Are principal writedowns bad for bank equity?

By Felix Salmon
July 13, 2011
Steve Waldman has a fantastic reaction to my post about principal reductions, saying that "accounting is destiny":

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

Steve Waldman has a fantastic reaction to my post about principal reductions, saying that “accounting is destiny”:

If a bank has a loan on its books valued at par, and it offers a principal reduction, it must write down the value of the loan. It takes a hit against its capital position, and experiences an event of nonperformance that even the most sympathetic regulators will have no choice but to tabulate. If a bank has purchased a loan at a discount, however, the loan is on the books at historical cost. The bank can offer a principal reduction down to the discounted value without experiencing any loss of book equity.

Of course this is a matter of mere accounting. Whether or not a bank takes a capital hit has no bearing on whether a principal reduction will increase the realizable cash-flow value of the loan.

But accounting is destiny. The economic value of a bank franchise, both to shareholders and managers, is intimately wound up with its accounting position. A bank whose books are healthy may distribute cash to shareholders and managers, while a bank whose capital position has deteriorated will find itself constrained. A well-capitalized bank is free to take on lucrative, speculative new business, while a troubled bank must remain boringly and unprofitably vanilla.

This is, basically, the muddle-through approach to troubled assets. If you keep them on your book at par, then you can claim to be well capitalized, and use all that capital to pay yourself well and expand into new businesses. Eventually, with luck, those businesses will be successful enough that they make enough money to cover the losses on the assets when they’re finally realized.

But the real world doesn’t always work like that. See David Reilly’s very good piece on Bank of America today:

BofA maintains it has ample capital and can grow into new, more stringent capital requirements by 2019. The trouble is investors might not be so patient…

BofA is juggling mortgage problems as the economy and housing again look shaky. Markets already have shown what they think: BofA’s shares trade at about 50% of book value and about 85% of stated tangible book value. This means investors think either its assets are overstated or liabilities understated.

BofA needs to put questions about its mortgage risks firmly to bed. Or it needs a thicker equity cushion.

In other words, there are two different ways of looking at bank equity. One is Waldman’s way, where you take the bank’s stated assets, subtract its liabilities, and are left with its equity. And then there’s the more common and salient way of doing things, which is just to look at the share price. While it’s true that regulators do worry about accounting equity, they pay attention to share prices too. (That’s one reason that they weren’t worried about banks during the subprime bubble: the shares showed very little risk there.) And when it comes to paying employees and opening new business lines, a healthy share price is in much more useful than any accounting fiction.

Investors know how much subprime junk is buried on the balance sheets of America’s biggest banks, and they take that into account when they value those banks’ shares. If the banks, through principal reductions, can increase the real value of that junk, Wall Street will likely reward them rather than punish them. Even if that means taking a write-down on assets which everybody knows are worth significantly less than 100 cents on the dollar.

More From Felix Salmon
Post Felix
The Piketty pessimist
The most expensive lottery ticket in the world
The problems of HFT, Joe Stiglitz edition
Private equity math, Nuveen edition
Five explanations for Greece’s bond yield
Comments
12 comments so far

I don’t see how you can write, with a straight face:
“That’s one reason that they weren’t worried about banks during the subprime bubble: the shares showed very little risk there,” and then one sentence later state, “Investors know how much subprime junk is buried on the balance sheets of America’s biggest banks, and they take that into account when they value those banks’ shares.”

Posted by MitchW | Report as abusive

Someone please explain to me why the share price of stock that had already been sold by a bank has any bearing on its lending capacity. Why do regulators care about the share price? How does the share price affect a bank’s ability to cover its liabilities?

And I would challenge the claim that investors know how much subprime junk is buried on the balance sheets of America’s biggest banks. We have no idea, which is bad enough in itself, but worse when the banks’ executives have no idea, either.

Posted by KenG_CA | Report as abusive

Another issue is you’ve ignored Steve’s optionality point. If investors know the balance sheet is lousy, then they also know that the bank has the option of writing down that principal in the future if it wants to.

The real question is whether they’d rather have regulatory advantage from a puffed up balance sheet that gives them a strategic advantage, or if they’d like to make a few dollars from a writedown. Clearly the revealed preference so far is toward the former. And if the regulators decide to stop looking the other way, then of course the calculus would change (as it may well in the BoA case). But giving up that option while the strategic advantage is in place is (apparently) not rational for the banks.

Posted by absinthe | Report as abusive

A good analogy might be the issue with observations in quantum mechanics. In measurements of very small phenomena, the measurement itself changes the actual state of the particle and thus the observed measurement. We cannot know the actual state of the particle because any measurement affects the actual state.

Similarly, JPM can write down principal if the new principal is more than book value with no accounting issues. But if the new principal is less than book value, than there must be an accounting writedown and clearly banks do not want writedowns.

The cynical view of this behavior is that Dimon is cooking the books and is trying to trick Wall Street with an income statement far better than reality. However, the accounting writedown may have some real cash flow consequences. A bank may suddenly become less capitalized and thus must pay more for FDIC insurance. I don’t mean more capital, but the actual insurance payment which the bank will never get back. The ratings agencies may also downgrade, which again for regulatory reasons increases the cost of liabilities. Also, if they have to increase capital cushions due to the writedown, the bank will have to issue more stock but it can’t put the cash flow from the new equity to work. The new stock will do nothing except dilute future profits.

I’m sure that there are other real, cash flow consequences of write downs as well. I’m not really an expert in the FASB rules on securities accounting. The best course is probably what should have been done all along: force the banks hand to write down assets from par when they clearly won’t get all their money back. And force them to do it sooner rather than later.

