 I'm old and I'm from the Netherlands. Ciao! My name is Antonio Marci. I am 29 years old and I am from Italy. Ni hao! My name is Indio Chi. I'm 32 years old and I'm from China. My work studies monetary policy in economies with domestic and international production chains. I'm interested in the interaction between monetary policy and financial stability. My research is on the role of information choice in macroeconomics and finance. I'm interested in finance and development. I do research in banking and copy finance. My work is about the policies of the European Central Bank which directly affect the risk premium on the sovereign debt of peripheral countries. I study how long-term investors change their bond haulings after a shift in regulation and how these changes subsequently affected interest rates. I found that savers choose to get more information about which bank or product they should use for their saving in recessions. My work explores how voter preferences determine financial regulation, focusing in particular on the role of political connections in this process. Welcome back to the ECB Forum on Central Banking. Forum participants, please do remember to cast your vote in our Young Economist competition. But now it's time for our first panel discussion which will be dedicated to issues related to the ongoing review of our monetary policy strategy. And it is my pleasure to introduce Philip Lane, member of the Executive Board of the ECB who will be joining us remotely today. So Philip, the virtual floor is yours. Thank you, Gary. Good afternoon, everyone. And as was just indicated, actually Cintra 2020 is I think an especially important edition of this series because it's right in the middle of our strategy review. So we talk this year, but in terms of the design of the topics that are being covered in the academic sessions and now in these panel format, it was important to try and cover a wide range of topics that relate to the strategy reviews. So in this panel, we've collected three topics which of course are interconnected. One is, if you like, structural forces. So this morning or this afternoon, I should say, we already have had a very important and interesting discussions on globalization and climate change. But of course, there are other structural forces such as a digitalization, automation, the ongoing structural, long running structural trends in the economy, for example, the trend, transition from manufacturing to services. And of course, in the background, we have the demographic transition, the aging of the population, globally, the kind of movements in the average rate of productivity growth. And of course, within the financial world, the rising demand for safe assets, which if you take those together, one way those structural forces will influence our world is through low equilibrium real rates. So these forces influence how the economy operates. It influences the transmission mechanism and monetary policy. And it is also, I think, an interesting question to debate about whether it also influences the optimal target for central banks. Does it change how we think about the anchor of our monetary policy in terms of the type of inflation aim we should have? Now, of course, in addition to that, whether it's for the effectiveness of monetary policy or more simply, for any public institution, a core value should be transparency and openness. Central bank communication is so important. So I'm very hopeful that we're going to have a very interesting panel. The format is we have three, I think outstanding monetary economists. We're going to take them in the order of Jordi Galli, Volker Rieland and Annette Viss in Newarkinson just from the composition of the grant they're going to cover. Each will have 10 minutes to make their opening contribution. After that, we'll go to the raised hands from the virtual floor and I'm looking forward to engaging Q&A interaction. So with that, let me hand over to Jordi Galli. Okay, good afternoon. Good afternoon. Thank you, Philippe, for your introduction and thanks to the ECD for inviting me to participate in this great event. So let me jump right into my presentation. I'm going to address the issues as to whether the ECB should revise its inflation objective. This audience needs a little background on the subject, but let me just mention that the inflation target of the ECB was set in 1998 to be a year-on-year increase in the HICP for the euro area below 2% and that was clarified in the year of 2003 to be below but close to 2%. Now, what was the justification at that time for setting an inflation target that was positive, given that the mandate that the ECB had was one of price stability and the justification similar to that of other central banks was, well, first, that there was some measurement and also that by setting a positive inflation target, the ECB would maintain a safety margin against the risks of deflation and there were two types of risks of deflation that the ECB was pointing to at that time. First, deflation in each member country given the likely structural inflation differentials within the euro area due to different growth rates of productivity and also risk of deflation in the euro area as a whole with the consequent risk of hitting the zero lower bound, the ZLB and hence making it harder for the ECB to stabilize inflation and output. Well, it turns out that the first risk of deflation has been hardly a problem in practice. There have been very few cases of negative inflation differentials on average over the life of the euro area. Only four countries have maintained a negative inflation differential. Germany, France, Finland and that negative inflation differential has been very small actually less than a minus 0.3 percentage point. Now, of course, the risk of deflation in the euro area and the risk of hitting the zero lower bound has been materialized as we all know. Now, at that time when that inflation target was set there were a number of studies conducted within the ECB and which are were published in a volume called background studies for the monetary policy review which concluded that the available evidence suggests that inflation objectives above 1% provide sufficient safety margins to ensure against the risk of hitting the zero lower bound. Obviously, that prediction was far to some gain and the assessment at that time of course with the benefit of hindsight may have been distorted by a number of factors. First, the recent experience may have cost a bias in that assessment because there had not been large deflationary shocks in the euro area for a long time and also at that time there was no evidence of no discussion of a decline in our star, the neutral real interest rate. In fact, in the background studies of the monetary policy review and in 2003, a baseline value for our star of 2% was used in a number of papers that made up those studies and as it was discussed that estimate of 2% was at the lower end of historical estimates. Of course, many things have happened since then and now there is a widespread agreement among academic economists as well as policy makers that the long run neutral rate of our star has experienced a substantial decline in recent years and current estimates for the euro area are systematic and many of them are actually negative. Now, the factors behind the decline in our star are well known. They are largely structural long term factors that are unrelated to, are not caused by monetary policy. They pertain to lower