Aristóbulo de Juan (Former Head of Banking Supervision. Bank of Spain) | The current obsession of bank regulators worldwide is capital, reinforcing capital. For good reasons: capital is fundamental. But, under the surface, this appears to be a nominal rather than effective fixation, at least judging by the questionable quality of some of the components of minimum capital required by regulators in Europe and elsewhere. Moreover, paradoxically, the regulatory and supervisory emphasis on capital comes hand in hand with a trend, observed in a number of countries, to forbear on asset valuation, provisions and income recognition. As a result, the quality of accounting, the transparency and soundness of banking systems and the reliability of markets are seriously affected.
Let us first remind some fundamental principles. From an accounting viewpoint, capital is the equivalent to net worth, i.e. assets minus liabilities. The minuend should represent the real value of assets, after updated assessment and appropriate value adjustment. That is to say, capital is the value of assets – adjusted to reflect the repayment capacity of borrowers and the market value of traded securities– minus the liabilities. In other words, real capital is the residual amount of resources available after paying off creditors in case of liquidation. An alternative, typical way of describing the capital of quoted institutions is as the market value of its stock. In any case, if capital is to be effective to absorb losses, it cannot be made up of fictitious assets, and should not be onerous or callable.
Starting in 1988, minimum capital requirements for banks were sequentially established by Basel I, Basel II, and Basel III. In Basel I, the requirement was a proportion of total assets as per the bank’s books, with a weighting system based on the types of transactions, rather than the actual risk of each particular asset or group of assets. This simple mechanism was later replaced by Basel II and Basel III, which determine the minimum capital requirements as a proportion of risk‐weighted assets. In this complex mechanism, mathematical models are used to group assets according to their risk. While risk‐weighting was in theory a positive concept, it led to catastrophic consequences. First, risks were assessed by extrapolating past average trends (historical losses), which do not often prevail under different bank management, risk attitude or economic conditions. Furthermore, the models –and the risk assessment that results– are built by the banks themselves, with often limited supervisory oversight of their underlying assumptions given the substantial complexity. This implies ignoring the universal fact pattern that problem banks hide their problems from the supervisors and markets alike, including by fictitiously reducing their risk weights, but not only.
Problem banks do this by “working” on the two sides of the capital ratio: the denominator (assets) and the numerator (capital). The most traditional technique affecting the former is the refinancing –based on unwarranted financial terms– of the principal and interest of unrecoverable assets. In addition, the more sophisticated banks deconsolidate masses of bad assets through fictitious sales to “special purpose vehicles” or offshore institutions, the imaginative use of derivatives or cross sales of assets with other banks. While those and other similar financial engineering practices do not actually eliminate the exposures (nor the connected losses), they allow banks to avoid provisioning while showing lower risk levels, and thus remain a going concern with lower capital.
Regarding capital, problem‐hiding techniques focus on deferring losses or having higher interest income or gains flow through to retained earnings, in an effort to artificially augment capital. An example of how this is achieved is by accruing income on unpaid loan installments. Another example is the so called badwill, the difference between the fair value of an acquired institution ́s net assets and its purchase price, which reflects the very strange circumstance of a bargain purchase. Accounting standards establish a number of safeguards that should make it virtually impossible for bank managers to recognize (and for auditors to validate) such gains. A fair representation of those instances should lead to adjusting the fair value of the acquired net assets, or ultimately to the recognition of a restructuring provision. Yet these gains continue to be recognized in many instances, and thus keep flowing through to the retained earnings of many banks. Paradoxically, the lower the price a bank pays the more capital it claims to have, turning badwill into a perverse incentive to alleviate a bank’s capital shortfall through bad mergers.
As concerns lower losses, the avoidance or delay of loan‐loss recognition comes to mind instantaneously, but other techniques may be less apparent. For instance, while goodwill –the excess price paid by an acquirer over the fair value of an acquired company’s net assets– is correctly deducted from regulatory capital to reflect its dubious cash generating and loss absorption capacity, supervisors remain silent as to its permanence in the books of acquiring banks. Rarely do the goodwill impairment tests mandated by the accounting standards lead to substantial adjustments to the value of this alleged asset.
The case is similar for Deferred Tax Assets (DTAs), which simply attach a right to tax credits on accumulated past losses to the bank’s potential future profitability. While these assets are also deducted from regulatory capital, as a special assistance to problem banks some governments allow their long‐ lasting permanence in the balance sheet, despite the bank’s doubtful profitability prospects.
