 Good morning to everybody and in particular to my four panelists, you only see three of them but I actually now you see the fourth one so Ulrike unfortunately can only join us remotely but let me say we're very happy that you were able to still make it virtually. So those of you who attended our virtual ECB forum last year may remember the lively panel that we had back then on the future of inflation. Charles Gotthard criticized in his intervention back then what he called the bootstrap theory of inflation. Namely, and he explained it as follows, as long as inflation expectations are anchored, inflation will remain anchored. Ricardo Reif was in the audience at the time and I think it's fair to say that he strongly objected to Charles' fundamental critique arguing that and I quote again so you see whatever you say here may come back next year. Inflation expectations are very useful in so far as they reflect whether people believe in the central bank. This interesting debate gave rise to the idea to follow up on this important question in this year's ECB forum and I think it's fair to say that this panel could hardly be more timely. Back in September 2021, so that was a time of the last panel, HICP inflation in the Euro area was standing at 3.4% and it has now accelerated to a staggering 8.1%. Market-based measures of longer-term inflation expectations, the five-year-five-year inflation swap, stood around 1.75 in September last year, rose to almost 2.5% earlier this year and has now settled slightly above 2% in recent weeks. Survey-based measures of inflation expectations have gradually crept up towards our 2% inflation target or have even moved above 2% as the median in our consumer expectations survey. In the baseline scenario of our staff projections, inflation is projected to stay high for an extended period of time but to return to around 2% over the medium term. And these developments raise a number of important questions for policymakers and this is precisely what we would like to discuss in today's panel. So what is and should be the role of inflation expectations in monetary policymaking? Which inflation expectations do actually matter? Those embedded in market prices? Those by professional forecasters or those by households and firms? How can we detect risks of de-anchoring? And finally, how can policymakers ensure that inflation expectations remain well-anchored when inflation is expected to remain above target over a prolonged period of time? These are the questions that I want to discuss with my four excellent panelists which I will introduce one by one as they speak. And it is of course natural to start where the discussion ended last year. So I'm delighted to give the floor to Ricardo Reis who is professor of economics at the Land School of Economics and who last fall published a fascinating brookings paper on losing the inflation anchor. At the time of our previous ECB forum, Ricardo was of the view that inflation expectations in the euro area most likely would stay anchored. But he asked central banks to make clear that if inflation turned out to be permanent they would do whatever it takes to stay within their mandate of price stability. So of course we are excited to hear how your views have evolved since then. Ricardo, the floor is yours. Thank you. Wait, my glass is going to fall. Well, thank you so much for having me to talk here about inflation expectations. So let me try to give you where I think inflation expectations are in the euro area right now and what challenges they pose for monetary policy. But to start, how do we go about measuring expectations? As Isabella already noted, we have three options. We can ask people and for that the best publicly available data is the one produced by the Bundesbank which has an online survey on consumer expectations with data available since 2019 and soon we'll also have one conducted by DCB coming live. We all know, or now we all, some of you may know that when one asks surveys of people of expected inflation you always worry about the fact that people have biases, they overreact to news and more generally that over the last 20 years when inflation was very stable the signal to noise ends up being very small, partly because there is so much noise in this data. Moreover, not only we can ask people about stations one year ahead or we can ask them at long horizons, but whether you ask three, five or 10 years you tend to get the same answer. Indeed the Bundesbank asks five and 10 years ahead and pretty much gets the same answer showing that people really can't distinguish that much between these different horizons. If you ask markets you can, the best publicly available data is DCB's survey professional forecasters now complimented again soon with a survey monitor and analysts and that on the one hand has a lot of signal to noise and does quite well in the previous 20 years but suffer from strategic behavior. Many of these professionals want to look different from others, after all often they are paid for their advice and that advice better be a little bit different than everyone else but at the same time they suffer from conformism insofar as they spend a lot of time in conferences like this and then have been driven by the arguments in it and more importantly as in that Brookings paper I noted if you look at the record of when we had large turning points and we had good professionals data early 1970s US inflation going up, Volcker inflation going down the professionals were way off, they really were completely they were as surprised as Central Bank I guess is a way of putting it. Third option is to measure market prices using inflation swaps, options in the case of the US break evens in the URLSO where the difficulties there is that you need to take out competition for risk and we know how the price and quantity of risk in markets changes a lot as well as just noise having to do with liquidity and all kinds of trading frictions beyond adjusting horizons and payoffs. Let me do that, let me look at these three types. By doing that I'm plotting here from the Bundesbank one year ahead the expected inflation, important you need to adjust it and I have a simple formula there that you barely can read that comes from previous work that I had to adjust for a standard deviation in skewness which is a very simple way to back out the noise to take out some of the biases of people the markets in blue and the professionals in green. It's clear that we have it's clear that in 2021 households started moving faster a little bit than the markets did but both very much together and then in 2022 we have especially the market prices rising significantly more as perhaps you start having risk compensation increasing as the recession became more likely and high inflation starts being associated with the recession therefore making sense of why blue got a little bit steeper than red after they had moved together. The professionals were mostly useless in detecting the increase in the inflation one year ahead and notably in the last three months and as we expect the new wave of the Bundesbank server which will come pretty soon what we see is actually that from the markets although we have very steady market prices if anything a reversal as monetary policy tightens and again let's wait for what the people think. That's the one year ahead and one that we can judge and say well some people were policies for my rudeness a little useless in so far as we already know a little bit what the inflation was or is very likely to be but let's look at the harder question but maybe the more important question what if it looks five years ahead and where perhaps we can also average over the biases and over interpretation of people well five years ahead you see a very similar pattern where households increased markets much more after January and again the professionals lag behind. So if you want to think about whether inflation is anchored in the sense of the next five years you get a similar image of somewhat of a de-anchoring starting after I made those statements Isabel and speeding up at the beginning of this year. Now this if you and looks at all three of them but in trying to instead of pick between the three why not use a method that tries to combine them to come up with a method and here's a lot of equations convince you that I'm serious let's also say that these are a model that tries to extract some of the biases allows for overreaction sluggishness and essentially combines these three lines these three series with measures of disagreement and skewness that allow you to put more or less weight on the people versus the markets if you do that here is what fundamental or your own expected five year inflation is since starting in 2024. I purposely did not put a number in the vertical axis why because that's an assumption I have to start it somewhere if I started it to then expecting inflation to have risen to five started on one it will have risen to four what happened why does it increase in these two steps well first when it looks at the data to decompose what the model is doing is that skewness and dispersion increased at first in mid 2021 and then as the market prices rose the model says aha indeed expecting inflation must have increased first the skewness and dispersion then the asset prices finally the median changes what happened in the very last data point the very sharp increase in markets and given the difficulty in this model the counter for change of composition of risk maybe perhaps lead to an overstatement so I don't completely buy that last small number over there let's wait for a little more data it's probably half a percentage to 1% lower either way the new anchor if before let's say was 1% now seems to be 2.5 or maybe even as high as 3.5 or so according to what the data on expectations is showing why is that well the picture from the Buddhist bank survey shows it somewhat clearly this is the histogram of the answers that people are giving and what you see is that it's not just that stations