There are two types of people in the world: those who like to divide people into two types, and those who don’t.
Too simple? OK, try this one: some people find taxonomies very helpful; and others, not so much.
I really like taxonomies. They’re not perfect, but they help me to remember important aspects of an idea, and they are an effective pedagogical tool. For example, with the help of my friend Sri Ramamoorti, I came to realize that Donald Rumsfeld’s taxonomy of uncertainties is actually quite useful for assessing audit quality: if management’s estimates require consideration of Rumsfeld’s famous “unknown unknowns” (unk unks), or even things we know that we don’t know, how is that actually auditable? Think bankers’ loss reserves on debt issued by the Greek government.
Another example of a taxonomy, which may fit topics I blog about even more broadly, has recently emerged via Daily Show host Jon Stewart’s instant-classic swan song devoted to the “three basic flavors” of “institutional BS.” (Editorial note: I don’t know exactly why – maybe it’s because my wife would get really pissed at me if she found out – but I draw the line in this blog at actually spelling out profanities.)
Stewart’s Type I BS is making bad things sound like good things, e.g., Patriot Act instead of “are you scared enough to let me look at all of your phone records.”
Type I BS is extremely redolent to the principles-based accountant. Standards setters love to use terms from other disciplines to hide the BS in their rules. For example: “probable” from statistics (to hide utter vagueness); “translate” from linguistics (to hide a process for creating arbitrary numbers); “statement of financial position” from finance (to promise a lot more than the balance sheet actually delivers); and “due process” from law (to obscure the reality that public deliberations, exposure drafts and comment letters are mere charades compared to the influence of monied special interests).
Type II BS is a mountain of complexity to obscure a rotten core. More on this later, but the first two words I associate with Type II are “US GAAP” and “IFRS.”
Type III is the BS of “infinite possibility.” Stewart says that it manifests itself as “We can’t do anything because we don’t yet know everything…” In accounting policymaking, Type III BS may smell worst of all. How long do we have to “deliberate” until everyone is thoroughly browbeaten by made-up numbers for loan losses, pension costs, lease obligations, revenues, and compensation of executives with stock options? How long did it take to deliberate on the utterly ridiculous question of whether loan loss allowances go into one, two, or even three buckets?
* * * * *
I write the above, believe it or not, to introduce one of the most important books to come out of the Financial Crisis of 2008, The Bankers’ New Clothes,* written by Anat Admati (Stanford) and Martin Hellwig (Max Planck Institute). If this book had been written with Stewart’s BS taxonomy in mind, perhaps an even more apt title would be Cutting Through Bankers’ Type II BS. If you want a simple and virtually costless prescription for preventing another banking crisis, read this book. If you like to get the lowdown skinny in plain English, using simple and compelling analogies, fully supported by systematic reasoning and carefully documented evidence, read this book.
Admati and Hellwig are adamant, and have completely convinced me, that banks have too little equity. Period. None of the measures to enhance the safety and soundness of banks that were put in place after the Financial Crisis of 2008 change the fact that the dominoes have been set up once more for a massive chain reaction (i.e., “contagion”). For example, the U.S. now has laws to forbid government bailouts of bad banks; but because banks are overleveraged, the legislation doesn’t change the fact that excess leverage will trip the first domino of a systemically important financial institution. If governments will want to avoid the massive contagion that bailouts averted in 2008-09, and they surely will, these laws will have to be undone.
In a nutshell this is their reasoning for deleveraging:
“Because the use of deposits and other forms of short-term debt can give rise to inefficient [bank] runs, deposit insurance in the style of that offered by the FDIC benefits society. Central banks’ occasionally providing liquidity support to sound banks can also be beneficial. However, the banks’ safety net distorts the incentives of bankers and their creditors, inducing them to take or to tolerate excessive risks from borrowing and risky investments.
Requiring significantly more equity is the most straightforward way of counteracting these distortions; it simply asks banks to reduce the risk of becoming distressed and thereby harming others.” (p. 179, emphasis supplied)
Argue against that reasoning any way you want, but Admati and Hellwig have it covered. Will the Basel accords stop inefficient lending and restore capital to safe levels? No way. Are bank stress tests effective? Not even close.
