Jump to content
House Price Crash Forum

Let Them Eat Cake

Recommended Posts

Rajan began Fault Lines: How Hidden Fractures Still Threaten the World Economy with a chapter called "Let them eat credit" which talks about how governments can (for a while at least) paper over cracks by allowing the expansion of credit. Philip Aldrick (Economics Editor at The Times) is on the twitters flagging an ex-central banker, Jason Cummins, describing the central banking community as akin to the court at Versailles. Start at 33:00

"The public is fed up with what's coming out of the Frankenstein's lab of monetary policy. They've had enough. ... It's like the Court of Versailles before the French Revolution where it was so insular and incestuous; you talked only in your own way and you had no idea what was going on in the rest of the world"

Link to post
Share on other sites

Apart from him saying that central bankers are not the masters of the universe that they think they are, I can't say I like this guy. He goes on about how the biggest failure of the Fed is failing to meet the inflation target and says that there is no greater advocate of negative rates than him. Thing is, who actually cares whether inflation is 2% or 0%? Before fiat money came along, inflation averaged zero over thousands of years and that did not bother anyone.

He also completely lets central bankers off the hook as far as creating bubbles in asset prices is concerned, which has an economic cost in terms of misallocation of investment and a human cost in terms of people stressing about not being able to buy a house, stretching themselves beyond their capabilities and then losing everything.

His recommendation is 'hire a central banker who can does their job properly [to meet the 2% inflation target].' Frankly that's just arrogant and doesn't address the question of how you achieve the 2% inflation target without making the bubbles that central banks have created even worse. I'm not a macroeconomist, let alone a central banker, but I would suggest printing cash and giving it to the average consumer if you want to create consumer price inflation rather than asset price inflation. Thing is, I don't think that's any better than just normalising interest rates.

Link to post
Share on other sites

I really enjoyed it but I think the apparent implicit assumption that an abnormally low bank rate definitely isn't actively contributing to the observed change in saving and investment behaviour relative to overall wealth is problematic at best.

Do people always save with a certain percentage of current wealth, rather than a target retirement income (whether drawdown or otherwise) from savings, in mind? Do those who are currently assetless never save harder when confronted with inflated and inflating asset prices? Do companies with pension obligations never redirect funds away from investment and wages to cover shortfalls in those funds?

Assuming that all of these things are definitely true, or of minimal impact if they are not, without (as far as can be ascertained from the talk) at least seeking evidence to establish that as fact, and when admitting that you don't understand why the behaviour that you are seeing has changed in the first place, seems to me to potentially be as unintentionally insular as the kind of thinking that Cummins is himself railing against.

Link to post
Share on other sites

This is the sort of thing that I'm thinking of above, only I'm wondering whether some of these effects start to become apparent at prolonged periods with a historically low base rate, before it even enters into negative territory:

Claudio Borio and Anna Zabai, Unconventional monetary policies: a re-appraisal, July 2016

The experience so far suggests that modestly negative policy rates transmit to the rest of money market and capital market rates for the most part much like positive rates do. This is the case as long as contracts are sufficiently flexible to accommodate them and market practices can adjust, which need not always be the case.

Their transmission to bank rates, by contrast, has proved more problematic. In particular, such rates have only been parially transmitted to wholesale deposit rates and so far not at all to retail deposits. Ostensibly, banks are reluctant to do so, presumably out of concerns with the reaction of depositors. Moreover, at least in one case (Switzerland), banks have actually responded by raising their mortgage rates, in all probability in order to preserve their profitability and facilitated by high concentration among lenders. Indeed, concerns with the impact of persistently negative rates on banks’ profitability and resilience contributed to a major sell-off of bank shares in February 2016 (BIS (2016b)).

This points to the limits of the strategy as a means of boosting financial conditions through the banking system. If policy rates do not transmit to lending rates, they cannot boost the demand for loans. If they do transmit to lower lending rates but not to deposit rates, they squeeze banks’ profits, over time possibly undermining their willingness and ability to lend. And if they transmit to both lending and deposit rates, they risk unsettling the deposit base, making it harder for banks to attract funds.

