Month: June 2013

Fed chairman Ben Bernanke spoke of a surprisingly promising outlook for the US economy, of 3% to 3.5% growth in 2014 that, if it all materialises as predicted, would allow the Fed to taper off its securities purchase programme from September this year and to close down the purchases by late 2014 (currently of the order of $85bn a month).

The market listened and reacted in ways that were consistent with the prospect of faster growth in the US but they would not have pleased Bernanke. He was at pains to emphasise in his statements how conditional would be the direction of quantitative easing (QE), ie conditional on the actual improvement in the US economy (and the labour market in particular) to keep the market at ease. The severe bond market reactions were not welcome because higher interest rates (especially higher mortgage rates) may threaten the recovery itself.

Longer term interest rates moved sharply higher in reaction to Bernanke and the Fed. By the weekend the 10 year Treasury Bond yielded over 2.5%, compared to about 2% a week before. More economically significant was the move in inflation linked bonds (TIPS). These real yields moved even than did the yields on the vanilla Treasury bonds, from close to zero on 16 June to over half a per cent by the weekend. Higher real rates are consistent with an improving growth outlook, leading to increased demands for capital to invest in real assets. And so the gap between the vanilla yields and the inflation protected variety narrowed to less than 2%.

Bernanke indicated in his press conference that his target for inflation is 2% per annum – anything less in his view represents a deflationary danger for the economy. The market now expects inflation to be dangerously low – implying more, rather than less, monetary easing to come.
The negative reactions of the equity markets to the more promising outlook for the US economy were not as easily explained. The S&P 500 was down by just over 2% in the past week, having been very firm before the Fed statement. Stronger, more normal US growth of the kind the Fed is expecting drives earnings as well as interest rates higher – possibly enough to add rather than detract from the value of equities that remain (in our judgment) still undemandingly valued by the standards of history and the prospects for earnings.

These interest rate developments in the US had severe repercussions for emerging bond markets and emerging currencies, that until recently have been beneficiaries of a search for yield in a world of generally very low yields. Not-so-low yields in the US reversed these flows, leading to pressure on emerging market currencies and yields of all kinds. SA was not spared these withdrawals of cash from high yielding assets, though the rand and the rand bond market did less poorly than many other emerging market currencies and bond markets, subject as they have been, for example in Brazil and Turkey, to violent demonstrations on their streets.

The rand actually gained by a per cent or two against the basket of EM currencies and the Australian dollar in the week ending 21 June. The yields on both long dated conventional RSA bonds and the inflation-linked equivalents rose, though this yield gap widened slightly, offering more compensation for bearing SA inflation risk by the weekend.

The yield gap between RSA rand bond yields and US Treasury bond yields can be regarded as compensation for bearing the risk of the rand depreciating. This yield gap can also be described as break even rand depreciation. If the rand depreciates over time at a faster rate than implied by the difference in interest rates, it would be better to buy US bonds (and vice versa if the rand does better, that is depreciates on average by less than the difference in yields). This yield remained largely unchanged through the past week. While the rand has weakened against the US dollar it is not priced to weaken at a faster rate.

Expectations about the direction of short term interest rates in SA have been revised sharply higher to the disadvantage of all the interest rate sensitive stocks listed on the JSE, the retailers, property companies and banks etc. With a more sharply inclined yield curve the one year RSA rate expected in a year’s time was 5.13% on 30 April. It is now 7.6% (see below):

The state of the SA economy does not justify higher short term rates. The Reserve Bank is predicted to raise them notwithstanding. Our view is that the Reserve Bank will correctly resist raising rates until the SA economy has picked up momentum, rather than slowing down, as it appears to be doing.

The immediate future of the longer term interest rates in SA will take their cue largely from the direction of long term rates in the US. Better economic news emanating from emerging market economies (and China in particular), would also help the rand and the RSA bond market. There would appear to be some chance that the US bond market has over reacted to good news about the US economy – expectations that have still to be fully vindicated. Higher rates will also encourage the Fed to maintain, rather than slow down, the pace of its bond purchases. If so, the past week may well prove a temporary high water market for interest rates in the US and elsewhere.

An unfortunate history of exchange volatility

The SA economy is once more challenged by an exchange rate shock. As we show below, such exchange rate weakness – of the order of a 15% or more move lower in the trade weighted exchange rate, compared to a year before is hardly unknown. In fact the latest shock is the fourth since 2000. As we also show below, the rand had lost as much as 26% of its foreign trading value by May 2002. By early 2007 the rand was down by about 15% on its value a year before and then 20 months later had lost 19% of its value.

The rand on 31 May 2013 was about 14% weaker than 12 months ago on a trade weighted basis. It will also be appreciated that while the long term trend in the value of the rand since 2000 has been one of rand weakness, the direction is by no means one way. Weakness can be followed by strength of similar magnitude. These unpredictable shocks, in the form of large sustained movements in the value of the rand in both directions, can be of similar magnitude and can complicate business decision making and monetary policy. They are a most undesirable feature of the SA economic landscape.

The sources of exchange rate volatility have very little to do with monetary policy

These large exchange rate movements are a response to interruptions or disruptions in the flow of capital to and from South Africa. Increased demands for rands push the rand higher and less demand moves the price of the rand higher or lower when valued in other currencies. It is very much a market-determined and flexible – very flexible in both directions – rate of exchange. The SA Reserve Bank does not typically use its own stock of foreign exchange to intervene in the market for rands.

