we do today some of my most recent work on international monetary policy spillovers and I am extremely glad to give this BBVA lecture I think the motivation for for this work is relatively broad it has to do with understanding better the channels of monetary policy transmission internationally I am an international macro economists however thinking about that they’re also issues about international about transmission of monetary policy year period in cost economy as well which I think we still do not understand the fully applies to the international setting with international monetary transmission the International transmission and white or a policy is particularly interesting of course because it has to do or it is related traditionally to exchange rate regime and in particular to the Mendelian trait Emma a concept I will describe again for you a little bit later recently I have done quite a bit of work on something that I call the the global financial cycle which also I will describe in this in this lecture giving you some stylized facts and I think most of what I’m going to say is relevant for the possibly for a conduct and design of monetary policy and also of macro prudential policies which have become very important after the crisis and and with Basel 3 so this is the menu for today and in order to go for his points so unfortunately I won’t have time to discuss each of our findings maybe in in a lot of details but I’m going to be basing my talks on a series of paper so the paper in which I introduce the global financial cycle which is the Jacksonville paper of 2014 I’m also going to be drawing quite a bit on my mother framing lecture at the IMF of last year as well as some joint work with Sydney and Agra Pino and with again so he and some new work with Elena Gekko probably the two papers I will be citing the most are going to be the model fami lecture and the paper Weaver with Sylvia today discussing these issues of course is a bit of a challenge because as soon as you talk about global financial cycle and the transmission of monetary policy you find out that there is actually a huge amount of literature both on the econometrics side for a number of things I’m gonna do dynamic factor models monetary policy veers in various dimensions and then on the theoretical side with different models traditionally of monetary policy transmission Mundell Fleming when you open economy macroeconomic literature but also all the new models with capital markets imperfections which have become more and more prominent after the crisis so there is a huge amount out there and and obviously I cannot give justice to all the literature’s which are actually relevant for the type of issues that I’m discussing today so in this in this lecture I’m first going to show you some some facts very briefly on the global financial cycles so what is a global financial cycle so as I said introduced in in my 2013 Jackson Hole paper it’s common movements in gross capital flows career growth leverage around the world it is also some remarkable I guess correlations between those commitments and some indices of market fear such as vivix and there’s a couple of of these indices around but the leaks turns out to have a longer time series and shows remarkable correlations with a number of of time series I will show you that so for example if I if I look at gross capital flows this is a chart taken from my 2013 paper these are gross capital inflows across all asset classes so you don’t need to be able to read all these numbers if you see green it means we are positive and these are correlations and if you see ready things we are negative okay this is all that matters and this is a huge matrix and what is this matrix well these are correlations of equity inflows into various geographical locations FDI flows that flows credit flows so into North America Latin

America Central Europe Eastern Europe Western Europe emerging Asia Asia Africa so various regions of the world and you see that all these cells Morris are green which means the correlations are positive so there is a lot of commitments engrossing flows across asset classes and across geographic areas around the world so this is a first kind of fact here about this global furniture cycle the common movement in growth capital flows now second one I’m just going to show briefly is if we look at an aggregate of banks so World Bank’s leverage growth which is going to be the blue line here around the world so I’ve just aggregated leverage growth of banks with some GDP weights and I look at how these things can move with VIX which is the dotted line and you see here again we have some kind of negative commitments with Avex so when the VIX goes up so this index which is the implicit volatility of ESI p500 so often taken as I said as an index of risk aversion and uncertainty there is the negative commitments between leverage growth and the leaks so more uncertainty more risk aversion means less leverage and by the way with cross capital flows it’s also a stylized fact that if you look at the commitments of this growth capital flows with VIX when the big spikes again more uncertainty more risk aversion you see pure growth growth capital flows so this is also negative commitments with verdicts here now what I’m going to show you a little bit more detail here is that if you look at a very broad cross-section of risk sets around the world I’m talking about stock prices I’m talking about some commodity indices and some corporate bond indices what you find is that there is an important global factor which explains quite a bit of a variance of this huge panel of of risky assets around the world and so in order to show you that I will show you what is tomato dynamic factor model as a direct follow-up on this one may wonder whether this global financial cycle is flows this credit growth is leveraged with global factor invested in risky asset prices is driven in any way by u.s monetary policy and this is something that I will study using a large var here a large by by Asian var Samus is the last kind of set of findings that I will show you and invite B by H and V as I will be able to look at the traditional monetary policy variables but not only I will also be able to look at all these interesting new financial viable credit growth leverage with premier movement in weeks global factor and all that so that’s going to be the advantage of being able to run a big patient year so that’s the menu and I will end also by some you know various identification strategies for these videos and in particular I will also show you international monetary policy spillovers from the u.s. to some specific countries which probably you wouldn’t have thought would be very much affected by u.s monetary policy that is to say free floating exchange rate countries we have advanced economies advanced financial markets such as the UK Sweden Canada and New Zealand so that’s that’s that’s the menu after having seen this evidence on flows on on leverage very briefly so let’s have a look at this global factor in risky asset prices so in order to estimate this factor what what we have done is to take this large panel off of return of rich risky assets as I said stocks corporate bonds commodities we have fitted a dynamic factor models so composed of a global factor regional factors and then some idiosyncratic components so you can see it here in the specification the return is going to be a function of a constant term global factor market specific component and it using critique term obviously then we have to test for a number of factors and see how much of a variance is explained by this global factor which which is of interest to us so in order to construct this panel we take as long as possible a sample and as wide as possible a sample that means we have to sample because a lot of emerging markets unfortunately starts to have data theory started in the 1990s so the best over to sample is ver shortest one from 1990 till 2013 sample it’s the broadest one in terms of

