Written evidence submitted by Prof. Huw Dixon (MON0053)

1. Introduction.

Monetary policy since 2008 can be described as one combining negative real interest rates (where the policy rate is below the inflation rate) with very low nominal rates. This is sometimes described as financial repression.  In the years during and immediately after the crisis, maintaining a near zero nominal interest rate was justifiable even if it meant that the real rate was negative.  The Bank of England and Treasury were acting together in order to avert the worst case scenario of financial collapse.   By the end of 2014, with inflation around zero and a small but positive real rate, financial repression had almost ended. Then there was the Brexit vote, followed by a mistaken cut in the policy rate. With inflation rising rapidly we are now being plunged back into a period of financial repression.  In this evidence, I will seek to explain in some detail the adverse economic effects of financial repression and what an appropriate monetary policy is for the British economy in 2017.

In summary, I argue that given the economic growth over the last few years, with output and unemployment near their “natural” levels, there is no justification for a monetary policy to impose second round of financial repression.  The Bank of England should raise interest rates to a level equal to their inflation target (2%) plus the equilibrium real interest rate (1-2%), i.e. a policy rate around 3-4%.  This should be done in a gradual manner, as we can see the Fed is currently doing.

Quantitative Easing should be partly reversed as the banking sector run down their reserves held at the Bank of England, to prevent excessive growth in the money supply.

2. Financial repression and monetary Policy.

As a preamble to looking in detail at monetary policy, it is useful to bear in mind some general principles of monetary policy, representing the consensus of monetary theorists.

1. In normal times, monetary policy has no long-run effect on real variables (output, employment and so on). The ability of monetary policy to affect real variables in the short run is greater for unanticipated changes to policy, and greater when there is more nominal rigidity in prices and wages. The main long-run effect of monetary policy is on inflation.

2. Politicians are likely to have an inflationary bias. When politicians run monetary policy, they are likely to be tempted to stimulate output and employment above their natural equilibrium level resulting in excessive inflation.  One solution to this inflationary bias is to delegate monetary policy to a conservative central banker who prioritises keeping inflation on a target set by politicians.  Note that central bank independence is not an end in itself. It is a means to keeping inflation on target. Also, it is not an issue of technical competence: the setting of interest rates can be done perfectly well by the Treasury (as was the case for many decades).  The reason the Bank sets the rate is because it is better able to target inflation than the Treasury since it does not have an inflationary bias.

3. An exception to 1 is when there is a financial crisis resulting in a sharp contraction in the money supply (due to many banks reducing lending at the same time, a systemic crisis).  In such times of crisis, the Central Bank needs to act to stabilise the financial system by acting as a lender of last resort (providing liquidity) and possibly supporting asset prices.

The first principle tells us that we should not expect too much from central banks.  The ability of central banks to influence output and employment is limited to the short run.  The general path of output and employment in the medium to long-term is not something monetary policy can influence. This would have seemed self-evident before the crisis.  During the crisis, the Bank did have a major role to play as indicated by the third principle.  However, the crisis is over and we are returning to normal times where the Bank will have the simpler role of targeting inflation.  The second principle is also important. The Bank of England controls interest rates not because it has some technical superiority to the Treasury, but because being independent it can focus on inflation.    The independence of the Bank in setting interest rates is not a good thing in itself: it is a means to an end, the end being controlling inflation.  Keeping inflation stable and on target itself stabilises output (this has been called the Divine Coincidence).   If the Bank fails to undertake this task, it undermines the justification for its own independence.

In order to describe monetary policy since 2008, we need to introduce three key concepts:

• The real interest rate
• Financial repression
• Secular stagnation.

The first key concept is the Real Interest Rate (RR).  In its simplest form, you can think of this as the difference between the nominal interest rate r set by the Bank of England and the inflation rate[1] , defined using the famous Fisher equation:

The real interest rate is one of the key drivers of the savings decisions of households and investment decisions of firms. Whilst the news reports and and financial markets focus on the nominal interest rate r, for economists what really matters is usually the real interest rate.   The real interest rate adjusts the nominal interest rate for inflation: what matters for savings and investment decisions is the relative price of goods and services today to goods and services in the future.  The consensus view is that the equilibrium level of the real interest rate is a figure around 2%.  In practice, it has varied quite a lot over time. The actual RR we see in the UK is in chosen by the Bank of England: it sets the nominal policy rate given the level of inflation, so can choose the policy rate to be at any value relative to inflation[2].  The long-run real interest rate is the nominal rate over a long period (for example 10 years) and the expected inflation over the same period. This is not directly chosen by the Bank, although it is influenced by the policies of the Bank (the policy of Quantitative easing was aimed precisely at lowering the long-term real rate by lowering the long-run nominal rate).

When the real interest rate is below zero, you have a situation of financial repression (FR).  This name dates from the early 1970s when negative rates were observed in developing economies[3].  If nominal rates are below inflation, then lenders (for example governments, home owners with mortgages) are in effect having their debt liabilities written off in real terms, with lenders (for example government bond holders) receiving a haircut as the real value of their assets falls. The pejorative term “repression” is used because it is generally accepted that the equilibrium real rate will almost never be negative in a free market.  The real rate of interest in a free market is linked to the rate of return on capital, which is viewed as being non-negative (why would someone invest if that reduced output). Now, an exception could be that there is an unexpected large negative productivity shock: the capital you thought was going to be productive turns out not to be. However, even in this case, you can just “throw away” the capital to “jump” to the situation where the marginal product is negative. If capital is not freely disposable, you will have a situation where you just let the capital depreciate to its zero marginal product level.  FR can also happen with positive real rates that are believed to be below the equilibrium.

