Month: February 2015

All about risk and return

Risk and expected returns: The inevitable trade off and how to improve it to the advantage of the SA economy.

Equities are more risky than fixed interest (bonds) and bonds are more risky than cash. Hence equities must be expected to return more than bonds and cash, as compensation for investors willing to bear the extra risk associated with shares. The risk we bear when owning different assets is that we cannot be sure what they will be worth in the future when we might be forced to cash in – the next day, month, year or even, in the case of active traders, in twenty minutes’ time. Hence the more risk, the lower will be the prices attached to assets so that expected returns improve.

Past performance of SA assets provides very strong support for the theory that more risk is accompanied by higher returns. Shares in general have returned significantly more than bonds or cash (in the form of capital gains and dividends or interest received over successive 12 month periods) for SA wealth owners since 2000. But these higher returns have come with significantly more risk, as measured by the movements around the average 12 month returns, calculated monthly. The JSE All Share Index returned an average 16.12% over this period, with a Standard Deviation (SD) about this average of 17.8% p.a. In the worst month for shareholders over this period, February 2009, the 12 month returns on the JSE were a negative 43% while in April 2006 shareholders were up 54% on a year before. The bond market returned an average 11.3% over the same period, with a much lower SD of 6.7% and a worst month, April 2014, when the All Bond Index (ALBI) returned a negative 3.1% over the previous 12 months. The best month for the ALBI was the 27% annual return realized in June 2001. Cash on average returned approximately 8.2% p.a between January 2000 and 2015 with a SD of 2% p.a.

Since inflation averaged 5.9% p.a over the period, all asset classes have provided very good real returns, higher on average than could have realistically been expected back then and more than could be expected over the next 10 years. Such excellent returns, if they are to be repeated, would have to be accompanied by excellent management of the capital invested by SA listed companies and a lower SA risk premium demanded by foreign investors. It was this potent mixture of good management and less risk priced into the JSE share and bond markets that delivered such excellent past performance.

The more the cash in value of any asset varies from day to day, the more uncertainty about their cash in value and so the more risky this asset. In the figure below we show the daily percentage move in the JSE All Share Index and the S&P 500 Index. The price of any individual share included in this index is likely to be more variable in both directions than that of the average share represented by the Index.

This day to day volatility can be measured as a rolling 30 day average SD of these price moves. It will also be reflected in the cost of an option on the Index. The more volatile the market is expected to become, the more expensive it will be for investors to insure against such volatility by buying or selling an option to buy or sell the Index at an agreed, predetermined value. The cost of such an option on the S&P 500 is indicated by the Volatility Index (the VIX) traded in Chicago sometimes described as the fear index. The higher the value of the VIX, higher the cost of an option on the market and the more the fear. The VIX may be regarded as a forward looking measure of expected volatility and the rolling SD as a record of past volatility. Yet past volatility appears to strongly influence expected volatility as we show in the figures below. We also include in the figure the rolling SD of the euro/US dollar exchange rate. It should be noted that volatility appears to revert back to some long term average of about 12. It spiked up dramatically during the global financial crisis in 2008. It also spiked during the various phases of the European debt crisis of 2010 and 2011. Volatility in the share markets now appears as close to average. The VIX had a value of about 14 yesterday. See the figure below.

We provide a close up of recent daily volatility in a further figure below. Volatility moved higher in late 2014 and early 2015. In the share market, volatility has moved back to the long run average very recently while volatility in the key foreign exchange market, the euro/US dollar rate, volatility has moved in the opposite, higher direction.

The relationship between volatility or risk of holding a share and its price, that is the return realised for the owner, is almost perfectly negatively correlated. So if volatility rises, share prices will move in the opposite direction in a highly predictable way. The figures below for the S&P 500 illustrate this. The correlation between daily percentage changes in the VIX and the S&P 500 is a negative 0.84 for daily closing prices since January 2014. The negative relationship between the recent decline in volatility on the value of the S&P 500 Index (volatility up, share prices down), is also illustrated.