geographic coverage and to get a longer time series and to look at robustness we take a longer but also a narrow a sample which have mostly advanced economies now when we test for a number of factors in that sample so again I’m not expecting you to be able to digest immediately all these numbers here but obviously the numbers tell you is is a test for of a number of global factors and how much of a variance is explained by each other of these factors and so what the test or the test while I say that varies one global factor and the variance explained by this factor depends on the two sample and when our sample it’s a very high proportion of the variance by these two measures yes – fern Mallis or 60% on the broader sample it’s 20 to 25% still quite a bit of a variance of the this risky asset prices which is explained by one global component so here is the dynamic factor model that we are using I already explained a little bit with the setup so now what I’m going to showing you is the structure of other factors so we are allowing for as I said the number of factors we test for the number of and we find that one global factor is what the data is telling us various possibly some covariances across the factors so we don’t impose anything on recurrence metrics across the factors and receive idiosyncratic component for for each of our returns now here are the loadings each of these returns loads on the global factor so that’s by definition that’s what Google factories and when we have market specific factors it can be factors about the geography they can be we looked at factors about the sexual composition when different can use different combinations yeah and then you have video syncretic terms so that’s the model that has been estimated and here is what the global factor in visa set prices look like so here we go I think the most interesting one as I said because of a broader coverage of a sample is the the dark line here this is a global factor we will shade hidden beyond recessions and the number of you know events that might explain some of the spikes on this of this factor so we see in particular very precipitous for West lemon Bravo of course in in the time series so very impressive fall at the time of of the current financial crisis now what is interesting about what which common component of this global common component looks like maybe is that again and by the way here everything is expressed in dollar but it doesn’t matter if you express it in local currency it will look more or less the same what is interesting is that if we look at the correlation of this factor which is the dark line again again with the VIX okay which is with dotted line but without a line that’s the VIX there is a very strong negative correlation so we are talking about minus 0.8 or more here in terms of correlation so again there’s a bit of a mirror image of this global component and asset prices and and movement in vivix and by the way I’m using the again because it has the longest time series but if you were to look at wavy stocks which would be the European equivalent or the UK equivalent of a Japanese equivalent there is a huge amount again of common variations in visa in these indices of volatility and and risk aversion now the correlation with of a factor with other measures of risk premia such as the Gilchrist and there’s a cup-shaped spread is also high but not as high as with the VIX or another measure of risk premium commonly used with beaver Verbier eighth grade the corporate bond spread now in the paper we have a very simple asset pricing model and we use it to do a decomposition of a factor with global factor on realized volatility on the one hand and aggregate risk aversion as a residual on the other hand so our global asset pricing factor just tell us that this common component should more or less be the product of year aggregate effective risk aversion of a market and then volatility of risky asset prices so here we just project our global component on global Global realized science which looks like this and we recover as a residual the aggregate risk aversion contained in our global in our global price factor and what you can see

I think which is quite interesting is that from 2003 moreless onward you have this very strong decline in aggregate effective risk aversion in the market and when we have a general spike Quincy not exactly coincident but again at the time where volatility goes up you have this fight in aggregate risk aversion and then a little bit of a decline was still at at high level here but you do see this very you know downward trend in risk aversion between 2003 and 2007 so that’s the decomposition of our global factor now the next step is to ask a question or two to see whether monetary policy in the center country of the international monetary system that is to say the u.s. plays a role if any maybe it doesn’t in driving this global financial cycle and whether there is any role for financial intermediaries in transmitting internal so internationally u.s. monetary policy so this is the question I’m asking now and why am I asking this question well first of all because I’ve done quite a bit of work also on the international dimensions of the US dollar the importance of the dollar as an international currency and so with the dollar is at the center of international monetary system in particular it’s a very important currency in international banking it’s very important in financial intermediation it’s also an important currency in the international price system as defined by a vegetable peanut recently so it seems like a natural question to ask whether US monetary policy has any role in driving this miserable financial cycle now we do have some guidance regarding international transmission channels of monetary policy into a theoretical literature and I’m just going to do a very brief review of what what we know and what maybe we don’t know about about the possible channels of transmission and when I will show you the result of his of his big var I’ve been talking about of looking at the effect of u.s. monetary policy on a number of variables including my my global factor so here we go so what what do we know about international transmission channels so first of all we have a very traditional mostly you know very much used and very useful model of Mundell Fleming where exchange rates and capital flows play a key role in terms of shaping how monetary policy spillovers may exist or not exist and in particular we have this very important statement about the Mendelian twin AMA which is the impossible you may remember the impossibility of having at the same time free capital mobility monetary autonomy and a fixed exchange rate okay so it’s very clear but if you have a fixed exchange rate regime and free capital mobility you are not going to get monetary autonomy I think nobody is going to challenge that it’s it’s very clear now however the corollary of at Ridhima is to