Secular stagnation is the popular buzz word for the view that real interest rates have been declining in the past three decades.  After the lax monetary policy of the 1970s, when real interest rates were often negative despite high nominal rates, there was a “conquest of inflation”: tough monetary policy with the objective of bringing inflation down and taking a hit on the necessary (temporary) reduction in output and employment. The RR was mostly above 5% from mid-1982 to late 1992. It fell into the range 2-3% from 1994 to 2002, and then to 0-2% for 2003-2008, with negative rates since 2010 (this is RPI, the CPI RR is higher since CPI is usually one per cent or so below RPI inflation).  However, much of the change in RR is due to monetary policy itself.  The excessively high rates in the 1980s were surely policy driven, as were the low rates since 2003.  So, much of the secular stagnation argument is circular: it seeks to explain falling real interest rates by the falls in real rates partly driven by monetary policy.  However, almost no serious studies have argued that the equilibrium real rate is now actually negative. For example, a careful study by the Bank of England found that the equilibrium RR had fallen to perhaps around 1% low - but still positive[4]. 

So, with these three concepts, the real rate of interest, financial repression and secular stagnation in mind, let us look at a graph with inflation (RPI), the real rate of interest and the Banks policy rate from 1950 to 2016, Figure 1.  We can see that there have been two main periods of financial repression, leaving aside the brief period in the early 1950s resulting from the Korean war.  The first was in the mid-1970s: the policy rate was raised (as high as 14% in 1977), but failed to keep up with rapidly increasing inflation.  I think that at that time inflation was a new phenomenon to the generation of economists and politicians in power and the fall in real rates was not so much a conscious attempt at financial repression but rather a bowing to the political pressures of voters with mortgages (which were nearly all variable rate then).  We have to remember that there was a lot of money illusion at the time:  high nominal interest rates led to an apparent increase in the government deficit.  The fact that this was offset by the real debt decreasing was not really fully appreciated until the pioneering work of Threadgold and Taylor published in 1979 by the Bank of England which applied the principles of inflation accounting to public finances.

The next period of potential financial repression is the period under consideration: December 2009 to the present. If we replace RPI with the CPI inflation statistic currently targeted by the Bank of England, we see much the same except for CPI inflation being a little lower, resulting to the onset of financial repression in January 2008 and a period in 2015-16 when the real rate was zero or slightly positive.

Under current monetary policy, the rate is heading rapidly into negative territory once again. We can subdivide the post 2008 round of financial repression into three periods.

Period 1 (2008-2013): Mervyn King. The first period of financial repression lasted from January 2008 to the last quarter of 2013 when real GDP got back to its peak level form 2007, shortly after the start of Mark Carney’s tenure as Governor in July 2013. Interest rates had dropped to 0.5 by April 2009. Sterling devalued by about 20% against both the Euro at the end of 2007 and almost a year later against the US$.  These devaluations set in motion a period of high inflation, which remained over 3% from January 2010 to April 2012, peaking at 5% in late 2011. For over two years, inflation was not only above the 2% target, but also above the 3% level at which the governor of the Bank has to write a letter to the Chancellor explaining why they have failed to keep inflation within the 1-3% target zone.

Period 2 (2013-2016): Inflation remained below 3%  and fell to around 0 by December 2014.  From December 2014 to July 2016 RR was non-negative.  For a period, the financial repression was ended, although real rates remained very low.

Period 3 (July 2016 – present): Post the Brexit vote, the interest rate was cut from 0.5 to 0.25%. The reasons presented for this cut by Mark Carney and the MPC were that this was a response to the Brexit vote. Together, these two factors (the Brexit vote combined with the cut in nominal interest rates)  led to a significant devaluation in sterling, by about 14% from $1.45 to its current value around $1.25 (and a smaller devaluation of about 10% against the Euro). In the period immediately after the Banks interest rate cut, Sterling went even lower.

3. Financial repression: who borrows wins.

One view, put forward by Carmen Reinhardt[5] and Lawrence Summers, is that Financial repression is a valuable tool for governments having to deal with large debt-GDP ratios, such as the UK faces in 2017. The debt GDP ratio is defined as the ratio of nominal debt to nominal GDP. It can be reduced either by increasing nominal GDP and/or reducing the value of nominal debt.    If growth in real GDP is difficult to achieve, the use of FR is possibly less costly than Austerity (raising taxes, cutting expenditure).  In order to understand exactly how FR works, let us first consider how with a perfectly functioning market (fully anticipated) inflation would be neutral, i.e. not affect any real variables at all), including the debt GDP ratio.

