Source: Bank for International Settlements
Thank you for inviting me to this event*. I am glad to be speaking here again.
This session is about the ECB’s toolkit for normal and crisis times. I hope you will excuse me if I don’t speak about the ECB specifically, but about central banking in general. Of course, some of the points I’ll be making are relevant to the ECB too.
After briefly retracing the extraordinary monetary journey since the Great Financial Crisis (GFC), I would like to focus on three issues: the lessons, the caveats and the challenges. The bottom line is that there are tough policy challenges ahead, and the answers remain elusive.
So, let me recap the monetary journey.
It is a sign of the extraordinary times we live in that the central bank tools for normal and crisis times are increasingly hard to distinguish. In the “old days”, the picture was quite simple. In normal times, central banks would steer the market overnight rate within a positive range. Liquidity management operations would work in the background. They would be designed purely to steer that rate, and carried no signal about the monetary policy stance. In crisis times, central banks would actively use their balance sheet in order to stabilise financial markets and institutions, typically through emergency liquidity assistance to financial institutions, essentially banks. One possible exception to this neat distinction, at least for some of them, most notably those in emerging market economies (EMEs), was FX intervention. This is a type of balance sheet policy in all but name.
Then came the GFC, which upended this simple world. In its wake, central banks started to actively deploy their balance sheet in order to spur aggregate demand, given the proximity of the effective lower bound. The balance sheet became a key tool to set the monetary policy stance. Hence the large-scale purchases of public sector and private sector securities and, in the euro area, public sector securities of different degrees of credit risk as well as special subsidised lending schemes for banks. In addition, central banks began to rely heavily on forward guidance, extending way into the future, as a quasi-commitment device. And some of them also pushed interest rates into negative territory. This was something historically unprecedented and would simply have been unthinkable until then. The cross-country differences that do exist do not invalidate this general picture.
The response to the Covid-19 crisis is yet another step along that path. Central banks have done more, in terms of both scope and amounts; hence the more direct support for firms of lower credit quality. And more central banks have done so; hence, for instance, the unprecedented large-scale purchases of government securities in EMEs. In the process, central banks have crossed a number of red lines, and they have done so with their eyes wide open: emergency times call for emergency measures. We have described and analysed this in detail in a chapter of our latest Annual Economic Report.1
Looking forward, if the post-GFC experience is anything to go by, it is not inconceivable that some of these tools will survive and become part of the normal toolkit.
Now for some lessons.
Arguably, unconventional monetary policies (UMPs) have been much more successful than generally expected. I still remember the debates over whether large-scale asset purchases of government securities would succeed in reducing bond yields to any significant degree. The Federal Reserve would try to find clues in previous experience with Operation Twist. Then came Ben Bernanke’s famous quip:
“QE – works in practice, but it doesn’t work in theory.” So now we know! The instruments can have a substantial impact on financial conditions, and financial conditions are the channel through which monetary policy influences economic activity.
During the Covid-19 crisis, we have seen just how powerful UMP instruments can be. Through shock-and-awe tactics, not only did their forceful and large-scale deployment stabilise markets, it also triggered a strong market rally. As a result, risky asset prices are now broadly in line with pre-crisis levels, having on occasion been even higher. Indeed, this has led observers to wonder whether asset valuations are disconnected from underlying economic reality.2
Looking further back, there is little question that central banks’ accommodative monetary policy stance has been instrumental in supporting the economic recovery post-GFC, in what proved to be extremely challenging conditions.
Now the caveats. The main caveat is that nothing is a panacea or comes for free.
Let me highlight a couple of points.
Point 1: There are grounds to believe that the tools have diminishing effectiveness. After all, and without going into details, there are limits to how far interest rates can be lowered and credit spreads compressed. In addition, the compression of banks’ interest margins can weaken banks’ lending capacity in the longer term (the so-called “reversal rate”) even if asset quality is temporarily boosted. Indeed, some work under way with colleagues finds evidence that, the lower interest rates are, the smaller the impact on economic activity. Moreover, the impact of the duration of low rates is also worth examining.
Point 2: There is a consensus that, while effective, the tools have limitations. Again, without going into details, there is a consensus on four issues. First, unusually easy financial conditions can spur excessive risk-taking. There is no doubt that some of the financial vulnerabilities (outside banks) that prevailed
pre-Covid-19 crisis, and that amplified its damage, were in part due to the unusually easy and prolonged accommodative conditions that prevailed post-GFC. Second, unusually easy financial conditions can sap the resilience of financial intermediaries – not just banks, but also insurance companies and pension funds. Third, those conditions may contribute to the misallocation of resources, essentially by softening budget constraints. And fourth, they raise tricky political economy questions, not least for the relationship between the central bank and the government. As we have discussed in our latest Annual Economic Report, the risk of fiscal dominance and loss of autonomy is material. The real debate is about the strength of these effects, about how long it may take for them to materialise, and how far they can be effectively addressed through other means (eg macroprudential measures).
The challenges follow from the caveats.
The wide-ranging and forceful measures necessary to contain the damage of the pandemic have further narrowed the room for policy manoeuvre. An economy with small safety margins is exposed and vulnerable. That’s why policies in non-economic areas explicitly build in those margins (transport, health, energy, etc). Arguably, the challenge of the decade ahead will be to rebuild policy buffers – prudential, fiscal and monetary.
To be absolutely clear: withdrawing policy accommodation – the first step in the process – is not for today, tomorrow or even the day after tomorrow. The economy will require support for quite some time. Moreover, there is a natural concern that even talking about withdrawal could reduce the effectiveness of the policies in place by sapping confidence. But at some point, as soon as conditions allow, disengagement will be called for. Starting the debate now can get markets, and economic agents generally, ready.
