Being illiquid means that you don’t have resources available to meet your current obligations. Figuring this out is straightforward: either you can pay your bills or you can’t.

Being insolvent means that you owe more than you own. That too seems straightforward. But while solvency is often discussed as a binary condition — either you are or you aren’t — the reality is more complicated.

We’re reminded of this because of an intriguing article in today’s New York Times that reignites the argument about whether the government had the ability to bail out Lehman Brothers before it went bankrupt</a>:

Back in 2008, the Fed possessed broad authority to lend to banks in trouble. Section 13-3 of the Federal Reserve Act provided that “in unusual and exigent circumstances” the Fed could lend to any institution, as long as the loan was “secured to the satisfaction of the Federal Reserve Bank.”

In the eyes of the Fed, that means a firm must be solvent</strong> and have adequate collateral to lend against</strong>, and making that determination was the responsibility of the New York Fed, the regional Fed bank that had begun to assume responsibility for Lehman. On that September weekend, teams from the New York Fed were told to assess Lehman’s solvency and collateral.


In recent interviews, members of the teams said that Lehman had considerable assets that were liquid and easy to value, like United States Treasury securities. The question was Lehman’s illiquid assets — primarily a real estate portfolio that Lehman had recently valued at $50 billion. By Lehman’s account, the firm had a surplus of assets over liabilities of $28.4 billion.


While the Fed team did not come up with a precise value for Lehman’s illiquid assets, it provided a range that was far more generous in its valuations than the private sector had been.

“It was close,” a member of the Fed team that evaluated the collateral said. “Folks were aware of how ambiguous these values are, especially at a time of crisis. So it becomes a policy question: Do you want to take a chance or not?”</blockquote>
The basic business model of banks is to buy risky long-term assets using money borrowed in short-term.</strong> This is inherently risky since creditors have the option of redeeming their claims</strong> for cash on demand or at maturity.*

Most of the time, short-term creditors are happy to let their deposits earn interest and let the bank worry about guarding the cash in its vaults. But occasionally people will want to get more cash out of a bank than the bank has on hand. These “runs” are not random manifestations of the mysterious will of vengeful gods, but the result of unstable and pro-cyclical financial structures</a>.

Banks can sell their assets to raise cash, but this probably means taking a loss. Better, if possible, to use those assets as collateral to borrow from others. Practically every failed banker likes to imagine that he was brought down by a temporary panic and that his debts could have been repaid if only people trusted him about the underlying value of his holdings. The theoretical justification for having “lenders of last resort” (central banks) that are willing to make short-term loans in exchange for collateral is that at least some of these failed bankers were right.

Either way, the amount of cash a bank can raise depends on how much its assets are worth.</strong> The problem for a bank desperate for liquidity is that nothing is inherently “worth” anything (at least from a financial perspective). Value is contingent, not fixed.</strong>

The “fair” price of an office building or a mortgage-backed security depends on what happens with, among other things, technological progress, inflation, employment, population growth, regulation, migration, commodity supplies, and geopolitics. Anyone attempting to determine the “fair” price</strong> also has to contend with the risk</strong> that any of her embedded assumptions could be wrong</strong>, as well as the fact that she could be wrong about the possible size of those errors.

This creates problems for central bankers attempting to save institutions such as Lehman. As the NYT explained, there were significant disagreements about the assumptions one could reasonably make about the “true value” of Lehman’s assets — which is precisely why the bank was in trouble in the first place!</strong> Ultimately the choice ends up being political, which is why many think the government let Lehman blow up pour encourager les autres</a>.

As Brad Delong put it</a> (emphasis in original):

Bagehot is often glossed as if he had declared that a central bank in a financial crisis should lend to illiquid</em> but not insolvent</em> institutions. But it is difficult to see how any institution whose solvency is common knowledge could possibly be illiquid.</blockquote>
Contrary to what some of its defenders may think, Lehman didn’t die because
evil short-sellers</a> conspired to bring it down. Lehman’s collapse was a symptom of the economy’s overall weakness and expectations that things would only get worse. These macro forces had already wiped out the value of many private-label mortgage bonds and would soon obliterate trillions of dollars of equity wealth.

Given this backdrop, Lehman was</em> insolvent — at least at the specific point in time when its short-term creditors concluded that the firm didn’t have enough human shields</del> was insufficiently capitalised to protect them from losses.</strong>

After Lehman blew up, the other banks only survived because it was clear that the government would do whatever was necessary to prevent them from dying. It no longer mattered whether those firms were solvent or not, because the Fed would ensure that they could always roll over their debts while taxpayers would be on the hook for any losses. Creditors don’t need to care about a borrower’s net worth if they know they will get paid back in full either way.

The situation fundamentally changed in March 2009, with the announcement of a combined package of fiscal and monetary stimulus. Rather than force the market to liquidate excess debts, the government was going to use its own balance sheet to help restore as much of the pre-crisis economy as it could. It wasn’t a coincidence that the best-performing assets that year were the ones that fell the hardest in 2008.</strong>

Banks that would probably have been insolvent quickly became solvent. Had it been lucky enough to get bailed out in 2008, Lehman probably would have appeared solvent as well even if the same mix of assets and liabilities would have made it insolvent just a few months earlier.

The whole idea that policymakers should focus their efforts on “solvent but temporarily illiquid” institutions therefore strikes us as a bit confused.

Does the government know more about a bank’s assets than its creditors and shareholders do? Maybe!</strong> After all, they have supervisors inside the banks, which is a lot more than an investor could hope for. (Take that logic too far, of course, and you undermine the rationale for having a privatised financial system in the first place…)

The government could plausibly argue that it has the ability to hold onto risky and illiquid assets during crises and wait until “normal” conditions return, therefore allowing it to determine a bank’s solvency by generously assuming that it only has to operate in “normal” circumstances. But who defines “normal”? What if what was believed to be “normal” before the crisis was actually freakish?</strong>

And besides, why should banks get the benefit of ignoring downside risks when deciding how much they should leverage? By creating a floor on asset values out of thin air, the government transfers wealth to bankers and bank stakeholders from everyone else.

Banks, by virtue of their unusual business model, exist in a netherworld between solvency and insolvency. If you have to ask whether a bank can repay all of its creditors all at once, it probably can’t. Government guarantees can provide some stability to this inherently unstable system, but the side effects include excessive risk-taking and unfair transfers to the beneficiaries of those guarantees. We suspect it would be simpler to just change the business model</a>.</strong>

The problem with Lehman wasn’t that the Fed made a bad judgment call in the heat of the moment, but that things ever got that far.