productivity growth aging of the population. Now the COVID crisis may only make the economy more steeper to the extent that it raises the possibility of recurrence of such pandemic shocks increases uncertainty and hence increases precautionary savings. Now, the main implication of the decline in our star is that given an unchanged strategy that there will be a higher incidence of hitting the zero lower bound or the effective lower bound and hence more frequent episodes of that sort. So what one would think was needed in response to that change in the environment was either an adjustment upward in the inflation target given an unchanged rule or a modification of the rule given an unchanged inflation target or of course an adjustment of both the rule and the target. I'll take some minutes to show you the findings of some working in progress that were conducted with Philippe Andrade and Julian Materon which is based on an estimated medium scale model for the Euro area and in which we conduct a welfare based analysis of the optimal inflation target given alternative monetary policy rules and under the constraint of an effective lower bound of 0.5% for the policy rate. Now the main focus of that work and the one I will concentrate on in here is the relationship between our star and the optimal inflation target conditional on alternative policy rules. So let me show you a picture that summarizes our main findings. What you see here is on the horizontal axis the assumed value for our star. And on the vertical axis the optimal inflation target delivered by the welfare based analysis based on the estimated model for the Euro area. The dash lines correspond to the historical estimates of our star and you can see a value of 2.6% and it turns out that the model delivers an optimal inflation target which is about 1.7% actually close to but below 2% in that case. But obviously this value of 2.6% for our star is clearly obsolete according to current estimates. So we can see that if our star is below 2% the optimal inflation target would immediately rises above 2%. For a value of our star close to 1% the inflation target the optimal inflation target is closer to 3%. So it calls for a large adjustment upward in the inflation target such a decline in our star. Now of course there are ways of avoiding the need for increasing the inflation target and those are ways to take the form of that would involve changes in the monetary policy rule that would shift the schedule that you see on the screen to the left. That would make it possible to maintain 2% or a value close to 2% as an optimal inflation target consistent with or coexisting with a much lower our star. So the blue line corresponds to our baseline simulation which assumes a conventional Taylor type rule. One is average inflation target targeting a version of which has been adopted by the Federal Reserve recently. And here we consider two possibilities one more aggressive than the other. We assume a Taylor type rule but in the yellow line here we assume a Taylor rule that responds not to current inflation but to average inflation over the past four years and in the Turquoise line we assume a rule that responds to average inflation over the past eight years. So you can see that shifts the eight year the rule with an eight year memory actually makes it possible to maintain an inflation target close to 2% as an optimal target even for an our star of 1%. Here we have an alternative deviation from our baseline scenario. Here we allow the fiscal authorities to respond with an emergency fiscal package and the cumulative output gap reaches 6% and the fiscal response is 4% of GDP and you can see that this achieves this is a very aggressive counter-cyclical fiscal policy that achieves a similar objective to reconcile a 2% inflation target with a much lower our star. Now let me just briefly discuss these three options counter-cyclical fiscal policy of course it's beyond the ECB's control and it would involve the need to allow for escape clauses in this stability and growth pack to operate effectively and most importantly I think this would require is to prevent at all costs the need or the imposition of large fiscal consolidation programs in times of recessions the way in contrast with what we observed during the financial crisis. Of course we know that the fiscal policies beyond the ECB's control but of course the ECB's views are important on this matter and during the financial crisis one has to say that there was maybe too much cheerleading for an aggressive fiscal consolidation at the time when this could hurt the economies very much. Average inflation targeting is easy adoption and it can only help in the short run because if anything it may raise expectations and by doing so it may raise inflation itself I view it just as a formalization of forward guidance now an important caveat is that it's benefits hinge critically on the workings of anticipation effects and on the ability of the ECB in this case to steer inflation with high precision to the desired levels now given the flattening of the Phillips curve now have some doubts about that ability also deviations from rational expectations may also call into question the workings of those anticipated anticipation effects. Higher inflation target on the other hand is robust deviations from rational expectations and the only caveat that one can see is how that transition from the current under shooting would take place and from that point of view what I would suggest to follow an approach that emphasizes gradualism and opportunism gradualism in the sense that the ECB could now announce that sometime down the road in the future it may consider adjusting the inflation target so that it is clear that the current inflation target is not set in stone forever and also the principle of opportunism which calls for implementing that new inflation target at the same time when inflation has reached or is very close to the new inflation target not at the time when its inflation is much lower than the target as it is now now the three margins are not mutually exclusive and one may want to exploit them simultaneously and that may be the safest way to go so just to conclude my presentation let me say as a final remark a monetary policy framework is a framework that is built on some assumptions necessarily it's not out of the blue it's not built in or designed in a vacuum now when these assumptions are revised the framework must be adjusted there is one key assumption that was made when the current monetary policy strategy was designed that has been revised and that's the level of our star okay so that calls for a change in the framework otherwise one must recognize the consequences of not adjusting it in this case would be recognizing and making it clear that the incidence of zero lower bound episodes with its consequences on deflationary consequences and consequences on output instability will increase in the future now making the current inflation target is clearly not enough for that purpose it's obviously a correction of what too many of us seems like just a minor anomaly but it's not enough and a change in the target or in the rule seems warrant I'll leave it at this thank you so let's move straight across to Volker so over to you Volker thank you it's a pleasure to contribute to this session my discussion starts from a quote by President Lagarde where recently she said or she explained that we've observed a persistent decline in inflation in the