Email exchange on bank leverage, regulation, and economic growth

Emailed to me:

What do you think would happen if we returned to a world where commercial bank leverage was much reduced? (E.g. 2X max.) Or, maybe equivalently, if central banks didn’t act as a lender of last resort? Is that “necessary” for a modern economy?

Asset prices would fall a lot (presumably). What else? How much worse off would current people become? (Future people are presumably somewhat better off, growth implications notwithstanding—they are less burdened with the other side of all these out-of-the-money puts that central banks have effectively issued.) > > How should we think about the optimization space spanning growth rates, banking capital requirements, and intergenerational fairness?

My response:

First, these questions are in those relatively rare areas where even at the conceptual level top people do not agree. So maybe you won’t agree with my responses, but don’t take any answers on trust from anyone else either.

I think of the liquidity transformation of banks in terms of two core activities:

a. Transforming otherwise somewhat illiquid activities into liquid deposits. That boosts risk-taking capacities, boosts aggregate investment, and makes depositors more liquid in real terms. Those are ex ante gains, though note that more risk-taking, even when a good thing, can make economies more volatile.

b. Giving private depositors more nominal liquidity, but in a way that raises prices and thus doesn’t really increase real, inflation-adjusted liquidity for depositors as a whole. There is thus a rent-seeking component to bank activity and liquidity production.

Less bank leverage, you get less of both. In my view a) is usually much more important than b). For those who defend narrow banking, 100% fractional reserves, or just extreme capital requirements, a) is usually minimized. Nonetheless b) is real, and it means that some partial, reasonable regulation won’t wreck the sector as much as it might seem at first.

There is however another factor: if bank leverage gets too high, bank equity takes on too much risk, to take advantage of bank creditors and possibly taxpayers too. Or too much leverage can make a given level of bank manager complacency too socially costly to bear. This latter factor seems to have been very important for the 2007-2008 crisis.

So bank leverage does need to be regulated in some manner, and the better it is regulated the more the system can dispense with other forms of regulation.

That said, the delta really matters. Requiring significantly less bank leverage, at any status quo margin, probably will bring a recession. The recession itself may make banks riskier than the lower leverage will make them safer. In this sense many economies are stuck with the levels of leverage they have, for better or worse. It is not easy to pop a “leverage bubble.”

I don’t find the idea of 40% capital requirements, combined with an absolute minimum of regulation, absurd on the face of it. But I don’t see how we can get there, even for the future generations. We’ll end up doing too many stupid things in the meantime; Dodd-Frank for all its excesses could have been much worse.

I also worry that 40% capital requirements would just push leverage elsewhere in the economy. Possibly into safer sectors, but I wouldn’t be too confident there. And reading any random few books on “bank off-balance sheet risk” will scare the beejesus out of anyone, even in good times.

Now, you worded your question carefully: “commercial bank leverage was much reduced.”

A lot of commercial bank leverage can be replaced by leverage from other sources, many less regulated or less “establishment.” Overall, on current and recent margins I prefer to keep leverage in the commercial banking sector, compared to the relevant alternatives. It may be less efficient but it is socially safer and held within the Fed’s and FDIC regulatory safety net, probably the best of the available politicized alternatives. That said, there is a natural and indeed mostly desirable trend for the commercial banking sector to become less important over time, in part because it is regulated and also somewhat static in basic mentality. (Note that the financial crisis interrupted this process, for instance Goldman taking up a bank charter. I would still bet on it for the longer run.)

Obviously, VC markets are a possible counterfactual. This all gets back to Ed Conard’s neglected and profound point that “equity” is what is scarce in economies, and how many troubles stem from that fact. Ideally, we’d like to organize much more like VC markets, partly as a substitute for bank leverage and the accompanying distorting regulation, and maybe we will over time, but there is a long, long way to go.

One big problem with attempts to radically restrict bank leverage is that they simply shift leverage into other parts of the economy, possibly in more dangerous forms. Should I feel better about commercial credit firms taking up more of this risk? Hard to say, but the Fed would not feel better about that, it makes their job harder. This gets back to being somewhat stuck with the levels of leverage one already has, until they blow up at least. There are pretty much always ways to create leverage that regulators cannot so easily control or perhaps not even understand. Again this bring us back to “off-balance risk,” among other topics including of course fintech.