are an anchored in the sense of a flatter distribution distribution shifted horizontally to the right now to 1.5 to 2% higher if one starts from the perspective that now let's good and bad glass half full glass half empty is on the one hand they are stable now at somewhere like again 4% here once incorrect for biases, composition for risks and others one could say it's more like a 3 to 4% is the right amount and note before you start dismissing all households I go to too many conferences where people say oh these households they don't know what they're talking about remember how they were pretty much on track, not so the professionals when it came to the one year and now think about it if inflation is 8% this year 5 next year, 3 the other and 2 and 2 that averages to exactly 4 that does not seem like a crazy out of touch household may not be the outcome one wants but certainly one that seems somewhat plausible moreover perhaps German households are particularly sensitive and therefore this is an upper bound but this to say that this is a reasonable estimate on the other hand is this a worrying estimate perhaps if for 5 years average again having an estimate around 3.5% 4% seems reasonable if instead it was 4% in a 5 year 5 year horizon that would be worrying given a target of 2 but as I told you you really cannot torture the data to tell you that because the people just don't distinguish this very well and moreover it is not that now institutions are anchored they are anchored at some number above the 2% target whether that's 2.5, 3 I don't know but certainly somewhat above 2 if you want to look at longer horizons in a 5 year 5 year horizon then you really want the only way you can do is markets markets do separate 5 year 5 year from 5 years ahead and there like I said you struggle with separating for plausible risk premia this increase all in the eurozone 5 year 5 year expected inflation I think can be too glibly dismissed as purely composition for risk there would be a 1.2% in composition for risk change in inflation that's what some of the estimates from DCB models would say it's very hard to write a model that has a 1.2% increase in a risk premia with no change in fundamental whatsoever over this period but let me leave it at that to say instead that if your measures if one looks at the distributions or sorry I mean slip ahead if one uses distributions to back out instead these probabilities of inflation being high in this 5 year 5 year horizon what one gets is that DCB and takes that as a measure of credibility in so far as being this far horizon one gets a perhaps worrying perhaps not shocking increase from essentially 0 to around 6% around January December but since then very little change these are the very last numbers that I calculated 2 days ago for June is the last number you get again a 6%, 8% likelihood if you want then maybe DCB will not be able to control inflation 2% at this long horizon which again seems credible and again, glass half full, glass half empty but should you care? having looked at the data I conclude that expectations seem reasonable, sensible the data is giving sensible things and they say that perhaps the inflation expectations are a little above where the target is and with a risk of staying above higher should you care? well for that you need a model you need to write the simplest New Keynesian model the 3-Kerzykism model because I thought it would be the more familiar to this audience but I think the points I'm going to make are independent of really the model itself what happens when you have rational expectations well, expected inflation stays on target at the target being pi upper bar and the actual inflation deviates from that in so far as we have either shocks to potential what I call the A where if you have a very central bank with a high phi then you really kill those and you achieve the so-called divine coincidence and in response to the data supply shock to the gap to an efficient potential well inflation again rises but how much depends on how dovish you are and if you are infinitely dovish you really let interest rates respond a lot to an innovation of output gap then you end up with z coefficient there goes to 1 essentially all of the shock goes to inflation so then what happens if we have expected inflation, pi e that you measured the numbers I showed you and the turnout to have gone up well if they are pure noise you should not respond to them theta should be zero in the response function of the central bank of course if they are pure noise you're just introducing noise to monetary policy you would be tightening too much by responding to my things seem to be a little bit above 2 perhaps but what if people in all their stupidity wrongness answering things wrong act on it that is they from the first equation what if people expecting inflation set higher wages and or set higher prices in the case of firms what if people expecting high inflation therefore taking account of lower real returns on their savings decide to spend more and through hand to mouth end up again expecting inflation comes with more consumption spending well if that is the case on both of them then you won't definitely respond to expected inflation for even if this pay was a deviations as long as people respond to it they are pushing up agate demand they're pushing up agate supply they're just raising wages raising prices raising consumption as a response to it what if instead people when they're raising inflation rotations are overreacting the supply shocks maybe it's all about the supply shock z but that z sadly comes with again these very very terrible citizens that we have to shepherd coming with raising their inflation expectations with a coefficient beta there that can be maybe small maybe large some people often will say oh people just overreact supply shocks they shouldn't be doing it again if they do though they're gonna spend more set higher wages set higher prices at a minimum the central bank wants to overreact as much as they do and potentially more depending on how much you care about inflation output and how dovish or hawkish you are but if people overreact they act as they also they're not overreacting their survey answers they overreact in their behavior and if they overreact in their behavior then 12 set the inflation that comes from it one has to overreact as well and finally what if instead it's longer in credibility what if instead these overreactions are coming up as a deviation of expectations that persists over time in that case the central bank in order to keep inflation on target need not just react to it but react extremely swiftly and very aggressively this is sometimes what's seen as the an anchoring of longer expectations for that is the only way to prevent a prolonged and very large deviation of inflation from target as a result of those an anchored slowly moving expectation of inflation that arose so again now you want to have a very high theta a theta that's at least as large as delta that is the extent to which people are biopic in their expectations to conclude if policy is just not and only if you think that the drift of expectations is pure noise that no one is acting on nothing on the shore running longer and should you ignore it otherwise a movement up is a tightening and how much when tightens depends on whether you think it's driving actions responding to market shocks leads to more reaction and doubts and credibility to conclude can we measure expect inflation accurately no and yet we can look at data and get signals and the signals that come looking both now exposed to the performance of our people from expectations a year ago and now that we have a year of data to see the inflation tell us that and then contrast that with the five year expectations give a coherent account that my view makes it hard to dismiss expectations as being irrelevant or a bottle something theory I forgot what the name was that you said Isabel then Charles had said last year what is the best measure expected inflation Isabel last introduction none better to combine them combine them suggest an anchor that is somewhere around three to four percent on the one hand fine that may be consistent with even projects what inflation will do on the other hand concerning if it persists but with some upside risk a longer horizons is the cb target still credible yes it's a small probability of five six percent that is being revealed by the market prices but at the same time there is a little bit of that upside risk and so given the upside risk both in the short run and in the long run should a central bank respond to noisy upside risk and the answer is that unless you're very confident that this is truly noise then not even the guys saying are taking seriously they're not even taking themselves seriously the response to surveys otherwise it suggests tightening with a different vigor depending on whether you think it is driving the measures up and I've gone way over time apologies but for answer to Charles Goodheart there's just this matter for monitor folks thank you thank you thank you so much Ricardo for an excellent presentation let me now come to the second speaker so I'm delighted to welcome Loretta Mester who is president and CEO of the Federal Reserve Bank of Cleveland and a very vocal member of the Federal Open Market Committee not so long ago Loretta said in a speech delivered at another ECB conference that a risk management perspective argues for caution because inflation risks are to the upside and because the longer inflation runs above our goal the higher the risk that long-term inflation expectations will become unanchored thereby making the return to price stability much more costly we are very happy to hear more about how the Fed is using such a risk management approach so thank you so much for being here Loretta we are very much looking forward to your intervention well thank you Isabelle thank you fellow panelists and thanks ECB for inviting me to speak today I'm really happy to be here the