial statements.
At the simplest level, the amount of capital a bank has on its balance sheet is the value of its assets, net of the value of its liabilities. But the values of these assets and liabilities are driven by accounting valuations. This therefore raises the most fundamental question of accounting: How do you measure an asset? Some assets are marked to market, while others are not. All these decisions trickle down to make up the amount of capital that the bank has. That figure, in turn, affects how much lending the bank undertakes, meaning that it has microeconomic and macroeconomic consequences. So any real understanding of capital levels requires regulators to understand how assets and liabilities are being measured.
It would be dangerous to assume that by simply setting higher capital requirements, we have resolved the issue of too-big-to-fail financial institutions.
This demonstrates that accounting, finance, and economics are intrinsically interconnected. Sometimes accounting gets thought of as something of a veil, a dreary but necessary process of asset measurement that has no economic consequences. But the value placed on a bank’s assets and liabilities affects its capital levels, which, in turn, affect lending, borrowing, and interest rates in the economy.
Currently, however, regulation often proceeds as if setting rules for banking and setting rules for accounting were two unrelated, isolated activities. Banking regulators view accounting as something of a mechanical, back-office function, akin to the gears of a machine. Accounting standards setters view regulating banks’ abilities to withstand severe shocks as someone else’s problem. But improving oversight requires us to use insights from both to give a complete picture of a bank’s balance sheet. The focus should be on the interaction between accounting measurement and capital requirements.
On the accounting side of the equation, a key area to focus on is how banks measure the performance of their loan portfolios, because this determines the amount of capital they need to hold. Lenders that are proactive in recognizing losses on their balance sheets could potentially hold less capital without increasing risk. By contrast, those that are not proactive in accounting for losses should potentially be required to increase their capital requirements.
There are two implications. First, banking regulation needs to round out its focus on capital requirements with an equal focus on loan loss provisions and how those shape the bank’s capital level. Secondly, this challenges the idea of imposing a one-size-fits-all approach to capital requirements. There is some necessary flexibility in the interpretation of accounting rules, since each lender’s loan portfolio is unique. But there is no corresponding flexibility in our attitude about how much capital lenders should hold.
Given our one-sided approach to banking regulation, we should beware of complacency. It would be dangerous to assume that by simply setting higher capital requirements, we have resolved the issue of too-big-to-fail financial institutions. In fact, setting capital requirements without taking into account the type of loans that banks are originating could be dangerous. We might set high capital requirements, but banks might respond by originating riskier loans. Or lenders might originate loans that are not too risky but do not necessarily maximize the value of the bank. The effect would be to weaken rather than aid financial stability. Beyond the question of capital requirements, it is critical to understand whether banks are provisioning for loans appropriately.