VoxEU Column Financial Markets

Escaping from a Combined Liquidity Trap and Credit Crunch

The dizzying falls in equity prices seem to have stopped. If they restart, it may be time for radical measures. This column suggests one motivated by bubble theory. The Fed could temporarily guarantee a lower bound for the S&P 500 through targeted purchases of market portfolios via open-market operations and financed by injecting cash.

Between the collapse of Lehman Brothers on 15 September and the announcements of European and US bank recapitalizations on 13 and 14 October, stock prices fell daily, producing double-digit percentage-point losses in most major markets. The muscular interventions agreed on 13 and 14 October seem to have quelled the worst of the panic, but stock prices have not rebounded.

The current situation is worrisome. Private investors are liquidating all kinds of real-valued assets and are instead hoarding liquidity at banks considered safe due to government guarantees. Banks are selling their shares and are calling in debts to minimize loses and save their remaining equity and to simply remain solvent. They are parking their liquidity at central banks and are not issuing new loans.

Monetary policy is currently ineffectual. Interest rate cuts are not being passed on and the expansion of the money supply has also failed to influence credit terms.

We are in a credit crunch – albeit not for lack of liquidity in the banking sector, but rather due to banks’ naked fear of having to write off more debts and, thereby, of endangering their own solvency.

As monetary policy can presently neither reduce market interest rates nor stimulate the issuance of private loans, we are, furthermore, in a liquidity trap. The novelty here is that this liquidity trap is occurring at significantly positive interest rates, whereas it appeared in Japan through the zero bound on nominal interest rates. 

A Pessimistic Forecast

If the large scale selling of stocks continues, the assets of banks and insurance companies will continue to fall in value until these institutes are insolvent.

Banks and insurance companies may be nationalized as governments infuse these institutions with fresh capital in exchange for shares.

This will be financed through the issuance of new government debt that will be accepted by markets as investors are fleeing to safe nominal assets.

Should governments not be able to finance themselves on capital markets, due to a loss of confidence by the markets, they would still have access to a final recourse: the purchase of these government debts by central banks.

Alternatively, central banks could prop up banks by recapitalizing them. Here, a lasting expansion of the money supply would be necessary.

The last two possibilities would inevitably lead to higher inflation.

Textbook Wisdom

There are two ways out of a liquidity crisis:

  1. An increase in government expenditures (fiscal policy). This is already being accomplished indirectly through the partial nationalization of the banking sector. Additional government expenditures to stimulate the demand for goods could reduce the consequences of the crisis for the real economy, but would likely lead to large distortions and inefficiencies. Increased aggregate demand will not, however, stimulate investment if banks do not provide loans. Therefore, multiplier effects fail to unfold.
  2. Increasing inflation expectations. This would lead to a reduction in real interest rates even with constant nominal rates and normally would stimulate investment demand. In the current situation, however, the latter is unlikely, as investment demand is irrelevant for markets so long as banks are not issuing new loans.

Textbook wisdom on escaping a liquidity trap is unlikely to work in the current situation, because we are in the unique situation of a combination of a liquidity trap and a credit crunch.

Downward spiral

We are currently experiencing a downward spiral. Banks are selling their assets to maintain their solvency. As all banks are acting in the same way, this reduces the value of assets, destroying banks’ equity. As a result, banks are forced to sell even more assts. Private investors are also acting in the same way to limit their losses. Since October 13, there are signs, however, that the downward spiral is coming to a halt.

What to do if the downward spiral resumes

Real value underlies stocks and mortgages. The downward spiral leads to a negative price bubble as was last seen in 2002. The dividend returns then exceed the returns on fixed income instruments and the risk for long-run oriented investors becomes very small. Eventually, investors will shift their portfolios and take advantage of the depressed prices to enter the stock and real estate markets.

When will the flow of capital turn around?

It is inevitable that expectations of a future rise in prices eventually prevail over expectations of a continuation in price decreases. For this to happen, however, a significant disequilibrium would have to occur first.


We are seeing an increasing (positive) bubble in fix-income investments, government bonds and other investments considered “safe”. (As their nominal value provides the basis of measurement, this is equivalent to a negative bubble in real values.) Here, Abreu and Brunnermeier (2003) and Brunnermeier and Morgan (2005) are applicable: every market participant is aware of the mistaken valuation, the lack of common knowledge, however, hinders the bubble’s bursting. The turnaround in capital flows can be expedited by a common signal, a piece of information that provides the markets with common knowledge that investment in real-valued assets is worthwhile.

