While the Fed’s massive policy response to the Covid-19 shock was successful in reversing the financial meltdown, it did not prevent a dramatic collapse in the real economy. This column argues that the patterns observed are consistent with optimal monetary policy once the subtleties of the relationship between monetary policy, the stock market, and the economy are considered.
By Ricardo Caballero and Alp Simsek*
The blue line in Figure 1 shows the acute drop in the stock market index once the economic impact of the COVID-19 shock became apparent. The figure also shows the dramatic recovery of stock prices after the Fed announced its massive policy response (e.g. it cut the policy rate to zero and pledged close to 20% of GDP in asset purchase and credit support facilities). Other financial assets show a similar pattern. While the Fed was successful in reversing the financial meltdown, it did not prevent a dramatic collapse in the real economy. The red line (inverted scale) in Figure 1 shows that the US continued jobless claims rose substantially and remained high. This disconnect between the quick recovery of financial markets and the sluggish response of the real economy has been the source of much debate, as highlighted by the cover page of The Economist on 9 May 2020, “A dangerous gap: The markets v the real economy” (see also Igan et al. 2020). To explain this discrepancy, some observers point at what they see as irrational exuberance in financial markets, while others point at the ineffectiveness of the Fed’s policies in helping main street.
Figure 1 The disconnect between Wall Street and Main Street
Our main claim is that the patterns in Figure 1 are consistent with optimal monetary policy once we consider the subtleties of the relationship between monetary policy, the stock market, and the economy. First, the stock market is not representative of the average firm – let alone the average asset that matters for economic activity. In standard macroeconomic models, the Fed (and other central banks) can be thought of as stabilising risk-driven fluctuations in the price of the market portfolio (Caballero and Simsek 2020a). This portfolio includes houses and bonds in addition to the stocks of all firms – including those of non-traded small firms. While the stock market index can sometimes act as a proxy for the market portfolio, this is not the case for the COVID-19 recession. The major indices are overrepresented by sectors and companies which are not nearly as affected by the virus as the average firm, but which still benefit from the expansionary monetary policy aimed to contain the economic downturn.
Second, and more importantly, while asset prices have a large impact on economic activity, this impact takes time to build up. For example, Chodorow-Reich et al. (2020) find that stock price changes influence employment through a household wealth effect, but the response is sluggish and maximised at around two years after the price change. Carroll et al. (2011) find a similarly delayed response of consumption to housing and financial wealth changes. From a theoretical perspective, these delays are natural and result from a variety of frictions that range from fixed costs for durables consumption to habit formation and inattentiveness for broader categories of consumption. While natural, the delayed impact of asset prices on economic activity creates challenges for optimal monetary policy. Central banks not only must think about the current needs of the economy, but also about the needs in the future, by the time current policies have their maximum impact. In a recent speech, Chairman Powell (2020) emphasised the importance of transmission lags as follows: “Finally, we continue to believe that monetary policy must be forward looking, taking into account… the lags in monetary policy’s effect on the economy.”
In a recent paper (Caballero and Simsek 2020c), we analyse optimal monetary policy when asset prices affect aggregate demand with lags. We find that the dynamic aspects of optimal monetary policy give rise to potentially large temporary gaps between the performance of asset prices and the real economy. When applied to the COVID-19 recession, our results provide a rationale for the patterns in Figure 1.
In our model, output is determined by aggregate demand due to nominal rigidities. Aggregate demand depends on asset prices through a wealth effect on consumption (but the channel could be interpreted more broadly to include collateral constraints or the q-theory of investment, for example). We generate the lagged response of aggregate demand observed in the data by assuming that individual agents adjust their consumption infrequently. Monetary policy affects aggregate demand through its impact on asset prices; thus, the policy also affects consumption and aggregate demand with a lag. The central bank ‘sets’ the path of asset prices to minimise output gaps, taking transmission lags into account. For simplicity, we focus on conventional monetary policy: the central bank affects asset prices by adjusting the interest rate. However, our results extend to unconventional policy instruments such as LSAPs that enable the central bank to influence asset prices when the interest rate is constrained (e.g. Caballero and Simsek 2020b).
Sign up for the Eurasia Review newsletter. Click here to have Eurasia Review's newsletter delivered via RSS, as an email newsletter, via mobile or on your personal news page.