I view central banks as “lenders of second resort.” The first resort is the private sector, the last resort is Congress. I favor empowering central banks to keep Congress out of it. Central banks are actually a fairly early line of defense, in military terms. And I almost always prefer them to the legislature in virtually all developed countries.

I fear however that we will have to rely on the LOLR function more and more often. Consider how it interacts with deposit insurance. If everything were like a simple form of FDIC-insured demand deposits, FDIC guarantees would suffice.

But what if a demand deposit is no longer so well-defined? What about money market funds? Repurchase agreements? Derivatives and other synthetic positions? Guaranteeing demand deposits is a weaker and weaker protection for the aggregate, as indeed we learned in 2008. The Ricardo Hausmann position is to extend the governmental guarantees to as many areas as possible, but that makes me deeply nervous. Not only is this fiscally dangerous, I also think it would lead to stifling regulation being applied too broadly.

But relying more and more on LOLR also makes me nervous. So I view this as a major way in which the modern world is headed for recurring trouble on a significant scale, no matter what regulators do.

I am never sure how much of the benefits of banking/finance are “level effects” as opposed to “growth effects.” It is easy for me to believe that good banking/finance enables more consumption at a sustainably higher level, in part because precautionary savings motives can be satisfied more effectively and with less sacrifice. I am less sure that the long-term growth rate of the economy will rise; if so, that does not seem to show up in the data once economies cross over the middle income trap. That said, if there were an effect, since growth rates slow down with high levels in any case, I don’t think it would be easy to find and verify.


I am definitely for narrow banking, any benefit of increased liquidity by banks is massively outweighed by the moral hazard of TBTF of major commercial banks. Not only does this create bad behaviour by bankers it also provides ammunition for anti-capitalist campaigners like Corbyn and Trump who get to power thanks to public disgust against the "system". The consequences of this as we saw in places like Venezuela could be very serious. I would make it illegal for people to offer guaranteed debt returns with no notice period in all circumstances unless the cash was directly deposited in a government account available on immediate notice, of course in a fiat currency world that pledge can always be honoured. We managed to make Ponzi schemes illegal, I regard this as similar. In terms of the political will to do it, we just need one more massive bailout like 2007 and I think you will be there.

Ie, you would require everything be common stock, not loans?

Common stock may or may not pay dividends. No assurance you can get your money back, but you might get more.

In fact, banking was originally more like trading shares, and not at all like taking deposits and making loans.

I like ChrisA's thoughts of deleveraging banks, but I don't like narrow banking since it relies on government fiat money and government ledgers for record keeping. Though I do think money is neutral (hence narrow banking is no worse than any other kind of banking like fractional reserves banking), I don't like government in charge of the ledgers. My suggestion, that's politically more palatable than narrow banking, is a return to the gold standard. Keep in mind I think money is neutral, so don't tar me with the "goldbug" brush. But gold-backed money will decrease volatility, and also IMO implicitly reduce leverage.

You see the same tension between leverage/risk taking vs safety with bankruptcy laws. I think bankruptcy should be made harder to do, so we have less speculation. If in fact we have a Great Stagnation then we don't need more speculation, since it will lead to nothing. Instead of speculation, we should have a better patent system and more government R&D, which would move real GDP a lot more than speculation and leverage do.

Re your objections to "government fiat money" (i.e. base money), I don't see much wrong with that. What's fundamentally wrong with Fed issued dollars, or ECB issued Euros?

Re your objections to "government ledgers", I agree that that makes it easier for busy body government officials to spy on what you and me do with our money. But there's a solution to that, which is to have existing commercial banks do the detailed work (with check books, debit cards, etc) while those commercial banks simply remit to the central bank at the end of each day the total amount that their customers say they want deposited at the central bank, net of withdrawals.

My head spins with the concepts, reserve, leverage, equity.... Nothing need be that complex in finance or banking.

