How should a central bank react when it observes that a potentially dangerous credit and asset price boom is under way? Can monetary policymakers defuse rising financial stability risks by ‘leaning against the wind’ and increasing interest rates? These questions have sparked considerable disagreement among economists. Proponents of policies that ‘lean into the wind’ argue that tight monetary policy can rein in roaring financial markets, lowering the risk and severity of financial crashes (Stein 2013, Adrian and Liang 2018). Critics of such policies counter by arguing that monetary policy is ineffective in lowering crisis risk and that the side-effects are potentially severe (Bernanke and Gertler 2001, Gilchrist and Leahy 2002, Svensson 2017).
The issue looms large for current thinking about monetary policy. It has become exceedingly clear how large the economic costs of financial crises are (Jordà et al. 2013). Crisis prevention may outweigh the costs of tighter monetary policy. Yet empirically we do not know much about the effects of monetary policy changes on financial stability during financial booms.
Our new research aims to close this gap (Schularick et al. 2020). We systematically study the available evidence for the state-dependent effects of discretionary monetary policy on financial stability based on the ‘near-universe’ of advanced economy financial cycles and crises since the 19th century. The state we condition on is a financial boom, defined as a large and sustained deviation of credit growth and real asset prices from trend. Conditional on being in such an (observable) ‘boom state’, we estimate how unexpected and exogenous policy rate hikes affect the financial crisis probability and severity.
The effects of discretionary leaning against the wind policy
To assess how monetary policy affects crisis risk over a five-year horizon, we estimate crisis probability models in which the outcome variable is a crisis dummy and the main explanatory variable of interest is the central bank’s policy rate. More specifically, our empirical analysis is based on a local projection instrumental variable (or LP-IV) strategy that has recently been introduced by Jordà et al. (2019). The instrumental variable exploits a type of monetary policy variation that is not itself influenced by local economic conditions, namely, policy rate changes in small open economies with fixed exchange rates that are induced by the base economy.
The following episode illustrates our identification strategy. In the early 1990s, Sweden witnessed a credit and house price boom. When the German Bundesbank surprised markets in December 1991 and raised its Lombard rate to 9.75% (in response to inflationary pressures following German reunification) under the prevailing fixed exchange rate regime, it forced the hand of the Swedish central bank too.
At the time, the New York Times (1991) quoted a market economist: “This is the Bundesbank’s way of showing they will use their power and independence without regard to the economic conditions in the rest of Europe.” The Riksbank had to defend the exchange rate of the Swedish Krona vis-à-vis the German Mark. Following the Bundesbank, the Riksbank also increased its policy rate at a time when credit and housing markets in Sweden were booming. This episode provides us with a quasi-experiment for an exogenous change in monetary conditions at a time when credit and housing markets in Sweden were in a financial boom.
We bring this identification strategy to bear on a long-run dataset that spans 150 years and covers most advanced economies, including dates of systemic financial crises. The dataset contains 1,525 country-year observations of countries whose currency is pegged to a base country’s currency. Among those, we observe more than 170 credit boom episodes, of which 98 coincide with exogenous increases in base-country policy rates.
It is important to note that our analysis focuses on ‘discretionary leaning against the wind’ policy (or D-LAW) – exogenous and unanticipated monetary policy actions. Our analysis does not speak to the effects of ‘systematic leaning again the wind policy’ (or S-LAW) – monetary policy that predictably reacts to financial booms in a rule-based way (Woodford 2012, Filardo and Rungcharoenkitkul 2016, Gourio et al. 2018). Currently, most central banks do not follow an explicit systematic leaning against the wind policy rule. This means that any policy change in that direction initially resembles a discretionary policy change until the commitment to the new policy regime has been credibly established (Svensson 2016). So, while our study speaks to the effects of discretionary and not to the effects of systematic policy, it can also inform the debate about the transition to systematic leaning against the wind policies.
Leaning against the wind is more likely to trigger crises than prevent them
Can discretionary interest rate hikes diffuse crisis risk? The upper left panel of Figure 1 describes the effect of a policy rate hike on crisis probability over the full sample. A one percentage point policy rate hike increases financial crisis risk by 3.6 percentage points on impact. The size of this effect is substantial, given that average annual crisis risk in our sample is 3.4%. After the initial increase, crisis risk remains elevated for one year, before subsiding to its long-run average level. In other words, leaning against the wind policies are more likely to increase crisis risk, rather than decrease it.