Example: fully anticipated inflation is neutral. We will compare two cases: one with zero inflation and the other with 10% inflation.   Turning to the case without inflation, the government issues a bond that promises to pay £104 in a year’s time. The current market value of the bond is £100, implying a real (and nominal) rate of 4%. At the end of the year, the government repays the £100 which has exactly the same value as it did when it was issued (zero inflation) plus the £4. Now, suppose that there is 10% inflation.  The government now will issue a bond promising to pay £114.40, which it sells for £100, representing a nominal interest rate of 14.4%.  From the Fisher equation, the real interest rate is still 4%. The nominal interest rate of 14.4% divides into two parts: £10 to cover the falling purchasing power of money so that after a year the £110 has exactly the same real value as the £100 purchase price today; £4.40 representing the real return of 4% on the terminal value of £110. The government and the bond holder are unaffected by the inflation, since the real rate is unaffected. 

For a 10 year bond, matters are a little more complicated, but essentially the same.  With no inflation, there is a 4% real and nominal rate implying a £4 coupon each year until the bond is due to be repaid. In the last year the government repays the initial loan value in full plus the interest, a total of £104. With 10% inflation, the nominal rate would need to be 14.4% with a real interest rate of 4%.

In Table 1 I illustrate the two cases for a 10 year bond of value 1 issued in year zero (think of £100): the left hand Table 1A representing the zero inflation case, the right Table 1B the 10% inflation case.  The CPI column gives the level of prices. Deflator gives the reciprocal of CPI, the amount you need to deflate nominal values by in that year to adjust for the higher CPI.  Nominal gives household wealth on the assumption that the interest is reinvested. Real gives this adjusted for inflation (the real value of household wealth is the nominal value divided by the deflator).  In Table 1A, the CPI  and hence the deflator are always the same at 1.  Nominal and real are also the same. In Table 1B, with 10% inflation, we see the CPI rising and the deflator falling. Nominal wealth increases to 3.84 (£384), but this is deflated by 0.39 to give 1.48 in real inflation adjusted terms (£384 times 0.39).  The real wealth of the household is thus exactly the same in both cases.  But the real wealth of the household is simply the real value of the bonds it is owed by the government.  Hence the real value of government debt is exactly the same in the two cases.

This seems a bit counterintuitive.  The real value of the initial £100 bond is decreasing over the decade: at the end the £100 pounds paid to the household is worth only £39 in year zero prices. If the government issues no new debt, the real value of its debt will have decreased in real terms to just £39.  Yes, this is indeed true.  However, the key to understanding public finances under inflation is that if you issue no new debt, the government is on an accelerated debt repayment schedule. The government is paying the household £14.4 per year:  it is paying off its real debt each year and not just the real interest.  Just as we think of the household “reinvesting” its interest payments, we should think of the government issuing new debt to keep the real value of its outstanding debt constant.    In order to get this neutrality of inflation, the government needs to issue new nominal debt so that the real value of its debt remains constant. Thus, at the end of year 1, it pays the household £14.4 and issues new 9 year bonds of £10 with the same nominal interest rate. 

The example of the neutrality of fully anticipated inflation needs some caveats. The main one I want to highlight here is that in practice the tax system does not take inflation into account: nominal interest payments are taxed the same irrespective of inflation.  Hence, the household would be worse off with inflation, because the effective tax rate in real terms would increase.  There are other caveats (perfect capital markets amongst others).  The example is just to show that on its own, inflation does not need to redistribute income and wealth between borrowers and lenders. It need not improve public finances.

However, with financial repression matters are rather different. Let us replicate the two cases with a nominal rate of zero imposed by the Bank.   In the zero inflation case (Table 2A), the household is worse off: it has not been getting its compound interest rate of 4% over 10 years.  Household wealth is 68% what it would have been with the market real rate of interest. The government is better off: it has not had to pay the interest.  The real value of its debt has remained constant.  The money saved in interest payments could have been spent on current expenditure on health and education, or paying down the debt.

With 10% inflation, as we can see from Table 2B matters are very different.  Every year, the value of what it owes goes down by 10% in real terms.  If it does not issue any new debt, at the end of 10 years, the real value of the £100 has fallen to £39 (valued at period 0 prices).  The household is very much worse off: not only has it lost its compound interest rate, it has also had the real value of the outstanding loan eroded.  Compared to the free market equilibrium outcome (Table 1B), the household wealth has been reduced to 26% of what it would have been (£100 as opposed to £384). 

The choice of 10% inflation was simply to make the example easy to understand. What if we think of something more realistic?  This is shown in Table 3 which can be directly compared to Tables 1b and 2B.  The reduction in the value of the £100 is far less dramatic: it has fallen to just £82 in period prices. However, compared to the free market case, the household has still lost out significantly.   Even if we assume a more modest real rate of 2%, the compound interest over 10 years would be 22%.  So, in constant period zero prices the household would have £122 at the end of 10 years, rather than the £82 with financial repression.  The household is about 30% worse off in real terms.

As these examples make clear, one of the main losers from FR are savers.  We are all savers in terms of our pensions. When we start work, we start saving through our pension schemes.  We save for our pensions over a long period, the forty or fifty years of our working life[6].  In order to provide a decent retirement, pensions rely on the magic of compound interest: small amounts invested early on accumulate over the decades.  Since 2008, the magic of compound interest rates has been switched off by FR and put into reverse. £100 invested in 2008 is now worth less in 2017 than it was in 2008 even if the interest had been reinvested.  This affects all of us who are in pension schemes, whatever our age.  The pension crisis is usually looked at from the point of view of those actually retiring.  Despite having had the magic of compound interest working for them and building up large pension pots, if they have retired since 2008 they have little to show for it. Annuity rates have tumbled since 2008. Pension pots have seen the income stream they can generate fall by over a third. In 2008, £100,000 could buy a single person an annuity of almost £8,000; today the figure is around £5,000.   Defined benefit pension schemes are under great pressure: the amounts needed to honour existing commitments have gone up considerably, leading to a rush to the exit door by many companies wishing to close existing schemes or attempts to cut back the benefits.   

