 And welcome to our speaker, John Greenwood, who is going to give us quite a comprehensive look at the presentation, and it really is answering lots of the questions that I suppose many people have an interest in knowing. What makes QE work? Does it work at all? So looking at some of the examples from around the world of that, with a particular focus on the Eurozone and Japan. So John has been Chief Economist with Invesco for many years, spent a lot of about half of your career in Asia, and as you probably know Invesco is a very wide-ranging wealth management firm based in 20 countries, including here in Ireland. Including Dublin. So we look forward to your talk. Okay, well thank you Dan, and Tim for all the arrangements. Good to see you. My talk is built around the idea that QE's and the policies taken by the US, and I think it's fair to say the UK government, at least before Brexit, have worked out relatively well. But clearly something isn't working quite so well either in Japan or in the Eurozone. And I've thought long and hard about these issues, and gone into a lot of the detail. And I want to share some of that with you today. So I'm going to contrast the US and the UK on the one hand with Japan and the Eurozone on the other. But before I get into the nitty-gritty, I want to say something about what happens in the aftermath of a debt bubble or credit bubble bursting. And I've created this little diagram, it's not really a model, of typically what happens in a bubble and bust. So what we have on the vertical axis is a debt-income ratio. And what typically happens is when credit conditions become easy, so put yourself back in 2002, 2003, that sort of period, interest rates were very low and credit was easily available, the private sector, by which I mean households, non-financial corporations and financial entities, borrowed and typically would buy assets, could be real estate, could be commercial or residential, could be equities, could be commodities, a whole range of assets. But in any case, they would borrow and bid up those asset prices. The rising asset prices often also meant that people's net worth increased and they were able to go back to the banks to borrow more. So there was a fair amount of positive feedback in this rise in this ratio. But typically what we saw in many countries was debt ratio rising. And that happens in country after country until something stops it. Either the asset prices get too high, people say, there's no way I'm buying any more tulips at this price, right? Or the central bank raises interest rates and you have an inverted yield curve so it doesn't make sense to go and buy the assets. Or credit conditions are tightened or as in the case of 2008, you have a major accident like the Lehman bankruptcy. And then suddenly everything changes. Behaviour changes because asset prices fall, but debt doesn't go away. So people are left in a position of negative equity which you saw a lot of in this country and we saw a lot of in the UK in the early 90s. And basically they then try to reduce their debtiness, so reduce the debt-to-income ratio over a period of time. Now while they're doing that, they cannot spend at the normal rate. They cannot consume at the normal rate, they cannot invest at the normal rate. So GDP will not grow at the normal rate. So typically what you see is initially a recession and then a period of sub-par growth. And that's exactly what we've been through, what we've been through these past eight years or so across the developed Western world. Meantime governments step in trying to counterbalance that and they do that by fiscal spending either financed by borrowing, sometimes financed through central bank means, but nowadays that's severely restricted. And they do it by means of explicit fiscal stimulus programs or through automatic stabilizers. And typically because the economy is growing at a sub-par rate, treasury revenues are down, so the government is running a budget deficit. So if it adds to the deficit with stimulus programs, you see the budget deficit rising, so the government's debt-to-income ratio rising strongly. And you see this in country after country. Now the scales are not necessarily the same on the private and public sector. But I would say that no country has yet come out of phase two. All countries, no country has yet fully normalised. And phase three, this is not entirely over until the government has reduced its debt ratio. But of course governments never repay debt. All they do is run balanced budgets and try to grow the GDP, the denominator here. So gradually over time this would come down, but not through running budget deficits. So that's kind of the theory. The point I want to make, the point of showing you all this, is that during this period in phase two, while the private sector debt ratio is declining, the mindset is, I don't want to borrow. And as I'll show you, people don't want to borrow no matter how low the interest rates become. So during this period there's a complete lack of demand for funds. But we live in an economy, in economies where credit is created through private sector institutions. So if there is no demand for funds, then