Inside financial boom episodes, crisis risk increases even more in response to a monetary tightening (see the remaining panels of Figure 1). Against the backdrop of booms in credit, house prices, and stock prices, a one percentage point rate-hike increases short-term crisis risk by about ten percentage points. The empirical evidence lends support to some of the worst fears about discretionary leaning against the wind policy – that it increases (rather than defuses) financial stability risks in the short run (Bernanke and Gertler 2000, Bernanke 2002).
Figure 1 Effect of a one percentage point rate hike on crisis risk (95% confidence intervals)
Effect on crisis severity
In spite of the crisis trigger effect, discretionary policy could still be beneficial if, by causing a small crisis now, it prevents a much bigger crisis later on. In other words, by hindering booms from proceeding unchecked, discretionary leaning against the wind policy might limit the fallout from the subsequent bust.
We investigate this hypothesis by looking at whether such policies reduce the real GDP loss associated with financial crises. To do this, we compare real GDP losses across financial crises that were preceded by different degrees of leaning prior to the start of the crisis, instrumenting the central bank’s pre-crisis monetary policy stance. We find that real GDP falls by around 8%, regardless of whether monetary policy took a ‘leaning stance’ prior to the crisis or not (see Figure 2). As a result, it appears that discretionary leaning against the wind policy fails to redeem itself by lowering crisis severity.
Figure 2 Leaning against the wind and crisis severity (95% confidence intervals)
Monetary policy and financial stability
Whether monetary policy should be applied to address financial stability risks is a long-standing question in macroeconomics. Our empirical evidence substantiates concerns that have been voiced by the opponents of leaning against the wind: a discretionary monetary policy tightening appears more likely to trigger financial crises rather than prevent them (Bernanke and Gertler 2000, Bernanke 2002).
Our findings also add a new perspective to the current debate about the costs and benefits of macroprudential and monetary policies. While monetary policy ‘gets into all the cracks’ (Stein 2013, Adrian and Liang 2018), the empirical evidence points to severe side effects of discretionary leaning against the wind policies.
Adrian, T and N Liang (2018), “Monetary policy, financial conditions, and financial stability”, International Journal of Central Banking 14(1): 73–131.
Bernanke, B and M Gertler (2000), “Monetary policy and asset price volatility”, NBER Working Paper 7559.
Bernanke, B and M Gertler (2001), “Should central banks respond to movements in asset prices?”, American Economic Review 91(2): 253–257.
Bernanke, B (2002), “Asset-Price ‘Bubbles’ and Monetary Policy”, Speech before the New York Chapter of the National Association for Business Economics, New York, NY.
Filardo, A J and P Rungcharoenkitkul (2016), “A quantitative case for leaning against the wind”, BIS Working Paper 594.
Gilchrist, S and J V Leahy (2002), “Monetary policy and asset prices”, Journal of Monetary Economics 49(1): 75–97.
Gourio, F, A K Kashyap and J W Sim (2018), “The trade-offs in leaning against the wind”, IMF Economic Review 66(1): 70–115.
Jordà, O, M Schularick and A M Taylor (2013), “When credit bites back”, Journal of Money, Credit and Banking 45: 3–28.
Jordà, O, M Schularick, and A M Taylor (2020), “The effects of quasi-random monetary experiments”, Journal of Monetary Economics 112: 22-40.
New York Times (1991), “Bundesbank increases rates to highest level since 1948”.
Schularick, M, L ter Steege and F Ward (2020), “Leaning against the wind and crisis risk”, CEPR Discussion Paper No. 14797
Stein, J C (2013), “Overheating in credit markets: origins, measurement, and policy responses”, Speech at the “Restoring Household Financial Stability after the Great Recession: Why Household Balance Sheets Matter” research symposium sponsored by the Federal Reserve Bank of St. Louis, St. Louis, Missouri.
Svensson, L E O (2016), “Cost-benefit analysis of leaning against the wind: Are costs larger also with less effective macroprudential policy?”, IMF Working Paper 16/3.
Svensson, L E O (2017), “Cost-benefit analysis of leaning against the wind”, Journal of Monetary Economics 90: 193–213.
Woodford, M (2012). “Inflation targeting and financial stability”, NBER Working Paper 17967.