Good and Bad Deflation

by George Hatjoullis

Deflation, it seems, is like the curate’s egg; good in parts. On reflection, this makes about as much sense. Central banks have referred to falling food and oil prices as ‘good’ deflation. This makes some sense if a country is a net importer of both categories. It takes less GDP to pay for the imports (others things being equal) and this is good for the importing countries. However, for a country like the US which is largely self-sufficient in both this makes little sense. Consumers gain but producers lose and the net macroeconomic effect rather ‘depends’. For the global economy it is completely nonsensical unless the earth is importing food and energy from another planet. The idea of good and bad inflation in a global deflationary bias is thus silly.

Central banks clearly think a positive but stable inflation rate is a good thing. Otherwise why is the typical target 2% p.a. and not zero or -2% p.a.? Implicit in this target is the idea that a healthy economy generates a bit of inflation ‘heat’ and to set the target lower might stifle otherwise healthy growth. It also implies that the ‘natural’ tendency for economies is towards inflation. This would explain why CBs are typically doing their best to downplay the risk of a deflation. They don’t believe it is much of risk. They fear that if they are too accommodating they will ignite inflation so they are all keen to ‘normalise’ monetary conditions (i.e. raise interest rates) as soon as possible. The same central bankers ignore the obvious inconsistency in their own words. Janet Yellen, in the speech referenced in my recent blog “Understanding the Great Deflation Part 2: Janet Yellen’s Speech”, states;

“A second reason for the Committee to proceed cautiously in removing policy accommodation relates to asymmetries in the effectiveness of monetary policy in the vicinity of the zero lower bound. In the event that growth in employment and overall activity proves unexpectedly robust and inflation moves significantly above our 2 percent objective, the FOMC can and will raise interest rates as needed to rein in inflation. But if growth was to falter and inflation was to fall yet further, the effective lower bound on nominal interest rates could limit the Committee’s ability to provide the needed degree of accommodation.”

If it is easier to halt inflation that contain deflation why be so dismissive of the risk of deflation? The topic suffers form a tendency to focus on the observed changes in aggregated price indices or survey expectations of such changes. This is not irrelevant but not the whole story and can lead to some misunderstanding. The Bank for International Settlements (BIS) has published an empirical study, The Costs of Deflation: an empirical study ( that concludes the costs of deflation are quite variably and depend upon the underlying causes. In particular, the costs may be associated with asset price deflation and banking distress.

In an effort to gain some clarity I have used the blog space to discuss deflation in the context of portfolio behaviour. In conventional discussions the focus is on how expectations of aggregated price changes may affect the demand for money as an asset. In 1989 I visited Brazil for three weeks and got first hand experience what this means in practice. I made purchases with US Dollar traveller cheques as often as possible and spent any local currency by the end of every day. Locals had inflation indexed bank accounts and never carried cash.They used debit cards. Average cash holding was about half a day and then only if there was no other choice (hard to buy the kids an ice cream with $100 travellers cheques). Evidently central banks believe that if inflation is stable around 2% such considerations do not materially impact behaviour and they are correct.

The great achievement of modern central banking has been to remove the fear of inflation from the demand for money (except perhaps in Germany where collective counselling may be needed to exorcise the ghost of Weimar). This has allowed the positive attributes of money to come into focus. It is credit risk free, it is interest rate risk free and it is equity risk free ( see my previous blog ‘ Keynes and the Misspecified Liquidity Trap’ if this is not obvious). In the absence of (much) inflation, cash in your pocket is a risk free asset. In a deflation it pays to hold cash because you earn a return equal to the rate of deflation. From this simple observation I have suggested that one can characterise the inflation context by looking at attitudes to holding cash. Specifically, the willingness to hold cash is a good barometer of the deflationary bias within the economic system. It is not necessarily ‘causing’ deflation, though it could contribute. It is merely signalling that there is ,or is a risk of, deflation. It is another take on inflation expectations.

Most cash is, of course, not held as notes in the wallet but deposits with banks. Insured deposits are equivalent to cash. However, uninsured deposits are not. Traditionally all deposits were treated as cash equivalent because states operated a too-big-to-fail policy and protected uninsured depositors. This ended with the Cyprus banking crisis and eurozone banking reform. Large depositors are no longer quite sure that their uninsured deposits are de facto insured. This realisation has prompted some profound changes in portfolio behaviour. It also effectively reduced the money supply. Uninsured deposits are no longer ‘money’. They are senior bank loans on which one can draw. The behavioural importance of this has been acknowledged in recent reforms (see my blog ‘Bank Deposit Protection in the Uk’).

The 2008 financial crisis raised awareness of credit risk and increased the demand for risk free assets. Cash, insured deposits and some sovereign debt is the stock of risk free assets. The demand went up (as a proportion of portfolio holdings) and supply contracted. It contracted because uninsured deposits were no longer viewed as risk free. It contracted because banks contracted their balance sheets and unleveraged (and thus deposit growth of all types slowed). This was in part due to prudential regulation. It contracted because CBs, to stimulate money growth bought risk free sovereign debt and removed it from the system. Agents wanted more cash and equivalents and were given less. In effect monetary policy was tightened in the period 2008-20014. The widespread incidence of near zero and in some cases negative bond yields is clear evidence of this. The demand for cash and cash equivalents exceeds supply so investors are willing to pay for the privilege of lending to someone who will pay them back with certainty. Credit risk clearly dominates any fear of inflation.

It is also possible to give property  and stock prices a different spin in this context. Owning a stock or a property does not involve direct credit risk. There is no one to default to you. A London property might have been worth adding to a large portfolio to diversify away from cash in the bank, for rich people. Similarly, stocks paying sustainable yields of 3%+  would look interesting compared to zero to negative bond yields. Not all stocks of course but stocks, on average, and some specific classes of stocks. Credit risk does enter stocks indirectly depending on the nature of the business. Diversified portfolios of individually poor credit risks also proved attractive when properly managed but this tends to be a more specialised activity. Interestingly retail investors have moved into this via Peer-to-Peer lending. The point is not that all credit risk is now avoided but that credit risk became the focus of concern post 2008 and affects the distribution of portfolio risks and the supply of risk free assets.

The Quantitative Easing programmes are consistent with this credit risk hypothesis. They were justified as encouraging investors to move into riskier asset classes by reducing the risk free return. However, it did not work. Investors merely chose to hold risk free assets at the higher price (lower return) until yields on government bonds moved to zero and negative. The QE programme was successful but through a wealth effect. The lower bond yields made all assets more valuable and this had some positive economic impact and partly explains higher equity and property prices. So where are we now?

The Federal Reserve has ceased QE and is soon to raise interest rates. The Bank of England also ceased QE and may also follow the Federal reserve in raising interest rates. This will occur long before inflation is deemed to be a threat. The Bank of Japan and the ECB are still in an aggressive QE and easing mode. There is no evidence at the moment that inflation is a threat and deflation remains the risk. The first signs that deflation has ceased to be a concern among economic agents should appear in financial markets. A greater willingness to borrow on one side and assume credit risk on the other. A fall in demand for risk free assets. The implication of this blog (and the many blogs that precede it) is that CBs should hold off raising interest rates until long dated bond yields rise of their own accord. The counter argument is that this may take a while and encourage excessive risk taking that could destabilise the financial system. In other words they have no faith in their own micro and macro prudential regulation which is explicitly designed to make this impossible. They also have not grasped that currently risk free yields indicate excessive risk taking is not occurring. There is something not quite logical about the CB stance.