Euro area annual inflation averaged about 2.3% from 99 to 2008 but only 1.2% from then to 2019 and that we should analyze the forces that are driving inflation dynamics and how to adjust policy strategy so that's what I'll be addressing directly looking at averages of inflation I think it's interesting to split the period since 2009 in two parts namely the double recession till 2013 you know the Euro debt crisis followed the global financial crisis and the recovery and what you observe is actually that HICP inflation remained high so in the first period close to 2% 1.8 and then it went down to 0.9 during the recovery which is a bit odd because you think you know recovery inflation starts rising if you look instead at prices of all goods and services produced in the Euro area so that's the GDP deflator it makes no difference before the financial crisis about 2% but it went down during the double recession period to 1% and it went up on average in the recovery to 1.3% at least something and where is that coming from if you look at import prices well 1.6% before the financial crisis in the double recession period 2.1% in minus 0.3% during the recovery so the CPI that the ECB is looking at was held down by import prices and the GDP deflator you see some effect of the ECB policy and the recovery now going forward here is the time series you can actually see that for the GDP deflator so the domestic price inflation it's very similar pre-financial crisis but since then it was low in the double recession and rising back to the target it's very similar to what we designed for the HICP in 2019 so domestic price inflation at 1.8% in 2019 at least something worth looking at I think and to consider in the strategy there are other measures of inflation which are not fully covered in the HICP if you think of households a big part in the HICP it's very similar to what you're seeing on your home and of course there are rental costs included in the HICP here you see the orange line it's pretty flat in those periods here it's the HICP excluding energy right where it's low rental costs held up but they are underrepresented in the HICP because it's not covered or it's not even a rental equivalency is not included if you just look recently at owner occupied housing indices for example in Germany this rose between 3-5% in the last 3-4 years so that's excluded but it definitely affects people living there of course partly it's an asset price the land is gaining also construction costs so it's just becoming more expensive it's more variable of course in France it's all different but these are issues to look at and issues which give thought that maybe inflation is higher than measured by the current HICP for technical reasons it may be difficult to include in the HICP directly but it's something in the strategy the ECB could start looking at I'll come back to that looking forward I think it's of interest you talked about climate policies a while ago today looking forward this climate policies will also affect prices will also affect the HICP here's a study from staff of the German Council of Economic Experts which tells you the effect the new CO2 pricing for heating and mobility will have in Germany probably more than 1% on the HICP in 2021 and then quite a bit also in subsequent years and about one third of that will be on the Euro area measure of course that's not just inflation that's actually probably a cost push out so it's a new dimension I can't touch on here but definitely issues to look at in terms of the strategy now coming to the policy dimension I think it's important to take into account the actual policy instruments used I'm not sure that was fully into Jordi's analysis but what happens is when we hit the effective lower bound policy switches from industry or brings in in addition quantitative easing of course the quantitative easing effects are uncertain if they're not uncertain if quantitative easing would be a perfect instrument then you can hit the inflation target independent of what happens to our star but it's not perfect it's uncertain here's a simple linear process if you look at inflation here this is the affected by industry deviations from the equilibrium nominal rate the I star that involves the high star of the inflation target and the R star Jordi talked about but also inflation can be affected by quantitative easing here this parameter B measures the impact and that may be quite uncertain so over here there's some parameter uncertainty and the sigma B the variance accounts for that uncertainty about the effectiveness of quantitative easing and of course where we are in terms of inflation inflation expectations also but also where we are so lack inflation affects where we're going so let's look at the simple process let's look at a policy which tries to hit the target so minimize expected square deviations of inflation from target that has two components you want to get the expectation of inflation on target that's what Jordi talked about but also the variance of inflation is involved so the variance of inflation is influenced when you're uncertain about the effects of your policies we may not be so uncertain about the effect of interest rates but definitely in terms of quantitative easing and then as long as you don't hit the expected lower bound which is not zero by the way if you don't hit it so if the interest rate remains above this I L B here basically here you can implement this optimal feedback rule I made this very simple you can make it dynamic you can include expectations here it's very simple and basically the normal interest rate deviates from the equilibrium rate right here it's our star plus pi star whenever inflation is away from the target you lower when inflation is below you raise when inflation is above all this time under this policy here the quantitative easing stays at zero you don't use the quantitative policy and as a result of the optimal policy expected inflation is equal to the target now when you're hit by a shock or inflation is very low right then what happened okay then we're here at the effective lower bound so if inflation here is low enough like what you can see here can be defined numerically the interest rate will be at the lower bound and policy can switch and we've observed that in many jurisdictions to quantitative easing so there is a new feedback rule you see it's the same element your feedback you respond to inflation deviations from target accounting for the effect you get from the interest rate being at the lower bound but there's a difference if you look at the feedback parameter it's different it's not one over B it doesn't nail it exactly there is an attenuation factor because you're uncertain about the impact of the quantitative easing you respond by less right this is positive this is positive so the basically the coefficient is smaller as a result actually expected inflation is below target so the message you get is at the effective lower bound it may actually be optimal to have inflation converge more slowly to the target from below why because of uncertainty about the effects of policy about the quantitative policy which I'm not sure was included in Jordy's analysis there is a second reason why you may choose that route that's because there may be risks of side effects of quantitative easing people have worried that when the settle banks ramp up their asset purchases that this creates risks for the fragility of asset prices or they creates risks on the banking