Recently, I engaged in a Twitter exchange in which I asserted the financial accelerator model proposed by Bernanke, Gertler and Gilchrist was a complete piece of junk (another 4-letter word is actually more appropriate).  Naturally, this did not go over well with the PhD Economist, Christoph Thoenissen, who I was exchanging tweets with.  Especially because he had written a paper attempting to extend the applicability of this model.  The reason the model is fatally flawed is its lack of connection to the real world.

As I tried to explain, if you were a policymaker in the 1930s and were given the task of redesigning the financial system you faced a question.  Would you redesign the financial system to transmit problems in the financial system to the real economy or would you redesign the financial system to protect the real economy from problems that develop in the financial system?

If you were like the policymakers or myself, your answer would be to redesign the financial system to protect the real economy from problems that develop in the financial system.  This redesign was and is elegant.

Policymakers recognized so long as a bank is in business, it would continue making loans to support the real economy.  This was true whether the bank was solvent or insolvent.  Why can banks do this?  In the presence of deposit insurance and a lender of last resort, an insolvent bank can continue operating indefinitely until it is shutdown by financial regulators.  Is there a real world example of this happening.  At the time BGG were writing their paper in the 1980s, the Savings & Loans Crisis was going on in the US.  These insolvent banking institutions were well known to be making high risk loans in a desperate gamble on redemption.

Policymakers also recognized this ability to continue in operation even while insolvent makes banks perfect for absorbing losses (also known as credit market frictions/financial distress) on the bad debt in the financial system.

So policymakers not only understood how to trap all the bad debt in the financial system without impacting the real economy, but also how to keep lending going! [Unlike the EU/Japan/UK/US, both Sweden and Iceland used their financial systems as they were designed when facing a financial crisis.  Both the severity and length of their crises was minimized.]

One more thing policymakers understood.  By making bankers and the investors in banks responsible for the losses on the banks’ exposures, you gave the bankers and the investors an incentive to make sure the banks didn’t take on more risk than they could afford to lose.

Let me explain how this incentive works in the world of banking.  Before making a loan, bankers look to the big 3 Cs:  Character, Cash flow and Collateral.

Character comes first because if you lend money to someone who has no intent of repaying their loan, it is virtually assured they will not repay.

Cash flow comes second.  You don’t lend a borrower more than they can afford to repay.  When bankers ask how much can a borrower afford to repay, they don’t do this with the intent of lending them 100% of what the borrower can afford to repay.  Bankers know circumstances change.  So they build in what is known as a “Margin of Safety”.  The margin of safety is reflected by the bank’s being willing to lend a borrower up to say 50% of what the borrower can afford to repay.

Please note, by building in a margin of safety, bankers don’t have to be constantly changing how much they will lend to the borrower.  Hence, any borrower’s access to credit tends to be reasonably constant over time barring any significant changes in what they can afford to repay.

Collateral comes third.  All bankers know if you have to resell the collateral pledged to back a loan to repay the loan you are likely to lose money.  For example, the buggy whip manufacturing facility might be worth $5 million if buggy whips are still being manufactured.  Otherwise, it might be worth only $500,000.  Again, the banker will build in a margin of safety before making a loan.

So what does this have to do with Bernanke’s financial accelerator?  BGG’s paper starts out by making a series of assumptions that are inconsistent with and don’t acknowledge the existence of a margin of safety.

For example, they state if the value of collateral declines, so too does the amount of the loan.  For this to be true, the decline would have to exceed the margin of safety.  Is there any reason to think bankers would set the margin of safety so it is routinely inadequate for cyclical changes in valuation?  No.

They also talk about information asymmetry between the borrower and the lender.  If you are a lender and you don’t have access to all the information you need in order to assess the risk of a loan, would you make the loan?  If you did make the loan, how big of a margin of safety would you require?

Ignoring the existence of the margin of safety allows Bernanke et al to create a model with zero explanatory power including the idea of a financial accelerator.

Why?

Because the margin of safety means lending is simply not that sensitive to monetary shocks, changes in collateral value, etc.