Correlation does not however mean causation. The cause of share price or exchange rate volatility is in the degrees of uncertainty felt by investors about what the future may hold for the economy and for the companies and currency traders that are influenced by by these economic developments. If the world could be predicted with certainty, there would be no reason for prices to change; no reason for investors to change their mind about future prospects and so to force prices lower or higher to reflect their less or more optimism about future prospects. For every buyer who must think prices will rise in the future to provide attractive returns, a profit seeker willing to bid up an asset price, there will be a seller (a profit taker) who thinks the prices or the market in general will move in the opposite direction. The more uncertain the future appears, the more they will tend to disagree and the more prices will move widely in both directions from day to day to reflect their differences of opinion until something like greater calm in the markets resumes. This degree of movement in both directions, up then down, down then up, is what makes for volatility. The random price walk that always characterises asset price movements over time s can become abecome a wider or narrower path as prices fluctuate one with wider or lesser changes in both directions over time, as has been illustrated above.

ByG greater calm in the market place is characterized by smaller swings in prices from day to day or even within a trading day resulting in smaller moves, implies one associates with a higher degree of consensus about the state of the world and so the less reason incentive for market participants to push prices more sharply in one and the other direction, when markets are highly liquid and attract many buyers and sellersMore volatile markets, that is wider price swings in both directions, imply a higher degree of dissonance about future prospects.. Changes in market prices Volatility reflects essential disagreement between market participants about the state of the world and the prospects for companies. And prices fall as volatility rises in order to provide investors in general with higher expected returns to compensate for these greater perceived risks.

Thus it might help market participants trying to time correctly their entry into or exit from the market to ask a different question – not where the market is going, but rather where will volatility be going, that is will the world become more or less risky? That is because if it does become more or less risky as prices fluctuate over a wider range, the value of relatively risky assets (perhaps all assets other than the true safe havens – perhaps only US Treasury Bills and Bonds) will move in the opposite direction. Or in other words, the question to ask is not what is the likely outlook for the economy etc. but rather, what is the outlook for the economy and for listed companies, currencies and almost all bonds?

Another way of putting it is to ask whether market views will have reason to become more or less diverse. Will events evolve that become more or less easy for market participants to interpret? Will it become easier to solve the known unknowns that investors recognise as consistently driving valuations? A global financial crisis is a very unusual event, the outcomes from which were extremely difficult to agree about – hence the volatility in 2008 and 2009. A Eurozone debt crisis raises uncertainty about how it will all work out for share prices interest rand exchange rates – as would any breakup of the Eurozone system, another first time possibility for which history supplies little guidance..

Hence the obligation for policy makers to act as predictably as possible. Certainty in economic policy reduces risks for investors and helps raise values. Less risk means higher asset prices and lower expected returns. If SA wishes to attract foreign and domestic capital on superior terms, the aim should be to reduce the high risk premium attached to incomes dependent on the SA economy. High returns of the kind earned by investors in SA assets since 2000 might well be realized, should the SA risk premium come down rather than go up over the next few years. The SA government could do a much better job than it has been doing to introduce certainty in its economic policies and, as important, certainty that the right income enhancing policies are being adopted.

Point of View: Less obvious than they seem

Lower average inflation in SA is surely welcome – but will it make doing business or consumption less risky? The benefits of lower inflation in SA may well be less obvious than they seem.

The average price that SA consumers paid for goods and services actually fell by 0.2% in January. The Consumer Price Index (CPI) measured 110.8 in January compared to a level of 111 reached in December 2014 (based on December 2012 = 100). The CPI first reached the level of 111 in August 2014 and is now lower than it was five months ago. Thus headline inflation, calculated as the year on year change in the CPI, has fallen away sharply and, if present trends in the CPI were to continue (which is unlikely) , inflation in a year would be below 2%.