whole conference has just been wonderful so what I thought I would do in brief opening remarks is talk about inflation expectations from a practitioner's point of view and a practitioner from the US so I thought I'd talk a little bit about theory a little bit about practice and then sort of an approach so very simply inflation expectations have been a central feature in models of inflation dynamics since the 60's and 70's Phelps, Friedman, Lucas all talked about the role of inflation expectations and they do play a key role in the New Keynesian DSG models that are used to inform monetary policy at many central banks across the globe now many of the central bank inflation models some measure of a resource utilization gap helps to determine inflation lagged inflation included in the models to capture inertia in the inflation process and some measure of inflation expectations is included in the model and the different models put different weights on these three factors but household and business expectations matter since they affect wage demands and offers and therefore the firm's price setting behavior and Cleo and Fett has an institute on inflation and inflation research and they've been doing a lot of research and they've found that including measures of inflation expectations actually improves the inflation forecasting models although I know there are papers out there that find that they don't improve them that much anecdotal information indicates that businesses do base pricing decisions on their expectations about inflation and there's recent research that documents that high inflation expectations cause firms or lead firms to raise their prices the inflation expectations and this is something Ricardo talked about also provide an indication of how credible the public finds the central banks commitment to achieving its policy goals so if you look at the Federal Reserve's monetary policy framework it emphasizes the role of well anchored inflation expectations in helping to maintain and achieve price stability and it wasn't until 2012 that the FOMC actually first established an explicit 2% long run goal for inflation and then when we revised our statement on longer run goals and monetary policy strategy in 2020 it really emphasized the role that the committee puts on longer term inflation expectations that need to be well anchored at 2% and their contribution to helping us achieve both of our policy goals full employment or maximum employment and price stability so now the problem is there's various ways to define well anchored so here I'm going to be using well anchored to mean longer term inflation expectations that are insensitive to the data and are at levels consistent with 2% inflation and of course this definition then would depend on the public understanding the central banks inflation goal and how strongly it believes the central bank is committed to returning inflation to goal when it's deviated so what that underscores is the role or the very important role that central bank communications can take in these days in really keeping inflation expectations anchored and via that channel communications can help mitigate the persistence of shocks to inflation so if you go back to the three factors here inflation expectations that are anchored can help mitigate the pool of resource gaps on inflation and therefore the cyclical movements and interest rates that policy makers induce to maintain inflation expectations are not well anchored and I think this is consistent with what Ricardo just talked about and that's particularly useful with the zero lower bound when it constrains interest rates so arguably the US would have had much lower inflation during the great recession had inflation expectations not been relatively stable offsetting some of that influence the negative output got had on inflation and similarly in the face of today's very high inflation readings if inflation expectations were to become unanchored then their influence would offset the impact of any beneficial change in the output gap and monetary policy would have to act much more forcefully to return inflation to the goal so I think the theory is very compelling I mean I think theory is very compelling real world doesn't always cooperate so you remember in Japan for a long time inflation expectations have run well above their actual inflation rates for a number of years so one of the first thing a policy maker has to do is to confront that the theory speaks of inflation expectations but those expectations are not directly observable I mean in the model if you have rational expectations model the unobservability doesn't come into it because the inflation expectations are model consistency but in the real world that's not true so we look at a number of different measures and Ricardo showed a number of measures for the euro zone these measures differ by type of agent and time horizon and there's measures based on consumer surveys there's surveys of businesses there are professional forecasters surveys and then there's the measures derived from financial markets now in practice I think if you want to get an indication of where inflation expectations are and where they're going you need to look at all those measures you can't just pick one but when you do that you see that a clear signal isn't always forthcoming some of the measures move in different ways and the research just shows that different groups of agents can behave very differently from one another the literature has not informed us not firmly established whose expectations matter most for inflation dynamics and even within a particular group of agents even if you say consumers there's considerable heterogeneity across consumers so we're doing some measures which we call indirect consumer inflation expectations at the Cleveland Fed and basically it's an idea based on indirect utility so you ask the question okay given the basket of goods you buy so it's personal to the person answering the survey how much would your income have to rise in a year or a five year horizon for you to be as well off so you can still afford that basket so in other words rather than sort of ask them directly tell us what your expectations of inflation a lot of our work says people don't even know what inflation is what's this indirect thing if you look at that what we get in the data is these expectations of inflation from this indirect measure and this is in the literature elsewhere they vary by demographic and socioeconomic factors so women tend to have higher expectations than men you know more wealthy people have higher expectations so again that isn't really in the models that we have the prices of particular sailing items like food and energy also seem to affect inflation expectations they kind of outsize effects the empirical results also show that we don't really understand the direction of causality right it's not clear so you know do high inflation measures cause inflation expectations to rise or do higher expectations about inflation affect the decision making which is something Ricardo was getting at leading to higher inflation or is it both so it's a problem because you have to then infer something and it's not even clear from reduced form estimates which way it goes and the other thing we know is that while businesses are the ones that set prices we only have limited information on inflation expectations for businesses although work's being done both in your ECB is doing work and the Fed is doing work to try to increase our knowledge of business expectations I think another very practical consideration for policy makers is how to assess whether inflation expectations are becoming unanchored from the target and relatedly the level of degree or the level of the central bank's credibility in the eyes of the public so you can look at levels of longer term expectations rather the shorter term expectations to get some indication you know for example if longer term expectations are stable in the face of a positive shock to inflation the least inflation will come down although it need not indicate that they believe that monetary policy will be the main driver of that reduction and in addition to the stability of the mean or median level dispersion across survey respondents might also indicate how well inflation expectations are anchored with lower dispersion indicating better anchoring so I think we need to be looking at all those measures we have to look across different factors agents in the economy to understand their expectations and we need to look at more than just mean or medians we need to look at distributions policy makers also have to contend with the possibility that financial markets may have more confidence than the general public in the central bank's ability and commitment to bring inflation back to goal and I think that again it suggests that policy communications are going to be important for keeping inflation anchored so taken all together the research suggests that there's still much to learn about how inflation expectations are formed yet as a policy maker we need to make decisions based on the limited information we have and the recent data in the US indicate that longer term inflation expectations are still below current inflation readings so you might infer that the public expects inflation to come back down from its unacceptably high level but if you look at measures of longer term expectations they are rising at longer horizons and that's problematic and if you look at dispersion and this is a household expectations that dispersion across survey respondents is also increasing so this is looking at dispersion measured by the 25th percentile and 75th percentile in the units of Michigan inflation survey so I think the fact also that sailing in prices like food and gasoline remain elevated suggests that there is some risk that longer term inflation expectations of household and businesses will continue to rise from the current levels so what do you do on the current situation from a risk management