What surprised me this week about the capital calculation cock-up of Bank of America Meryll Linch were the reactions of the press. The financial press responded to this gaffe by highlighting the difficulties of calculating capital. Reuters, for example wrote: “The announcement illustrates how difficult it is to determine appropriate capital levels for the biggest banks.” Capital specialist Mayra Rodríguez Valladares also highlighted the complex nature of the issue.
Whatever led Bank of America to making this mistake, it was not the result of complexity. In fact, it very much relied on an inconsistent accounting rule.
Actually, this part of the regulatory capital calculation is simple, it is intuitive. I will explain why. It is also well-known. I will also explain why the particular adjustment to capital makes sense in bad times, but not always in good times. Lastly, I will show that Basel III will help us not to make such mistakes again.
Adjusting for changes in own credit standing: Noblesse Oblige.
To calculate capital ratios, banks apply adjustments to the book value of their equity. For example, for the end of 2013, Bank of America reported an equity figure of $223bn. However, after applying the regulatory adjustments, Tier 1 regulatory capital of Bank of America is about $162bn. Divide this amount by total assets and you will obtain a Tier 1 leverage ratio of 7.86%. The graph below shows the effect of the most significant regulatory adjustments, expressed in basis points of equity, for Bank of America.
One adjustment is the one for changes in own credit standing, see the dashed lines in the graph. Actually, this adjustment is intuitive: If you borrowed money from your parents, then you should know how this works. At first, your parents must have been lenient when it came to repaying the loan. That is, as long as you were young and unemployed. However, that changed when you landed the great job with a salary to match. Your parents noticed your improved credit standing. As a result they wanted you to pay back the money much more quickly.
An improved credit standing therefore comes at a cost: cheaply borrowed money becomes more expensive. This is perfectly acceptable: noblesse oblige.
Bank regulators, in dealing with changes in own credit standing, act like responsible and unforgiving parents. That is, in particular in bad times.
When a bank’s credit standing deteriorates, the regulator forces the bank to ignore the gains caused by the deteriorated credit standing. This is very prudent, it requires banks to keep in mind that they have to repay debt in full, even if the value of their debt is temporarily depressed. So, when Lehman collapsed, many banks suffered from deteriorated credit standing that lowered the value of their debts. This led to a perverse effect on equity: lower debt values increased equity values. Banks gained from their deteriorated credit standing! However, thanks to the system of regulatory adjustments, banks were not allowed to have this gain reflected in their regulatory capital or their capital ratios.
So far so good.
What goes up must come down. When credit standing improves, a borrower loses. It shall pay back any loans in full, not at a discount. But hold on, banks already were told that they should pay back their debts in full, see the previous paragraph. So it makes sense to ignore any losses that are the result of an improved credit standing, else one would be double counting.
But here is where the system of regulatory adjustments appears to fail. Suppose a bank acquired the debt securities of another bank at a low value. This is what happened with Bank of America when they acquired Merrill Lynch & Co., Inc. in 2009.
The value of the acquired debt increases if the credit standing of the acquiring bank improves. The value of accounting equity then deteriorates. But the system of regulatory adjustments requires a bank to ignore the effects of an improved credit standing. Losses resulting of an improved credit standing shall be ignored. In other words: banks add the loss back to regulatory capital ratios.
Temptation does the rest …, I mean what would you do if you could improve the capital ratio of your bank if you could do so by adding the value increase of debt securities that you obtained at a discount five years ago? Would you let go of these securities if they contributed to capital?
Despite noblesse oblige, Bank of America conveniently forgot to let go of the resurrected capital securities.
Accountants and bank regulators knew.
They were aware of the treatment of own credit standing. It sparked controversy at the introduction of International Financial Reporting Standards in Europe, in 2005. Initially, under pressure of some European bank regulators, it led to a “carve-out” of the fair value option from European IFRS. But shortly after IFRS came into force in Europe, and after close coördination between the European Commission, the International Accounting Standards Board, and the Basel Committee, this carve-out was eliminated from IFRS. Bank regulators, however, responded by introducing their own version of this carve out. They decided to ignore any changes in the value of debt caused by changes in own credit standing. They introduced the regulatory adjustment for changes in own credit standing.
It is therefore unlikely that auditors and regulators were blissfully unaware of what happened at Bank of America.
In fact, the auditors relied on an inconsistent standard.
The capital securities that Bank of America inherited from Merrill Lynch were marked to market, by virtue of the fair value option. This option can only be used by specific assets and liabilities. According to the standard governing the Fair Value Option, this is the option’s stated objective: ” … to improve financial reporting by providing entities with the opportunity to mitigate volatility in reported earnings caused by measuring related assets and liabilities differently without having to apply complex hedge accounting provisions.”
Now, U.S. accounting standards do not have formal eligibility rules that block the entry of securities to this category.* However, applying the stated objective to Bank of America’s capital securities leads to an inconsistency: How can a bank mitigate volatility in reported earnings with capital securities? Mitigating volatility would, if the standard would put money where its mouth is, allow Bank of America to trade the capital securities. This of course is a stretch, capital securities are not to be traded. These securities are meant to be permanent and loss absorbing. The idea that the securities can help mitigate volatility is just wishful thinking.
Therefore, the capital instruments should have been valued at par.
The future looks promising.
Basel III luckily helps preventing gaffes like these to happen again: the Basel III disclosure requirements require banks to disclose everything about capital securities: Par value of securities, Amount recognised in regulatory capital, Unique identifier (eg CUSIP, ISIN), Original date of issuance, Original maturity date, etc, etc.
Chances are that mistakes like the one made by Bank of America will become visible quickly, and will not affect the ratios for many years.
* IFRS, on the other hand, requires liabilities to be managed on a fair value basis, which means that Bank of America should be able to trade these securities. (IFRS also renders the embedded derivative attached to them pretty much valueless). Note that until April of this year, the FASB contemplated adopting eligibility criteria for the Fair Value Option that would likely exclude these securities. Luckily the FASB decided in Bank of America’s favour. I credit Matt Levine for pointing this out to me.