What could such a signal look like?

The coordinated lowering of target interest rates on October 8 was apparently not a sufficient signal. From theoretical research and experimental results regarding coordination games and bubbles, we know that a common signal has an impact on behaviour and beliefs about others’ behaviour only when the signal by itself has an impact on expected payoffs.

That means:
  1. A signal must change the relative return expectations even in the absence of immediate changes in behaviour.
  2. Due to the changes in return expectations, there is an increase in market participants’ propensity to change their investment positions.
  3. As the signal is common knowledge, market participants expect that others will also change their positions.
  4. Due to strategic complementarities, participants react more strongly to the signal than they would have solely due to 1.

The coordinated lowering of interest rates did create common knowledge, but failed to fulfil condition 1, since changes in central bank rates are currently ineffectual or are considered ineffectual by the markets.

A coordinated recapitalization of banks, as was decided during the week from October 13 to October 17, might not suffice, as banks have no interest in buying stocks or expanding lending under current conditions. Although banks are required to expand lending to the real sector as a condition for recapitalization by the government, it is not yet clear whether this stipulation will have any impact on the credit crunch and lead to the desired turnaround in market expectations.

Concrete suggestions:

A turnaround in the dynamics could be achieved by several signals:

  1. Several large private investors, whose actions comment sufficient attention, invest their liquid funds in stocks. Warren Buffet already did something along these lines with his investment in Goldman Sachs. This was, though, only a recapitalization of one bank and, so had only a small impact. What would have been necessary was the purchase of a broader set of assets on a much larger scale. I do not think that private investors hold enough funds to invest on the necessary scale.
  2. Sovereign wealth and pension funds engage in large-scale purchases of stocks. These funds, like private investors, have no interest in coordinating their purchases for the common good, as they would thereby eliminate the bargains they would otherwise be able to get. If only in the interest of their investors, pension funds are not in a position to make coordinated purchases.
  3. China uses its foreign exchange reserves to buy stocks on US markets and makes its intentions publicly known. One could also question China’s interest in announcing its intentions, but China would be the hero of the US financial system if it were to succeed. It could, furthermore, increase its strategic holdings and finally get rid of its exposure to Dollar-denominated government bonds. It would be a realistic chance for China to participate in stock markets without provoking knee-jerk political reactions. Thus far, however, China does not appear to be seriously involved in discussions.
  4. The Fed could buy up the excess supply of stocks and guarantee a clearly defined lower bound for the major stock indices for a limited time. This could be easily achieved through the targeted purchase of market portfolios in exchange for cash (Concrete example: the backing of the S&P 500 through the sharing of costs and acquired shares by the participating reserve banks).
  5. A credible announcement by the Fed of inflationary policy for a limited time.

Options 4 and 5 seem the most realistic, so I focus on them and, in the end, recommend option 4.

Temporary guarantee of a lower bound on stock prices:

I assume that the guarantee will be at a level that is clearly below the present value of expected future revenues (fundamental value). Thus, it should only be effective if there really is a negative bubble; and, in the long run, potential purchases are expected to be profitable.

Pros: The announcement alone should prevent further sales from occurring. The Fed would actually only need to engage in limited purchases in an initial period to ensure the credibility of their announcement. By eliminating the downward risk, the flow of capital would reverse immediately and prices would advance significantly past the announced lower bound.

Cons: A guaranteed lower bound creates moral hazard problems. Thus, any such guarantee must be temporary in nature.

Inflation expectations: As the expiration of the guarantee nears, there are three possible scenarios to consider:

  1. The crisis has come to a halt. Stock prices are sufficiently above the announced lower bound, and the expiration date has no significant effect on the level (apart from some temporarily higher volatility).
  2. The crisis has come to halt, but markets expect that the guarantee is still a binding restriction. Stock owners will sell to the Fed well before the expiration date and the Fed is forced to acquire a large amount of stocks.
  3. The crisis continues up to the expiration date of the guarantee: In this case, the Fed may decide to extend the guarantee or let it expire and stand ready to acquire a large amount of stocks.

If the Fed really is forced to buy a large amount of stocks, this will be conducted through open-market operations, increasing the money supply, and, thereafter, increasing inflation. As there is an ex ante positive probability that one these negative scenarios realizes, inflation expectations should increase immediately after the credible announcement of such a guarantee.