The CPI is but an indicator of the average prices paid by the average SA household for a fixed representative basket of goods, as pre-determined by Stats SA based on its surveys of household spending patterns. As we are all well aware, any average can hide a large dispersion about the mean. The old saw about feet in the fridge and head in the oven yielding a moderate average bodily temperature makes the point. Some of the goods and services included in the CPI may be rising at a much faster rate than others – some important items may even be falling, helping to reduce the CPI and the average inflation rate. This has been the case over the past 12 months.

The different components of the household budget have realised very different inflation rates. The average price increased by 4.4% since January 2014. Food and non-alcoholic beverages, with a large weight in the average budget of 15.41%, rose by an above average 6.5% over the past 12 months. Inside the food trolley, dairy products, milk, eggs and cheese rose by as much as 12.1% year on year. The goods helping to hold down average inflation in 2014 were petrol, with a 5.6% weight, was down 17.6% over the 12 months, while the prices of so called private transport, with a weight of 7.25%, fell by 13%. Telecommunication equipment, presumably high quality or computer power adjusted, was estimated to have fallen by 12.1% in 12 months.

While the quality adjusted prices charged for cell phones and the like may well fall further, the chances of fuel prices declining further seems remote – since even if the rand price of a barrel of oil were to decline further, National Treasury is bound to levy a higher excise tax on petrol and diesel.

Clearly the all important relative prices – the price of food relative to the price of transport – changed quite dramatically and can be expected to continue to do so. Businesses have to be constantly aware of the changing relationship between the prices of the goods and services they buy, including labour services, and the prices they are able to charge in their market places and adjust accordingly.

Presumably households consume more of the relatively cheaper goods and less of the more expensive stuff. They may well trade down – that is sacrifice quality for price as goods or services become relatively more expensive. Stats SA only periodically (every five years or so) adjusts its CPI trolley of goods and services for such shifts – that may be influenced by price as well as by innovations on the supply side of the economy.

Another way of measuring prices is through the use of deflators, as used in the National Income Accounts to convert the value added in money of the day prices to their real equivalent. A deflator takes current consumption or expenditure patterns and converts them into their constant price equivalents. In other words, it calculates what the goods and services bought today would have cost in some base year. Changes in this deflator then offer an alternative view of inflation.

A comparison between the Household Consumption Goods Deflator and the CPI, based on 2010 prices as well as the respective inflation rates, is shown below. The trends are similar but not identical.

Using the deflators can demonstrate just how much relative prices have changed in SA over the years. Deflators are available for a large number of items included in total household consumption. The deflators for the main categories – household spending; non-durable goods, mainly food and beverages; semi durables, mainly clothes and footwear; durables, namely vehicles, furniture and appliances; and household services, utilities, restaurants, entertainment and domestic service – are shown below and are based on 1990 prices for purposes of comparison. As may be seen, the prices of food and services have increased at a much faster rate than the prices of semi-durable and durable consumer goods. Food prices have increased by over seven times since 1990 and clothes and footwear by only two times, with services increasing at almost the same rate as non-durables. We also demonstrate how much more relatively expensive food and services have become.

A large part of the theoretical case made for low rates of inflation is that low inflation helps stabilise relative prices. Such greater certainty about relative prices – or the relationship between the prices of the goods and services we sell and those we buy – would be helpful to producers and consumers. It would help to reduce uncertainty about relative prices and so reduce the risks of undertaking consumption and production over time, which would thus be to the advantage of economic growth.

Unfortunately there is no evidence that lower consumer goods inflation in SA has in any way reduced the dispersion of the prices of goods and services consumed by households about their average. According to the deflators, the rates of inflation of the many goods and services consumed by households differ now by as much as they ever have, as we show below.

The benefits of lower inflation in SA may well be less obvious than they seem. Lower inflation does not appear to have reduced the risks in consumption and production. Relative prices remain as variable as ever. Nor does it appear to have stabilised interest rates after inflation or the rand exchange rate (once adjusted for differences in inflation between SA and our trading partners).