credit doesn't grow. You get either a slump in money and credit growth, or you get a contraction in money and credit. And that has very severe consequences. But just to illustrate the point and validate this framework, here's the data for the United States. In the United States you can see the long build up in the debt ratio prior to the lean bubble bursting. And then the abrupt deleveraging in the early stages of the downturn and more recently a more gradual reduction in the debt ratio. So the US, I think you could say, conforms to that template that I just showed you. And actually the US has done better than any other economy. So the debt ratio today has come down by some 65, 70 percentage points, excuse me. And the debt ratio has now returned to roughly where it was in 2002, reversing about two thirds of the build up in leverage, which is pretty good progress. The public sector, on the other hand, that ratio remains at its peak. And you can see one could draw that second vertical line that I had in the previous chart, but I'm not drawing that second vertical line yet because this flattening out of the debt ratio is occurring at a time when interest rates are still incredibly low. And if interest rates were to rise, I suspect there might be more deleveraging by the private sector. So that's kind of the framework. Let me make a few general propositions before I get into the detail. First, that, this, let me go back to this chart. This vertical line, the first vertical line after phase one is often known now in the literature as the Minsky moment after the American economist Hyman Minsky. This is the point when behavior changes. And the point is that after the Minsky moment, deleveraging becomes the mindset. Instead of leveraging up and hoping that asset price increases will bail you out, there's a complete change of psychology and all people want to do is to repay debt or at least to restore a positive equity position. Now that deleveraging has blocked credit and money growth, slowing normal and real GDP. Now my proposition is that QE can solve that problem. I mean it can create money independently of the banking system. Now creating money may not necessarily be achieved in this deleveraging environment, simply by lowering rates. And yet what we see is some central bank simply lowering rates and waiting for the banks to lend. But if customers don't want to borrow and banks don't want to lend, then there's no assurance that those very low rates or even negative rates will achieve anything. So what I'm going to show you is how QE done properly can create new money and furthermore can assist in the deleveraging process because the creation of new money comes without increasing private sector leverage. So that's what I'm now going to show you in a few slides. First though, the general, the orthodox view is to treat QE as a method of not only providing funds but more a way of lowering the yield curve to obviously the central bank sets the short-term interest rate typically. But purchases of assets, securities, long-term securities by the government, by the central bank, excuse me, tend to push down the yield curve along its length. So the traditional view is that QE helps mainly by lowering interest rates. But if you look at the record, for example in the United States, each time QE was announced, that's the first vertical line in each episode, there was an announcement effect. But once the purchases went into operation, rates rose. And that happened on each occasion, there was some decline here, and then a rise in rates. So, and if you think about it, I mean, a successful QE program would restore economic growth, would raise inflation expectations, and therefore you would expect interest rates to rise anyway. So measuring QE by its impact on interest rates, I don't think it's very helpful. More fundamentally, I think that the right way to approach this is to think about QE in the context of what happened in the United States in the 1930s. Now, as you may know, after 1929 and the stock market crash, the volume of money in the United States contracted sharply. In fact, the cumulative decline in M2 between 1929 and 1933 was 38%. The cumulative decline in M1 over the same period was 32%. And the decline in lending was even larger. Total lending by lending to the private sector declined by 54% over the same period. In other words, because of the shock, not only was there a reluctance to lend, but there was a determination on both sides to shrink balance sheets. So households and corporate customers reduced their borrowing by paying back, and banks were forced to shrink their balance sheets as well. So liabilities, which is deposits, shrunk, hence the decline in the money supply, and loans also shrunk. So we had this disaster in the US to which the Fed did not react. Essentially, the Fed did nothing until after 1933. Now, the problem in 2008, in a sense, was to avoid this, how to prevent this from happening again. And I think that we can see that very clearly in the data for US bank lending. Here's total bank lending by commercial banks in the United States. In 2008, where this chart starts, and you can see it was growing at a