say that you can have monetary autonomy if you have free capital mobility and a flexible exchange rate so in other words a flexible exchange rate is enough to insulate you from international monetary policy spillovers you can you can hit your domestic target in terms of monetary policy if you have a flexible exchange rate so that’s a very powerful statement of a trillion mind which is this kind of view of a Trillium at I’m going to be challenging if we look at open economy neo Keynesian models so here we have a trad very traditional trade-off between output gap stabilization and the terms of trade in English literature and more or less the bottom line of all the four models in that literature is that if you if this is were relevant trade off for international monetary policy transmission there is not much spillovers quantitatively so in a sense that that literature concludes that you know if everyone’s house is in order there’s not much gains from international cooperation in other words Padova’s are not very important and this goes back actually so this is within this kind of New Keynesian Europe an economy New Keynesian models there’s also an older literature and particularly by Ralph Bryant and a number of people who have shown more as the same thing using non macro funded model but but very much invest in the same spirit so not much where in terms of gains from international cooperation not much in terms of monetary policy speed over so this is one set of of

models different set of models which most of them have been developed in closed economy settings are models with capital market imperfections so in particular models with strong type of collateral constraint network constraint balance sheet constraint something like that okay so here we start there’s a so called credit channel by Bianca Kepler and a lot of recent work by a particular get ricotta key karate etc recently we’ve seen some of these models in in open economies variant warning of have worked on that our key and coffers also I’ve worked on that and there sergeant Acosta are important inverse models we can apply to to banks or to non banks this literature started we’ve actually collateral constraints on firms and households and not on financial intermediaries that have been imported to the financial intermediation side well you can see that varies so visa agency costs result in some kind of wedge an external finance premium she’s going to give you the cost of funds in some sense and an abyss external finance premium can be affected by monetary policy so there is more scope for you know quite a bit of action in terms of international spillovers even though it has not been developed very much yet in an international context various there is more going on there are also so this is this is one very important class of model but another class of models which is emerging a little bit is also model about risk-taking risk-taking channel of monetary policy and here we go back to some description by volume and zoo and also bruno and shin here is a narrative of this model so they differ a little bit by the way these models think about the constraints usually we are talking here about models we evaluate with constraint with neutral intermediaries and the narrative goes in between good times asset prices are high spreads a compressed measured risk is very low and so in good times because you have evaluate this constraint your value of this constant gets relaxed because of this low measured risk and somehow this is going to lead to financial intermediary to borrow more as the prices are going to be even higher and we simplifies the boom so as kind of amplification mechanism also in these type of models so this is one also existing literature now that may be very relevant as well because what we do know from quite a long historical studies now about financial crisis is that credit booms are actually one of the best indicators of early warning indicators of financial crisis we don’t have very very good indicators of financial crisis we’re so quite a high noise to signal ratio but credit bond is one of the best the best actually probably right now the literature would argue if very stating channel is relevant and if we see some international transmission via this channel this can be also this can be potentially to to be effect okay unlike when you Canadian literature etc so these are a type of international monetary spillovers that the current theoretical literature may suggest so on the one hand we have all these models with credit channel the balance sheet effects which which are going to work in particular so in India yet all model is going to work for foreign debt on the balance sheet we have excessive risk-taking channel which we just discussed and we have also other types of spillovers yeah so if we think about the Neo Canadian literature it may work for aggregate demand externality as in ferry and warning and incompleteness of market or it may work for picking our externality that will be more of a models with again collateral constraints so these are the type of speed of us that may be or may not be relevant from the theoretical literature so in all these cases we are in some sense for mendelian truth emma is going to fail because domestic monetary policy even if you have a flexible exchange rate will not be able to which the optimum using only the interest rate as mendelian true demand we tell you if you have your flexible exchange rate you can reach you optimum with your interest rate but that’s fine you’re fine all these kind of spill overs are going to prevent you from doing that there is no divine coincidence so floating exchange rate is not going to do everything for you now let me qualify immediately I’m not here saying absolutely not that floating exchange rates are irrelevant or useless I’m just saying they are not enough to insulate

your economy from international monetary spillovers and so they are not enough to allow you to use your interest rate and hit your target okay how bad the issues are will depend on which of these actual channel is actually relevant and how big the welfare roses are if you don’t have any other tools than your interest rate even with your flexible exchange rate to achieve your optimum monetary policy and this is to a large extent still an open question so I’m going to try I’m trying to make progress on this question and show you some I’m going to show you some evidence using this var to try to get precisely at these questions what is you know what type of effect seems to be relevant and with each envelope to to improve our theoretical modeling of these issues to finally envy and address the importance the quantitative importance of these international military reverse so I think this is a very important set of empirical issues and and questions here and and in the end it’s about how we think we should model the transmission channel of monetary policy internationally but that means also includes the economy but I think there again the debate is not totally settled ok so to back up my my claims about you know the dollar is is important so I told you that it’s important in the center of international monetary system it’s important currency in international banking that was heart of international price system so if we look at the data it’s kind of interesting to see that the dollar credit extended to non-financial boroughs so even non-financial Brewers outside the u.s. is approximately 13% of non-us world GDP