[. . .]

But more generally, there is clearly a limit to how far risk premia can be compressed, expectations guided and interest rates pushed into negative territory. As those physiological limits are approached, policy loses effectiveness and trade-offs worsen.

An obvious such example is the measures’ impact on the financial system’s profitability, resilience and hence ability to support the economy. We have already discussed the effect of persistently negative policy rates on banks (eg Borio et al (2015)). And such rates, and ultra-low yields more generally, also have a bigger debilitating effect on insurance companies and pension funds, whose liabilities have a longer maturity than their assets (eg EIOPA (2014)). The plight of pension funds is especially telling. It makes much more transparent the need for households to save more for retirement, which could weigh down on consumption (Rajan (2013)),21 and for sponsoring companies to replenish any underfunding, which could weigh down on investment.22

The pension fund example highlights the broader possibility of counterproductive effects on confidence – hardly captured by the formal macroeconomic models used to estimate the effects by extrapolating from more normal times. Take, for instance, forward guidance. Even when it is fully understood and credible, it may have unintended consequences. For, in order to convince markets that interest rates will remain unusually low for unusually long, the central bank may need to paint a rather bleak picture of the outlook. Paradoxically, the more credible the central bank is, the larger this effect is likely to be. Or take negative policy rates. The adoption of such extraordinary measures is unlikely to convey a reassuring message about the state of the economy, especially to those economic agents unfamiliar with the typical economist’s way of thinking – thinking in which real (inflation-adjusted) variables are the only thing that matters for behaviour (Box 1).

21 For instance, a recent survey indicates that only a small percentage of households would spend more if faced with negative rates and a similar percentage would actually spend less. No doubt, here confidence effects are at work too. See Cliffe (2016).

22 The macroeconomic models used nowadays simply assume that lower (real) interest rates always raise consumption by making consumption today more attractive relative to consumption tomorrow (the ”intertemporal substitution effect”). They rule out the possibility of depressing consumption as agents need to save more in order to get a given income in the future (the “income effect”). See Woodford (2003, Chapter 4) for a discussion of how consumption depends on the expected future path of real interest rates in textbook New Keynesian models.

Box 1: Delving into negative interest rates and “money illusion"

It is common in economics to assume that agents’ behaviour is, or at least should be, only a function of real (ie inflation-adjusted) variables. This is modelled, in particular, by assuming that they derive utility exclusively from real variables, such as real incomes, real money balances and the like. It is also common to treat departures from this assumption as “irrational”. For instance, an agent should be indifferent between receiving the same real income regardless of whether in one case prices have risen and in the other fallen. If the agent prefers the outcome with a higher nominal income but the same real income, then he/she is said to suffer from “money illusion” (Fisher (1928)).

Do agents disregard real quantities? Is this necessarily irrational? And what are the implications for negative interest rates? In what follows we will argue that nominal quantities do appear to matter over and above real quantities, that this need not be “irrational” but may reflect at least in part the fundamental role money plays in our economies, and that this has significant implications for the impact that negative nominal rates may have on behaviour.

There is, in fact, a voluminous literature indicating that agents behave as if they had money illusion. The evidence takes a variety of forms. One is surveys designed to tease out money illusion (eg, Shafir et al (1997), Shiller (1997)). Another is controlled experiments in the form of “games” played out within groups (eg Fehr and Tyran (2001), Noussair et al (2008) and references therein) or of a neurological character, monitoring the brain’s reactions to specific stimuli (Weber et al (2009)). Yet another is econometric studies, for instance those that examine the behaviour of asset prices, such as equity (Modigliani and Cohn (1979)), house prices (Brunnermeier and Julliard (2007)) or bonds (Shiller (2015)). In all of these cases, nominal variables, such as nominal interest rates, appear to play a role over and above real ones.