Start with a zero profit banking model, remove time, and see if you can get a simple math. Like, for example. Real growth is about 2-3%, the implicit deflator about 1.5%, so let the currency banker run a bounded 5% variation between loans and deposit balances, assume it is automated; solve that model. Then add in all the bells and whistles to conform to our distorted system.

To avoid LOLR-squashing and leverage-level regulation, add more skin-in-the-game requirements? E.g. require rating agencies to write CDSs on the debt they rate? Make banks pay FDIC rates in line with the banks' ratings?

While this could be seen as just pushing the problem onto Insurance -- and how do we know what level is growth-optimal -- imagine insurance markets going through the democratization / transaction cost reductions / innovation that the equity markets have gone through in the past 20 years.

More skin in the game is selling shares instead of borrowing money, buying shares instead of lending.

"Islamic" banking retains this historic, biblical standard, once part of Christian and Jewish doctrine.

Assume that you don't know or can't enforce the optimal level of leverage, what can you change?

As TC points out, the debt cat's already out of the bag, and it's not going back in without a recession.

Is commercial bank leverage substatialy different now compared to the times prior to central banks? Does anyone have access to some commercial bank balance sheets from around 1900?

According to Mervyn King (former governor of the Bank of England), in Walter Bagehot's day (around the middle of the 1800s), bank leverage was much lower then - about six to one, he says. See p.3 here:

Thanks for the link. Interesting read.

I think illiquid industrial continuity often seems riskier than High Bet, and often seems healthier

More leverage – less leverage “these questions are in those relatively rare areas where even at the conceptual level top people do not agree.”

May I get in?

The problem now is that with risk-weighted capital requirements banks are authorized to leverage differently; more with “safe” assets, less with “risky” assets; which allows banks to earn higher risk-adjusted returns on equity on “safe” assets than on “risky” assets.

This has two major problems:

First, it distorts the allocation of bank credit to the real economy… affecting growth (World Bank issue)
Secondly, it is absolutely useless for purposes of financial stability (IMF issue); as you should be able to figure out when I tell you that Basel II assigned a meager 20% risk weight to the AAA rated, that with which banks could create dangerous exposures to; and a whopping 150% to the below BB- rated, that which banks wont even touch with a ten feet pole.

Here I link to a growing list of mistakes.

Keeping leverage in the commercial banking sector (where it can be regulated and risks can be mitigated) is certainly safer that most alternatives, but didn't "bailing out" the banks and bankers help produce the election of a demagogue as president, something that may turn out to be far more damaging and long-lasting than a financial crisis? Here is Robert Shiller's latest on risk in the financial markets: Shiller's point isn't that today's level of risk is appreciably higher than normal but that "[t]echnology has made viral rumor transmission much easier" (as compared to 1987 when the Dow fell 20% in one day). [An aside, the preference for debt over equity had its genesis in the 1980s and was promoted by financier Michael Milken, including debt issued by the parent company of many banks (S&Ls) - the banks themselves had an all-equity capital structure, hiding the risk that was in plain sight. My point is that "creative" financiers make it impossible to regulate risk in the financial sector by keeping leverage in the commercial banking sector. Indeed, today (as compared to the 1980s) lying is the coin of the realm.]

You might have said more about the role of federal deposit insurance, though I realze that getting rid of it seems politically impossible.

Agreed. Money market funds are basically banks without FDIC and they get by just fine without any equity at all.

"Money market funds are basically banks without FDIC and they get by just fine without any equity at all."

Until they don't. Wasn't one of the critical points in the great financial crisis the busting of the $1/share for a couple of money market funds back in 2007? Are you overconfident that the money market fund managers really know what they are doing?

Strikes me that the stakes that savers have in "floating net asset value" MMMFs are actually equity. Like equity, the value of those stakes floats and come a total and complete disaster, the value of those stakes is wiped out.

Maybe the answer should reference the relevance of tax policy (and Modigliani/Miller)? And the empirical work?