perspective I think it's important for policy makers to ask which situation would be more costly erroneously assuming longer term inflation expectations are well anchored when in fact they're not or erroneously assuming that they're moving with economic conditions and in this case moving up when they're actually anchored and if you do simulations using the board's service model they suggest pretty strongly that the more costly error is assuming inflation expectations are anchored when they're not if they're drifting up policy makers treat them as stable what ends up happening is policy will be set to loose inflation would then be moving up and this would be reinforced by increasing inflation expectations on the other hand if the expectations are actually stable and policy makers view the drift up with concern policy will be initially reset tighter than it should and inflation might move down and might even go below target but not for long because inflation expectations are anchored at the goal so those simulation results coupled with the research suggesting that persistent elevated inflation poses an increasing risk then inflation expectations could become unanchored strongly argue against policy makers being complacent about a rise in longer term expectations I think it's important to remember that inflation expectations are determined not only by movements in inflation but also by policy makers actions to follow through on their strongly stated commitments to return inflation to the longer run goal thereby justifying the public's belief in the commitment so obviously the current inflation situation is a very challenging one both here in Europe and in the US central banks are going to need to be resolute and they're going to need to be intentional intentional in taking actions to bring inflation down if you think back to the low inflation we had in the pre-pandemic expansion that really led to considerable research the ECB is at the forefront of that research as is the Fed on how low equilibrium interest rates and the zero lower bound can create a downward bias to inflation and inflation expectations and the policy implication that some drew from this research was that if policy had to air it should air on the side of being too accommodative since it would be easier I'm going to emphasize that it would be easier to address high inflation than low inflation I think the current situation very challenging one in which a sequence of supply shocks to contribute to inflation being at a 40 year high belies that view I think it also calls into question the conventional view that monetary policy should always look through supply shocks in some situations such supply shocks could threaten the stability of inflation expectations and they would require policy action so my hope is that this situation that we're in just as a period of low inflation generated important research that will have a new burgeoning research and we saw some of it actually this morning to help us actually understand more about high inflation the role of expectations and how supply conditions matter as well as demand side factors so thank you very much thank you thank you very much Loretta so I must say you were very consistent with what you said six weeks ago but I'm not surprised so our third panelist is a professor of finance and economics at the University of California at Berkeley so I'm very happy that you can join us remotely this is very much appreciated and as you probably all remember Ulrike has not only explained to us why we pay for not going to the gym but she has also done some fascinating research on the formation of inflation expectations and she is well known for thinking out of the box and so I'm very curious to hear your insights on inflation expectations so Ulrike it's great to have you here even if only on the screen thank you so much for including me in this super interesting and topical panel I hope the wifi will hold up where I got stranded so what I'd like to do in my brief statement today is to draw your attention make your pause and think about what the current experience of living through an inflationary period will do to our beliefs and ultimately our decision making where we as both households and firms and actually monetary policy makers and then as a result what this means for inflation expectations as a variable of interest for monetary policy the starting point is of course the research Isabel alluded to that we have now accumulated lots of dispersion within a population it turns out that the personal experience of past inflationary periods or high stability periods has a lasting and very strong impact so here for example is an updated graph using a US data from the Michigan survey the dots just show one year ahead inflation expectation separately for people below 40 40 to 60 in red above 60 in blue after taking out the population means was showing the disagreement and what you can see first of all if you just focus on those dots as well sometimes older generations are more pessimistic sometimes younger generations are more pessimistic sometimes the different generations pretty similar views on inflation and sometimes there's huge dispersion and noteworthy in particular in the period of the 70s up to 1980 there was increasing dispersion up to 3 percentage points at the time if you're trying to predict these cross-sectional differences after accounting for your usual suspects in terms of demographic you know wealth education stuff but even monetary policy with fixed effect that individual level lifetime exposure to inflation is enormously powerful those are the fitted lines I drew in here the solid and the dashed lines it turns out that if you know you give me your birth year I get data on the inflation you have I average it with somewhat declining rates there's some reasons you bias I get a lot of predictive power so our personal experiences seem to stay with us we tend to put more weight on realizations experience over our lifetimes than other historical data in some sense there's a relative of the adaptive learning models but the big difference is that it's crucial not just to say oh people over age recent realizations everybody puts too much weight on the last year or whatever you prefer no it depends if you're 20 year old you have seen you know what happened in the last year or two and then before that you have not seen it if you are 60 year old you remember of the 1980's we have a slightly broken connection so maybe it's better if you turn your video off I think then we can still see the slides right that's possible maybe that's better because sometimes lose your voice unfortunately thank you okay and please interrupt again I'm happy to kind of just go back a slide but what I was trying to say here is that there's lots of data now on past realizations having an over proportional influence on our expectations actually not only in inflation also in the stock market for example in unemployment experiences stay with us we remain cautious spenders for years to come even if life cycle consumption savings models would say otherwise and these effects not only affect what we answer in the survey question of inflation expectations or other expectations they affect actual decisions as also Ricardo alluded to you know in this case investment long term bonds are you shying away from it do you buy a house or not do you use it fixed rate or variable rate mortgages etc. I'm using here on households I do find I do want to pause and say it's not only households who exhibit this behavior in one of their New Zealand manager papers this one the Brookings paper by Kuma, Fuzi, Krabion and Gornet Senko they asked managers of the New Zealand firms how do you typically form your inflation expectations and the top four answers included shopping experience and prices of competitors and suppliers so in other words the prices they see around themselves right now in the recent past have an over-proportional impact on their forecast of future inflation there's also meetings and discussions so information you get from other people but it's not financial advisors it's not monetary policy experts it's actually people like me co-workers and family and it seems to be the case that information coming from people who identify with information that resonates resonates seem to have an impact and maybe we can get back to that later in the discussion thinking about what the central banks could do but finally after covering households and firms I do want to talk also about monetary policy makers themselves being strongly affected by their prior experiences so my favorite example is a guy born as Heinrich Wallig in Berlin in Germany in 1940 into a family of bankers lived through Germany's hyperinflation in 1923 and then emigrated to the US in the 1930s where he had a very successful career in the Fed system first in New York Fed got a PhD at Harvard Fed governor from its 70s to 80s to my knowledge he still holds the record in a Federal Reserve history of dissenting the proposal of the chairperson and kept warning people that you know they don't understand the dangers of inflation how it's around the corner it's imminent etc so why I love this anecdote so much and I think right now Henry Wallig would appreciate being cited during the current period is that this is clearly a highly educated person has all the inflation data at the fingertips models available knows that he's in a different country in a different time period and still cannot shake that experience but I do not have to go to him I can also look at the same annual monetary policy reports and the forecast when you had forecast members make in those if you correlate them you plot them again their personal experience based inflation forecast so I get you know Loretta's and everybody's birth year look where they lift and then calculate their lifetime average of inflation experience I get to predict why they deviate from their stuff forecast and very strongly so why is that why do our personal experiences stay so strongly with