in the banking system on the in terms of interest rate risks or fiscal dominance risks whatever very simple way here to include it this is my side effects variable q here the quantity of easing c is the potential risk of side effects so I'm taking the position no side effects are expected but so the parameter is zero but there's some uncertainty I could be different from zero what happens then well if you include the side effects in the policy you get a second interesting tradeoff namely in response here to putting some weight on the risk on on the side effects and what comes out is that the reaction function for quantitative easing changes it looks the same but here the reaction coefficient is different right there's a new element there's the the sigma c the potential risk of side effects in there so that reinforces the tradeoff and basically is another reason why it may be optimal for you for you to converge slowly to the target from below now concluding in the simple framework I've included the constraint on interest rate policy the effective lower bound but I've also included a monetary policy response quantitative easing which has been used extensively and then let's look at now the choices the other strategic choices how they affect the effective lower bound now there's a plus so it means there's a non negativity constraint but it doesn't mean that the zero low it's a zero low about right there's an ILB which measures how far below you can go now as Jody rightly pointed out if our star now I've broken down the I star in the equilibrium real rate and inflation target right if the R star goes down interest rate policy is going to be more constrained that's certainly correct now in terms of a tie star it's a little tricky I mean Jody is absolutely right on average or if you start from steady state it gives you more room but when you are already at effective lower bound as we are currently and you announce a higher inflation target basically you also widen the distance you have to cover and this requires immediately even easier policy so it's a little tricky at this point only if you have an immediate massive impact on inflation expectations you gain more room if you're a little more taking into account that inflation is a sticky process it's actually a tricky issue on the other hand something the ECB has done in contrast to the fit exploring where actually the lower boundaries on the short rate or medium rate right the ECB is now implementing TLTROs with minus one percent that makes policy less constrained and it works one for one to the R star so yes the R star may have come down that's still a bit more uncertain I didn't mention that research but it's a bit more uncertain than it sounded but yes assuming it came down we also got more room than we thought when we did the studies in 2003 and finally if you look at inflation measurement I think it pays off not to focus only on the HICP because if you think of inflation being closer to the target of course that also effectively means either giving more room or effectively means that policy is having a bigger impact concluding what did I want to say well in terms of strategy I think it makes sense to consider inflation measures and more inflation measures more broadly in policy communication not just the HICP and actually it doesn't need a massive change in the strategy currently there is a objective of below but close to two percent regarding the HICP and that that could be interpreted as a range for example 1.5 to 2 percent something used by others but a range always gives you flexibility and in this case it gives you flexibility in policy communication to also point out what's happening with other inflation measures finally just to summarize at the effective lower bound it may be optimal to have inflation return to target more slowly from below due to uncertainty of the effects of monetary policy quantitative easing and potential side effects raising the inflation target yes that can help but it's actually when you're right at the effective lower bound you may want to explore other options other instruments just as quantitative easing maybe other instruments and actually the lower bound was not as binding as we thought back in 2003 so that's an issue to consider by the way it's also something in terms of side effects of QE a lot of banks affect their profits particularly in Germany I always tell them look you make profit normally on a boring short taking deposit at short rates in landing out long you would more easily make profits if you have a steeper term structure so don't complain so much about the negative short rate but I'll leave that for discussion so that's my point thank you for giving me this time thank you Volker let me just turn over to Annette just make the point for those of you who are thinking of questions or who want to offer comments raising your hand in a virtual format doesn't cost you anything it's not going to cost a lot of energy to you to hand up so I'd encourage you because it makes it easier for us to form the queue of people who want to make a contribution the quicker and earlier you raise your hand out the better okay with that Lazo let me turn over to Annette so over to you Annette thank you very much for the opportunity to speak about central bank communication today I'm going to try to convince you that informal communication including what I'll call unattributed comments play a central role by an unattributed comments I mean comments that people don't put their name on sometimes refer to as off the record or on background or not for attribution I'll show you some facts and I'll talk you through the benefits and costs of using this kind of communication as an institution or as an individual policy maker and then I'll finish with some suggestions to reduce the use of unattributed communication including a discussion of recent ECB efforts so here are the three facts the facts argue that information flows from the Federal Reserve to the stock market at unexpected times whereby unexpected I mean times where there's no formal fed communication of releases or speeches the first fact is from Luca and Monk they used data post 1994 to document that the average US stock were turning the 24 hour period from 2pm to 2pm before and not including the announcement was about 50 basis points they view that as a puzzle since money type policy news coming out had to be systematically positive and this is part of the blackout period and the view leaks as being unrealistic from an institutional viewpoint the second fact is from just like Morrison visiting also in post 94 data there we show that looking over the whole cycle in between FOMC announcements that stock returns over bills are much higher on days that fall in even weeks relative to the announcement you can see that in the left graph here which graphs the average 5 day excess return on stocks over bills against the number of days from the FOMC meeting so not only as we know from Luca and Monk is there high returns right around day zero or before the announcement but there's an additional three quite strange looking peaks in these average realized returns now in the paper we argue that this pattern actually is due to money type policy news a key argument for that is that prior to 1994 ints meeting target changes were common there were about two thirds of the changes timing when the Fed does its thinking and its decisions and you can see in the graphs to the right that using now the pre 94 period