Interest rates: A play on the rates

The JSE as a play on interest rates. The scope for still lower long term interest rates in SA

The importance of movements in interest rates for share prices over the past 12 months has never been more obvious on the JSE. Interest rates turned out to be significantly lower than expected early in 2014 and a group of large cap interest rate sensitive stocks, banks, retailers and property companies, have accordingly performed outstandingly well.

Since 1 February 2014 to 30 January 2015 our market cap weighted Index of large cap interest rate sensitive stocks generated a total return (including dividends) of 48.7%. The Global Consumer Play Index, also market weighted and one that includes Naspers and Aspen, while also performing well returned a lesser 42.9% while the JSE All Share returned 16.8%. The S&P 500, the best performer of the developed equity markets provided a 12 month return in rands of 19.6%, a highly satisfactory outcome, but less than half the return provided by the SA interest rate plays, as may be seen below.

(The index of Interest Rate Plays is made up of the following 30 companies: BGA, FSR, GRT, INL, INP, IPL, MSM, NED, RMH, SBK, TRU, CCO, CLS, CPI, FPT, HYP, NEP, PIK, RDF, RES, TFG, WBO, MPC, WHL, CPF, ATT, PSG, RPL, AEG and FFA. The Global Consumer Plays are: APN, BTI, CFR, MDC, MTN, NPN, SAB, SHF, NTC and ITU)

These interest rate sensitive stocks on the JSE should be regarded as demandingly valued by the standards of the recent past. They were priced at January month end at a well above average 16.9 times trailing earnings. They surely have benefitted from unexpectedly lower interest rates.

Presumably these interest rate sensitive stocks will remain so, making the further direction of interest rates in SA of great importance in stock selection and asset allocation. Long term interest rates in SA will moreover continue to take direction from interest rates in the US and Europe. Furthermore the value of the rand is bound to be strongly influenced by the self same interest rate trends.

When interest rates in the developed financial markets decline, all things remaining the same, especially country specific risk factors, funds will tend to flow towards less developed markets where yields are higher. The search for yield in a low interest rate world will tend to compress yields and yield spreads everywhere, so adding to the demand for emerging market currencies that supports exchange rates, including the rand. And where the rand goes will influence the outlook for inflation in SA and so the direction of short term rates. Over the past year lower euro yields have been very strongly associated with a stronger rand vs the euro, as well as a weaker euro and rand vs the US dollar. A weaker euro, both against the US dollar and the rand, has come with additional demands  for and lower yields on RSA long dated government bonds.

The wider spread between US and German yields shown in figure 1 has clearly helped to add to US dollar strength and can be expected to continue to do so.

Given the freedom to move capital from one market to another, it is clear that interest rates in Europe must influence rates in the US and vice versa – rates in the US must influence rates in Europe as well as SA and elsewhere. It seems as clear that, were it not for the weakness of the Eurozone economies and the threat of deflation there, as well as the promise of European Central Bank (ECB) quantitative easing on a large scale (and so very low Eurozone interest rates), long term interest rates in the US would have been a lot higher than they now are. The leading force in the longer end of the global bond market may well be European deflation rather than US economic growth and the reactions of the US Fed. The US economy seems firmly set on a good growth path. The impressive growth in the numbers of workers employed in the US is ample testimony to the strength of the US economy. The latest employment numbers have been revised sharply higher for 2014, when an extra 3.04 million employees were added to private payrolls. The response of long term US interest rates to these bullish developments has been quite muted.

These employment numbers may well encourage the US Fed in June to raise its own key short term Fed Funds rate from its current zero level, as is now widely anticipated in the money market. But longer term rates may still take their cue from rates in Europe and stay where they are, with 10 year Treasury yields staying closer to the current 2% level than the 3% level, which might be regarded as more normal. In such a case, long term interest rates in SA will also not move sharply higher any time soon. If however rates in Europe trend still lower under pressure from aggressive QE interest rates in Europe, the US and SA can still surprise on the downside. A stronger dollar would press on both US inflation and growth rates and weaken the case for higher short term rates.