that’s very large and that’s even non-financial borrowers the top three stocks of dollar credit inferior as where you probably wouldn’t think that much about famous dollar rise Daria maybe you would think of China as a dollar rise area but you wouldn’t think when you re a new UK as dollarized area and yet we are in the top three in terms of stocks of their credit this is according to a to be is number on this study by McCauley & Co offer as documented also by the BIS and by and by Shin in particular is a very important role for global banks in particular European Union based in all this period between 2003 and 2007 2008 we have seen a tiny bit the role of global banks in international capital flows going down since the crisis and asset managers coming up the dollar is also widely used around the world by asset managers though we have less precise numbers at least at this point but what we know is that the global asset management firms have won more funds under management and that again the dollar is one of a preferred currency in in this industry now let’s have a look at this cross-border and a local position in foreign currency so this is unleash foreign currency by reporting banks of over bis and so what we see here is that with the currency the only currency which is very widely used abroad is as you can see this blue is big blue thing here that’s the dollar you see some use of of a euro as well which is in green but it’s not nearly as big as as dollar use on trying markets and when the rest is very small sterling Swiss Hong yen we also see by the way but there has been a tremendous increase in this in this dollar use in all this period since its 2000 and then we see some kind of level down a bit with the crisis the stocks are still very high so this is to motivate this is a few just a few facts to motivate again why I’m looking at the effect of u.s. monetary policy as a possible driver of the global financial cycle now so how how do we usually look at effects of monetary policy in macro well we we run the years and there has been very many monetary policy videos being run so I’m going to actually show you various ways of doing identification in English literature etc most of the studies that have analyzed the links between a u.s. monetary policy and over some financial factor which is a very

recent thing to do with a paper by backyard here or some recent work on between linking monetary policy and capital flows have used small-scale years so five seven variables most of the time and of course this is because we cannot estimate a very large number of of parameters if we use classical methods but also means that there is always a concern about omitted variable bias in this more Cavalier and in particular it has been part quite salient in the monetary policy literature the early work on the years have shown that were number of puzzles coming up in small-scale viras price puzzles price going in via other way that theory would predict and it was typically dealt with by by adding commodity price indices in order to spend more more information to put more information in var so here the way I guess we the innovation here in the work is that what we really want to do is yes to study the effect of u.s. monetary policy on with on the standard economy real economy variables inflation investment GDP etc just to make sure that we’re we are capturing what we think we are capturing but also we really want to include those financial variables and what global financial cycle variables so that means we want to include credit growth we want to include capital flows to banks to non banks we want to include the global price factor some measures of risk premium maybe so we want to include leverage so we want a lot of things so a way to do that is going to be to use biasion estimations by relying on some new work which has been done and what I will describe and of course relying on some priors and by the way we can increase the number of eyeballs that we can put in in verveer now the first set of results I’m going to show you I will have two types of identification schemes I will have traditionally in the literature what has been used a virtual s key identification scheme with block ordering of eyeballs into slow moving and fast moving ones there are some issues when you look at financial viable because you want financial variables to be able to react very very quickly and therefore I’m also going to use a second identification scheme which is going to be an instrumental scheme if effectively and I’m going to use V Roemer and Romer narrative approach as an instrument for u.s. monetary policy shock so that’s going to be my first kind of set of results with these two different identification schemes fortunately you will see what I will get very similar impulse responses so that’s going to be the good news when I will move to more country specifically ours I’m also going to use completely different identification scheme an identification scheme based on high frequency data on Fed Funds future surprises which has been used by York and I can Co offer and buy a guitar and keratea recently yes I’m going to have three types of identification in my ears so here are again don’t worry I really I will show you big figures and you will know what it is but it’s just to tell you to give you the idea that there’s a lot of variables in this in this video and in particular where’s the whole top panel stuff is about real economy variables so from GDP to disposable income we are private fixed investment housing starts and then you have the price indices GDP deflator what social expenditure and then all of that are about with financial variables global domestic credit inflows so credit flows to bank or to non-bank various measures of US banking sector leverage but also remember we are looking at the International transmission here so global banks as I said are important in terms of channeling dollar liquidity so there’s going to be also European Union disaggregated actually also in UK and your area banking sector leverage I’m going to have also my global factor the risk premium of dividends architects etc so a lot of furniture viable as well the setup of a var is a priori quite standard so the only thing is that we

are going to where we are going to be able to add all these variables because we are going to use Bayesian estimation and and where we are relying on recent work by a particular de mogi an onion heist in and genuine islands and prima cherry in order to choose the prior in an optimal way so we we completely build on their work here so we import that technology and so that it doesn’t sound too mysterious so what type of what type of priors what does it mean to use v priors well more or less the type of thing that you are going to to use for priors are very intuitive so a lot of macro variables are a very persistent so you’re going to start with random walk with with drifts or for cointegration and when you are going to say okay you know four legs which are recent legs have more information than distant legs and legs on your own via balls you should have more information but legs of cross fire balls of other variables so this is the type of priors but I used to do this estimation and the tightness is optimally chosen so whatever the results and I will show you I will show you all the graphs but I’m going to give you a little summary so we look at hundred basis point increase in the effective Fed Funds right what we are going to