us even if we are highly informed know the models etc why does it on the margin which has quite strongly in one direction or the other the message today might be that we should pay a little more attention to neuro rather than knowledge whether we have the information we should acknowledge that as you know living and breathing organisms our brain keeps changing as we walk through life every new experience leads our brain to form a new connection between neurons the synapses that tell our body to react the world how to react to the world around us importantly for the inflation how often we make an experience matters so Loretta mentioned repeatedly gas and food prices which get this over proportional weight that's exactly consistent if you have repeated stimulation or particular over prolonged time that causes long-term potentiation every time you go to the gas pump you see the higher price who can just not stop your brain from forming a strong association between gas and high inflation fears in fact the word fear is important here because neuroscientists talk not only about synaptic tagging but emotional tagging emotional events attain privileged status and memory so if this would result in some kind of panic and anxiety fear that's how we ensure that's really deeply ingrained in the population which of course we would want to avoid. Learned knowledge you know as much as we finance professors think we can fix this by explaining the seasonal AI form model better and that's what we're going to do here. So Loretta mentioned earlier that sedition has very limited power to undo this I think a good reference is the literature on trauma how synaptic changes are caused by traumatic stress and it's a good reference also because trauma is not only the big T trauma the you know traditional I mean in the neuroscience literature the war experience adverse childhood experiences to increase prices even if that doesn't completely destroy your life you know daily worry about food insecurity unemployment insecurity which reshapes and reforms our brain and use us to think differently about the world one of my favorite examples are those gender differences which I saw Loretta also mentioning on her I think third slide there is now I think over 50 years of evidence maybe starting from the Swedish data that women tend typically to have higher inflation expectations so when we did a survey we found this even during the low inflation period this data up to in 2015-16 even within households the male and female heads of households different than inflation expectations controlling for everything. Where does that come from? Well you know it's not literacy it's not education it turns out that the prices males and females see in their daily life have a lasting and strong impact so what we did for example is to look at who does the grocery shopping in your household so traditional gender roles makes that very much the woman still and however in those households where the men indicated that they do at least some of the shopping the gender differences in inflation expectations completely disappears and that of course then reflects the highly volatile food prices where people tend to latch on to the increases rather than decreases so to conclude in terms of implications for monetary policy the experience effect perspective kind of highlights the three or maybe four things first frequency of being exposed to signals is really important that explains the over proportional roles of food and gas prices personal shopping so from a monetary policy perspective it will be important to acknowledge it and understand that that's a reality of consumers rather than just saying well those those high volatility items they really just obfuscate actual inflation trends let's just focus on poor inflation if we do want to understand where expectations come from we need to acknowledge frequency more similar to duration inflationary experiences are extra powerful if they remain high for a long time the effect will last for a long time so you might want to account for that in your policy choices that once we are back to before we are back to 2014-15 people will still be different they will still make different planning decisions and form different expectations this is particularly true if these events get emotionally anchored panic means strong anchoring in memory so obviously it is a key for the monetary policy perspective to try and combat that but not by kind of trying to teach people oh food and gas that's only 15% of those typical urban consumers consumption but I'll acknowledge that's their lived reality and reassure it maybe use more of a resonant channels of information conveyed by people which know their world with its managers which live in their managerial world face those prices to convey this information and then finally I also wanted to talk about the anchoring the anchoring topic which is of course on our mind and to some extent the perspective of experience effect breaks a little bit the link maybe to the usual way we think about credibility right so normally we think of inflation as well anchored but we don't say the 2% target oh that means the central bank is credible and can you know pursue effective monetary policy the perspective I have been giving you says well their expectations very much reflect what they have experienced so if that says 2% they will answer 2% and if that says 5% they will say 5% it does not necessarily undermine the credibility in the sense that the central bank is not clear on their triggers for decisions and it's not going to go etc so expectations are important in that they show whether the central bank is doing its job but just in the sense of fighting inflation reality then the credibility as a result some of the monetary policy topics such as forward guidance might not be as powerful as we have now stop here. So thank you thank you so much Ulrike I think I didn't I didn't promise too much so I very much hope that you are also going to talk in the discussion a bit about the kind of proposal that you had how we what we can actually do to also you know reach these portions of the brain that may at the moment be affected by high inflation so we are now moving to our final speaker who is Erik Nielsen a group chief economic advisor at Unicredit so he is the person sitting every weekend in a Berlin cafe to write his very popular Sunday rap to which one may not always agree but which is always thought provoking and we are of course very curious to hear your views a bit also from the markets perspective thank you very much for being here. Thank you thank you thanks for having me thanks for inviting me to central again and thank you for inviting me to this panel very honored to be here. I am as you said I'm or like Ricardo said I'm from the dark markets who are not only paid to be different but also always long I think you said so let me try giving the perspective from the market and not only markets in terms of financial markets but also markets in the corporate sector we are corporate bank so also my most interactions with clients are actually with European corporates and what we hear from them so I try to give you a sense of this certainly one thing I would say is that the conversations in markets both financial markets and corporate markets picked up very dramatically this year I mean this is the discussion of inflation expectations was not something we did talk a lot about before way back to some extent but certainly this has picked up so this is a really hot topic and very important obviously if I can just summarize the key message I hope you take away from what I say today is the following inflation expectations certainly the medium term to longer term inflation expectations are definitely really important for the credibility of the central bank no question about it I would not doubt that one moment but my sense in my assessment this year the amount of the impression I get is that the use the references to inflation expectations on a monthly basis even sometimes more frequent in terms of not setting the directional monetary policy but sort of moving or indicating changes to the steps I think is excessive there's a risk of being too pointy monetary policy I don't need to tell you but monetary policy is more an art than a science and yes data dependent but my key message here is that I worry a little bit that there's paid too much attention to it and let me try to illustrate that very quickly by by sort of two segments on what I'm going to say first the market based inflation expectations I have sort of two key concerns number one what Ricardo talked about with a 5 year 5 year a little bit dramatic how it's gone up what I see in the 5 year 5 years volatility it came down as as Ricardo said I'll just remind you we were down at 175 earlier this year then it jumped up or moved up to two and a half and then less than three months ago and now it's back at round number two or there about which if I were a central bank I say very good no drama people are believing in us and rightly so but it's volatile right so and remember I don't need to remind you but monetary policy has a policy relevant horizon which is not the next couple of months so for me it is a it's too volatile if I may say so the second point that I have for the worry about the market based on what we read in the markets is what you have said in several of your speech and Philip Lane also and others which is that it's kind of a messy instrument right it's if you have to try to take the liquidity and risk premium out and I could have shown this chart that you had Isabel a couple of months ago in one of your many fantastic speeches where you looked at the inflation risk premium on a one year three year ILS and what I saw from that was that that premium moved from deducting 80 basis points early in 2020 to adding 40 basis points at the time you spoke and it's about the same place now I think that's a whopping 120 basis points shift in that noise inside what would have been the policy relevant