pretty healthy rate, 10% to 13% per annum. But then after the Lehman episode, prior to that, there had been the Bear Stearns bankruptcy as well, but lending then plunged and contracted. And over the entire period, from the middle of September 2008 to 2011, the cumulative decline in bank lending was about 14%. So if you like roughly half of the decline in the money stock in the 1930s, in other words, so what that means is if the authorities had done nothing this time around, we would have had a slump at least half that severe. Now, what happened in the 1930s was there was a huge contraction in GDP. Unemployment went to somewhere in the mid-20s, 20, 25%. And of course, there was deflation as well. So huge impact on society on misery because this was not prevented. This time, however, through the Fed's QE and the Bank of England's QE, that effectively was prevented. And this is how it was done. In the US, what I've done here is I like these episodes of QE. But instead of looking at the amount of securities purchased by the Fed, I'm looking at the rate of change of banks' reserves and their contribution to the money supply on the other side of banks' balance sheets. If you want to get into the detail, I'm very happy to explain it, but basically what you do is you take the stock of money and explain its growth rate by the growth of the assets on the other side of the balance sheet plus any residual items. And I've reduced those to three groups. So we've got the broad money supply, which is the black line in the background, which is shaded. And then the three main contributory items. First of all, there's the lending, which you just saw contracted sharply. Now, if the money supply had followed bank lending, there would have been a sharp decline. But that didn't happen because the Fed stepped in injected reserves, which counterbalanced in terms of the bank's balance sheets. It prevented the assets from shrinking. And the counterpart on the other side was that the deposits were also able to continue to grow. Now, they did slow down quite a lot for a while, but the Fed's QE was able to prevent that complete contraction. So there were three episodes of QE, as you know. QE1, QE2, QE3, and each time you get this sort of surge of growth of bank reserves, these are actually not the growth rates. These are contributions to the rate of change of money. So if you add up the blue, the red, and the green line, the net result gives you the black line. Anyway, you could see that while lending was very weak, QE prevented a complete collapse of the money supply in episode 1. In episode 2, lending was still very weak, and the injection of funds enabled the money growth to pick up a bit. And again, in QE3, lending was still very, very weak, it didn't really pick up until the beginning of 2014. And then again, the injection of new funds enabled money growth to continue. And today, what we have is our QE is obviously terminated in the US, so this is now negative or zero. And what's happening is banks are once again lending, and you can see that we have a normal situation in the banking system. That is that bank lending is roughly the equivalent of the growth of bank deposits on money supply. So the black line and the red line are in line together. So there's a sort of template, if you like, of how to do QE and get it to work so that you don't have a monetary disaster as we saw in the 1930s. Now from that, I've derived two basic rules for successful QE programs. The first is to buy the securities from non-banks. I haven't really emphasized the detail of that yet, but I'll come to that in just a second. And the other thing to do is to buy long-term securities, not short-term securities, because if the central bank buys short-term securities, it's clearly going to have those rolling off its balance sheet maturing, and then it's going to have to replace them. And that's going to dilute the effectiveness of its program. So let me deal with the first one first. Well, I've got one other thing to say for that. When the ECB announced its QE program in January of last year, January 2015, they said that they would be buying 60 billion euros a month, and I think the initial program was going to last until March of this year, and then subsequently it's been extended, and this number has been increased to 80 billion. No matter. So what I did was to apply the principles I'd been observing in the US and calculate what that would mean if the program was successfully implemented, what that would mean for the rate of growth in stock of money in the eurozone. This spreadsheet I put together in the chart from the spreadsheet combined with January 2015, I haven't changed it since then. The ECB balance sheet was going to do that, growing at roughly 30% per annum, and broad money, M3 in the eurozone, would grow from very low rates of growth. Actually, it was already growing at 4% or 5% here on the right-hand scale, but it would end up growing at about 10% per annum. But in fact, that hasn't happened at all. Money growth has been much lower than that, only about 5% per annum, and lending, of course, has hardly picked up at all. So the question is, what's different? And this