find our literally textbook results for the real economy viable so if I want to teach the effect of monetary policy to my students I can use was was graph because you have negative effect on on production on inflation whichever way you measure it no price puzzle an investment on housing starts employment but interestingly and this is the new part we also have the effects on the financial variables so what are these effects so we have a negative effect on the global component of of asset prices we have an increase in risk premium when increasing the volatility measured by the realized violence we have a decrease in the bank leverage both in the US and the European Union and we have a decrease in global domestic credit we’ve all without the US and of cross-border credit to bank or to non banks so these are the provider balls that we return Vuitton you to the analysis and what we find okay so now in order for you to be able to read both graphs easily so remember it’s a hundred basics point including the effective federal fund rate and responses are going to be expressed in percentage points so that’s going to be very easy to read now you may not evolve be able to see this so I will blow them up okay I will blow a few but but so what to make sure that I’m not hiding anything this is everything is here and okay we have done several baseline with from 22 to 25 via balls this is one of us based on cases I can pick up for example if we look at we really connect I what you have a GDP variable here but with your line you see that we have a decline no price pass all this is with GDP deflator this is here see what is your so decline and then we have all the financial vehicles but these ones I’m going to show you bigger bigger graph so here is what happens to global domestic credit to cross-border credit to banks and cross-border credit to non banks so you see a significant decline in all these credit measures and visa in percentage points so these numbers are quite big because we are talking about massive amounts here so this is this is what what we find for the credit variables this is what we find for the financial price style viable so we have in particular the global asset price factor here so the decline we have the excess bond premium increase you have the volatility measured by the global realized violence so you see the effect on visa on this global financial cycle viable here so here this is the measure of leverage which is based directly on the balance sheet of globally systemically important banks for au Aria and the UK and in the US which is us Broker Dealer data from the flow of funds so what we find and so here the leverage responses are kind of of interesting I think because what we we tend to find is but the broker dealers and the global systemically important banks these are the banks which have important capital markets

divisions so we are talking about banks such as Societe Generale BNP paribas Barclays Visa Visa in the list of the 50 most important banks we tend to behave very similarly in terms of quickly leveraging and then bouncing back here I mean more less significant the bouncing back but they were quick the leverage is similar there are fewer banks in the UK so it’s not as tightly estimated but there is a kind of similarly de leveraging which is reasonably quick now if we take broader measures of leverage if broader banking aggregates with more banks so not only the big guys we tend to see Valda leveraging later in the day so there is a kind of difference in timing in this Invista leveraging now we also do some additional exercise you could wonder if ever decreasing global domestic credit is entirely driven by the u.s. or not and you could also wonder what happens to the monetary policies in the UK and also in the euro area as a response to what’s happening in the US so here are the answers to was the worst question so here is the split between US domestic credit and global credit excluding the US and you see that in both cases this goes down so the results are not driven only by the three domestic credit in the US this is were split between cross-border credit to banks and two and two non banks again significant decrease and here is again my de leveraging graph together with what’s happening will be EUR Aria rate and will UK rate and in both cases you see an endogenous response now it doesn’t mean better necessarily at all that there is fear of floating or whatsoever if you do an endogenous response of a you aria rate and of the UK rate you see some increase but of course less than 1 then 1 to 1 okay so this is the endogenous response of V of a main central banks so this was one set of result so just to sum up showing you very briefly some Co movements about the global fund your cycle then the estimation of a global component of risky asset prices ask the question could it be VAT u.s monetary policy is one of a driver is a several theoretical reasons to think about that so and vidura is very important so let’s have a look that was the Bayesian var with a lot of viable we do find there seems to be something going on in terms of of u.s monetary policy and all these variables that that are linked to the global financial cycle including the global asset price factor but also these credit flows also these leverage variables in the in the UK in the euro area however we want to be sure that we are not picking up some spurious things and so we are doing the second identification scheme which is using some instrumental variables instead of actual st decomposition to identify a monetary policy shock so the principle here is very is very simple and again we build on the work of Matins and Ravin recently also get Lauren Kennedy and we the idea is that you can instrument the monetary policy shock by some proxy variables and run the VAR with this instrument the conditions for identifications are a very simple if you have one shock of interest it’s the simplest case here we have only one shop at the monetary policy shop which is of interest so you need to make sure the instrument is correlated it’s strong enough so you’re going to have to do an F that as well and and but it’s uncorrelated with the other variables so that’s these are the conditions for identification we use here one more and Roma narrative instruments as I said later on I will use high-frequency instruments so this is the set of instruments were using I am not I don’t have time to go over the Romer and Roma methodology but this is exactly what we are what we are using here here are the narrative instruments for monetary policy we extendable Romer and Ramos sample to 2009 at which point was Euro lower bound is hit so we stop there and here again don’t worry what you have to get from this picture is just that I’ve superimposed with two sets of impulse responses the ones that I just showed you and the one with these the instrumental variable identification and what you should get out of these graphs is that we are more or less the same not exactly of course but we are mall is the same all the stuff goes in the same way the variable which is maybe