period right from 2022 into 2020 to into now so I interesting and it's important work but I take away from this that this is a again a as I said too volatile and too many distortions and I know I'm stepping into dangerous territory now but it's looking at that as a guide for policy direction it's a little bit like the focus on the neutral rate neutral rate fantastic economic concept wonderful exposed is and to be don't really know so this is a really I think it's an important thing to do or to worry about in other words and this comes now to Ricardo's point of my professions having spent my life I if you adjust the the market based expectations for the noise it really becomes not much more than a weighted average of professional forecasters and we have not been very good the fact is that I've done this for 25 years and I will humble pie with anybody around here we've been terribly wrong the fact of the matter is the market prices something they expect to know anything about the future more than we do or you do so to be too using as a tool I worry about now that leads me then to just a more anecdotal side to this it's a good time to tell you that since the ECB which is part of the major central bank I spend more time thinking about with clients from financing conditions as a key measure that was always measured or talked about to normalization and now inflation expectations an enormous increase in conversations picked up the clients called from the financial markets for corporates and what have you an enormous rich discussion pursued and the only thing I can take away from even market professionals suffer from the same thing as research shows that the general public and corporates and households suffer from which is basically a lot of lack of knowledge properly many people don't really know what inflation is I mean so the index we are talking about what's in it generally until a couple of months ago overestimate what inflation actually is they also do not seem to know most people that generally speaking inflation expectations have been higher than actual inflation turned out to be so there's a lot of confusions out there that was the first segment that I am right on time the second point I want to talk briefly about is the survey based and I'm only talking about the medium longer term I think short term should not really worry anybody and I have three observations the first one they are also prone to change or revisions and I will say with all respect that when the University of Michigan revised its inspection number last Friday I have never spent so much time talking to financial market participants about a revision of a number of from 3.3 to 3.1 and we know why so this was become the only conversation the market has had for a good long time so that is one thing to maybe keep in mind one number then there is a second question which is reflecting what I said before that is from my conversation with all corporate clients who are often part of the surveys not completely clear that they know what they are being asked or what they reflect on there is often confusion between micro and macro do they talk about input prices or can they put on to output prices now there are questions about the index which one it is, is it an index they care about or the personal or the business index and for bigger or more senior corporates or clients it is often then confused with asset prices also so there is a lot of noise in this that makes me somewhat nervous about the usefulness of it so this leads me to my final point which is the only reason why inflation expectation may matter obviously is the risk that they lead to higher wage fees and then the self-fulfilling story and that is certainly something that I respect and I know is very important I am not so sure I mean we know economic research shows there is a good correlation or a strong correlation but I am still somewhat nervous now but I see the ECB is concerned about the upcoming wage negotiations in particular Germany and the effect of the indexation of minimum wages in France and other places and then the spill-through of course I respect all this I suspect and I would love to be told I am wrong but I strongly suspect that when you see unions now ask for higher wage growth or in negotiations which is nothing with inflation expectations and everything to do with past inflation and an attempt to recover some of the loss that their members have had and that of course leads to this very difficult question that we had a massive supply shop that is a tax on society paid to somebody else outside the Eurozone and it is a question of who carries the cost how do we distribute it ultimately and that is a policy decision that will be on my pay grade but this is ultimately where it comes down to so here is my conclusion monetary policy is immensely complex I have studied it for most of my life it is obviously as I said it is more an art than a science it is obviously it has to be data dependent but it is in my opinion dangerous and it may be very tempting if one has a bit of that incentive function that previous speeches agree with and I can understand the incentive function to be on the cautious side before getting it wrong of not doing enough could be very damaging for the institution's credibility I understand that but maybe I fear a little bit from what I sense from my standpoint is that there is a little bit too much attention to try to pick numbers that sort of gives you the story that explains it that one picks individual numbers that are very, very volatile and maybe not so meaningful so my two cents worth would be that we all understand it is very, very difficult to run monetary policy and I wouldn't set policies on the basis of one, two individual numbers thank you very much so thank you very much Erik actually I have a short question in common on what you said so I was surprised that you said that risk premium are noise and are not informative I would have said they actually contain useful information about the uncertainty about inflation going forward and therefore I would not say that we can say it simply noise and the question related to that is a couple of years ago when inflation expectations were going down did you have the same interpretation of the risk premium to be honest I don't really know I didn't think that much we were more worried about deflation and we were looking at this but it reminds me there was a lot of criticism I think among many when at this place in 2019 another dose of QE was announced very much on the back of a five year, five year and I shared that view that was that's one indicator that starts to look of course there's a risk of deflation so at that time personally I was not as aware of the risk premium of the movements as I am today so that's a short answer alright so thank you very much actually I think we can say that we all agree that inflation expectations in principle matter and to be fair to Charles Gotthard who is of course an excellent economist whom I admire quite a bit I mean he's of course he never said that inflation expectations don't matter but he said there is a certain over emphasis of inflation expectations in the discussion I actually heard that you share that view and the other point you made is that we have to think about the causality and that was the point that you also made so normally I wanted to have around now among the four of you but I must admit that you all took so much time for your presentations there I think we cannot do that otherwise we won't have questions from the floor so I would like to move directly to the floor I will connect collect maybe three questions oh my god I have many questions so okay so let's start on that side so Markus please thank you very nice panel nice discussion I have three questions one the first one is to Ricardo if you think about de-anchoring would you look at the headline or the coinflation index and in particular if households de-anchor do they care not more about the headline than about the core so and often when we do monitor the policy we focus too much on the core and too little on the de-anchoring another question to Loretta in terms what wasn't emphasized so much is communication how much of the communication should be targeting the market the financial markets how much the households and how much the firms are we overemphasizing how much we talk to the markets and too little to the real economy and then I would like to ask Ulrich of course within the euro area we have of course different countries which have different experiences before the euro and inflation experiences have you looked at some within the euro area that people who have experienced high inflation before the euro came into place and they also have high inflation expectations now and hence they're more hawkish at this stage thanks thank you very much so take me to the middle so James please Jim Bullard, St. Louis so I think there was emphasis on the panel about discounting past observations and I think that's certainly the case when you get practical about measuring inflation so does the panel think that inflation expectations are at more risk of becoming unmoored in the current environment where you're starting to discount some of the very valuable data that we had in the pre-pandemic era and putting more weight on the very recent outcomes and that seems like that could be very risky for both the US and for the UCB thank you very much okay so Pierre please thank you very much one question in terms of particularly household inflation expectations and the policy responds to a potential the anchoring or too high inflation because does it matter you know a lot of central banks often take confidence or the UCB in particular to say well market pricing has moved so there is obviously a tightening of financial conditions and so we are already on a good way but the question is do households do regular people really notice that if the actual policy rate is still negative in this case and might be negative in September still until