is really what triggered the analysis that I've been doing. Now, in the case of the Fed and the Bank of England, the Fed bought long-dated US Treasuries and a small amount of T-bills which it later reversed with that majority extension operation or operation twist, as it's sometimes called, and they also bought mortgage-backed securities. The vast majority of these securities were purchased from non-banks. Similarly with the Bank of England, they bought long-dated gilts, some commercial paper, but almost entirely from non-banks. But if you look at the QE programs of the Bank of Japan and the ECB, most of the purchases have been from banks. The Bank of Japan has bought a small amount from non-banks, ETFs, j-reads, but also starting in June, the ECB has been starting to buy corporate bonds, which again, it is buying from non-banks. But the vast bulk of the purchases government securities have been from banks. Now, what's the difference? Why does it matter? And this is my analysis. So here's how the Fed and the Bank of England were doing it. When they bought government securities, they added to their assets and they purchased the government securities from non-banks. So here we've got the balance sheets of banks, sorry, central banks, banks and non-banks. Assets on the left, liabilities on the right, the American Convention. So step number one, the central bank buys government securities from non-banks. You could use mortgage-backed securities, it doesn't matter, the result would be the same. In exchange, the seller, which could be an insurance company, a pension fund, a money manager, obtains a check essentially, nowadays it's an electronic payment, but it's a check from the central bank, which the seller lodges with its bank. So the seller now has a new deposit and the commercial bank essentially completes that process by taking the deposit, the draft to the central bank, which is credited to the reserve account of the commercial bank. So we now have an increase in reserve deposits of commercial banks held at the central bank. So the net result is central bank balance sheets, central bank balance sheet has expanded, commercial bank balance sheets have also expanded, but on the books of the non-bank public, this is households, firms and non-bank financial institutions, they now have more money at less government securities. If we fill in the remainder of the, some of the assets and liabilities of the commercial banks and the non-banks, what you can see is that the money supply, which consists of banknotes and coin plus deposits held by the non-banks have increased. So this method of implementing QE incontrovertibly increases the stock of money. Now of course it's possible that non-banks may be repaying debt. So they're burning up these deposits by repaying loans that they've previously drawn down. But this would nevertheless be a substantial increase in the rate of growth of deposits in the system compared with what otherwise would have happened. So the other thing to point out about this set of transactions is that deposits have increased without loans growing. The counterpart asset to the deposits is new reserves for banks at the central bank. The fact that loans have not grown means that those loans which are reflected here as borrowing liabilities of the non-bank public enables or assists in the deleveraging process. It means that instead of creating deposits by making new loans, banks have had an injection of funds into their balance sheets. And as I say, the money supply has increased without any increase in loans and therefore in leveraging the private sector. So this kind of QE assists in that deleveraging process that I talked about at the beginning. Now let's look at what's happened in the Eurozone and in Japan. Here again we've got the same balance sheets that central bank, commercial banks and the non-bank public. And initially what happens is that the central bank typically is buying securities, government securities from banks. The banks in exchange receive a draft payment from the central bank and that's credited to their reserve accounts. So on the books of the central bank there's no difference. Both sides of the balance sheet have expanded by equal amounts. But on the books of the banks they now have reserve deposits where they previously had government securities. But if we fill in the rest of the items on the balance sheets of the commercial banks and the non-bank public, you can see that there has been no impact whatsoever on the money supply at this level. The money supply to remind you again is banknotes and coin plus deposits held by non-banks. So there's no liquefaction, no injection of liquidity into the real economy. If you like, this is the real economy, this is the financial sector or the banking part of the financial sector. That's why I call this an asset swap. Essentially what's happened is that the central bank has swapped assets with the banks but there's been no impact at this level. And that I think is the fundamental explanation of why money growth in Japan and the Eurozone has not been as rapid as it otherwise should have been. In addition, let me highlight again that if at this point