What The Fed Should Do (Mishkin 1997 & 2007)

Frederic Mishkin, who, along with Don Kohn and Ben Bernanke, is one of the major macroeconomic heavyweights on the Board of Governors, recently gave a speech on the Outlook and Risks for the U.S. Economy in which he highlighted a paper he wrote in 1997, when he was Director of Research for the FRB of New York.

That paper (The Causes and Propagation of Financial Instability: Lessons for Policymakers (pdf)) lays out the circumstances under which asymmetric information problems can magnify and multiply into a full-blown financial crisis. The bad news is that those preconditions describe the current situation disturbingly well. The good news is that Mishkin tells us what the Fed should do in response.

Here's the abridged version of Mishkin's top four causes of financial instability:

Four factors typically help initiate financial instability:

(1) increases in interest rates,
(2) a deterioration in bank balance sheets,
(3) negative shocks to nonbank balance sheets, and
(4) increases in uncertainty.

Countries often begin experiencing major bouts of financial instability when domestic interest rates begin to rise, often with the rise initiated by interest rate increases abroad. For example, most financial crises in the United States in the nineteenth and early twentieth centuries began with a sharp rise in interest rates that followed interest rate increases in the London markets. ... these rises in interest rates increased adverse selection problems in the credit markets. The rise in interest rates also increased moral hazard problems because the resulting decrease in cash flow hurt the balance sheets of nonbank firms. In addition, the increase in interest rates weakened bank balance sheets because of banks’ maturity mismatch....

Figure 1: Propagation of Financial Instability in Industrialized Countries

Also characteristic of the early stages of financial instability is a deterioration in bank balance sheets because of risky loans that have turned sour. In the recent Mexican episode, the source of these weakened balance sheets was financial liberalization that led to a rapid acceleration of bank lending, in which bank credit to the private nonfinancial business sector rose from 10 percent of GDP in 1988 to over 40 percent in 1994. This lending boom, which stressed the screening and monitoring facilities of the Mexican banks, along with the inability of the National Banking Commission in Mexico to adequately supervise these new lending activities, led to growing loan losses in the banking sector. ... The deterioration in bank balance sheets decreased the ability of the banks to lend because efforts to improve their capital ratios required retrenchment on lending.

Stock market crashes are also typically associated with financial instability.... The declining net worth of nonfinancial firms then increased adverse selection and moral hazard problems in financial markets because the effective collateral in the firms had decreased, while the decline in net worth meant that the incentives for borrowers to take on risk at the expense of the lender had increased.

The fourth factor that frequently appears when there is financial instability is an increase in uncertainty, whether because an economy is already in recession, or because a major financial or nonfinancial firm goes bankrupt, or because of increased political instability.... Increases in uncertainty make it harder for financial markets to process information, thereby increasing adverse selection and moral hazard problems and causing a decline in lending and economic activity.

If any of the four factors in the top row of the figure occurs, it can promote financial instability. If all of these factors occur at the same time and are large, the situation is likely to escalate into a full-scale financial crisis, with much greater negative effects on the real economy.

That was what he said ten years ago. Now here's what he said last week:

...an increase in uncertainty and concerns about the quality of information can lead investors to pull back from financial markets and restrict productive lending--with potentially adverse implications for real activity. That is essentially the story I laid out in a paper delivered at the Kansas City Fed’s Jackson Hole conference about ten years ago.⁷ In my view, such an increase in uncertainty is an important part of what we have observed recently.

That takes care of the diagnosis; now for the prescription (again from 1997):

The asymmetric information analysis thus suggests that a lender-of-last-resort role may be necessary to provide liquidity to nonbanking sectors of the financial system in which asymmetric information problems have developed....

One problem in deciding whether to engage in the lender-of-last resort role is to recognize that for it to be effective, it has to be implemented quickly. Less intervention is required the faster the lender-of-last-resort role is implemented because once market participants know that liquidity is being injected into the system, uncertainty in the financial markets will decrease. Thus, the Federal Reserve’s actions during the stock market crash of 1987 are a textbook case of how a lender-of-last-resort role can be performed successfully. The Fed’s action was immediate, with an announcement that operated to decrease uncertainty in the marketplace. Reserves were injected into the system, but once the crisis was over, they were withdrawn. Not only was a financial crisis averted, but also the inflationary consequences of this exercise of the lender-of-last resort role were quite small.

However, the need for the lender-of-last-resort action to be quick does mean that central banks may not be able to wait until all the information is in that tells them a financial crisis is about to occur or is occurring. To wait too long to implement a lender-of last resort policy could be disastrous.

Sounds to me like a 50bp cut on Sept 18, with a neutral bias after that, which implies they'll have a hiking bias once the downside risks from the financial market turmoil recede.