The effects of financial repression are quite far ranging.  Borrowers will gain, lenders lose out.  Thus although most working households are made worse off because of the cumulative effect on their pensions, those that have mortgages have a compensating gain.  It is now possible to take out a fixed rate mortgage at just under two per cent, which means that with inflation over 2% the borrower is being paid to borrow the money (the mortgage plus interest will be worth less than what was borrowed after inflation).  However, since we all save for pensions, I believe that most households will be net losers from FR.   Only those who are net borrowers have gained.

Other effects of low interest rates include the boom in car finance in the form on PCPs (personal contract purchase): the car buyer pays a fixed monthly sum for three years (or some other fixed period) and then has the option of either returning the car or paying a lump-sum to buy the car outright.  The monthly payment will cover interest payments, depreciation of the car and the various commissions involved in setting up the loan.  With interest rates of say 1%, the interest element has become quite small. For example, on a £40,000 car the annual interest is just £400 or less than £35 per month.  This means that Ling’s cars is able to offer a BMW series 5, 2 litre diesel estate for just under £600 per month over 4 years.  A Citroen C1 can be obtained for less than £150 per month. If interest rates were at 5%, then the interest element of the BMW loan would be about £170 per month, meaning Ling’s cars would need to charge £735 per month for the same vehicle. Not surprisingly, these PCP deals are very popular in a low interest environment and have led to sales of more cars and better cars. The volume of PCPs was over £31bn in 2016. Some have likened this boom to the “new sub-prime” crisis, as loans are being made to people with worse and worse credit ratings. In my own opinion, it is something of an exaggeration to call this a looming crisis: at worst the car companies underwriting much of this might take a hit, along with financial institutions holding the assets based on PCP loans.

The most important effect of low interest rates is to boost asset prices. By definition, a low interest rate means a high value of bonds. For a one year bond promising to pay B with an annual interest rate of r the current market value V is:

If r=0, then V=B;  as r increases, the current market value of the bond goes down.  A hundred pounds in a years’ time is worth what you have to invest now to get the £100 in five years.  High interest rates mean you have to invest less now to get the same £100. However, this inverse relationship also holds for the stock market and more importantly for securitisation.

Securitisation takes the form of selling the rights to future streams of income (for example the stream of repayments from mortgages and car loans).  These flows of income are converted into a current cash value (its present value).  Now, the calculations for this conversion depend on risk factors and so on. But, in all securitisations, the interest rate is central. If we take the case of a near certain flow of repayments (prime mortgages, for example) which lasts forever, then the market value of that unending stream of B per year is[7]:

Now, if nominal interest rates are at 1%, with r=0.01, we can see that the present value V of B pounds per year is 100 times B. With 5% interest rates, r=0.05 it is just 20 times B.  The incentive to securitize is thus much greater when r is very low.  Take leasehold ground rents.  A ground rent on a 999 year lease might be for example £400 per annum.  With nominal interest rates of 1%, this has a securitization value of £40,000.  House developers can sell a house as leasehold, and then sell on the freehold (in effect securitizing the flow of ground rents). No wonder that house developers in recent years have been taking advantage of this: you can sell a house for £300,000 for example, and get an additional £40,000 by securitizing the freehold. The incentive to do this is much less when interest rates were at the normal levels. At 5% our securitised £400 ground rent will sell for only £8,000.

The value of a stock is basically the securitized value of future income streams (dividends and capital gains).  Of course, these income streams are highly volatile, but exactly the same principle holds. Other things being equal, low interest rates will tend to make share prices higher.    This was the essential mechanism behind the infamous “Greenspan put”: when something adversely affected the US stock market, Alan Greenspan would intervene to cut interest rates to help keep the market value of stocks high.

So, we come to another big winner from low nominal interest rates: hedge funds, investment banks and wealth management funds. These gain because the assets they manage will have a higher value and hence the commission will be higher (commission usually includes a percentage of the asset value being managed).  However, these same institutions are often the creators of asset backed securities (ABS).  Again low interest rates mean ABSs have higher value when they are created, so that again more commission is charged at the point where the asset is created.  Low interest rates create an environment which is highly conducive to the creation of ABS which increases the commission income of investment banks.  For example, the PCPs for purchasing cars have become an important source for ABS. In addition to all of this, Hedge funds and investment banks make a further gain: the Bank of England policy of low interest rates and QE means there is plentiful supply of cheap liquidity, so that it is easier and cheaper to leverage their investments. 

Whilst hedge funds and Investment banks gain from the low interest rates imposed under FR, ordinary retail banks do not prosper.  An important part of retail banks is the profit they make from current accounts.  These accounts typically do not pay interest, so that the bank can lend on the basis of this “free” money.  Its profit margin is then driven by general interest rates, which determine the rate it can lend at.  Low rates mean that retail banking tends to have low profits under FR.  

