Tweet
My reading of Merv King tonight:-
So let me now move into broad exploration mode and give three examples in which a trade-off between monetary and financial stability might arise, and which could in theory justify a policy of aiming off the inflation target in order to reduce the risk of future financial instability, before I turn to whether such a policy would have been appropriate before the crisis.
The first is where misperceptions about future incomes persist and are embodied in key prices, such as the exchange rate and long-term interest rates. Households, businesses, and banks can all make big mistakes when forming judgements about the future, and make spending decisions today which they will come to regret when their true lifetime budget constraints are revealed. There is no mechanism for ensuring that misperceptions about the sustainable level of spending are corrected quickly. It may take many years before those beliefs are invalidated by experience. So an equilibrium pattern of spending and saving can emerge that is stable temporarily but not sustainable indefinitely. And misaligned prices may reinforce mistaken beliefs if people are using market prices to extract signals about future incomes and consumption opportunities. Evidence of the persistence of misperceptions can be seen in the imbalances in the world, and especially the European, economies.
I do not mean to imply that when economic agents make these mistakes they are behaving irrationally. Rather that in a world of intrinsic uncertainty it is far from obvious how to make decisions. The assumption of rational expectations is very helpful for economists when trying to understand the implications of their own models – it is a discipline to prevent the drawing of arbitrary conclusions. In practice, however, households are on their own in a highly uncertain and complex world where they are learning from experience. When it comes to decisions about how much to spend and how much to save, expectations of future incomes are crucial. In the absence of a complete set of markets for future consumption goods – and labour – there is no mechanism to ensure that decisions today, and so the implied plans for tomorrow, will be consistent with the possibilities available in the future. If revisions to expectations of future incomes are uncorrelated across households, then aggregate spending will be relatively stable. The problem comes when many households have similarly over-optimistic views about the future. Aggregate spending and borrowing can then be unsustainably high and lead to an inevitable correction at an unpredictable date when reality dawns. Financial markets both reflect and propagate that common degree of optimism. Sentiment and animal spirits can change very quickly.
Examples include the extrapolation of past growth rates of incomes or asset prices into the future when in fact they reflect an adjustment of the level of income or asset price to a new equilibrium. At the time, the MPC argued that the rise in the ratio of house prices to incomes in the years leading up to 2007 reflected a fall in long-term real interest rates – in other words, an adjustment to a new equilibrium house price to income ratio. But if households extrapolated past increases in house prices into the future, then they may have mistakenly inferred that future incomes too would be higher, and so spending and borrowing more than could be sustained. Similar arguments could be made about the reaction of businesses and households to the rise in the sterling effective exchange rate in the late 1990s, and I shall return to this later.
Since long-term interest rates in financial markets are, if anything, even lower today the question of sustainability has not yet been resolved. Misperceptions mean that unsustainable levels of spending, and associated levels of debt, can build up over many years. When those misperceptions are eventually corrected, they lead to sudden large changes in asset values, a synchronised de-leveraging of balance sheets, a large downward correction to spending and output, and defaults.15 Keynesian policies to smooth the path of adjustment by supporting aggregate demand can help in the short run, but their effectiveness is limited by the fact that a significant adjustment to spending – from consumption to investment – is required.
If policymakers can, first, identify misperceptions, and, second, correct them by changes in monetary policy – both highly uncertain empirically – then there is indeed a trade-off between hitting the inflation target and reducing the chance of a financial crisis down the road. But are central banks less prone to misperceptions than others?
My second example concerns what Masaaki Shirakawa, Governor of the Bank of Japan calls the ‘cycle of confidence’. He argues that success breeds confidence, and eventually over-confidence and complacency, leading to collapse. Such ideas are closely associated with the work of Hyman Minsky and others. Minsky set out a ‘financial instability hypothesis’ in which a period of stability encourages exuberance in credit markets and subsequent instability.
My second example concerns what Masaaki Shirakawa, Governor of the Bank of Japan calls the ‘cycle of confidence’. He argues that success breeds confidence, and eventually over-confidence and complacency, leading to collapse. Such ideas are closely associated with the work of Hyman Minsky and others. Minsky set out a ‘financial instability hypothesis’ in which a period of stability encourages exuberance in credit markets and subsequent instability.