the less well estimated in all these var and which you know gets the more valuation is actually the exchange rate so whether we use multilateral exchange rate measures or bilateral measures this is a viable that gets the most the wider bands and does more different things between the two actually identification schemes all the other variables that I have discussed give similar patterns so this is I’m just again showing you some for example here we see the domestic credit estimation for the u.s. this is we don’t global credit estimation so again not exactly the same but qualitatively it looks similar this is the estimates for the excess bond premium for a volatility for the global asset price factor for the term spread so again very similar and this is the estimate for the leverage and for the interest rate responses so again this gives us some confidence as I said the exchange rate is not behaving very well however so what do we get from that well so one of the things I think we get from from this exercises but u.s. monetary policy does seem to affect variables linked to the global financial cycles does seem to affect this global factor in in asset price credit flows leverage so in Part II the global financial cycle is driven by u.s monetary policy of course is not the only driver either okay and in the paper you you can find a variance decomposition of the whole variables so for the global financial factor I don’t have the exact number in mind but you will be able to see but depending on the horizon you’re going to be able to explain maybe between two and ten percent but of course this is only about shocks right so as usual variance decomposition is not going to tell you anything about the systemic component so it’s not a small number to be able to explain between two to ten percent it explains quite a bit actually up to something like fifteen percent of her credit flows so this is not the whole story by any means but it seems to be but it seems to be there now does it matter well what what does it tell us well what fundamentally what it tells us is that all these countries are importing financial conditions that may not be correlated with their cycle with their economic cycle so if everybody was perfectly correlated and u.s. monetary policy is kind of a spillovers when it doesn’t really matter but if you are different economic cycles or different financial cycles and you import and these pillars are important then this may matter if we believe in some of this transmission channel I love you via veterinary externality or financial stability channels these balance sheet channels this is where the models will be very important and this is where we need to sort out how big this potentially visa these effects are from the point of view of welfare so it isn’t it is possible that countries import monetary and financial conditions even conditions leading to asset price bubble which are not appropriate for their current you know financial cycle so it is possible that we spillovers affect financial stability now you could say if done this big via US vs. rest of a world global financial cyclic cetera but maybe most of these results are actually driven by fixed exchange rate regimes or so my fixed exchange rate regime for which we know that there is no independent monetary policy right nobody challenges that so it’s worth digging father and take specific examples of countries from which we think a typical example of inflation targets three floaters we’ve advanced you know advanced economies with a lot of of developed financial markets and to see whether we find anything for this set of countries so ready to take one extreme of a distribution and so this is this is why I have am now investigating whether it is financial spillovers or not in New Zealand Sweden Canada the UK I could do more but just a question of of finding the data and then I’m using yet this other identification scheme okay again for more robustness and this is surprising Fed Funds future prices in 30 minutes windows around monetary policy announcement around FOMC meetings so this is exactly why identification schemes that have been used in the literature by your clinic and by Gatlin

Kennedy and I’m using these shocks to instrument again for some measure of monetary policy I have a Fed Funds of a one-year or could you could you could liquid instrument different different measure of course you have to test for the strength of instrument for each of these for each of these instruments I’ve been using and this depends on the variable you have in Verbier okay so you always have to check with validity of instrument now so the first result in order to get an idea of a magnitude is to use this identification scheme on the US itself and here it’s more or less replicating the results of get locality except using also the VIX in inver var as an indication of one of the global financial cycle variables so here what do we see we see that if you have a twenty basis point increase in the US one year we see so what we expect to see on CPI and an industrial production we see an increase in VIX and what we see so here are two of the external financial premium one is based on mortgage spread and one is based on the commercial paper spread and I’m gonna use mostly mortgage spread because the data is been more widely available in my cross-section of countries so I’m going to compare the magnitude of a response of mortgage spread of US military policy on US mortgage spreads and I’m going to do the same thing for this set of countries that I just discussed so you see here that the response of a mortgage spread is approximately the point estimate is probably around seven eight basis points okay for a 20 basis points increase in value ace one year so that’s the order of magnitude for this channel of monetary policy transmission for the u.s. now let’s have a look what do we find for Canada well if we look at Canada so interestingly what we find is this as an endogenous response of policy rates the policy rate in Canada and what we find is the mortgage spread in Canada raises by about nine basis points also following a u.s. monetary policy shock so here again I’m not looking at Canadian monetary policy I’m looking at the u.s. monetary policy you can instrument it by these high-frequency instrument F start above ten so this instrument is is valid what about Sweden effect is a bit more muted you do see an effect on the mortgage spread but it’s a smaller magnitude maybe four by this point five basis points on impact what about New Zealand so here its quarterly data because unfortunately New Zealand doesn’t have monthly data for some of these time series so to change the setup here it’s a forty basis point increase so it’s a double the shock and it’s here you have a maybe a 15 basis points effect or something like that but so you have to divide by two so it’s it’s about again maybe seven basis points response of the UK so interesting level UK again you see an effect here and it’s a relatively strong effect so the peak of a response the bar twelve basis points for the mortgage spreads in the UK I mean there’s no doubt that you know I’ve run quite a few of his videos but one can run more that’s no question but at least the current state of play seems to suggest that these responses roughly of the same order of magnitude than the domestic us responses to US military policy shock so with responses on the mortgage spreads which is you know one of these measure of the external finance premium even though all these countries have our free floaters so how big that is is it a little big number to have you know four basis points ten visible well it depends a bit of what you you think whether the US number is already a big or not so if you think this channel is irrelevant in the US when it’s it’s not very relevant either internationally if you think it’s a big deal in the US which I think we still don’t this is the type of thing we still have to ask ourselves then internationally at least it seems we get the same type of order of magnitudes so when we should not discount that but now is it a big deal or not but very important questions that I can’t at this stage really answer okay and I and I need a model for that however what this type of exercise to my mind clearly shows is that whenever we want to test for monetary policy independence if we run only regressions on correlation of short-term rates which