We first analyse a ‘recovery’ scenario in which potential output has just recovered from a large decline in productivity (e.g. an effective vaccine has just been discovered following the COVID-19 shock). However, aggregate demand, dragged down by a recent history of poor economic conditions, is initially below potential output. Over time, demand gradually catches up with potential output and closes the negative output gap. In this context, our main result shows that the equilibrium features asset price overshooting: asset prices are initially high (above their levels consistent with potential output) even though output is low. A central bank that dislikes output gaps boosts asset prices to close the output gap as fast as possible. This boost creates a temporary disconnect between financial markets and the real economy, but it also accelerates the recovery. In fact, the asset prices boost (and the disconnect with the real economy) is greatest at the beginning of the recovery phase, when the output gap is widest.
We then consider a (supply) ‘recession’ phase that precedes the ‘recovery’ phase. In the recession, potential output experiences a deep contraction but is expected to increase according to a Poisson event—in which case the economy transitions to the recovery (e.g. an effective vaccine to COVID-19 is anticipated but not yet discovered). The overshooting result also applies in the recession: if output is below its (already low) potential level, then the central bank boosts asset prices to close the gap. But the result goes beyond this, the equilibrium in the recession also features preemptive overshooting: the central bank boosts asset prices even if output is at its potential level. This preemptive overshooting is stronger when the central bank expects a faster transition to recovery. Intuitively, the central bank anticipates that the ‘recovery’ will start with a large negative output gap due to the inertial behavior of aggregate demand. Therefore, the central bank acts preemptively to boost asset prices and aggregate demand during the recession—in order to ensure that aggregate demand is not too depressed during the early stages of the recovery phase.
Figure 2 Equilibrium output and asset prices with optimal monetary policy
Figure 2 illustrates these results using a particular parameterisation of the model. The thin dotted lines in the top panel show the path of potential output, which starts low and is expected to recover. In this simulation, potential output recovers at time t=2. The solid blue lines show the paths of actual output and asset prices (in deviations from their potential levels in recovery). The equilibrium features preemptive overshooting: the central bank sets relatively high asset prices, which induces output in the recession to exceed its potential. This ensures the economy enters the recovery with a smaller output gap. After transition to recovery, asset prices increase even more, which helps to close the output gap. For comparison, the figure also illustrates what happens when the central bank does not overshoot asset prices (dashed-red lines), setting them equal to the level consistent with potential output in each corresponding state. In this case, output does not start to recover until productivity actually recovers. As a result, the economy enters the recovery with a greater aggregate demand gap and closes this gap more slowly.
To be clear, we are not arguing that this overshooting asset pricing policy and the inequalities that it creates are the best overall response to the a COVID-19 shock. Instead, we are arguing that it is the best a central bank can do, given a fiscal policy path (our model does not feature fiscal policy). Moreover, by speeding up the recovery of output, the gaps in Figure 1 arguably help most people – including individuals that do not invest in the stock market. Conversely, a counterfactual in which the monetary policy reduces asset prices to line up with the low levels of economic activity (e.g. by reversing the LSAPs or raising the interest rate) would slow down the recovery and hurt most people.
Finally, while we have demonstrated that the broad features of asset markets during the COVID-19 episode are consistent with a well-managed monetary policy framework, we
do not wish to imply that there are no anomalies or pockets of irrational exuberance in some markets. Having said this, the logic of the model suggests that experiencing an episode of irrational exuberance during a deep recession with a difficult recovery ahead has a positive dimension, since it reduces the burden on the central bank to engineer an overshooting.
*About the authors:
- Ricardo Caballero, Professor of Economics, MIT
- Alp Simsek, Rudi Dornbusch Career Development Associate Professor of Economics, MIT
Caballero, R J and A Simsek (2020a), “A risk-centric model of demand recessions and speculation”, Quarterly Journal of Economics 135(3).
Caballero, R J and A Simsek (2020b), “Asset prices and aggregate demand in a “Covid-19” shock: A model of endogenous risk intolerance and LSAPs”, NBER working paper No. 27044.
Caballero, R J and A Simsek (2020c), “Monetary policy and asset price overshooting: A rationale for the Wall/Main street disconnect”, NBER working paper No. 27712.
Carroll, C D, M Otsuka and J Slacalek (2011), “How large are housing and financial wealth effects? A new approach”, Journal of Money, Credit and Banking 43(1).
Chodorow-Reich, G, P Nenov and A Simsek (2020), “Stock Market Wealth and the Real Economy: A Local Labor Market Approach”, NBER working paper No. 25959.
Igan, D, D Kirti and S Martines-Peria (2020), “The Disconnect between Financial Markets and the Real Economy”, IMF Special Note Series on Covid-19.
Powell, J H (2020), “New Economic Challenges and the Fed’s Monetary Policy Review”, speech delivered at “Navigating the Decade Ahead: Implications for Monetary Policy,” a symposium sponsored by the Federal Reserve Bank of Kansas City, Jackson Hole, August.