The factors that lie behind this type of behaviour are not fully understood. As often argued, psychological biases or cognitive limitations may well be at work. For instance, the price level is in fact an index computed based on the prices of a basket of goods and services, which are weighted differently depending on the intended use. Coming to grips with that concept is more complicated than dealing with dollars or pounds. But more fundamental factors may also be relevant.

One has to do with the choice of price level. A question to ask is “whose price level?”. In a world in which economic agents differ widely, the general price level need not correspond to anyone’s needs in particular. It is well known that consumer spending patterns differ greatly, depending on age, income and the like, that individual producers care about different prices, and so on. The mental experiments that lie at the heart of money illusion and are embedded in benchmark models ask “what would happen if all prices changed by a certain amount?”. Zero inflation in this sense is not the same as a much more realistic situation in which prices change by varying amounts, including in opposite directions, and they just happen to cancel out in aggregate for a particular index. In this second case, some economic agents gain, others lose. In such circumstances, even if agents do not experience money illusion, they will behave as if they do. Moreover, whenever agents do not have accurate knowledge of the behaviour of all the relevant prices, it is natural for them to rely more heavily on nominal rather than real variables.(1)

Such considerations are likely to be especially relevant in the proximity of price stability, ie when the aggregate price level, as conventionally measured, rises or falls gradually. In fact, in this case it is almost a certainty that some prices will rise and others will fall so that, strictly speaking, there is no general increase (inflation) or decrease (deflation) as envisaged in the standard counterfactual questions asked in the models and in mental experiments.

A second factor has to do with the critical role money plays as a “unit of account”, ie as the unit in which all prices are measured, contracts are struck, and assets and liabilities denominated. Money acts as a standard unit of measurement for all the (relative) prices in the economy, greatly reducing information costs (Brunner and Meltzer (1971)). By quoting all prices relative to the same unit – the “numeraire” – the number of price quotes is dramatically cut, with clear efficiency gains. This greatly simplifies what needs to be known and communicated, much like a standard unit of measurement helps in other scientific fields and walks of life. Moreover, once the same unit is also naturally used to denominate deferred cash flow receipts (assets) in contracts, it makes sense to denominate liabilities in the same unit in order to reduce “exchange rate” risk (Doepker and Schneider (2013)). Much like, say, it makes sense for oil producers whose revenues are denominated in US dollars to incur liabilities in the same currency: by so doing, they hedge their risk. In turn, it is natural for the unit of account and medium of exchange (or settlement) to coincide.

Once contracts as well as assets and liabilities are denominated in nominal terms, then they take on a life of their own in influencing behaviour. The corresponding nominal “inertia” and “rigidities” are, in fact, a reflection of the deep and pervasive role money plays in our fundamentally monetary economies.(2) The strength of the underlying forces helps explain what would otherwise be a puzzling phenomenon, ie the limited reach and fragility of indexation mechanisms (Shiller (1997)).

The analysis also sheds light on why nominal interest rates may be relevant for behaviour quite apart from real interest rates. And it indicates that the instinctive aversion to negative interest rates has deep roots in how our economies work, rather than being just a sign of irrational behaviour. Likewise, zero may not just be a point along a seamless continuum, but have a significance of its own: it marks the difference between receiving something in exchange for the sacrifice of parting with one’s money and paying for the corresponding “privilege”, between encouraging the use of the unit of account and discouraging it.

One possible implication is that negative rates may adversely affect behaviour in ways that would not be understood if one reasoned purely in real terms. Regardless of whether inflation is positive or negative, negative nominal rates are likely to be perceived as a “tax” – a more visible one than that associated with inflation, as conventionally measured. This also raises huge communication challenges for policymakers who resort to them in order to boost inflation. One rationale for adopting low inflation objectives is precisely to reduce the hidden tax inflation levies on the population, as it erodes the purchasing power of income and wealth. With negative interest rates put in place in order to boost inflation, it is as if economic agents were taxed twice. The justification relies on models in which negative real rates would generate better economic outcomes and whose rationale, therefore, needs to be convincingly explained.