the target changes are more common in these same even weeks in FOMC cycle time so the graph shows the 5 day probability of a target change graphed against the number of days since the FOMC meeting now we argue also that over this post 94 period money type policy news actually was systematically for example Fed funds futures yields fell on average in these even weeks and even week stock returns were particularly high after the low prior returns so that would be consistent with the surprisingly strong Fed put we link the even week timing back to the timing of the frequency of discount rate requests from the reserve banks those are ways for the reserve bank to affect the policy making they have to be made at least once every two weeks which would make a bi-weekly decision or discussion frequency meaningful the final fact I want to talk through is from a new paper by more than missing audience which is based on a freedom of information request for the calendars for calendars of Fed governors based on six calendars and data since 2007 we categorize as many as I think 30,000 calendar items into many different categories and it turns out that the interactions between the governors and the reserve bank presidents are what's tightly linked to information flow to markets so we document that these even week returns are particularly high on days where the governors and the presidents interact either at the FMC events or in private phone calls or meetings and that's even more so if there's informal communication either via on the record public commentary using FMC speak data or via on background media interviews for which we obtain data from the governor's calendars so that rules in the role of informal communication and as I said before all three factor unaffected by controlling for form of that communication so I want to focus the rest of my comments on attributed communication since a lot has already been said about the public cacophony so a bit more evidence for this kind of communication you may remember actually the FMC statement emerged in 1994 after congressional pressure for transparency after a series of newspaper stories revealing that information furthermore if you read from a governor Myers book he notes that the use of reporters as part of the fed's signal core is not official border FMC doctrine the public affairs staff and the chairman like to pretend it doesn't happen I put a quote here from a top fed reporter Greg Gibb saying you know I don't want to get into the background interviews with the fact finally I have in another paper documented many discussions of leaks in the FMC documents or in FMC transcripts although I'm going to argue in a second that these are not leaks many of them are not leaks in the traditional sense of the word all right so how does money say policy communication normally work is it in this unattributed way well no for policy effectiveness in the sense that the impact of forward guidance and QE and longer to demonstrate depends on the public understanding that these policies are going to be in effect for some time so in that framework communication should be public and on the record so why might central banks or central bankers use unattributed communication it could either be from the perspective of the institution or it could be from the perspective of an individual so let me talk through each of those in turn so starting with the institution use of unattributed communication I think we can learn from political science what the benefit of this might be so post in 2013 documents that many White House leaks are not leaks in the traditional sense revealing on our information they're more authorized plants of information in media outlets these plants allow the sender to quote impart information about executive branch policies without officially acknowledging those policies and thereby inviting unwanted forms of accountability or constraints so let me map that to central banking so in our world communication to some extent ties policy makers hands if the public doesn't fully understand that policies stay contingent perhaps in that setting unattributed comments retain a bit more flexibility than public communication or of course less than no communication at all furthermore explaining policies without distribution or background time-consuming public debate unattributed communication may be a convenient way to get support outside the central bank for a particular policy change but there are also costs most importantly I think loss of credibility unattributed communications is clearly the opposite of transparency and accountability today we are part of an ECB listening event there are other listening events before that let's ask ourselves what the engagement citizens would think if we told them that there's a lot of use of this kind of communication and that it has big effective massive prices I suspect they would be a bit worried about their pension fund managers doing well in that information game with others a second cost is that once the institution uses unattributed communication that facilitates use by individual policy makers in the sense that when there's less formal guidance about a sense that leaves a bit more room for the individuals to try to move market expectations now as you can see there's pros and cons of using this kind of communication for institutional purposes I'm going to argue though that it's used for individual policy maker purposes is clearly welfare reducing so let's turn to that so as we all know there's lots of disagreements between individual policy makers and I think in that setting communication is not only an institution but also individual so it becomes about driving market public expectations in order to improve your marketing position in policy negotiations there's different ways one could try to do that posturing is one approach that just means making firm statements of what policy you prefer to make opponents think that you're not going to budge in negotiations that's probably best done through on-the-record communications and speech interviews and then influencing so try to convince the public that you have some good arguments or perhaps even spin which is a little sneaky and involves distorting the public's assessment of what the likely policy decision is now critically if the information used for influencing a spin is confidential then it has to be done using on attributive communication we can think in our context as central bankers of staff protections internally liberations and so forth that are often confidential until the decision has been made so let's just do a two slide version of how this game works out between the different policy makers I'm going to call this the quiet cacophony since it's done on attribute so think of two policy makers who are choosing what to reveal to the public about their policy preferences at an intermediate date in between policy decisions they care about not being viewed as flip-flopping in the sense that both will incur a loss if somebody is communicating and the chosen policy rates deviates from market expectations formed based on any disclosure so if nothing is revealed if no one reveals information ahead of the meeting well then they can set the policy rate equal to the average preferred policy rate across the policy makers but if someone communicates now the hands are partly tied and the chosen policy rate is going to be a weighted average of what they prefer and what they have made the market expect policy