While the level of RSA rates will respond to the directions of global markets it may be asked what should be regarded as the normal level of interest rates in SA? The Reserve Bank has spoken of the normalisation of SA interest rates, implying higher rates should be expected, though in its latest Monetary Policy Statement it referred to a likely pause in rates given the much improved inflation outlook. Normal must refer to rates after inflation or, when longer rates are interpreted, it would be by reference to market rates after expected inflation, that is to say real rates.

In figure 7 below we compare RSA 10 year nominal bond yields with their inflation protected alternative yield since 2005. Nominal RSA Yields have a daily average of 8.13% p.a since 2005 with a high of 10.9 % p.a. in August 2008 – and a low of 6.13 in May 2013. Inflation protected real yields averaged 2.4% p.a with a high of 3.65% p.a and a temporary low of 0.38% p.a. in May 2013. The daily volatility of both these yield series, measured by the Standard Deviation (SD) of the daily yields about the average, was about the same, 0.69% p.a.

The difference in these yields, nominal and real, may be regarded as compensation for bearing the inflation risk in vanilla bonds in the form of higher yields, has averaged 5.8% p.a with a SD of 0.61% p.a. Inflation expectations revealed by the RSA bond market appear as highly stable about the 6% p.a level, which is the upper end of the Reserve Bank’s target range for inflation. Headline inflation in SA calculated monthly has not co-incidentally averaged 6.1% since January 2005. Thus normal long bond yields might be regarded as 6% for inflation plus 2.5% p.a as a real return, summing up to approximately 8.5% p.a yield on a long dated RSA bond.

The evidence is that inflation compensation in the bond market follows the inflation trends with a long lag. It will take a sustained period of well below average 6% inflation to reduce the expected inflation priced into nominal bond yields of about 6% p.a. It will take faster growth in SA and globally to raise inflation linked 10 year real interest rates in SA meaningfully above their current 1.72% p.a. This seems an unlikely development in the short term. The equivalent 10 year real inflation protected (TIPS) yield in the US is only 0.28%, offering investors in inflation linked RSAs a real yield spread of 1.5% p.a. This real spread appears rather attractive in current global circumstances and may well decline. This real spread can be compared to a nominal yield spread of 5.44% p.a. in favour of 10 year RSAs on 9 February 2015 (that is the RSA at 7.38% – US Treasury at 1.94% = 5.44%, which is very much in line with the trends in this spread since 2008).

The case for JSE listed interest rate sensitive stocks at current demanding valuations could be based on the prospect of a further decline in SA interest rates. In the first instance this is on US rates rising less than the currently modest 20bps expected by the US Treasury bond market in a year’s time. The market is expecting the 10 year US Treasury Yield to rise from the current 2% p.a to approximately 2.2% p.a in a year. This expected increase should not be regarded as a grave threat to the SA bond market. It could take lower rates in Europe to deny such expectations or any softer actions or words from the US Fed regarding its Fed Funds rate.

The other hope for interest rate sensitive stocks on the JSE would be a decline in the RSA real rate, which appears quite high, compared to real rates elsewhere. A modest decline in the real rate would help depress nominal rates. A more likely, but more and more potentially significant decline in RSA bond yields could follow any decline in inflation expected. This could occur if SA headline inflation stays well below the 6% mark for an extended period of time. Clearly, with interest rates and the valuations of interest rate sensitive stocks where they are, there are upside as well as downside risks to interest rates in SA and to SA interest rate sensitive JSE listed companies.

Hard Number Index: Picking Up Momentum

A dispatch from the economic front: vehicle sales and supplies of cash are picking up momentum

Sales of new vehicles by SA dealers in January 2015 some 52306 units of all sizes were good enough to keep the new vehicle sales cycle on a recovery path that began in mid year. If recent trends are sustained, the network is on track to sell over 700 000 new units in 2015, close to the record levels achieved in 2006.