Having looked at the winners and losers of FR, we can maybe look a bit more closely at the implications for government finance and debt.  Is financial repression a useful tool for reducing debt-GDP ratios as some have suggested?  I would argue not in the case of the UK.  FR imposes distortions and redistributes income and wealth in undesirable ways.  There need to be clear benefits to offset these. In order to see FR in action, we can look at how it impacted the UK government finances in the Mervyn King years (period 1 above), when inflation peaked at 5%.  I will take the example from a talk I gave to the students of the UCL Finance and economics society in 2011 (although the figures have been updated)[8].

Here we see for each year the inflation rate, the debt-GDP ratio and the deficit-GDP ratio.  These are taken straight from the IMF. The next column is the “inflation adjustment” made for financial repression: it is the effect of inflation on reducing the debt-GDP ratio.  Under FR, this is approximately given by multiplying the inflation rate by the Debt-GDP ratio. The last column gives the deficit-GDP ratio adjusted for financial repression: we subtract of the fall in the value of outstanding debt as a result of holding interest rates low.  This is an illustrative figure to show how large the effect of FR can be: as we show below, Quantitative Easing (QE) greatly reduces the figure. We can see that in the year 2011 (when I gave the talk at UCL), the figure was quite large, inflation at 4.6% implied that the adjusted deficit -GDP ratio was almost half the unadjusted figure.   What does this mean? Well, the headline deficit-GDP ratio is driven by the gap between government spending and its revenue. Our adjusted deficit figure takes into account that under financial repression, the government is paying off some of its debt in that year, and subtracting it from the headline deficit.  We can see that this FR effect is quite significant: at the end of the four years 2010-2013, the debt-GDP ratio in 2013 was possibly over 10% lower than it would have been without FR.

How might the figures stack up today, with inflation heading to 3% or more in 2017 with interest rates held at 0.25% as the Bank seems to intend.   The IMF figures I used to illustrate the UK experience are gross-debt.  The ONS headline figures for debt-GDP are net, subtracting off foreign exchange reserves and some other items. However, what the ONS does not do is consolidate the Bank of England into the public sector accounts: under QE the Bank currently holds £435bn (22% of GDP) of the government debt: with gross debt-GDP of 89% in 2016, the gross debt net of QE is 68%. If we take the ONS definition of net debt –GDP (currently at 80%), netting out QE takes it down to 59%.  FR does not operate on that part of the debt held by the Bank under QE, since the public sector is both the borrower and the lender.   This argument would also hold for the illustration I gave in Table 4: I assumed that FR operated on all of the government gross debt, whereas QE had already kicked in by 2009.  If we consolidate the public sector, then the impact of FE on public finances is greatly diminished due to QE.  Using the same method as before, even inflation of 4% would only reduce the FR adjusted net deficit-GDP ratio by 2.4%.  The problem with FR is that it has very large effects when it is imposed and the gains in terms of lower debt-GDP ratios will only show up in the longer term as it is applied over a prolonged period.  In the post-World-War 2 period, the main factor driving down the debt-GDP ratio in the UK was economic growth.  The period of FR in the mid-1970s did play a role, but it was short lived.

I think that Reinhardt is simply wrong to view FR as a useful tool for current UK fiscal policy. If we net out QE the net debt GDP ratio might be as low as 60%. The huge costs in terms of reducing the future pensions of British households are for me far more important than the gains in terms of improving public finances.

4. Quantitative Easing:  zombie debt.

QE was introduced in March 2009, as a method of reducing long-term interest rates.  The official policy “Bank Rate” is a short term interest rate.  This was reduced to 0.5 in 2008.  QE was seen as a way of forcing down longer run interest rates: by purchasing bonds it raised their price and hence lowers interest rates. This was often reported in the press as “printing money”: the bonds were purchased by money created by the Bank[9].  This could have led to a big increase in the money supply and hence inflation.  However, it did not. The reason for this is that the financial crisis had led to a collapse in commercial bank lending, lending being the main driver behind creating money (we are talking about M4, the broad money supply that includes bank deposit accounts).  The “money multiplier” is the process by which the banking system creates money:  deposits are made by firms and households into banks, which are then lent out again by banks (subject to some reserves held by banks to cover depositor’s cash needs). The banking system takes a relatively small amount of cash (the narrow definition of money M0) and turns it into the broad money supply.  In 2006, the size of the money supply M4 was £1.5 trillion, equal to the size of nominal GDP.   When banks stopped lending (both to each other and to firms and households) during the financial crisis this created a gap into which the government could step, buying up its own debt without causing an explosion in M4.

First let us see what happened with QE.  This is shown in Figure 3.  The stock of bonds[10] is given by the brown line.  As we can see, at various dates the MPC decides on an increase in QE leading to a the second in late 2011 early 2012. Now, something quite amazing happened.  The money that paid for the bond purchases largely found its way back to the Bank of England.  The blue line is “Balances held at the Bank of England” by banks[11].  As was can see, with QE these reserves deposited at the Bank of England have mushroomed to a level only a little smaller than the level of QE. In 2010-11, they were 50-60% of QE, rising to over 86% by the end of 2016. Rather than lend out the money to firms and households, banks prefer to deposit the money with the Bank of England where it is safe and earns 0.5% (or currently 0.25%). The money held in these reserves by commercial banks is taken out of the money creation process: unlike loans to firms and households, a “loan” (deposit) at the Bank of England is not then recycled as a loan to firms and households.