makers can within limits of course spin market expectations of policy preferences by selectively revealing information okay so think here of the hawks would want to reveal the hawkish internal information and commercially for the dogs one can then show quite easily that if there's enough disagreement relative to how much news might arrive before the policy meeting and sufficient spin is possible well then there's a unique national equilibrium where each policy makers would spin in order to improve their bargaining position therefore the spin cancels assuming there are equally many hawkish and dovish arguments to let out there but each awards off from a lost policy flexibility relative to non-disclosure you'll see that this is exactly the same as the prisoner's dilemma where both prisoners would be better if neither confessed to get a reduced sentence but both confessed in equilibrium now in the in the game I'm assuming the spin is entirely one could think of cases where there's only information going one way in that case the conclusion remain in the sense that each side will gain as often as they lose and they still have a loss from flexibility I put here two recent disclosures of ECB staff projections to illustrate to the left adorvision to the right of hawkish league that obviously excess ECB projections to show future growth barely above 1% thus underpinning the ECB's plan to approve more stimulus and the more recently to the right is a hawkish one now there are a plus to an attributed individual communication than lost flexibility it damages again the credibility of the central bank I put a quote from Bernanke stating that and also important there's harm to the decision making process both in terms of collegiality but also importantly information withholding in order to prevent leagues avoid as article in 2014 noted that several ECB sources said that Draghi had cut back on circulating policy papers papers in advance of council meetings apparently out of concern that opponents notably the Bundesbank were leaking to try to block this credit decision so let me finish up with a couple of simple suggestions so first I think it's important to resist communicating through these very expensive fat watcher ECB watcher newsletters whether for institutional or individual reasons of course it's easy to talk to experts and they can move market expectation quickly but so can the best financial newspapers thinking back to the medley scandal at the fat bag in 2012 this is very harmful to the fat's reputation because it be enforced perceptions of unequal access to information after all the newsletter is 120,000 per year and central bank independence of course especially in these times is not something that we can take for granted my second and final suggestion will be to stop this prisoner's dilemma of disagreement driven on attributed communication by seeking consensus so that's a very nice Reuters article from earlier this year describing how president Lagarde is doing precisely that putting a great emphasis on consensus getting more voice to critics more time listening more time for deliberating in return asking for discipline or to keep internal disputes out of the media this maps very nicely into the standard solution for the prisoners dilemma which is to consider repeated game of the version with quote unquote punishment and the consensus building fits very nicely into this in that now the president has something to remove namely influence if discipline doesn't follow of course it's hard to identify who's leaking but I think with an improved culture leaking is less acceptable among colleagues who may know the identity of a potential or actual leaker and therefore colleagues can help informally discipline those who may consider leaking and I put a quote from again political science that while leaking is leakers are rarely disciplined in public they are often disciplined in form I'll finish it there thank you very much very good thank you on this so the way we're going to run this panel is we're going to go straight to Q&A I'm going to take a few questions together before returning to the panel members so I'm very pleased to see on the first list in front of me and my former colleague Vitor Constancio so first of all over to you Vitor thank you Philip it's a pleasure to be able to participate my point is about the vocal's proposal to have more than one price index to assess the objectives and the performance of monetary policy to those indexes you mentioned one could have of course the index for core inflation which has been most of the time well well below any reasonable target but the point is that this would be detrimental to the anchoring of inflation expectations and one of the objectives of going back on the review of the objectives and the strategy in general for major central banks as we have seen with the Fed and now with the ECB is precisely the point that the underperformance in reaching the target over so many years is affecting inflation expectations anchoring and that there is the risk there is an unhankering of inflation around the target and that expectations could be anchored at the much lower level than what is desirable and that of course will make it much more difficult for policy to normalize towards the target the inflation level the other point is that precisely because of that the decision that the Fed made recently to go for flexible average inflation targeting is very well justified because is an enhancement and new commitment of the central bank that will aim at inflation during a number of years above target in order to offset the previous underperformance and this is more than just being symmetric is really adopting a policy towards that averaging inflation targeting which was very nicely presented by Jordy and that showed that in my view central bank should go for the longer period of averaging without fixing the number of years my question to all members of the panel is the following what is your assessment in terms of the effect in anchoring inflation expectations the proposal by Volcker to to use several price indexes both to define and assess the objective of monetary policy thank you thank you next yes thank you I have a question that would relate to all three panelists and it regards the form that the inflation target should actually take there seems to have been some disagreement and there's also some sort of let's say fuzziness in the way ECB has up to now defined target with a price that is the same between 0 and 2 and the target or objective somewhat close to but below 2 but not precisely defined now Jordy seems to have advocated if I understood him correctly a single numerical value and Volcker seems to have favored a zone of indifference based on the argument of flexibility and I was just wondering how the panelists would sort of learn concerns namely that for one a zone would be somewhat harder to communicate and it would also given the results of Annette leave room for individual communications and disagreements in the background to that about where exactly in the zone we should be anchored and it is also probably less good at pinning down inflation expectations and at detecting potential deviations from the pinning of inflation expectations so I would be curious to hear what the panelists think about what form the inflation target should optimally take thank you thank you class so again in gathering this round of questions let me turn to my colleague Ignacio Visco yes thank you