the commercial banks wanted to increase deposits or policy makers wanted money to grow, what would have to happen is that the commercial banks would have to make loans to increase deposits here. And that of course would add to leverage in the private sector or at least it would prevent leverage from coming down. So this process is not only not liquidity enhancing, it's detrimental to or discouraging to the whole deleveraging process. And what an economy needs after a debt crisis is balance sheet repair and relinquification. This approach to QE is not generating those results. So what's the evidence for this? Here's the data. This is what I've done is I've taken the consolidated data for commercial banks in the Eurozone. So that's all the countries of the Euro area. These are securities they held in trillions of euros on the left-hand scale, no I'm sorry, right-hand scale I think. The scale is for the central bank, the scale is for the commercial banks. So Euro area banks were holding roughly 3.1 trillion euros worth of securities as of March 2015. With the start of the ECB's QE program, look what's happened to their holdings of securities and meantime of course the ECB's holdings of securities have gone up a lot. Now these two sets of numbers don't quite match because actually what's happening is that the banks are consistently buying securities in the open market, replenishing their magazines, if you like, and then selling the securities back to the central bank. Now they may be securing those by buying they may be buying them in the market from the public or they may be buying them at new auctions. But either way, all I can tell from the data is I've got the absolute level of holdings. But I suspect that this number is the gross sales of securities by commercial banks is a lot higher than this number over the period. After all the ECB over this particular period had bought 723 billion euros worth of securities but the headline number of securities held by banks had only gone down by 287 billion. But as I said they'd been replenishing their stocks. So it's clear that the ECB was buying has been buying from the banks. And as I said, entitled the ECB has purchased at least 40%, that's this number as a fraction of that from other banks. Incidentally, the same thing has been happening in Japan. Here's the bank of Japan holdings of securities on the left hand scale. And here's the banks holdings of securities going down again by a smaller amount. But in this case, roughly 30% of estimated purchases just at the headline level. Again, I can't get the gross sales figures by the banks. So clearly, the operation is different. Now in Europe that's been justified by saying, well, but we are a bank based financial system. We don't rely like the Americans do on the capital markets. I think that's an excuse. I don't think it's a valid reason. There are plenty of securities held by non-banks and arrangements could be made to purchase those securities. More broadly, I think Europe has quite a big problem in that banks have not yet started to expand their balance sheets. Here we see Italy. Total assets of Italian banks still decline and their holdings of securities declining quite rapidly. It's a similar story for Spain. Although the Spanish economy is recovering, Spanish total bank assets are declining, loans are declining and holdings of securities are declining. So across the board, we're seeing zero growth or even contractions in the individual country's bank assets. That's not desirable from the point of view of fostering a recovery. I think that even the other part of the ECB's program, the LTRO program, was not very successful. The ECB increased, it made loans to banks through the LTRO program starting in late 2011, soon after Mario Draghi became president. So the ECB's balance sheet expanded then from 2 trillion euros to just over 3 trillion euros. But at the same time, this was lending to banks against collateral, but at the same time, total lending by banks declined from roughly 3% or 4% year on year to about minus 4%. I really don't think you could call that much of a success. It was accompanied by a shrinkage of banks' balance sheets. And essentially, it's the same problem that the ECB was making cheaper funds available to the banks. The banks were substituting loans to existing borrowers with funding from the central bank, but overall they were contracting their balance sheet. And I suspect the same will happen with the new targeted LTRO program. A little bit more specifically, these are the Japanese bank holdings of securities that I talked about earlier. You can see how they've come down very sharply since Mr. Koroda's scheme was started. Japan, of course, has done two episodes of QE. It did QE in the early 2000s. There was a brief period here after the Lehman bankruptcy under Shirakawa when they did a very half-hearted sort of QE, but they didn't really start in earnest until Governor Koroda was appointed in April 2013. And since then, we've had a very expansionary bank in Japan balance sheet. Financed in the same way, largely by purchases of a company security. And they have also made the other mistake. So instead of buying