A second implication is that the prospect of setting interest rates at any negative level as a means of boosting output would risk backfiring and undermining the functioning of the economy. This option has been proposed, in particular, to address the debt overhang problem (Rogoff (2014)). More generally, it is the logical implication of models in which it is only real variables that influence behaviour. Our analysis points to potentially far-reaching consequences of the prospect of charging a tax of, in principle, an unconstrained size on the holdings of the settlement medium which, in turn, coincides with the unit of account.(3) This is the medium on which the whole economic edifice is built.

(1) Lucas (1972) famously developed a model along these lines, in which agents could only observe a sub-set of prices (from their “information islands”) and hence sought to distinguish general from relative price changes based on the imperfect information available to them. (2) The depressing effect of persistently negative interest rates on bank profitability in the text is just one example of the far-reaching effect of these forces. (3) This argument is distinct from the other implications resulting from the mechanisms needed to implement the measure. One that has attracted much attention is the potential loss of privacy linked to the elimination of cash (eg Cochrane (2014)).

This analysis points to a couple of conclusions.

First, it suggests that, over time, as the power of the measures through domestic channels diminishes, policy may end up de facto relying more on exchange rate depreciation – not necessarily by design, but simply by default. This appears to have happened. It is no coincidence, perhaps, that exchange rates have been increasingly prominent in central bank statements (BIS (2016a)). Globally, however, this has presented problems of its own. Depreciation may result in unwelcome appreciation elsewhere, especially in countries struggling with the similar problems or worried about strong capital inflows as they seek to constrain the build-up of financial booms (eg Borio (2014b), Rey (2013), Bruno and Shin (2012)). Indeed, there is evidence that large-scale asset purchases have affected not just exchange rates, but also capital flows and asset prices in other countries (Table 8). In turn, unwelcome exchange rate appreciation can induce other countries to ease policy in order to fend it off. Hence the frequent references to “currency wars” and “competitive depreciations” (eg Rajan (2014)). This helps explain why monetary policy has looked unusually easy globally, regardless of benchmarks (eg Hofmann and Bogdanova (2012), Taylor (2013), Borio (2016)).

Second, it also suggests that, over time, the balance between the benefits and costs of the measures deteriorates. The effectiveness tends to diminish, especially as the policy room for manoeuvre narrows. And any side-effects tend to grow – on the profitability and resilience of the financial system, on risk-taking in financial markets and on the global balance of policies (eg Borio (2014a), Borio and Disyatat (2014)). Ultimately, this may undermine the credibility of central banks.23

The combination of context and measure-specific characteristics raises serious exit issues. On the one hand, it is compelling for central banks to press further on the accelerator in order to achieve their objectives. Naturally, if inflation remains stubbornly below target, they may hardly see any alternative. Failing to press further could be viewed as signalling the measures’ ineffectiveness and that central banks have given up on their mandates. Markets could then lose confidence in central banks’ ability to deliver and governments could accuse them of undermining their own policies. The increasingly insistent discussion of helicopter money, regardless of its inherent merits, is just the latest example of the incentives at work (Box 2). On the other hand, unless the measures do prove sufficient or other factors come to the rescue, sooner or later their limitations would become fully apparent. The consequences would be similar. The market turbulence in February 2016, seemingly exacerbated in part by a loss of confidence in central banks’ powers, highlights the dilemmas involved (BIS (2016b)). Markets’ dependence on central bank support is hard to shake off (eg El-Erian (2016)). And even short of such loss of confidence and credibility, the narrowing room for policy manoeuvre would make it harder to address the next recession, whenever it arises, unless central banks manage to “reload the gun” ahead of time.

23 There is, indeed, some evidence consistent with this view (see BIS (2016a), Chapter IV). The impact of the measures on output and inflation appears to have declined in some countries (eg Hofmann and Weber (2016)) while that on the exchange rate has not. Similarly, there is evidence that the impact of interest rates on lending weakens as they fall to very low levels and squeeze net interest margins (Borio and Gambacorta (2016)).