September if we take the current communication and doesn't it mean that for example for floating rate mortgages is there still extremely low so does it really have an impact on these expectations I guess on the on the Fed side you could argue that the most prominent mortgages 30 years they are now around 6% so obviously there has been a very big change and so there it matters but on the Euro area I'm not so sure thank you actually I will take three other questions so Michala Sylvia Wittor so thank you very much and thank you for three very interesting presentations it seems to me that fiscal policy is also a key factor in what's happening today in the policy mix and I wonder when we think about inflation expectations to what extent people's expectations on the fiscal policy stance would be playing in it seems to me that especially with the pandemic the previous expectation on fiscal discipline has really changed in the minds of households so I was wondering if this is something that you have thought about and how it feeds in thank you so we need the mic here so one of you first thank you very much for this very interesting panel my question to Ricardo is thinking behaviour to expectations and I would say also vice versa one on wage formation I think we have some data that show that workers are getting higher wage increases for these current years but they accept lower wage increases for year two or year three multi-year negotiations what you make of that and second when you look at household surveys of consumption they all expect prices to keep increasing in 12 months horizon but you have not seen an increase in consumption 7M has shown the consumption data even the high frequency are disappointing and on expectations of future consumption they're also lower so how would you link this to your model thanks can you pass on the mic to Vitor please well thank you I am on the side of Charles Goodhart in some skepticism about the importance of expectations and I would like to ask Ricardo and Loretta if they can clarify to us how exactly they think that expectations and expectations of um really influence price formation in our economies one stylized fact that economy is dominated by imperfect competition markets with pricing powered by firms also many service show that more than 50% of the firms use a markup model to take decisions on prices and so how exactly are the mechanisms that lead say from forecasts of expectation of inflation by households and the end price formation one thing is clear the expectation of households about inflation affect their behavior on consumption and savings and what not but not necessarily on price formation because when Phelps and Friedman introduced expectations in the Philips curve they were thinking about the importance of collective wage negotiations the expectations and information on both sides and all of that that has over decades disappeared in advanced economies trade unions weighing even in Germany only 15% of the workforce belongs to unions currently so after the late 60's you don't find examples of the importance of those wage negotiations in advanced economies especially those there I would like to draw attention to a paper by the ECB law of last year 2006 and 4 at the working paper that asks do inflation expectations improve forecasts of inflation which is a different question it's not about how expectations impact the formation of prices but about if they help to predict inflation and very briefly they conclude they use 18 models reduced forms and three Bayesian VARs to predict inflation and they conclude first that expectations coming from surveys either of households and firms do not contribute to improve the forecast of inflation and also comes from extracting from financial markets also doesn't improve only what comes from the survey of professional forecasters and consensus improve by about 10% in what regards forecasts of airline inflation and up to 20% in what regards forecasts of core inflation so that introduces some skepticism from this angle which is different from the first I asked about this angle of expectations contribution to improve forecasts of inflation thank you thank you very much so maybe you can try to be very concise because I have more questions in the audience and I would like to give everybody the floor wanted to have it so we go in reverse order maybe so Eric whatever you want to respond to maybe it's there was no specific otherwise I think on consumers I think this is a I don't know of any there is no real research that gives credit to the importance of it but you said something else which I triggered my mind which is that the ECB has actually already implemented a very significant monetary or will do all of stimulus by a forward guidance so the market has reprised dramatically and since you were so kind to to refer to my beloved Berlin on Saturday I was at my hooffest in Berlin and three of my neighbors when they heard I was going to Central and I have to explain what Central was were telling me about their reset of their mortgages which was close to a dobling of their monthly payment and one of them two of them were small business people so it's hitting already and this is in Germany even so you've done very well in getting going but it's so it's felt and that was on households, small business but think of them as households details upon request Ulrike Yeah just two or three quick things first on Markus's question there is indeed research on the different inflation experiences across different countries and others which shows that that has still impacting inflation expectations today and explains some of the cross-country differences we've also shown that this in turn shapes housing markets so people who have people who have experienced high inflation will look for hedges and one of the hedges is kind of fixed rate mortgages and that those past experiences strongly affect the rate of people living in their own home this then links to some of the questions about market might move by the households will the regular consumer be affected I think the third question and indeed that would be my perception exactly as you stated Ulrike with what mortgage rate you are experiencing what you're seeing in the prices is I think what will be in the minds of consumers and that will persist for a long time which briefly links to Sylvia's remark that consumption spending continues to be kind of disappointing that's exactly what this inside of really long lasting effect of past experiences would predict we also saw it after the Great Recession in the US and other countries thank you Ulrike Loretta? Yeah so it's just a couple of things so markets it's communications not only to the market matter so we're you know the communications it's very important to talk to households so they understand what the policy goals are and what we're trying to do with policy so I agree totally with you that we not just focus on communicating with market participants but also communicating with households and business people who are the price setters. Jim you know the models that I've seen the forecasting models of inflation you know there's different measures you have forward-looking expectations or backward-working all the estimates I've seen there's much less weight on the backward-looking than the forward-looking so I take I think your point is right that in this environment and this I think fits with Ulrike's results is that the current situation has a higher potential destabilizing inflation expectations and Vita your question about forecasting models I mean all research coming out of Cleveland says it does reduce the forecast errors on inflation if you look at inflation expectations yeah but I mean it's still like a benefit to that in terms of the actual mechanism right again the the research doesn't really tell you what the direction of causality is when you talk to business people one mechanism is they do think about markups because they're experiencing higher costs there itself but how much they mark up when you talk to them depends on what they think inflation is going to be doing going forward so how much do they think the Fed's going to get back down to 2% and that's informing how they're setting their prices now whether that's sustainable in a perfectly competitive market is debatable but in the moment that's kind of what they're doing yeah no okay a few points first is it hard to measure expectations of course it is that was my very first point by the way it's pretty hard to measure output gaps in case you haven't noticed and then you don't pay attention to those when you're setting monetary policy and if you're worried about picking numbers boy look at what happened with picking overreacting to numbers on inflation measures over the last few months Eric where people are looking at different measures you go up for some or more so of course it's hard and I hope that I made that very clear in my presentation it is absolutely hard and no that's why also I even took numbers out of the axis to say and always said maybe between 0.3% changes in expectations two though I don't know what this is second though should you not respond again to some of those measures again as I showed if markets expectations they move all the time I would never respond to them there are a lot of noise and this is why as I try to emphasize these expectations are useful in turning points they're useful when you see very large deviations when you see them all moving as has happened the professionals of the delay but also happening and so the point was that it's in turning points not in a setting of policy month to month that one looks at expectations and that was very much the point I try to make third during the professionals the professionals as Vitor reminded and absolutely is correct precious are very useful in forecasting they add the forecasting power in normal times but in spotting turning points they in the three or four examples of history they got it wrong in the same way that the type of model the central