Perhaps the experience of unprecedented stability in the UK and world economies before the crisis dulled the senses and bred complacency about future risks. I talked about this when I christened the period leading up to 2003 the nice (non-inflationary consistently expansionary) decade. The point of that speech was that the following decade was unlikely to be as nice. And, of course, it wasn’t. But the point didn’t get home, and the financial system became more and more fragile as the leverage of our banking system rose to unprecedented levels. The experience of continuing stability may have sowed the seeds of its own destruction.
That idea has been explored recently in an interesting new book by Nassim Taleb.18 He argues that the opposite of fragility is not resilience or robustness, but “antifragility”, that is a state in which people or institutions thrive on volatility, shocks to the system and risk. We go to the gym to stress our muscles in order to strengthen them; occasional seismic activity may prevent a more damaging earthquake. Frequent exposure to shocks and surprises may improve the way people learn about and manage risks. In a complex world, we are “better at doing than we are at thinking”, in Taleb’s words. Unless we train and practice at coping with bad outcomes we may fail to respond in the right way to adverse shocks when they come. “Antifragility” does not imply that it might be desirable to engineer small recessions in order to head off a deep depression. We know far too little about the economy to attempt any such strategy, and in a world of intrinsic uncertainty we rely on heuristics – simplified rules of thumb – to guide our behaviour. But it offers a warning of the dangers of believing that the role of monetary policy is to offset all shocks. Rather than pretend that we can forecast the future, a more intelligent response is to reinforce the resilience of those parts of the financial system that we cannot permit to fail and encourage entry and exit in a free market in other parts. It is clear that we need to understand more about how stability affects risk-taking, leverage, and the ‘cycle of confidence’.
My third example relates to the so-called ‘risk taking’ channel of monetary policy.19 Short-term policy rates, especially when they are, as now, exceptionally low, may encourage investors to take on more risk than they would otherwise wish as they ‘search for yield’.20 Financial institutions with long-term commitments (pension funds and insurance companies, for example) need to match the yield they promised on their liabilities, with the yield on their assets. When interest rates are high, they can invest in safe assets to generate the necessary revenue. When interest rates are low, however, they are forced to invest in riskier assets to continue to meet their target nominal rate of return. That tends to push down risk premia and lower the price of borrowing. Other investors too find it difficult to accept that in a world of low nominal and real interest rates equilibrium rates of return will not meet their previous expectations.21 If these mechanisms are important, the financial cycle may be heavily influenced by monetary policy, especially when interest rates are low. That also creates the possibility of a trade-off between monetary and financial stability.
All three examples suggest that the conventional analysis of the trade-off between the volatility of inflation and the volatility of output is likely to be far too optimistic. Does this add up to a case for ‘leaning against the wind’ of rising asset prices rather than waiting to ‘mop’ up after the bust? Certainly we have seen that monetary policy cannot fully offset the effects of financial crises for two reasons. First, crises may impact output before the response of monetary policy is felt. Second, crises typically reduce potential supply growth, for example by disrupting the supply of credit to productive firms. A failure to take financial instability into account creates an unduly optimistic view of where the Taylor frontier lies, especially when it is based on data drawn from a period of stability. Relative to a Taylor frontier that reflects only aggregate demand and cost shocks, the addition of financial instability shocks generates what I call the Minsky-Taylor frontier, shown in Chart 5. This reflects the influence of misperceptions, financial cycles and the search for yield. On the Minsky-Taylor curve, for a given degree of inflation variability, output is more volatile in the long run than on the simple Taylor curve. Ignoring financial instability might mean choosing a policy reaction function that is believed to imply a trade-off at point O in Chart 5. In fact, the true trade-off is given by point P. Once that is understood then the optimal policy reaction function might well change and correspond to a trade-off at point Q.22
The examples I have given suggest the possibility that there is a trade-off between meeting the inflation target in the short run and reducing the risk of a financial crisis in the long run. To shed light on whether that possibility warrants a change to the way we implement inflation targeting, I want now to conduct a counter-factual thought experiment and ask whether monetary policy before 2007 might have moderated the crisis if it had not simply pursued a target for inflation.