The overall behaviour of the broad money supply is depicted in Figure 4. The left hand axis gives magnitudes in millions. The dotted line is nominal GDP, which simply serves as a reference value. On the bottom right corner you can see the same amounts for QE and reserves as were shown in Figure 3.   We show the M4 quantity as the grey line, and the blue line is M4 net of QE.

As we can see, although there was a big increase in M4 during the height of the crisis in 2009 up to a peak in the first quarter of 2011, after that the level of the money supply fell and remained more or less constant until 2016 during which it has resumed growth.  If we net out the money generated by QE, we can see that since the crisis the money supply has been fairly level at around its pre-crisis value.  

The level and growth of the money supply have not been targeted by the Bank of England since the monetarist experiment in the 1980s, but are an essential statistic.  The money supply captures the creation of money via loans within the banking system and also injections from the public sector.  The money supply is increased by the public sector when the fiscal deficit is not funded by issuing new bonds (a process called monetization), and more recently by QE.   If the money supply increases at a sustained rate which is significantly greater than the growth of nominal GDP, this can lead to future inflation, both as reflected in CPI (the rice of goods and services) and also asset prices.  Even if CPI inflation is not generated, asset price inflation can lead to bubbles (for example in housing or the stock market) which can, as we have learned, have adverse consequences.

This means that QE may be storing up problems for the future. If banks decide to withdraw their deposits at the Bank of England and lend the money to households and firms, then without policy changes there could be an explosion in the money supply.  This could happen very quickly.  Thankfully, QE can provide its own solution to this problem.  The Bank of England can sell back the bonds it owns in order to maintain a target growth in the money supply.  For example, since inflation is 2% and the real economy is growing at around 2%, nominal GDP should grow at around 4% per annum.   This would indicate that the annual money supply growth should be kept (for example) to no more than 5% - nominal GDP growth plus 1% (as is clear from Figure 4, there is a long-run trend for M4 growth to exceed nominal GDP growth, reflecting natural variations in the “velocity of circulation”).  The exact “upper bound” or ceiling for monetary growth would need to be decided by the MPC, the figure of 5% is simply a ball park illustration. The rate at which QE is unwound will then depend on the behaviour of the private sector – essentially bank lending[12].  Note that QE may not be completely reversed.  It may be that due to changes in regulations and lessons learned, the money multiplier has permanently decreased.  In that case, part of the QE assets may remain on the Banks balance sheet for a long time (until they naturally mature).  They will become a zombie debt buried in the deep recesses of the Bank and largely forgotten.

5. Monetary Policy 2016 onwards: back to the future.

Monetary policy since 2008 has been characterised by Financial Repression: a negative or very low real interest rate.  FR had been reducing prior to mid-2016 as inflation fell to zero. Two things have changed since then: first, the Brexit vote and then the policy response of the MPC in cutting the nominal rate to 0.25% and expanding QE by £60bn to its current level of £425. The result of these two events was a dramatic fall in the value of sterling, which is part of the explanation for the increase in inflation that got underway later that year.

I for one see the MPC’s response as a mistake, a “Greenspan put” just as wrong as the post 9/11 cut by the Fed.   Firstly, Brexit itself did not require any immediate action from monetary policy.  I have been openly critical of Brexit and believe that it will reduce British economic growth.  However, this will be a long term factor, spread out over many years as long term investment decisions by firms respond.  Even the most critical Remainers would predict a level of output maybe 5% lower in ten years’ time, indicating a slower growth rate spread over several years (half a per cent lower spread over 10 years). However, maybe the Brexiteers were correct and the new opportunities for trade will outweigh the lost market access to Europe. Only time will tell. Brexit itself required no immediate response from the MPC.  Also, whatever the effect of Brexit will turn out to be, this is a real factor in the economy and monetary policy can have no lasting effect to alter it (see principle 1 above).

Monetary policy needs to be forward looking, but primarily as far as inflation is concerned. As Mrevyn King would often say, the interest rate policy now needs to be looking two years ahead in terms of inflation.  Brexit did imply that inflation was going to increase. Thus the most obvious policy response was to increase interest rates.   A simple increase of 25 basis points would have signalled to markets that the Bank was going to be serious about trying to keep inflation to target.  Instead the MPC did the exact opposite: it sent the signal that (nominal) interest rates were not going to be raised for the foreseeable future.   They were happy with imposing another period of repression on the British economy, with all of its adverse effects.

Am I ignoring the weak state of the British economy? No, I think not.  Most economic forecasters believe that British output is close to its “equilibrium”.  The consensus is that since 2015, the output gap is quite small: the actual level of output is less than 1% below potential.  Furthermore, growth has been reasonable: 2014 3.3%, 2015 1.8%, 2016 2.1%.  These growth figures need to be seen in the context of estimated growth rate of potential output being less than pre-2007, largely because of smaller growth in the labour force. Both the IFS and OBR point to a growth in potential output in the next few years of around  2.1%, so that actual growth will be a little above 2% .  Brexit may dampen growth prospects, but the effect will be gradual and spread over many years.  Unemployment at its current level of around 5% is indicative of being around the natural rate.