To the extent that there is immediate action and a well-defined expiration date, the suggestion comes close to a foolproof way of escaping a liquidity trap in the spirit of Svensson (2003).

Note, however, that in the period leading up to the expiration date, inflationary expectations will exceed actual inflation. According to textbook wisdom, this should have contractionary effects on the demand side and tend to raise savings. In the current situation, however, with a liquidity trap occurring simultaneously with a credit crunch, such reductions in demand will hardly affect GDP, because current GDP is restricted by credit, not by demand. This is the same reason why fiscal policy is unlikely to have multiplier effects (see above).

Long-term impact: this action would continue to create a moral hazard problem in the future, as market participants would expect that the systemic risk in future crises would be borne by the Fed. On the other hand, current actions aimed at refinancing and recapitalizing banks have the same effect – but even more, because they are directly aimed at helping those institutions that created systemic risk. The latter creates a bailout arbitrage in which institutions have an additional incentive to magnify systemic risk.

A guarantee of a lower bound on stock prices is appealing in comparison with the original Paulson plan of buying troubled assets from banks, as the former exploits market mechanisms for injecting equity into the banking sector. There is no asymmetry distorting relative prices and the moral hazard problem is limited to systemic risk. Buying troubled assets, instead, preferentially rewards those institutions that caused the crisis by accumulating these assets.

Announced temporary increase in inflation targets:

Pros: -A temporary increase in inflation targets is credible under current circumstances.

If the downward spiral continues, an increase in inflation is already inevitable.

Higher inflation expectations lead to a shift in real return expectations from fixed-income assets to stocks and real estate and make real assets more attractive in one fell swoop.

A joint signal along these lines could serve to coordinate market participants’ portfolio realignments and, thereby, lead to a turnaround in the flow of capital. The turnaround would become self-propagating and continue until the negative exaggerations in stock markets have been corrected.

As a large fraction of Dollar-denominated debt is held abroad, the associated burden (inflation tax) will be borne by foreigners.

Cons-1: Consequences for monetary policy

A temporarily higher inflation target must be accompanied by a monetary policy that actually delivers the increases in price levels associated with the changed target. This would create the credibility that will be necessary for the subsequent period of disinflation after the cessation of the temporary target. Additionally, undercutting inflation expectations would lead to contractionary effects in the real economy that would necessarily have to be avoided as they would otherwise lead to a strengthening of the recession that is already expected.

In the long run, markets would have to accept and price the possibility of inflation temporarily exceeding its target. This possibility would have to be weighed against the probability of a systemic banking crisis (that will no longer be neglected in the future). This would lead to an inflation-risk premium on long-term debt obligations and would, above all, make financing government debt more costly. A higher inflation bias in the sense of Barro and Gordon (1983) is not likely though, as wage contracts generally have shorter maturities.

Cons-2: Political consequences

A temporarily higher inflation leads to a de facto one-time real devaluation of savings deposits. This places the burden of financing the financial crisis on those parties who were least responsible for its inception, including older people whose savings are to a larger extent invested in nominal bonds.

I cannot foretell what consequences this might have for politicians and decision makers. To me, it would not seem to be a popular way to solve the crisis.


I recommend policy 4:

In the event that the downward spiral in asset prices resumes, the Fed should agree to temporarily guarantee a clear lower bound for the S&P 500. It should arrange the implementation of this guarantee through targeted purchases of market portfolios. This action should be carried out by open-market operations and financed by injecting cash.

The portfolios thusly acquired can be sold for profit at a later date, reducing inflationary pressure in the long run, or to repay the government debt that has already been accumulated in various bailouts.

For a more elaborate descriptions of a downward spiral, see e.g. Adrian, T., and H.S. Shin (2008), Liquidity and Financial Cycles, BIS working paper No. 256.

Abreu, D. and M. Brunnermeier (2003), Bubbles and Crashes, Econometrica 71, 173-204.

Brunnermeier, M., and J. Morgan (2005), Clock Games: Theory and Experiments, working paper.

Svensson, L.E.O. (2003), Escaping from a Liquidity Trap and Deflation: The Foolproof Way and Others, Journal of Economic Perspectives 17, 145-166.

Barro, R.J., and D.B. Gordon (1983), A Positive Theory of Monetary policy in a Natural Rate Model, Journal of Political Economy 91, 589-610.