has been one of a favorite test of Mendelian trina– MA then we cannot get the right answer because the short term interest rate is not the sufficient statistics in order to discuss monetary policy independence and we have seen that some of these countries were actually responding some somewhere not some short term interest rate does different things and yes we yet we have some spíritus through this external financial premium there are more questions meant even when answers as usual in this this type of research agenda what we can say is that u.s monetary policy is one of the drivers of credit growth certainly in the US but also abroad of course water credit flows of leverage of US banks but also of the European banks but it is a driver of a global factor in asset prices and of mortgage spreads including in free-floating countries such as Canada Sweden you can use Zealand now it doesn’t mean so everything that I said is kind of independent of a very interesting literature by Calvin Reinhardt on the fear of floating so these offers are valued that very often actually central bankers say they are floating but they are not really floating because that afraid of the exchange rate appreciation so they actually do something to mitigate that obviously if you do something to mitigate that then you lose some monetary policy independence now I’m not saying this this does not exist the two things actually reinforce each other but with channels of transmission these P levels have been discussing they’re distinct from the fear of floating of course so these two channels may kind of definitely reinforce each other as there’s no there’s no issue about that now I think the very big question we have to answer is given this set of apparel evidence which of these models do we do we favor in terms of transmission of monetary policy both internationally but I guess also domestically and based on that obvious type of estimates a big deal or not so how big an issue is it that in a wave a Trillium is failing that even with floating exchange rates you cannot reach your optimal domestic target using your your interest rate so if this international credit always taking channel or both are really important then what it points to is that you cannot reach this optimum with your one domestic instruments and this is where macro prudential policies come in potentially to restore some monetary or police or monetary autonomy so you need you have several issues to deal with we need several tools and the interest rate is one and when you in you need to introduce this move is a macro financial ones so I do think that in order to assess you know how important confidential policies are and how best to do them we have to set on what type of models we actually want to work with and I think this is where we are now and that’s quite an exciting year with our agenda but I don’t know the answer yet and I’m very happy to to take questions I think we have ten minutes left yes so if you have any questions I’ll be sure the first question so the identification is precisely supposed to take you the unexpected component so what I’m picking up here is everything that is uncorrelated with all these viable that may be describing what is going on in the world and I have quite a few of those because I also have very different as you’ve seen identification schemes and so varis is very even expected component of a monetary policy shock in the US and I’m looking at what it does well in the US or what it does on this on these countries so if we believe identification makes virginity of instruments then we should we should take care of that on the second question so how big is big I think that’s precisely that’s a very good question because in a way that that’s asking from a global financial psych factor exactly

what also have been asking at the end about easy to big deal or not we anyway to answer this question is to to have a model so how big is B I can only give you the statistical answer about the variant the compositions about etc and whether it really matters or not in the end well you will boil down to the type of model we are going to be writing to feed worse facts and then what type of welfare implications we are going to be getting from these models and this is the part of I guess of the research agenda which is which is which is still missing so you know we can start with very different priors so some people may think that indeed the world is very it’s very integrated and this is no big deal if write a lot of commitments in all these things this is reflecting some some shocks and in an optimal way some portfolio balancing mechanism that is going to equate returns or or induce capital flows in a way that but doesn’t necessary effect welfare in any significant way everything is optimal now if you have a model in which vrv is for example financial stability concerns so if you have a model which you know for some frictions in capital markets and you start seeing countries which are in a boom and still importing via some kind of ivory stating externality or something like that this kind of financial condition in the US that are very loose and and and induce even more in stating at a time where your economy is already booming when you may you may think that you might have several you might have serious welfare consequences because financial stability may become a serious threat okay so you might you might think you know booming economy importing an asset price bubble on top of our current economic conditions may not be a good idea and and we’ll end up you know in a financial crisis and you may interpret maybe some of the evidence during that last crisis in this light but I don’t know this may not be the right model we’re really bad but we have a work I think it’s a very open debate but I mean one side of whoever has kind of very important consequences question about semantics is that there are these financial shocks and that creates potentially twin dots between financial stability and price stability and as a result if monetary policy also cares about interests ability it may not but and then so that’s where having another tool it doesn’t mean you lost your autonomy in terms of monetary policy because in principle you can still most of the fathers who could still focus your multiples instrument on the price stability objective it maybe it will come at a cost in terms of financial stability but that’s that’s a policy choice so again it’s more semantics the question is more of the sort of the big question about what do you think your results mean for the dimensional financial system and cooperation I’m not sure these fees of because of course if you’re looking at you identify particular things just to see whether these are heavy I guess for that you would have to basically know what are the other banks yeah yeah so on the first comment so I would almost will review but not quite because in the sense that it is absolutely true but if the issue is going to be financial stability so we are going to be talking more about should monetary policy care about it or not so first of all so when there’s an issue of mandate and and trade-offs between between the two but maybe the mandate of monetary policies is most clearly linked to inflation and you can think of macro-prudential policy more about financial stability but I can also and so that would that would be you can yo kind of comment and but but of course there are two problems with that so one could be that in a dynamic sense you know financial stability is also going to matter for your traditional monetary policy viable so your inflation rate your output your econ develop economic activity etc so even if you cared only about these variables but in a