Treasury report n. 18:

"Literature reviews conducted by Oliver Wyman and staff at the OFR, IMF, and Federal Reserve generally review a similar set of studies that attempt to estimate the costs of higher bank capital requirements. See Oliver Wyman. Interaction, Coherence, and Overall Calibration of Post Crisis Basel Reforms. August 9, 2016, available at:
See also Jihad Dagher, Giovanni Dell’Ariccia, Luc Laeven, Lev Ratnovski, and Hui Tong. Benefits and Costs of Bank Capital. IMF, March 2016, available at: See also Federal Reserve Board. An Empirical Economic Assessment of the Costs and Benefits of Bank Capital in the US. March 31, 2017, available at: These reviewed studies generally found evidence of higher credit costs to borrowers and reduced lending capacity. The magnitude of the higher cost of credit upon borrowing costs generally ranged from 2 to 21 bps for a 1 percentage point increase in bank capital requirements. Oliver Wyman estimated that range would imply a gross increase in credit of between 15 to 109 bps (assuming an increase in capital requirements of 5.3% in the U.S.). The estimated range of credit contraction (see IMF table 4B) is focused on reduced volumes of bank credit during the implementation phase of higher capital requirements."

As memory serves me, back in the 1970's money and banking courses viewed the health and credit worthiness of banks via various ratio analyses. The Fed clearly saw bank leverage increasing prior to the Crash and it was fully in its domain to tell the Banks to sell assets or raise equity as they saw leverage increase meaningful beyond historical standards, but did not.
Depositors can easily be protected and socialized costs minimized if not eliminated by ring fencing the depository side of a bank. Insurance companies do this to safeguard policy holders and regulators can sell off a sub quite easily leaving equity holders and bond holders at the holdco level to restructure assets not in the policy holding sub.

Who is to blame for the 2008 financial crisis? Here are the results of a poll conducted by the IGM Center at the Univ. of Chicago of leading American and European economists: What's striking to me is that the economists have such differing opinions. Far down the list is savings and investment imbalances; it's not surprising that it's far down the list but that it even made the list. Of course, "savings and investment imbalances" is a euphemism.

What a fascinating meditation. Much to think about. Thank you.

Proponents usually view the abandonment of a) a feature rather than a bug, and growth and required adaptation as an inconvenience. They believe themselves entitled to riskless positive returns however impossible that is.

Here's something I deduced from this TC post that none of you probably thought about: the author of the email was a high-status individual, like say FT's Martin Wolf. You can deduce this from the lengthy and thoughtful reply by TC, like he's talking to one of his peers. When I email TC with my brilliant thoughts, he replies, if at all, with a few words answer, sometimes canned, like "thanks I'll read". :-)

Bonus trivia: reading "Dark Money" by Jane Meyer, surprisingly good with scuttlebutt on the Koch family like: "[Fred Koch] employed a German governess for his first two sons, Freddie and Charles. At the time, Freddie was a small boy, and Charles still in diapers. The nanny’s iron rule terrified the little boys, according to a family acquaintance. In addition to being overbearing, she was a fervent Nazi sympathizer, who frequently touted Hitler’s virtues. Dressed in a starched white uniform and pointed nurse’s hat, she arrived with a stash of gruesome German children’s books, including the Victorian classic Der Struwwelpeter, that featured sadistic consequences for misbehavior ranging from cutting off one child’s thumbs to burning another to death. The acquaintance recalled that the nurse had a commensurately harsh and dictatorial approach to child rearing. She enforced a rigid toilet-training regimen requiring the boys to produce morning bowel movements precisely on schedule or be force-fed castor oil and subjected to enemas."

This would make a great movie. Nazi governess, I'd watch that. Maybe it would be better as a porn flick than a documentary.

Tyler says: And reading any random few books on “bank off-balance sheet risk” will scare the beejesus out of anyone, even in good times.

What books is he referring to?

Tyler says “Requiring significantly less bank leverage, at any status quo margin, probably will bring a recession.” Well that’s easily dealt with via standard stimulatory measures, monetary and/or fiscal. Personally I prefer a combination of the two, which essentially equals “the state prints money and spends it into the private sector”.