We combine this statistic with another very up to date hard number, notes issued by the Reserve Bank in January 2015, to establish our Hard Number Index (HNI) of the state of the economy. The HNI for January 2015 indicates that economic activity in SA continues to grow at a modest pace.

Furthermore the pace of activity that appeared to be slowing down in mid-year has gained some momentum and is forecast to sustain this rate of growth in 2015. The HNI may be compared to the coinciding business cycle indicator of the SA Reserve Bank. This economic activity indicator, based on a much larger set of mostly sample surveys (not actual hard numbers) is also pointing higher, suggesting a pickup in growth rates, but is only updated to October 2014. It should be noted that the turning points of the HNI and the Reserve Bank indicator were very well synchronised when the economy first began to recover from the post Global Financial Crisis recession, in 2009.

In the figure below we track the two separate growth cycles, unit vehicle sales and demands and supplies of real cash – the note issue – deflated by the CPI. Both series are pointing higher. This upward momentum will be sustained by less inflation to come and the relief lower rates of inflation provide for interest rates. The lower inflation might, in due course, possibly only in 2016, mean lower (not higher) costs of financing vehicles and will help the vehicle market and the economy generally.

Monetary policy: The big bad wolf

Deflation is the big bad wolf threatening monetary policy. The logic of QE

Public enemy number one for central bankers in the developed world is deflation. When the CPI declines we have deflation, when it rises we have the opposite, inflation. Prices in general fall, when aggregate demand in the economy exercised by households, firms and governments, fails to keep up with potential supply. Prices rise when demand exceeds supply in general. The economic problem in the developed world and in much of the less developed world (including SA) is too little rather than too much demand that has called for highly unconventional monetary policy.

These central bankers, with modern Japan very much in mind where prices have been falling and economic growth abysmally slow since the early nineties, are convinced that deflation depresses spending and by so doing serves to prevent an economy from achieving its growth potential.

They are therefore doing all they can to stimulate more spending to arrest deflation or a possible decline in average prices. What they can do to encourage reluctant spenders is to create more money and reduce interest rates. Inflation, as they used to say, was too much money chasing too few goods. Deflation may be said to be caused by the opposite – too little money chasing too many goods. Inflation calls for less demand and so less money creation. Deflation calls urgently for the opposite – more money creation.

But while these central bankers have the power to create as much extra cash as they judge appropriate they have had to overcome a technical problem. The extra cash intended to encourage more spending may get stuck in the banking system. And when it does there may be no additional demand for goods and services and the extra money created (at no cost) may not encourage spending or lending and provide no relief of the unwanted deflationary pressures. The banks may hold the extra cash as additional cash reserves or use the cash to pay off loans they may have incurred previously with the central bank.

The mechanics of money creation go simply as follows. The cash the central banks create (or more specifically the financial claims they create on themselves called deposits with the central bank) when buying securities from willing sellers in the financial markets (typically pension funds or financial institutions of one kind or another) reflects immediately as extra privately owned deposits with private banks. In the first instance, as the proceeds of a sale of securities by a private bank customer are banked, the private bank will have raised an additional deposit liability and will receive in return an asset in the form of additional claim on the central bank. That is, the asset to match their additional deposit liabilities will take the form of an increase in their (cash) reserves at the central bank. Normally the banks will make every effort to convert this extra cash into a loan that earns them more interest. Such loans would then lead to more spending. Normally banks keep minimal cash reserves in excess of the regulated requirement to hold cash reserves in fractional (say 5%) proportion to their deposit liabilities.