thank you very much thank you for the insights which will be very helpful also in our strategy revision and I will not say much on that but I would like only to make a similar point to what put forward by Vitor hello we can hear you Vitor Ignacio do you hear me do you hear me now hello do you hear me Philip sure absolutely perfect no because I got a message which was confusing thank you very much but what I was saying is that similar to what Vitor has said Valker's proposal or the revision that perhaps the convergence to the inflation target should be slower when we are defective lower bound misses the point that if this is so then there will be an effect on inflation expectation that is a procrastination of going back to the target may at the end feedback into maintaining inflation expectations de-anchored and therefore making the convergence more difficult rather than possible the other point is that also in Valker derivation is that there is in your point an uncertainty about the effects of QE on inflation but the issue here is also to take into account the effects that QE has on the macroeconomy at large if there is no prejudice to price stability we have to favor the improvements of the macroeconomy the reduction of financial instability and so on therefore that should be taken into account in measuring the effects of QE finally I share instead the point that if we are far away from the target then Geordi proposal to raise perhaps the inflation target would have some effects and credibility that is the possibility of reaching the target might be considered to be more difficult and again these might feedback in the reaction of the economy indeed I think that perhaps in Geordi framework the inflation target could very much be path dependent rather than being fixed thank you very much Thank you Ignacio, next Alan Blinder Okay am I unmuted now? Yes. Okay good well thank you very much I just in two seconds those were three really fascinating presentations and thank you all but I have a particular question for Geordi about average inflation target so in the economic literature what is now called average inflation targeting is a close cousin if not almost a twin brother to price level targeting and the idea is to tighten up the grip so to speak of the central bank my screen just went funny am I still on Phillip? Yeah we can hear you Alan please Okay to tighten up the grip on inflationary expectations and indeed and especially to drive them up when they're too low and that is preventing the central bank from getting up to its target so my question has to do with flexibility which could also be called vagueness the Federal Reserve adopted this idea sort of essentially but in an extremely vague way so that nobody knows over the what period inflation is supposed to average 2% how quickly or slowly the FOMC intends to approach those higher levels when necessary and so on and some of us obviously not the consensus on the FOMC some of us think that vagueness makes this really undermines the basic idea of average inflation targeting so that's my question for Jordan Okay thank you I think it makes sense given the time we're going to run over a little bit but I have two more on my list so I'll take these two last questions and then I'll return to the panelists and you can basically decide how you want to respond so next I have Zeal Malek Hello Philip Can you hear me? Yep Thanks a lot to the three panelists I have a question actually to Falker and his point on inflation measurement and his focus on introducing or raising the weight of house prices in particular in the measurement would you agree that it would be also a case to look at another source of divergence between inflation perceptions and measured inflation for instance you could consider that the hedonic correction to prices create the divergence between measured inflation and the way it's perceived by households why shouldn't we look at that obviously it would take it in the other direction it would actually raise the prospects of begin deflation if Falker would agree with looking at it in this symmetric way rather than just focusing on something that would make the inflation measure go up Thank you Thank you for that The last question before we have to close is Harold Ulick Harold over to you Thank you I wonder whether the time has come to go to a price level price path targeting it's an education tool it's the further development from inflation targeting it's a bit like inflation averaging but maybe a bit more precise one could announce a price path one could also announce how speedy one wants to get to that I mean that would apply if an undershoot sometimes one overshoot some other time when Egerton Woodford proposed this in the Brookings paper in 2003 as a powerful tool to address issues if it's yellow bound people thought at the time that inflation targeting is simply easier to communicate but now we have gotten forward guidance and all the rest of it seems to me that that would be the logic the next step to take I wonder about that and it's a question mostly to Geordi and to Falker Thank you Harold Now I'm going to come back to the panelists but I suppose in the tradition of running panels I'm going to reverse the order so first of all Annette whether Annie the Crest since you got there some of your fellow panelists you would like to add to Yeah I want to agree with both Klaas Adam and Alan Blinder that any sort of vagueness in what the central bank's mandate is of what the specific goals are is going to result in just infighting a disagreement and complicated communication so I would agree that we should pick something a little bit easier to communicate for those reasons Okay Very good, thank you Falker Thank you for the questions, that's great The mandate is price stability inflation so it's the ECB's choice what measure to pick and the question correctly pointed out by President Lagarde is why has HICP inflation been so low and that question has to be explained what are the forces behind it and as you saw from the comparison between domestic price inflation which has been rising and the HICP that it must have to do with import prices so I think Vitor Constancio I think we have to look at that we want to have an answer to what's going on and not just say let's just look at one thing and no matter what happens the other thing which is good about that is because it actually gives some hope that the policy easing by the ECB which has heavily relied on negative interest rates and quantitative easing has had at least some effect domestically and by the way domestically are a lot of rigid prices sticky prices which New Keynesian theory tells us that's why we need inflation targeting or price stability so I think it makes sense in a strategy review to look at broader measures to Klaus I did not necessarily or also to Vitor and Klaus I did not say that we need to switch the inflation measure I just pointed out or I didn't mean to say that we need to switch to a larger range I just pointed out that the current system gives you some flexibility and that should not be totally disregarded or kicked out at least we should note that it has some advantages back with Orfanides I did a paper on optimal inflation zone targeting there can be some good reasons also related to the trade-offs in the Phillips Curve but I just wanted to point out this new trade-off that you can actually better communicate what else is going on with inflation by other measures so that's that in terms of