The issue now is that if only JPM writes down par-value mortgages, then JPM is at a competitive disadvantage. They have to keep them at par-value as long as possible to not suffer real, cash flow consequences. But if all banks have their hands forced by regulators, then banks can write down principals when prudent more easily and get past the toxic subprime stuff on their books.

Posted by mwwaters | Report as abusive

You have basically embraced the ‘bite the bullet’ approach to mortgage bubbles: Write down the debt today, take the hit, carry on – the markets will reward you.
But look at the experience of Japanese banks after their commercial real estate boom (never bit the bullet) and American banks after their mortgage boom (haven’t bitten the bullet) and, hell, even the European banks after their PIGS lending (haven’t bitten the bullet). All of them delay and deny.
If biting the bullet is no worse than – and perhaps better than – delay and deny, wouldn’t someone, somewhere have done it?

Posted by RZ0 | Report as abusive

absinthe, that is not how I interpreted Waldman’s admittedly cryptic point about optionality. Here is what I think he was saying.

A bank has a fictive amount of equity that is reported in its accounting statements. It also has a “true” amount of equity that reflects the “true” values of its assets and liabilities. This true equity is unobservable and in principle might be either higher or lower than the fictive equity. In extreme circumstances, an event may occur (inability to pay) that forces fictive and true equity to converge, but in ordinary circumstances it is fictive equity that matters. A high fictive equity benefits bank shareholders and managers because it allows the bank to engage in activities from which it would otherwise be barred, and these activities are (at least fictively) highly profitable; the profits accrue to shareholders and managers.

Now, suppose the bank has the option to execute a transaction that is profitable in real terms but has the effect of diminishing its fictive capital to the point where it must reduce its other profit-making activities. The “strike price” of this option is the foregone profits. This is not the same as the option to write down principal in future. For one thing, that option is continually decaying: foreclosing or foregoing cashflows on a viable modified mortgage results in value destruction for the bank. But even if that were not the case, it would not be the option to which Waldman is referring.

Posted by Greycap | Report as abusive

Sounds like an argument based on the world as it is, and the world as it should be. Waldman makes a great deal of sense based on how the world is. Your argument has more merit based on the world as it should be.

Posted by Sechel | Report as abusive

Felix,
It’s undoubtedly true that SOMEONE will likely get the benefit of an already written-down loan, but it is unlikely to be a distressed homeowner except in very rare instances.

Take a credit-worthy borrower who has suffered the dread double-wammy: terribly underwater and experiencing job/income loss. Let’s say, further, that his original lender went BK and the note was acquired at steep discount (it may even have been bought by B of A). Does B of A cut the borrower some slack, saying “there’s a good fellow, let’s split the difference” or words to that effect? No. It repossesses the house, sells it at a price that is below market value but perhaps above the written-down level, and, voila, it has its money back and one less dead-beat to worry about. Lather, rinse, repeat. Crisis? What crisis?

Somebody gets a break, but it will rarely be the original mortgagee.

Posted by LadyGodiva | Report as abusive

Reply to RZ)

“If biting the bullet is no worse than – and perhaps better than – delay and deny, wouldn’t someone, somewhere have done it?”
The reason someone hasn’t “done it” is that the “someone” who would have to decide to do it is bank management. And while biting the bullet would certainly be better for the economy, the taxpayer and usually the shareholders, it would not be better for bank management who can continue to extract rents as long as they can delay and deny.

Posted by chris9059 | Report as abusive

Let’s start from the perspective that the banks you’re talking about were technically insolvent since before the system seized up. A lot of them probably still are if they had to admit what they expect to realize from their crap books. So what they’re doing is restructuring (re-creating) their equity through an orderly liquidation of assets whose accounting value is greater than their actual value. All this is basically an accounting sideshow for the benefit of some regulators who know better and sponsored by the US taxpayer. Their different stock prices indicate extremely painfully which banks are good at making money and which ones investors expect to be good at making money in the future. The stocks of banks who aren’t expected to perform well are valued at about the price of a night at the roulette table with an accountant spinning the wheel because that is game at that bank. If you have the accounting staff calling the shots at a bank, or any other business, that business is destined to shrink because the accountant will have the staff whittled down to nothing and the cash flow reduced to a trickle because that is what they do. Anytime the accountants are running the business, you’re screwed ..

Posted by Woltmann | Report as abusive

Felix, capital ratios for banks are (broadly) calculated as capital over assets. Keeping non-performing assets at inflated par value increases the denominator of the equation and actually deflates capital ratios. What on earth are you on about when you write, “If you keep them on your book at par, then you can claim to be well capitalized, and use all that capital to pay yourself well and expand into new businesses.”

Furthermore, a well run bank (and I admit that assumes a lot with some of these players) will have prospectively taken the hit to capital by recognizing future writedowns as part of their provisions for bad debt, which fund the allowance for loan loss accounts. This is a bit of an oversimplification, but it should suffice to say that the decision matrix for something like a principal writedown is fairly complicated.

This muddled treatment of notions about “equity”, “share price” and “capital” fails to recognize that, ultimately, capital ratios do matter for regulatory purposes (and to stock prices), and that writing down principal is almost always a bad move from a bank’s perspective.

Posted by DLSabo | Report as abusive

“Felix, capital ratios for banks are (broadly) calculated as capital over assets. Keeping non-performing assets at inflated par value increases the denominator of the equation and actually deflates capital ratios.”

Accounting. Balance sheets balance. When assets are written down, a matching entry needs to be written down on the liabilities & equity side of the balance sheet. That writedown is to retained equity, which is part of capital. Since assets are larger than capital, an equal decrease to both assets and capital results in lower capitalization ratios. Felix is right.

Posted by TurtleBay | Report as abusive
Post Your Comment

We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see http://blogs.reuters.com/fulldisclosure/2010/09/27/toward-a-more-thoughtful-conversation-on-stories/