This market, a deep one at that, regularly transacts over US$15bn worth of rands every trading day, according to the Reserve Bank on the basis of information provided by the trading banks. Three quarters of the trade is conducted between third parties without a direct connection to SA trade or finance. They presumably trade and hedge the rand so actively as a proxy for currencies that are less liquid. As we show below there is no obvious relationship between the trade weighted value of the rand and turnover in the currency market.

The one highly predictable impact of an exchange rate shock- more or less inflation

The one highly predictable influence of an exchange rate shock is that more or less inflation will follow in the opposite direction. More inflation when the rand weakens – less when it strengthens. It is most important to recognise that for SA the exchange rate leads and the inflation rate follows. In conventional monetary theory it is faster domestic inflation (caused by easy monetary policy) that leads to a weaker exchange rate. The weaker exchange rate then should help to maintain the international competitiveness of exporters and firms that compete with more expensive imports priced in the weaker domestic currency. In the figure below we identify the timing of the shocks that have sent the rand weaker and show how the trend in the inflation rate has followed these shocks consistently.

In the figure below we show the results of a very simple model. The trend in inflation is very simply explained in a single regression equation by the annual movement in the trade weighted exchange rate, lagged by six months. The model does well in predicting the direction of inflation in SA and also its level. The explanatory power of the model is rather good – explaining over 60% of the inflation trend. As the chart also shows, there is somewhat more to inflation than the exchange value of the rand. The model significantly underestimated inflation in 2008-09 and has less significantly underestimated it recently.

Among other forces moving SA prices and inflation are trends in global prices, particularly in the prices of grains and other soft commodities in US dollars that influence the domestic price of food when translated at import price parity into rands. The global price of oil is also very important in this regard. Clearly independent of the value of the rand, global inflation or deflation (including oil price increases or decreases) will influence prices in SA and their rate of change. The pace of administered prices increases in SA (taxes by another name) will also have an influence on the CPI. So will the strength or otherwise of domestic spending, supported more or less by the growth in money supply and credit and interest rates, that is by monetary policy.

Monetary policy is largely impotent in the face of exchange rate shocks of this order of magnitude

It should be very obvious from the recent history of inflation in SA that there is little the Reserve Bank or monetary policy can do about inflation because it cannot influence the variable exchange value of the rand in any predictable way. The same monetary history tells us that raising or lowering interest rates have simply no predictable impact on the exchange rate. Monetary policy is largely impotent in the circumstances of exchange rate shocks of the order of magnitude suffered by SA. Inflation targeting, to which SA subscribed in the early 2000s with all the sincerity of the newly converted, had as its justification the conventional wisdom of monetary policy of that period. The presumption of inflation targeting was that a politically independent central bank would target inflation with its interest rate settings in a sound way and inflation and the exchange rate would behave itself in a predictable way.

The unpredictable nature of exchange rate shocks – global or domestic in origin

That presumption has proved to be a false one. The exchange rate has not behaved itself and so measured inflation has remained largely outside the influence of monetary policy. Monetary policy and interest rate settings can clearly influence domestic spending. But maintaining the balance of domestic demand and potential supply does not at all necessarily secure exchange rate stability as we observe.

The exchange rate can have an unhealthy life all of its own, responding as it does to global forces, as it did during the Global Financial Crisis of 2008-09. This crisis increased the global demands for safe havens and reduced the demand for riskier emerging market assets and their currencies, including the weaker rand, and led to more inflation in SA.

The other shocks to the rand we have identified are much more SA specific in their origins. We can identify such SA specific risks driving the rand by comparing the behaviour of the rand to other emerging market or commodity currencies over a period of rand weakness or strength. The forces driving the rand in 2001-02 and in 2006-7 were largely SA specific in their origins. The rand has weakened by significantly more than its peers over these periods of weakness.

The only time the Reserve Bank may be held responsible for rand weakness was in 2006-7. Then the bank adopted interest rate settings that were too severe, that threatened the growth prospects for the SA economy and frightened capital away. The 2001-2002 weakness was an unintended consequence of partial exchange control reform – that led to panic demands for foreign currency by local wealth owners and fund managers. The latest burst of rand weakness that began in August 2012 is associated clearly with labour relations on the mines and elsewhere that threaten mining output and exports that are so important to the trade balance of the rand. Foreign and local investors have been discouraged by the political responses to this crisis.

Nothing for the Reserve Bank to do but watch the economy ride out the storm

As clear as are the political origins of the latest exchange rate shock is that the Reserve Bank and its interest rate settings can do nothing now to meaningfully assist the rand. It is out of their hands. Raising interest rates would further weaken domestic spending, that cannot be regarded as excessive. Still slower growth in domestic spending following any imposition of higher interest rates would if anything further undermine the case for investing in SA and could lead to a still weaker rand. Indeed, were it not for rand weakness, interest rates would have been reduced to encourage domestic demand. But such action might well be regarded as less than responsible in the circumstances.

The best the Reserve Bank can do in these difficult circumstances is to do very little. The economy must be left to rise out the exchange rate shock and the temporary increase in inflation that is likely to follow. The weaker rand will encourage production for export and for the domestic market as prices and profit margins for exporters and those competing with imports have improved. Hopefully the mining sector will be allowed to benefit from these price and profit trends. Hopefully too, the politicians can help the industry. Higher prices, especially for goods or services with high import content, will discourage consumption. There is nothing the Reserve Bank can usefully do to slow these inflation and relative price effects down. Raising interest rates would damage the economy further. The best monetary policy can do in response to an exchange rate shock (that is not of its making) is to do nothing at all – but also to explain why doing nothing is the best policy.