from non-banks, they were buying from banks. But in addition, they were purchasing not only government securities, the black part of these bars. Bank of Japan was also purchasing short-term securities, so-called Tegata or financing bills. That's the red part. So those were continuously maturing and running off the balance sheet. And that result for Japan is that although the monetary base or the balance sheet of the Bank of Japan, essentially, has expanded massively from an index level of 100 to over 300, there's been barely any growth at all in money or bank lending. And I can show you that in a more conventional way. These are the rates of change of money in black and bank lending in red. You can see that during Japan's bubble in the late 1980s, we had double digit growth of money and credit. When the bubble burst, they had that big change in behavior. Nobody wanted to borrow, banks didn't want to lend. And lending and money growth collapsed. And since then, they've been in this prolonged period of very slow growth of money and credit. And that's the primary explanation for their deflation. And even this very aggressive QE hasn't really worked very well because, as I say, the Bank of Japan essentially made both mistakes. It bought securities from the banks and it also bought short-term securities. This is what's been happening on the books of Japanese banks more generally. So under QE1, their holdings of securities increased, their loans declined. Net result was that they're holding a total bank credit, which is the sum of these two declined. And the same, roughly, although it's slightly reversed, basically this time securities have been declining. Lending has been growing just a little bit, but the net result is that bank credit has been declining. So all in all, it has not been a very happy tale. So what's behind all this and what's the problem? Now the normal kind of interpretation is that monetary policy works through interest rates. More specifically, QE should work primarily through, that's the interest rate is the direct pass-through effect, lower interest rates. But as I pointed out, that doesn't always happen. But more specifically through portfolio re-balancing effects. That is, after the stock of money is increased, then the institutions that acquire those deposits are likely to reinvest them in equities, corporate bonds, real estate, and so on. And that triggers that portfolio re-balancing effect. Of course it pushes up asset prices, but that in turn should induce some increase in spending. And given that QE purchases from non-banks helps with deleveraging, then there should be some good results. However, if you look at the Bank of Japan's analysis, and this comes from one of their bulletins in May last year, it's their diagram. All of the emphasis is on interest rates. Their view of the mechanism of QQE, as they call it, qualitative and quantitative easing, is that pledges of JGBs, Japanese government bonds, would lower normal interest rates. The announcement effects would reduce inflation expectations, or raise them, sorry, in this case. And then the net result would be lower real interest rates, which would encourage borrowing and spending. So it's all interest rates. No mention here of the stock of money, the quantity of lending, et cetera. In a broad sense, what has happened to conclude my talk is that the central banks in the eurozone and in Japan have relied on a transmission mechanism for money and credit through the banking system. But the problem is that that transmission mechanism is broken. Banks are reluctant to lend because they're repairing their balance sheets, building up capital, improving the quality of their loan portfolios, et cetera. And on the other side, households and firms have been reluctant to borrow. So the standard transmission mechanism isn't working. So the solution that I'm advocating is that the QE program should be redesigned so that it bypasses or circumvents the banks if the central bank were to buy securities from the non-banks, then that will directly create money in the way that I've described and that would boost the amount of purchasing power or spending power in the economy. So design of QE is really important. The Bank of Japan's QE program in 2016 failed and similarly the ECB's LTRO program and its QE program has so far been disappointing. As a result, we've only got at best growth rates in these two areas, the Japan and the eurozone, of one one and a half. Of course individual countries like Spain and Ireland grow much more rapidly, but for the eurozone as a whole where there's been almost no deleveraging in Italy or France for the eurozone as a whole, the results have been disappointing. By contrast, the Fed and the Bank of England asset purchases I think were largely successful because they focused on long-term assets and they acquired these securities from non-banks. That directly injected new deposits into the financial system and that has helped support not only purchasing power but it has also helped support the deleveraging process. So that's my thesis. I'm sure some of you will have other opinions but I'd be very happy to take your questions. Thank you.