banks in turning points often get it wrong because they're very biased on past data that is the point I made not the professionals are never worthwhile I didn't say that I said in turning points the professionals expectations tend not to go your right inflation risk premium if your model is telling you that inflation risk premium 120 basis points at a same time where we see nothing comparable in equity risk premium or bond risk premium there's something wrong with your model inflation risk premium not that the data is an informative and it's all noise and that's a point that I made very very quickly also my talk I think the models that we use to measure inflation risk premium are just throwing a lot of stuff out and Loretta also mentioned it as well fifth and importantly really to agree with the way Loretta put it which to emphasize how consistent it is with what I tried to do with my table Loretta noted risk management my table of shocks and how they work that's the typical risk management table and the point there is maybe you want to ignore it if you focus on the first row but if you take a risk management all the other ones point in a certain direction that if you want to interpret this somewhat as a danger then you want to respond and luckily in this time all of them were actually pointing in very much a similar direction and then briefly two more points on VTAR on the expectations of firms we have I don't have data on expectations of firms in the euro area but when we look at expectations of firms and I think it was Ulrico mentioned in her talk we have not found and again preliminary research we meaning the community of researchers not myself individually have not found that firms are actually all the different households they seem to move with the households or if there's anything more with the professionals so not having data I'm going to look at the households and say it's closer to those of the professionals that's what the papers by Varneychenko and many others have found so far subsidize of that how does the price formation work? just like I already told you but I'd even go a little further back just like Friedrich Hayek or Ken Arrows taught us in general equilibrium you respond sometimes to the observations of what you see not necessarily to some expectation of how the equilibrium works firms respond because all of a sudden they're facing higher wages all of a sudden they're facing less demand for their product or more demand for their goods that's what they're responding to why did we have higher wages why do we have more or less demand because some households start expecting higher inflation but the channels are there in general equilibrium it's basically the fact that it's not that you say I'm responding exactly to someone's expectation it's the fact that those in affecting shopping behavior like Ulricho is noting in affecting the way in which you choose a mortgage this or that that is going to affect relative demands and as a business all we have to do is Hayek and Arrows insight all I have to do is look at how many people are working through my shop what is happening to my costs I'm adjusting my prices and lo and behold what's happening to expectations and households animal spears as I said even if so are being reflected in what I do and that's why I even covered the animal spears and finally on Sylvia on the future expectations of course also when looking at expectations one has to I think take into account what we learned in the previous paper by Shabnam and others which is there is a supply shock going on and therefore given a supply shock expecting both high inflation and expected to spend less in the future is if anything a sign of some rationality of these consumers which is not to say that the counterfactual which is if you had expected no higher inflation would you have consumed a little more or less and finally on the wages which you also mentioned of course it is consistent and possible not to mention something to take into account that as households are forming expectations of how life is going to be hard given the supply shocks how unemployment may raise and has prices may come households may choose to choose the wrong word we may end up with wages rising less than inflation in many ways that fall in real wages may contribute to having less of an employment next year that is not inconsistent with being afraid of inflation with monetary policy having to be tighter as a response to those expectations meaning it is not because we don't see wages going up by 8% next year that shows ah ha those expectations of 8% were not merited no because we have the supply shocks that should lead people to be expecting a real wage fall as I would expect the question of course is whether it's very large or very small sorry thank you very much so we started a couple of minutes later I will have two more questions I think John was very early and ok Sasan but please in the questions and in the answers try to be as concise as possible just a quick comment on Ricardo Ricardo's model which assumes that real rates always have a negative effect on consumption and that isn't in fact true I mean theory says it could be it's ambiguous and in fact in the Euro area German households take liquid assets minus debt to income German households have the biggest ratio of net liquid assets in the major economies and negative real rates have a very negative effect on the expenditure of these households basically older people were saving deposits whose deposits are vanishing in front of their eyes or falling in value in front of their eyes so the expenditure effect for Germany on aggregate data actually suggests there's a small perverse effect of real rates other things being equal to German consumption then over there please thank you very much Sasan Garamani SGH macro advisers I'd like to pick up on some of the strands including Jim's comment here a question and ask about non-linearity the economics profession tends to try to smooth out and look at sort of averages I'll keep it brief and simple does volatility matter in an environment like this just volatility per se especially in the instance where we have current expectations forming a lot of our actual behavior now and I hear talk about looking at the five-year and settling in on an average of 8, 5, 4 whatever to 3.5, 4% but isn't there a lot more dispersion and risk here that this taking signals from this could be dramatically wrong at this point in time thank you very much not everybody has to respond so who wants to volatility matters hugely and that's why in those measures I told those algorithms try to really down weight depending on variance skewness as well as volatility of the series what they are and absolutely you do not want to respond to any day-to-day week-to-week even month-to-month measure in those I try to be very careful with the movements up and down month-to-month of my own estimates because I don't think you should in any way respond to those and on the non-linearity turning points that's precisely the point I've been trying to drill in which is these data are useful to show you there's a switch here there's a non-linearity in terms of some regime changes on the just on the real weight perverse effect John I mean I wrote down it wasn't my model it was just the standard New Keynesian model that I was trying to make on whether it's perverse then I guess ECB then changing normal interest rates is going to have a perverse effect on Germany as well because the normal interest rate the real rates that you said is normal interest rate minus direct inflation direct inflation is perverse so is the normal interest rate perverse so that's I think more to the monetary exports on whether raising or lowering normal interest rates has perverse effects in Germany or not Thank you very much. Anyone else? Can I add something? Data dependency at least in my conception of what we mean by data doesn't mean you react to every data point you're inferring what's going on underneath in the economy and the things that you should care about right so that you Michigan survey wasn't the precipitating event of raising interest rates at the last meeting it was really based on looking at the trends in the data and whether we were seeing any easing off of that inflation at all and you know we take all the data into account at least when I'm setting my own policy views it's based on not just one data point it's important to know that all right thank you very much so I I think we should come to an end of a fascinating session so let me just try to wrap up a bit so I think we all agree that inflation expectations matter but that we should not over emphasize them we should certainly look at different measures taking into account the weaknesses and strength of the different measures the dynamics of different measures can be quite different so market based measures may be overly influenced by current oil prices while the household expectations may be very much influenced by the lifetime inflation experience we are still debating on whether professional forecasts are useless or not I think we haven't made up our mind yet so there is a lot of heterogeneity across different people I think it's also clear that the anchoring of inflation expectations opposes a threat to central bank credibility and I think the measures that Ricardo has developed on that are extremely useful our commitment to price stability certainly matters communication matters and I had hoped that Ulrike would mention it but she was actually suggesting that your president should start to sing your monetary policy statement next time to get to the emotions of our audience was a bit exaggerating but so finally once inflation expectations are de-anchored it's very hard to re-anchor them again because part of the inflation experience may actually be built into the brain and I think this is something we should take very seriously but with that I wish you all a very nice lunch thanks to our excellent participants for a fantastic discussion thank you