At that point, rebuilding policy buffers will be a major challenge. Let me briefly elaborate with reference to the case at hand: monetary policy.
In order to succeed in normalising, monetary policy will need to address two issues, economic and intellectual.
The economic issue is well known and fully appreciated. It is the limited responsiveness of inflation to monetary policy that has prevailed for so long.3 Especially since the GFC, many central banks, including those in the leading economies, have tried very hard to push inflation up to target, and they have failed.
Two factors underlie the difficulties central banks face.
First, inflation has proved very unresponsive to economic slack. In other words, the Phillips curve has proved to be very flat, and indeed very hard to estimate. That’s why, in its recent review, the Federal Reserve has downplayed the role of an unobservable equilibrium rate of unemployment in setting policy.
Second, there is an increasing recognition that inflation expectations are backward-looking. This is indeed one reason why central banks are very concerned when inflation remains persistently below target. The concern is that inflation expectations may become unanchored, as economic agents are convinced only by outcomes, not by announcements.
Looking ahead, the picture is unlikely to change significantly. Disinflationary pressures will probably prevail for quite some time. From a cyclical perspective, economies may well operate persistently below full capacity. Above all, from a secular perspective, some of the forces that have weakened the bargaining power of labour and the pricing power of firms are still with us: globalisation (albeit somewhat in retreat), technology (in full swing) and demography (very slow-moving).
The intellectual issue is possibly less well appreciated. The main element here is the prominence of the notion of the natural interest rate, or r*. This is the real interest rate that defines equilibrium in the goods market and that is generally regarded as fully independent of monetary policy.
This notion, in effect, puts central banks in a straightjacket. It implies that the only way to gain policy headroom is to raise inflation so that nominal interest rates can increase alongside it. In other words, central banks have no option but to cut rates (ease monetary policy) today if they want to raise them tomorrow. Thus, paradoxically perhaps, gaining policy headroom on a sustainable basis tomorrow requires lowering it today.4
This notion is especially powerful when coupled with the view that the long-term side effects of unusually and persistently easy monetary policy are not significant or can be effectively managed through other policies. In some respects, this view about the significance of the side effects is not surprising. The costs of failing to rebuild buffers are not highly visible – either ex ante, as they materialise only in the longer term, or ex post, as it will be hard to attribute the costs (eg financial vulnerabilities, notably higher debt, private and public, as well as lower growth) to previous monetary policy decisions. But these vulnerabilities do weaken the economy’s ability to withstand higher rates – a kind of “debt trap”. In the case of public debt, this can give rise to challenges for the central bank’s independence5 and credibility.
The implication is straightforward. With inflation rather unresponsive to monetary policy, the risk of depleting buffers is material. The post-GFC experience confirms this.
What does all this mean for policy? I would suggest that it points to the need for a broader view. We need to recognise the limits of monetary policy as well as the importance of flexibility in the framework, which would allow sufficient weight to be placed on the longer-term factors on which monetary policy has a significant influence. And we need to think of what other policies can do. Hence the need to ensure that, for these policies too, adequate buffers are in place.
This applies to both prudential and fiscal policies. Pre-existing buffers in both areas have been instrumental in enabling the necessary policy support in the response to the Covid-19 crisis. Strong bank capital and liquidity buffers have allowed supervisors to encourage banks to keep credit flowing, and those countries with higher fiscal headroom have been able to respond more forcefully.6
At some point, though, there will be a need to rebuild the buffers. This is true for banks, as the crisis transitions from the liquidity to the solvency phase; and it is true for fiscal policy, as the imperative is to ensure that it remains on a sustainable path, which is essential for financial, macroeconomic and price stability.
Last but not least, while policy buffers promote badly needed economic resilience, the key to more robust and sustainable growth is structural reforms. Unfortunately, after a brief phase post-GFC, they have lost momentum. The current crisis offers an unexpected opportunity to regain it.
To conclude: building policy buffers is essential – in monetary policy, just as in other areas. The challenge ahead is how. After all, if something is valuable, it must be worth paying a certain price for it.
* The views expressed are my own and not necessarily those of the BIS.
1 BIS, Annual Economic Report 2020, “A monetary lifeline: central banks’ crisis response”, Chapter II, June 2020.
2 See BIS Quarterly Review, “International banking and financial market developments”, September 2020.
3 For an analysis of the factors behind this development that places an emphasis on benign supply side factors such as globalisation and technology, see C Borio, “Through the Looking Glass”, OMFIF City Lecture, London, 22 September 2017.
4 For a critical empirical analysis that casts doubt on the prevailing notion that the natural interest rate is independent of monetary policy regimes, see C Borio, P Disyatat, M Juselius and P Rungcharoenkitkul, “Why so low for so long? A long-term view of real interest rates”, BIS Working Papers, no 685, December 2017. For a theoretical model that illustrates how this can be the case, see P Rungcharoenkitkul, C Borio and P Disyatat, “Monetary policy hysteresis and the financial cycle”, BIS Working Papers, no 817, October 2019.
5 For an analysis of central bank independence, past and future, see C Borio, “Central banking in challenging times”, SUERF Annual Lecture, Milan, 8 November 2019.
6 See BIS (2020), op cit; and C Borio, “The Covid-19 economic crisis: dangerously unique”, speech at the National Association for Business Economics, Perspectives on the Pandemic Webinar Series, 2 July 2020 (forthcoming in Business Economics).