Any of these techniques have the effect of hiding the losses, but bring about an additional deleterious side effect: hidden losses are in fact non‐performing assets, and as such generate no income; but their financing conveys real payment of interest, meaning a financial cost and cash outflow –and thereby net additional losses– every day.
A number of regulators and supervisors forbear on the above practices, inadvertently or intentionally. They seem to favor a sort of constructive ambiguity to maintain an illusory sense of stability, instead of unveiling the problems timely and prioritizing solvency. This delaying and praying tends to accompany periods of general economic uncertainty, and relies on the hope of prompt recovery, ignoring the fact that, most of the times, the deterioration of these problem banks runs faster than the general recovery of the macro context. The reckoning usually comes at the turn of the cycle, when access to the capital markets becomes more expensive, thus contributing to a procyclical spiral.
Furthermore, there are a number of aggravating factors. First, the fact that two alternative ways of estimating bank capital coexist introduces a considerable degree of confusion, often leading bankers and supervisors to ignore the real value of assets and actual exposures, and to instead focus on the more theoretical, more discretional and less demanding approach: the Basel capital requirements. This allows effectively bust banks to appear safe and sound as per Basel rules, until they inevitably become illiquid and have to be resolved at much higher cost.
A second aggravating factor is how the regulatory authorities now think about recovery and resolution. It is commonplace to consider that the best solution to fill a “hole” in a bank’s financial position is to raise capital for the hole’s estimated amount. While this might make the entity look sounder and more liquid, new capital issued to cover losses would not formally accomplish the goals capital purports to meet, as it would not be available to cushion unexpected losses or to make new placements.
Furthermore, the quality of the “concrete” used to fill the holes might now have deteriorated, as new types of instruments issued to raise investor appetite are becoming a sizeable component of regulatory capital. These new instruments, such as those contingently convertible into capital –so called “Cocos. are deemed to encompass the same permanence and loss absorbency features of capital. However, even if the issuer has discretion to cancel related coupon payments, these instruments have not yet been sufficiently tested in real stress. Wouldn’t banks be economically compelled to keep making payments to avoid being punished by the markets? Furthermore, it remains unclear how some of these instruments would behave in case of insolvency. Would their holders actually be last in the creditor hierarchy line, pari passu with the shareholders? Judging by the prospectuses, this seems far from clear in many cases. Some things are certain, though: these instruments prove to be very onerous and a burden for the issuing bank’s income statement; and use of the conversion feature will tend to be delayed by the supervisor – perhaps indefinitely in the case of systemic banks, which are unlikely to ever be resolved.
The prospects turn even more sour when realizing about the current demise of traditional and successful supervisory tools such as on‐site inspections, provisioning or the suspension of interest income accrual for underperforming or non‐performing assets. Based on the assumption that asset valuations are unreliable and subject to managerial discretion, those tools are now considered outdated. Besides, inspection is being effectively transferred to external auditors. Not only do these professionals pursue private interests based on a different skillset as that required of banking supervisors; they have also shown a rather questionable performance over the past few years. Furthermore, even if the lack of reliability of asset values unfortunately proved true in many cases, value estimates that support the capital issuances required to refill the holes are equally discretional.
By refusing to use the more traditional instruments, both managers and supervisors are tacitly encouraged to use cosmetics and maintain risky policies. Also, they might be blind to instances of undercapitalization and therefore forgo prompt corrective action. Instead, they sit idly for as long as markets keep providing liquidity. Only when markets realize the situation –frequently as a result of sudden mistrust following an external shock– does the entity grow irremediably illiquid and a decision is made to restructure or resolve, but obviously at a much higher cost. At that point, some might blame the situation to a liquidity problem, excusing managers and supervisors. In fact, the problem is one of solvency: a growing mass of non‐performing assets and continuous poor management that gradually drains liquidity out of the bank.
As a practical exercise, it would be interesting if supervisors or analysts selected a sample of banks and made all the necessary adjustments to eliminate the impact of techniques such as those described in this article. The exercise would bring much needed clarity to capital and earnings figures, and would raise the alert on lower‐than‐reported profits and meagre capital.
Going back to the start, the whole issue of capital, and particularly the supervisory fixation on a particular characterization thereof, should be completely revisited. So should the mechanisms and strategies of supervision, remedial action and resolution. Let alone, of course, accountability and responsibility of managers, auditors, regulators and supervisors.