What about private sector debt? If we look at private household debt relative to GDP, we can see that its value in 2016 is 88%, back to what it was in 2005 (from a peak of 97% in 2010).  This is higher than it was in the 1990s (when it was around 60%).  However, with real interest rates negative, FR is itself encouraging people to borrow and save less.  If we look beyond households, we find that total (non-financial sector) private debt is less than 155% - again, back to its 2005 level[13].   The level of private sector debt does not seem to me to be extraordinary and certainly does not provide the justification for further FR.

What about the rise in inequality? This is definitely one of the main challenges for current economic policy.  However, it is not really a concern of monetary policy.  FR itself has been an important driver in generating inequality.  When FR takes the form of negative real rates and very low nominal rates (as in has since 2008), this leads to high asset prices. High asset prices tend to increase the wealth and also the income of people who own the assets.    Raising real and nominal interest rates will tend to reduce asset prices, reversing to some extent the effects of FR since 2008 on increasing income inequality.

Overall, there are possible arguments that the long-run real interest rate may have fallen:  as we saw earlier, the Banks own researchers look at several factors and come to a figure of around 1%.  With an inflation target of 2%, this implies that nominal interest rates should be at around 3%.  How should current policy get to this level?  We can see the Fed doing this now under the wise leadership of economist and chair Janet Yellen. I would recommend a sequence of 25-50 basis point rises to take us to 3%.  Because of the “mistaken” reduction to 0.25% in 2016, the first step should be 50 basis points to 0.75%.  Assuming no unexpected changes in economic circumstances, we should be at a 3% policy rate with 2% inflation by early 2018.

The psychological effects of a return to normality would be hugely beneficial.  Keeping unnaturally low interest rates (both nominal and real) creates an artificial world and an impression that the economy is somehow “sick” and in need of nursing by the MPC.   Also, households might be able to see how they can realistically save for their pensions and have a reasonable income at retirement.  Being able to save for the future will bring a great relief to households and raise their well-being.

Some have argued for an increase in the target level of inflation to 3%.  This is not a good idea and has been wisely rejected by Janet Yellen in the US.  Inflation has costs: it generates arbitrary price dispersion which makes the price mechanism less effective from a societal point of view: prices reflect costs less systematically.  Inflation also undermines the role of money as a store of value and medium of exchange. As Milton Friedman argued, with flexible prices the optimal rate of inflation is negative so that people are able to hold as much money as possible. When prices are rigid, the argument is less valid. A positive but small inflation rate allows the economy to respond well even when prices are rigid (and in particular rigid downwards).  Lastly, monetary policy becomes more difficult when you have higher trend inflation: it becomes more and more difficult to get a policy that keeps inflationary expectations tied down[14].  The target of 2% seems to have worked well over the years and given the costs of higher inflation it would not be wise to raise the target. 

The only argument for a higher inflation target is that we are less likely to run into the zero lower bound for nominal interest rates.  However, as we have seen, it is the real interest rate that matters and this can be negative if required by having positive inflation even when interest rates are zero (this would be a rare event). Negative inflation might cause problems, but negative inflation has almost never occurred in Britain in the last half century.   It is costly to have higher inflation at 3% for most of the time, when the problem of the zero lower bound on nominal interest rates is a rare event and can be solved when it happens with a short bout of higher inflation to get the real rate down.

6. Conclusions.

  Overall, we can characterise monetary policy in the years 2008 to present as the years of financial repression with negative real interest rates.  This was not inevitable: with nominal interest rates at 0.5%, the degree of financial repression might have been small if inflation had remained low.  However, inflation remained over 3% for a long period, peaking at 5%. The Bank of England perhaps had little choice in the aftermath of the crisis. Its priority in the years 2008-13 was aiding recovery.  However, the economy in 2017 is very different and in terms of growth, unemployment and the level of output we are well past the crisis.  Matters are still far from perfect: low productivity growth, rising income inequality and falling real wages for most of the population are not good. However, it is not realistic to think that monetary policy can solve the UK’s current economic woes. I think that in terms of monetary policy we are out of the crisis period and policy needs to return to its normal focus on inflation, the role the MPC was created to fulfil in return for its independence from the government.  The real problems of the UK economy need real government policy in the form of stimulating productivity, innovation and dealing with the pension crisis partly generated by the decade of financial repression imposed by monetary policy.  Monetary policy needs only to contribute to this by keeping inflation to its target of 2% and having an appropriate real interest rate.  With interest rates at sensible levels and inflation under control, the UK economy will be much better placed to face the future and any challenges it may bring.

April 2017

Appendix:

A1: Real Rate.

Here I will briefly outline different ways of thinking about the real rate of interest. The headline inflation rate is a backward looking measure of how much prices have risen in the last twelve months.  The real rate is a forward looking concept: it compares prices over the period of the loan. For example, a one year loan would look at the expected inflation over the coming twelve months. This will generally be different to the headline inflation rate, as inflation may be expected to increase or decrease on the coming year.  The Banks policy rate is a short term interest rate.  To get the expected real interest rate, it would be best to use the expected inflation rate over the next month or quarter (annualised so as to be comparable).