But, how much will it cost to provide more equity to banks? The authors convincingly argue that the benefits of adding equity to banks is enormous, and the costs borne by banks are tiny. First, although self-interested banking oligarchs claim otherwise, it stands to reason that changing the mix of bank funding (the right-hand side of a bank’s balance sheet) doesn’t change the amount of funds available to be invested (the left-hand side). And, the authors show that saner incentives should result in the banks’ investment portfolios becoming more efficient. Second, if there are costs to increasing equity, those costs will be paid for by reducing excess compensation to executives who could be inappropriately rewarded for taking excess risks (owing to the mix of government safety nets and high leverage). Again, in a nutshell:
“Among the advantages to the stability of the financial system of banks’ operating with much more equity is the fact that losses to banks’ assets depleted equity much less intensely and thus to not require as much of an adjustment as when banks have less equity. A loss of 1 percent in the value of a bank’s assets wipes out fully one-third of the bank’s equity if it has only 3 percent of its assets in equity but reduces its equity by only 4 percent if the bank’s equity represents 25 percent of its assets. The bank wants to sell assets to restore the relation between equity and total assets or for other reasons following a loss, it must sell 32 percent of its assets if the initial equity was 3 percent of its assets but only 3 percent of its assets if the initial equity was 25 percent. The contagion effects of deleveraging through distressed sales after losses is much smaller if the initial equity is much higher.” (p. 181, emphasis supplied)
Which finally leads me to have to explain why a book that deals almost purely with financial matters is appropriate for the theme of this accounting blog. Even though the authors explain that sufficient bank equity is essential to a safe and sound banking, they have extremely little to say about actually measuring equity:
“Since 2010, when we became more outspoken about the need for an ambitious reform of capital regulation, we had engaged in many discussions on the subject, yet we have never received a coherent answer to the question of why banks should not have equity levels between 20 and 30% of their total assets. [Footnote omitted] (A caveat on providing specific ratios is that their meaning will depend on accounting conventions.)” (pp. 181 – 182, emphasis supplied)
For such an amazing book, that’s pretty weak tea. How should loans and other investments be measured? Which, if any, derivatives should be presented gross, as opposed to net? How should regulated banks present assets that have been “transferred” to shadow banks if the regulated bank shields the shadow bank from losses through guarantees, derivatives or some other form of contracting? Some of these questions should be easy, and some of them are hard, but they all need to be answered, and in doing so there is a lot of Type II BS to shovel through. I’ll have more to say about these questions in subsequent posts; but for now, I’ll provide one caveat that relates to some of my more recent posts:
Measures of a bank’s assets must not depend, as they do today, to any degree on management’s estimates. Admati and Hellwig demonstrate that management has incentives to overleverage banks for their personal benefit. These incentives positively interact with the opportunities for the management of many public companies to manipulate accounting earnings. And, as accounting gets more complex, those earnings manipulation opportunities become even more valuable to managers. Accounting for banking enterprises is the sine qua non of complexity.
To be fair, Bankers’ was written for lay audience and it tackled a lot of very subtle topics about pricing of risk and cost of funding, and much more. It is understandable that discussions of accounting issues were finessed, but they kept coming up. The book contains numerous indications throughout that the authors are aware of just how slanted the accounting rules are towards the interest of banking executives. Yet, they decided to focus on the absolute necessity for deleveraging, and for their arguments to be as convincing as possible, they had to unravel the Type II BS purveyed by banking oligarchs in defense of their miniscule capital ratios.
Jon Stewart says, “if you smell something, say something.”
Admati and Hellwig sure do smell the accounting BS, but they leave it to someone else to explain to regulators that faulty measures of bank capital are inimical to the public interest.
* Another one that I highly recommend is 13 Bankers: The Wall Street Takeover and the Next Financial Meltdown (Simon Johnson and James Kwak, 2011)