forward-looking way it might affect you will connect with the represent this is true for national yeah yes yes yes but the bit but it may be more international in nature is that you can probably you know think of models in which when you have a dollar ization of a balance sheets actually even if you don’t care about these kind of financial stability issues necessarily you are going to have much more trouble to to hit your domestic target in terms of input in terms of output inflation trade-offs because of its dollarization and so you know interest rate again it’s not going to be enough there because you’re going to have these these issues with how you deal with dollarization of your balance sheet I think if you want a current kind of model in progress about that reality model is is a good one so it does affect directly your ability to actually hit your your target good always no no no you can’t resolution no no you can’t yeah well that’s what you know but precising one of the ratio but it may not be the right model but that’s what one can see that there are issues like that that are going to come up when you when you start having dollarization so that’s that’s one thing and the second thing is what does it mean to vie for the international financial system so yeah we really first we have really need these cultural models for that if we take these things seriously if we have anything very issues then you know we always talk about international Marie Tori cooperation in the sense of central banks talking with each other too much liquidity too little equality but we could also talk a lot more about cooperation in in macro prudential policies depending where in perfection lies you might think actually some macro prudential policies have to be set by the recipient countries and some have to be set by the the countries that that sound actually capital flows as well and having here a little bit more cooperation and thinking could be one way forward the way this will look in terms of optimal policy tools will entirely depend on the friction so this is where we need to make for progress but I can see that there’s going to be some more fruitful cooperation and the macro prudential side maybe even when traditionally on on the global liquidity side put attorney by successful installation packages so we keep the dinosaurs and I guess it super well at this book essential brands they can affect short-term interest rate but when it comes to only maturity if you looking at mortgage raiser that surprises a beautiful beautiful markets so I’m not sure with what and new you bring to this so again I think the issue is going to be whether these transmission channels that you get through financial spreads so how how big an issue away for your economy so if if your mortgage market is is going to be affected in a way but down the road will you lead to a crisis again you might find you might think that you want to use additional tools so macro-prudential tools are going to be the ones probably that you’re going to be thinking of using but certainly you’re not going to be able to say that having a forty exchange rate is able to insulate that from to insulate you and and conduct in your inflation target business as usual is going to is going to deal with that issue so we go back to the problem of what are the financial market imperfections and how threatening are wait for our financial stability or in some extreme cases maybe how frightening are they for even hitting your target it might not be the case for each country because it might depend on the degree of dollarization or whatever of typical you know structural balance sheets but for some can treat my deal case yes over is again a very wide debate on capital controls so

you know so a lot of studies also at the IMF around with issues these days so here I think my view is more like if you identify exactly what the problem is or where we perfection is you’re probably gonna get more targeted instruments than broad-based capital controls to deal with it and so it’s going to be better to use targeted macro-prudential tools to kind of deal with for example too much in flows in a specific financial asset or something like that and we are countries which have experimented with such tools such as South Korea both quantitative and and also taxes actually imposed at the same time so whether we have been useful or not again opinions my diverged but some people say we have actually worked I think we need more longer time series to be bit more affirmative about that capital controls might be used if you if you don’t have the availability of both macro-prudential tools for institutional reasons or if you really cannot get them to be effective you might you might try to to to deal more with market-based finance using capital controls because they are going to be broader than van van tools which are yet on yet the banking system I think it’s a very you syncretic thing or what I’m saying here is it has to depend so much on these that it should not set up on the supervisors etc but I’m very reluctant to give very you know broad-based answers to this question I think we really need was longer time series and assess what has been done and and learn from that well so our church’s economy it would be very interesting to do actually I would expect to find quite strong with us for for turkey because it seems to me were a lot of transmission channels with a lot of dollarization as well in the Turkish economy and I’m not sure I would have to look precisely at the at the central bank policies in terms of fear of floating issues and and and and and stuff like that but my my feeling is that there would be quite a lot in terms of there’s actually a nice paper by Shevlin calumny or Scania trying to look at similar issues but we very detailed firm based data is a very granular data and seemingly finding some evidence of quite a bit of is taking being driven by a by external by US military policy but it’s being yeah I think it’s it’s work still in progress so we’ll have to we have to see but seems to point towards the same type of effect so to look at other central bank’s no so I would love to do I’m sure that must be some effect by the ECB for example but maybe even the Bank of England but I so far I haven’t managed to do that yet I’d like my strong prior is that the u.s. is going to be strongest because of all the thing I’ve said about the importance of a dollar in the international monetary system but I would be surprised if I find nothing at all for for the ECB and and the UK I just haven’t done it and probably if you need long time Therese for a lot of this thing and as we know the ECB is a relatively new comer thank you very much