But ever since the Global Financial Crisis of 2008 this has not been the case at all. The central banks, led by the US Federal Reserve Bank, have bought trillions of dollars worth of financial securities and the private banks have added trillions of dollars to their cash reserves. In the figure below we show how the assets of the major central banks have expanded over the years. Their extra assets have consisted of securities issued by governments of government agencies- for example in the form of mortgage backed securities issued by government backed mortage lending agencies Fannie Mae and Freddie Mac in the US.

It should be noted that the ECB balance sheet has been shrinking in recent years, unlike those of the other central banks. This decline in ECB assets and liabilities (mostly cash reserves of the banks) has represented additional deflationary pressure on spending in Europe. The European banks have been repaying ECB loans previously made to them and the ECB has now responded by initiating a further programme of QE, that is to say security purchases, intended to inject an extra EUR60bn a month into the banks of Europe over the next 21 months. That is to encourage bank lending and spending to counter deflation.

The important point to note is that creating money in the form of a deposit with the central bank by buying securities in the market costs society almost nothing. Creating money that is an asset of the banks and an addition to the wealth of the community does not require any sacrifice of consumption or savings, as would any other form of wealth creation. It is literally money for nothing. But in normal times too much money means too much spending and inflation. Hence the resistance in normal times to creating money – because it normally leads to inflation.

But the times have not been normal. The extra supplies of money have been accompanied by extra demands by the banks to hold money and too little rather than too much spending has remained the economic problem. Hence the case for creating still more money until deflation is conquered and central banks can get back to worrying about too much money and inflation.

The extra liabilities issued by the central bank matching the extra assets were mostly to private banks in the form of cash balances in excess of required cash reserves. In the US the banks have received 0.25% p.a on these cash reserves.

It should be notie that the US banks have begun marginally to reduce their cash reserves with the Fed and that the assets of the Fed are rising more slowly with the end of QE in October 2014. QE ended because the recovery of the US economy is well under way and presumably does not need further encouragement in the form of further increases in the supply of cash. The banks presumably have more than enough excess cash to meet the demands of borrowers from banks. It could be argued that QE at least in the case of the US – the economy is recovering – the dangers of deflation have receded and the danger of inflation taking over is judged to be absent.

There is very little inflation priced into the yields offered on 30 year US Treasury Bonds that offer yields below 3% and less than 2% more than inflation protected US obligations with the same duration. Japan and Europe, it may be presumed, will be doing still more QE or as much as it takes to get the banks to lend out more of their cash.

The problem of deflation is complicated by very low or even negative interest rates. Cash has one advantage over other assets. The income return on cash can only fall to zero. Other assets, for example government securities or bank deposits, may only offer negative interest rates. The German government, for example, can now borrow for up to 10 years, charging rather than paying interest. In other words, it can take about 105 euros from you and promise to pay you 3 euros interest and repay you 100 euros in a year’s time. In other words the transaction will have cost you 2 euros and this, alas, is the best risk adjusted return you can get, though, if prices in general decline by 2% over the year, your 100 euros will be worth as much as 102 a year before adding to your real return. But you would have done better holding cash. 100 euros in cash will still be worth 100 after 12 months and will also buy you more.

Hence deflation forces down interest rates and encourages the demand for cash (and safe deposit boxes) and discourages the demand for bank deposits and for goods and services that may be expected to become cheaper. Deflation also encourages banks and other lenders to hold cash rather than to lend it out. Hence further reason to fight deflation (and too little rather than too much spending) by flooding the system with additional cash so that the supply of cash can eventually, even with very low or negative interest rates, overwhelm the demand for cash.

Until Europe succeeds in overcoming deflation and stagnation we can expect interest rates in Europe to stay low and for the euro to stay weak; and the US dollar (offering higher interest rates) to stay strong. We may also expect deflation and low interest rates in Europe to help hold down long term rates in the US and elsewhere as European lenders seek higher yields abroad. The weak euro and stronger dollar (and perhaps also stronger emerging market currencies, including the rand) may also help restrain any increase in short and long interest rates globally. It may take some time before we can say with any confidence: goodbye to deflation and welcome back inflation.