inflation expectations well inflation expectations are what they are people will do their best to figure out what the measure is going to be that they're looking at if they think that they have to forecast the HICP and this is important for some reason because in some contracts they will forecast the HICP and if the HICP is actually driven by import prices in your area in the recent years they will forecast that the HICP will be lower so this is not necessarily a signal that policy has been ineffective and so it can only help if you do the best you can to explain all the sources and effects going on in terms of average inflation targeting price path targeting yes I think price path and price level targeting I think in terms of models I've looked at that also in 2003 this can work quite well but it requires a lot of credibility and the problem with price level targeting is that it may be fine now when the question is to basically keep policy rates lower longer or keep the easing more aggressively but at some point we might end up above the price path and then it will be hard to communicate to raise rates to keep on raising them or to come back to the path so that's the problem but I think it's a very interesting strategy to discuss in terms of the question on perceptions I agree we need to look at household perceptions and my understanding is there's quite a bit of research that they somehow perceive much higher inflation so there is a lot of explaining also to do where we can get better so apparently the strategy where the ECB only talked about one measure of inflation hasn't really gotten all the way through to households so there is some education to do thank you Volker and that of the authority yes well very interesting questions let me go quickly over some of them I would definitely support the idea of having alternative in the sense of several inflation measures that the ECB should target however I sympathize with Volker in the sense that the HICP is not necessarily the best the best one I think some measure of domestic inflation in particular I would say more than domestic GDP deflator inflation core GDP deflator inflation that ignores the items that are subject to large fluctuations would be a more desirable practice this is particularly true to be honest from a theoretical point of view for economies that are small for which import prices may distort a lot the HICP but I think it should apply to the ECB now I'm skeptical that the change of the sort would be adopted that the ECB or any other central bank given that everyone seems to be focusing on HICP the exception though is the Fed that looks at the core consumer deflator which is closer to what Volker is saying now Klaus mentioned the form of inflation targeting I mean in our research we just derive what the optimal inflation target is and that's just one number but for communication purposes I think that it would not be a bad idea to establish a band it could be a narrow band and more than anything to make precise what the ECB believes it what the ECB means by being close to the target so in addition to having a symmetric target I think having a band that would make that a bit precise what is meant by meeting the objective or being close to the target I think would be a good idea on price level targeting the problem of course price level targeting just an extreme it's the limiting version of average inflation targeting where when memory goes to minus infinity the problem with price level targeting is that in the real world it may not work as well as it does in our models in which there's full credibility rational expectations and so on and with price level targeting a period like the one that we have experienced in which inflation is persistently it can be persistently below the target and substantially below the target may call for reaching inflation levels persistently above the target that may not be credible and that the central bank may think twice about trying to implement those so having average inflation targeting puts a bound on the extent to which that will be necessary and that's why I think it's more desirable so a four year or six or eight year average inflation target I think would be desirable also it doesn't have to be a strict targeting the the ECB should not try to hit that target no matter what okay the idea is it would be a reference point that would guide its decisions I agree that the the Fed announcement on average inflation target was I guess trying to be flexible was vague and that's probably why it didn't have much impact on markets I think it would be important to be if the ECB were to adopt the average inflation target it would be important to be a little more precise in particular on what the relevant horizon is so that you know people can actually compute what's the relevant relevant average inflation at each point in time that has to be compared to the target and not just one thing to conclude Annette's presentation was very interesting now I'm not sure her recommendations at the end Annette you were calling for or you were pointing to the importance or the desirability of consensus in such an environment you know consensus of course sounds like a word but also consensus depending on how it is implemented leaves much less room for individual policymakers to express their views and that either before or after policy meetings and you know that may also be harmful from many points of view okay thank you Jordan Annette because you didn't use it too much time earlier on if you want to have any reaction to that quick reply to Jordan so I think that individual policymakers should definitely express their views but they should do so you know on the record and in public and that's a very important part of the debate and that leads to great decisions what I'm arguing against is all this kind of behind the scenes stuff for you talk to word as I talk to Bloomberg I think very little good can come of that and we should try to clamp down on that and and I think consensus building more consensus means that everyone has a bit more at stake in kind of upholding good norms very good so let me just close by saying we're in a ECB listens mode so of course there's so much covered in this session but very much we will be we are discussing internally but it's fantastic to have external voices participate so let me thank the panelists and of course in every conference overruns that happen so I do apologize for overrunning a little bit but I think there was no loss of efficiency there all the material was so rich it was worth the overrun at least from my point of view so with that let me hand it back to Cieri and again to thank you to all the panelists thank you well thank you very much Philip thank you very much to all the panelists all the participants certainly very stimulating to have a debate which is concluding our program for today two points of info the videos of today's session along with the papers the various presentations are being made available on the ECB website so do keep an eye on that and also please remember to submit your questions for tomorrow's policy panel which will feature Christine Lagarde president of the ECB but also our prominent guests Jerome Powell chair of the Federal Reserve governor of the Bank of England so please do send in your questions via the twitter hashtag ECB forum or for participants using the dedicated email channel well at this point the point I want to make is that the forum will resume tomorrow at 2pm