I use the headline inflation rate for my illustrations.  The main difference between the headline and expected inflation comes down to timing. For example, take the recent increase in inflation after the cut in interest rates to 0.25%.  If you use headline inflation, the real rate did not fall until inflation started to increase.  However, expectations of future inflation increased almost immediately.  Hence there is a few months difference between the timing of real interest rate changes depending on whether you use headline or expected inflation.

This evidence does not deal with the month by month details of policy, but the overall character of monetary policy. Hence I used the simpler headline to show the general outline of policy.

A2: Can there be an equilibrium real rate of interest that is negative?

In general, the answer is no. If we think of a steady-state where consumption is constant, then because the household discounts the future, the real rate of interest has to be strictly positive. The real interest rate corresponds to the marginal product of capital. In a representative agent economy, a negative real interest rate is possible as a transitory phenomenon, and would correspond to a decumulation of capital indicating that the capital stock was too large and hence the household would seek to reduce it by maintaining a high level of consumption with possible dis-saving (consumption in excess of income). A negative real interest rate would be a temporary phenomenon on the path to steady-state: along the path, as the capital stock is reduced, the real interest would get back into the positive territory. In the Ramsey model, a very large initial capital stock yielding a negative marginal product would result in a high level of consumption which fell over time, with dis-investment.

Matters are different in “OLG models”, which are not in general dynamically efficient. In a simple exchange economy without production, it is possible to get a negative real interest rate in equilibrium. If current consumption is cheaper than future consumption, you need to give up more now to get less in the future. The key assumption needed is that there is no storage or capital: one generation trades with another. Eggertson et al (2017) have a model where people live for three periods: they have endowments middle age and when old: they borrow when young. Assuming that the endowment is largest in middle-age, consumption smoothing indicates that they will borrow when young, and save when middle aged to augment their retirement consumption (the old consume everything they have). At any time, there are all three generations living together. The middle aged at time t can only save for when they are old in t+1 by lending to the young at time t, who will repay the old at t+1. Here, the young demand loans (consumption) from the middle aged; the middle aged lend to them so that next period they get paid back and their old aged consumption is increased. The real interest rate here can be positive or negative, depending on the balance between the supply and demand for loans.

In order to link this monetary policy, we need to introduce nominal wages, nominal prices and a nominal interest rate. Making various assumptions, Eggertsson et al show that there can exist “a unique, locally determinate secular stagnation equilibrium” (Proposition 1, page 21. Figures 4 just above the proposition make the essential role of deflation clear). However, the secular stagnation equilibrium must have deflation: negative inflation. If inflation is positive, then there will be full employment. This is because the mechanism reducing output is the increase in real wages. So, in a secular stagnation equilibrium, the nominal interest rate is at the ZLB (zero lower bound), output is below full employment, inflation is negative and real wages above their full employment level (due to downward rigidity). The actual real interest rate is positive (equal to minus the deflation rate): it is a hypothetical real rate that is negative (the real rate that would restore full employment). The ZLB does not lead to an equilibrium negative real interest rate: it prevents the real interest rate from becoming negative when inflation turns negative.

Have we observed negative inflation? In the UK and the US just the occasional month in 2016 and (depending on whether you use CPI or RPI) perhaps for a month or two at the height of the crisis. Japan has had more disinflation since the late 90s (disinflation “peaked” at just over -2% in 2009). The Eurozone is a mixed bag: the aggregate inflation rate has mainly been strictly positive with a few exceptions as in the UK and US. For individual countries the story is more heterogeneous. So, if we look at the major economies, there is no evidence of sustained disinflation that might give rise to the high real rates required for Eggertsson Stagnation. In fact we find the exact opposite. The ZLB is combined not with negative inflation, but positive inflation. Rather than positive real rates, we find real rates are negative.

This brings us to the most important an obscure part of the paper: section 8, the model with over 100 equations. Here there are lots of generations and capital is introduced. The key equations are buried in the appendix: A81 and A82. The marginal productivity for capital A81 is the usual: the marginal product of capital will be strictly positive. Then there is A82. This is a little different: there is a price of capital goods term. Greg Thwaites has developed a model of falling real interest rates driven (in part) by the falling price of investment goods. There has been a downward trend in the prices of investment goods (relative to consumption goods), which means that savings leads to more investment (but possibly lower investment expenditure). This can drive down the marginal product of capital. However, in Thwaites model the real interest rate may be low, but is always positive. So what is it in the Eggesrtsson model that can give you their figure 7: secular stagnation with strictly positive inflation (recall, this was impossible in the world of proposition1). I must admit, that I have read the paper and am none the wiser about how this might be possible. The paper just presents a calibration and reports that this is what happens. Unlike the world of proposition 1 there is no clear story or intuition.

I do not doubt that with sufficient inventiveness a model with equilibrium negative real interest rates can be constructed. But it would not be a basis for monetary policy. Monetary policy needs to be based on robust models that have passed the test of time, not on exotica. I will continue to believe that real interest rates should be strictly positive in equilibrium.

Gauti B. Eggertsson, Neil R. Mehrotra, Jacob A. Robbins.  A Model of Secular Stagnation: Theory and Quantitative Evlauation, NBER Working Paper 23093 (2017).

Greg Thwaites , Why are real interest rates so low? Secular stagnation and the relative price of investment goods,. Bank of England working paper 564 (2015).