Next, Tyler says “I don’t find the idea of 40% capital requirements, combined with an absolute minimum of regulation, absurd on the face of it. But I don’t see how we can get there, even for the future generations.” Well Milton Friedman advocated something more extreme than 40% capital: he advocated 100%! As to how quickly he thought that could be implemented, he said, “There is no technical problem in achieving a transition from our present system to 100% reserves easily, fairly speedily and without any serious repercussions…” (Ch3 of his book “A Program for Monetary Stability”).

Finally, Tyler seems to suggest in his final paragraphs that 40% or 100% capital requirements involves more reliance on LOLR. Quite the opposite! Under 100% reserves LOLR is no longer needed. Reason is that depositors who want total safety simply have their money lodged at the central bank where there is next to no risk. As to those who want their money loaned out, they carry relevant risks. E.g. if depositors want their money to fund NINJA mortgages, and the value of those loans drops to 50% of book value, then the value of those depositor’s stakes in NINJA funds drops to about 50% of book value. No need for taxpayer funded rescues!

No sooner do financial experts plead ignorance to avoid prosecution or being held responsible for egregious bullshit justifying their own actions that led to the 2008 financial crisis than some people are quoting them as disinterested experts on the need for liquidity in banking. It was the fear of Narrow Banking that led bankers to endorse the banking restrictions put in place in the 1930s. What the banks fear, we want. But no. Once again, time will erode restrictions on banks as they puff their self-serving nostrums up the arses of the very important people. Narrow Banking, Vested Interest, Prosecution of Individuals, disallowing further growth and metastization of large banks, etc., will give way to fear of slower growth based upon? Keeping puffing men.

I offer one book...
Ronnie J. Phillips. The Chicago Plan and New Deal banking reform. Armonk, NY: M.E. Sharpe, Foreword by Hyman P. Minsky., CO State U, 1995

It's all there. Meet Frank Knight, Henry Simons, Irving Fisher...We have not seen their like again.

Bank capital requirements are not the only or the best way to reduce leverage in the economy generally and make it more robust. We can't prevent business cycles, but if our economy is less leveraged, we can make recessions less painful.

How do we achieve that without bank regulation? With an economist's favorite tools: incentives and alternatives.

* We have had a tax code that systemically favors debt over equity for a very, very long time. It is not hard to change this tax code bias. We could make dividends paid deductible to C-corporations equalizing the tax treatment of debt and equity. We could eliminate the tax deductions for interest on new mortgage debt and new student loan debt - if we want to subsidize home ownership and education there are more direct ways to do that than subsidizing debt financing of those investments. We could create a deduction for mortgage insurance that due to its regulatory structure discourages housing bubbles more than tiers of first and second loans that were common in the financial crisis. We could limit depreciation deductions on real estate to no more than the amount of principal paid on mortgage loans so taxable profits would more closely match cash flow instead of creating a tax preference.

* We could reduce the cost of small and medium sized public offerings of equity - perhaps with more safe harbors of some kind. We could finance higher education with a percentage of future income instead of as simple debt (Australia does it). We could experiment with mortgages that have lower interest rates and smaller payments, but share proceeds on sales and losses with the "lender." We could move to a single payer system where you give up a percentage of your income for health care, but don't have a debt-like large monthly insurance payment that you have to pay to get health insurance. We could require limited liability entities big and small to be bonded as to their trade creditors so that only bonding companies and long term financing lenders with the clout and incentive to make a consciously contemplated and researched credit decision would bear the risk of a default on their debts. We could give collection costs, late charges and default interest charges in excess of non-default interest rates subordinated debt status in bankruptcy relative to principal and non-default interest to discourage charges that cause a debtor to fail due to a minor stumble that gets out of control pushing creditors into races to inflate their debts for the purpose of getting a larger share of available assets in a creditor v. creditor fight in bankruptcy. We could create an "express bankruptcy" option that converts long term finance debt and bonding company debt into equity, cancels existing equity, and doesn't affect day to day operations at all.

One more key alternative: encourage more mutual companies in the financial sector (i.e. investor/customer owned) and discourage shareholder ownership of economically important, highly leveraged businesses. Companies organized that way systemically set appropriate reserves for themselves because they haven't set up a heads I win, tails you lose situation.

Comments for this post are closed