So let me now move into broad exploration mode and give three examples in which a trade-off between monetary and financial stability might arise, and which could in theory justify a policy of aiming off the inflation target in order to reduce the risk of future financial instability, before I turn to whether such a policy would have been appropriate before the crisis.
The first is where misperceptions about future incomes persist and are embodied in key prices, such as the exchange rate and long-term interest rates. Households, businesses, and banks can all make big mistakes when forming judgements about the future, and make spending decisions today which they will come to regret when their true lifetime budget constraints are revealed. There is no mechanism for ensuring that misperceptions about the sustainable level of spending are corrected quickly. It may take many years before those beliefs are invalidated by experience. So an equilibrium pattern of spending and saving can emerge that is stable temporarily but not sustainable indefinitely. And misaligned prices may reinforce mistaken beliefs if people are using market prices to extract signals about future incomes and consumption opportunities. Evidence of the persistence of misperceptions can be seen in the imbalances in the world, and especially the European, economies.
I do not mean to imply that when economic agents make these mistakes they are behaving irrationally. Rather that in a world of intrinsic uncertainty it is far from obvious how to make decisions. The assumption of rational expectations is very helpful for economists when trying to understand the implications of their own models – it is a discipline to prevent the drawing of arbitrary conclusions. In practice, however, households are on their own in a highly uncertain and complex world where they are learning from experience. When it comes to decisions about how much to spend and how much to save, expectations of future incomes are crucial. In the absence of a complete set of markets for future consumption goods – and labour – there is no mechanism to ensure that decisions today, and so the implied plans for tomorrow, will be consistent with the possibilities available in the future. If revisions to expectations of future incomes are uncorrelated across households, then aggregate spending will be relatively stable. The problem comes when many households have similarly over-optimistic views about the future. Aggregate spending and borrowing can then be unsustainably high and lead to an inevitable correction at an unpredictable date when reality dawns. Financial markets both reflect and propagate that common degree of optimism. Sentiment and animal spirits can change very quickly.
Examples include the extrapolation of past growth rates of incomes or asset prices into the future when in fact they reflect an adjustment of the level of income or asset price to a new equilibrium. At the time, the MPC argued that the rise in the ratio of house prices to incomes in the years leading up to 2007 reflected a fall in long-term real interest rates – in other words, an adjustment to a new equilibrium house price to income ratio. But if households extrapolated past increases in house prices into the future, then they may have mistakenly inferred that future incomes too would be higher, and so spending and borrowing more than could be sustained. Similar arguments could be made about the reaction of businesses and households to the rise in the sterling effective exchange rate in the late 1990s, and I shall return to this later.
Since long-term interest rates in financial markets are, if anything, even lower today the question of sustainability has not yet been resolved. Misperceptions mean that unsustainable levels of spending, and associated levels of debt, can build up over many years. When those misperceptions are eventually corrected, they lead to sudden large changes in asset values, a synchronised de-leveraging of balance sheets, a large downward correction to spending and output, and defaults.15 Keynesian policies to smooth the path of adjustment by supporting aggregate demand can help in the short run, but their effectiveness is limited by the fact that a significant adjustment to spending – from consumption to investment – is required.
If policymakers can, first, identify misperceptions, and, second, correct them by changes in monetary policy – both highly uncertain empirically – then there is indeed a trade-off between hitting the inflation target and reducing the chance of a financial crisis down the road. But are central banks less prone to misperceptions than others?
My second example concerns what Masaaki Shirakawa, Governor of the Bank of Japan calls the ‘cycle of confidence’. He argues that success breeds confidence, and eventually over-confidence and complacency, leading to collapse. Such ideas are closely associated with the work of Hyman Minsky and others. Minsky set out a ‘financial instability hypothesis’ in which a period of stability encourages exuberance in credit markets and subsequent instability.
My second example concerns what Masaaki Shirakawa, Governor of the Bank of Japan calls the ‘cycle of confidence’. He argues that success breeds confidence, and eventually over-confidence and complacency, leading to collapse. Such ideas are closely associated with the work of Hyman Minsky and others. Minsky set out a ‘financial instability hypothesis’ in which a period of stability encourages exuberance in credit markets and subsequent instability.