The political economy

Unconventional monetary policy measures also raise delicate political economy issues.24 Here, we focus on two – the policies’ perceived impact on inequality and the complications linked to balance sheet measures.

It has long been recognised that monetary policy has an impact on income and, in particular, wealth distribution. The policy’s incidence differs across segments of the population, depending on their sources of income, the amount and structure of their wealth, and their exposure to unemployment. To cite probably the best-known case, the differential impact between debtors and creditors has been analysed extensively by those exploring the political economy of monetary policy.

At the same time, as long as economic performance is satisfactory and monetary policy measures evolve within a narrow range, these issues can easily fade into the background. This is very much what happened pre-crisis, during the period of the so-called Great Moderation, as growth was generally strong and inflation low and stable.

Things have changed post-crisis. On the one hand, the financial crisis has put the spotlight on the growing inequality that had been developing for decades (eg Piketty (2014)). On the other hand, the extreme monetary policy settings have focused attention on central banks’ role, not least given their greater and explicit reliance on boosting asset prices and given the impact of persistent exceptionally low interest rates on savers. Not surprisingly, central banks have been drawn into the debate (eg Yellen (2015), Draghi (2015), Mersch (2014) and Haldane (2014)). Once again, the genie is out of the bottle.

Formal evidence on the impact of monetary policy on inequality is limited. To illustrate the issues involved, we draw on a recent study by Domanski et al (2016), which considers the link between monetary policy and wealth distribution. Through a set of simulations based on household surveys, the study concludes that wealth inequality – as measured by the difference between the wealth of the 80th and 20th percentiles in the distribution – has increased post-crisis. The impact of monetary policy is harder to pin down. But given typical portfolio configurations, it tends to raise inequality by boosting equity prices but may lower it by boosting house prices. France and, especially, Germany are exceptions in this respect, since the 20th percentile holds a smaller proportion of its wealth in both equities and real estate than the 80th percentile. In these cases, inequality increases on both counts.

This illustration highlights the communication challenges central banks face. The direct impact of policy on asset prices, especially equities, is quite visible. And if one focuses on the richest (eg, the top 1 per cent) and poorest segments of the population, it is hard to argue that policy does not raise wealth inequality. The defence has to rely on the argument that policy is sufficiently effective in boosting output and employment – a key source of income and, over time, wealth (eg Bernanke (2015), Mersch (2014)). But, regardless of its merits, it is politically difficult to counter an argument based on immediate observation with one based on a counterfactual.

24 In order to keep the paper manageable, we do not discuss those that arise in the context of credit policies during the management of crises, involving the role of central banks in emergency liquidity provision. Rather, we focus on the deployment of tools in the performance of traditional monetary policy functions.

What Else Can Central Banks Do?

Lambert here: Not exactly “This is fine,” but still…

Yves here (1:15 PM Sunday). I know we have to run this sort of thing occasionally to keep an eye on orthodox thinking, but this piece is based on the loanable funds theory, which was challenged by Keynes and debunked definitively by Kaldor in the 1950s. The key assumption here is that if you make the cost of money cheap enough, people will borrow and invest (or spend). But any competent businessman will tell you they don’t expand their business for the hell of it, they do so if they see an opportunity. So the cost of money can constrain investment, but making it cheap won’t create new markets or more demand. The exception is businesses where the cost of money is one of the biggest business expenses….such as financial speculation and other leveraged investing.

[There follows an article by some of the authors of the Geneva Report and a link to the report itself.]

For more on the impact of interest rates and bond yields on pension funds see here, and for more on the impact of pension funding on wages and the economy see here.

Also, AIUI, the lower down the income scale someone is the higher the percentage of their income that is likely to be spent in consumption, so any policy that tends to overly favour the already asset rich, and is poorly transmitted to financial products used by the average individual, may well operate as a drag on overall consumption when measured relative to overall wealth?

Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

  • Recently Browsing   0 members

    No registered users viewing this page.

  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.