Perhaps the experience of unprecedented stability in the UK and world economies before the crisis dulled the senses and bred complacency about future risks. I talked about this when I christened the period leading up to 2003 the nice (non-inflationary consistently expansionary) decade. The point of that speech was that the following decade was unlikely to be as nice. And, of course, it wasn’t. But the point didn’t get home, and the financial system became more and more fragile as the leverage of our banking system rose to unprecedented levels. The experience of continuing stability may have sowed the seeds of its own destruction.
That idea has been explored recently in an interesting new book by Nassim Taleb.18 He argues that the opposite of fragility is not resilience or robustness, but “antifragility”, that is a state in which people or institutions thrive on volatility, shocks to the system and risk. We go to the gym to stress our muscles in order to strengthen them; occasional seismic activity may prevent a more damaging earthquake. Frequent exposure to shocks and surprises may improve the way people learn about and manage risks. In a complex world, we are “better at doing than we are at thinking”, in Taleb’s words. Unless we train and practice at coping with bad outcomes we may fail to respond in the right way to adverse shocks when they come. “Antifragility” does not imply that it might be desirable to engineer small recessions in order to head off a deep depression. We know far too little about the economy to attempt any such strategy, and in a world of intrinsic uncertainty we rely on heuristics – simplified rules of thumb – to guide our behaviour. But it offers a warning of the dangers of believing that the role of monetary policy is to offset all shocks. Rather than pretend that we can forecast the future, a more intelligent response is to reinforce the resilience of those parts of the financial system that we cannot permit to fail and encourage entry and exit in a free market in other parts. It is clear that we need to understand more about how stability affects risk-taking, leverage, and the ‘cycle of confidence’.
My third example relates to the so-called ‘risk taking’ channel of monetary policy.19 Short-term policy rates, especially when they are, as now, exceptionally low, may encourage investors to take on more risk than they would otherwise wish as they ‘search for yield’.20 Financial institutions with long-term commitments (pension funds and insurance companies, for example) need to match the yield they promised on their liabilities, with the yield on their assets. When interest rates are high, they can invest in safe assets to generate the necessary revenue. When interest rates are low, however, they are forced to invest in riskier assets to continue to meet their target nominal rate of return. That tends to push down risk premia and lower the price of borrowing. Other investors too find it difficult to accept that in a world of low nominal and real interest rates equilibrium rates of return will not meet their previous expectations.21 If these mechanisms are important, the financial cycle may be heavily influenced by monetary policy, especially when interest rates are low. That also creates the possibility of a trade-off between monetary and financial stability.
All three examples suggest that the conventional analysis of the trade-off between the volatility of inflation and the volatility of output is likely to be far too optimistic. Does this add up to a case for ‘leaning against the wind’ of rising asset prices rather than waiting to ‘mop’ up after the bust? Certainly we have seen that monetary policy cannot fully offset the effects of financial crises for two reasons. First, crises may impact output before the response of monetary policy is felt. Second, crises typically reduce potential supply growth, for example by disrupting the supply of credit to productive firms. A failure to take financial instability into account creates an unduly optimistic view of where the Taylor frontier lies, especially when it is based on data drawn from a period of stability. Relative to a Taylor frontier that reflects only aggregate demand and cost shocks, the addition of financial instability shocks generates what I call the Minsky-Taylor frontier, shown in Chart 5. This reflects the influence of misperceptions, financial cycles and the search for yield. On the Minsky-Taylor curve, for a given degree of inflation variability, output is more volatile in the long run than on the simple Taylor curve. Ignoring financial instability might mean choosing a policy reaction function that is believed to imply a trade-off at point O in Chart 5. In fact, the true trade-off is given by point P. Once that is understood then the optimal policy reaction function might well change and correspond to a trade-off at point Q.22
The examples I have given suggest the possibility that there is a trade-off between meeting the inflation target in the short run and reducing the risk of a financial crisis in the long run. To shed light on whether that possibility warrants a change to the way we implement inflation targeting, I want now to conduct a counter-factual thought experiment and ask whether monetary policy before 2007 might have moderated the crisis if it had not simply pursued a target for inflation.
Comment