T Mboweni: Oxford and Cambridge Business Alumni of South Africa Risks on
inflation outlook

2 August 2007

Honoured guests
Ladies and gentlemen

Introduction

I would like to thank you for the opportunity to address such an illustrious
gathering. I was not fortunate enough to attend either the University of
Cambridge or the University of Oxford but should state that I followed the
rowing competition held annually between the two universities with keen
interest. I have always wondered how my Alma Mater, the University of East
Anglia, would perform in this competition. I am sure the students at the
University of East Anglia would relish such a challenge.

Members of the Oxford and Cambridge Business Alumni of South Africa play an
important role in the economic and social life of the country. Members of the
organisation include Chief Executive Officers of banks, investment bankers,
investment analysts, professors, bishops and headmasters, to name a few.

My role this evening is to talk about the more mundane issues of monetary
policy. I will share with you some of the issues that are of concern to the
Bank. These concerns are not new and they have been raised in recent Monetary
Policy Committee (MPC) statements, but I thought I would elaborate on some of
them in more detail.

The inflation target: where do we stand?

Before highlighting these issues, let me first remind you where we stand
with respect to our monetary policy objective. Our mandate is to keep CPIX
inflation within the inflation target range of three to six percent on a
continuous basis. This we successfully achieved for 43 consecutive months
between September 2003 and March 2007. In April 2007 Consumer Price Index
excluding interest rates on mortgage bonds (CPIX) inflation breached the upper
end of the target at 6,3 percent, and remained outside the target at 6,4
percent in May and June.

The major drivers of these developments were higher food and petrol prices.
In these months, had food and energy prices been excluded, CPIX inflation would
have averaged around 4,6 percent. Nevertheless there is evidence that
underlying inflation is becoming more generalised as evidenced by the fact that
this narrower measure of inflation has been trending upwards in recent months
from a low of 2,5 percent in June last year. In response to these adverse
inflation pressures, monetary policy has been tightened and we have increased
the repo rate by a total of 250 basis points since June 2006.

It is of concern to the Bank that inflation has breached the target range.
However I should emphasise that this should not be seen as a failure of
monetary policy. As I noted, the main drivers of the recent surge in inflation
were petrol and food price inflation. It is important to emphasise however that
monetary policy reacts with a lag, and that it takes approximately 18 months to
two years for the effects of an interest rate change to be reflected fully in
inflation. This implies that there is little monetary policy can do about
current inflation, which is an historical number, particularly if it has been
caused by unanticipated external shocks. In setting the monetary policy stance
we therefore have to focus on the medium to long term.

This does not mean that we can be complacent about near-term developments,
as inflation expectations and monetary policy credibility are determined in
part by these short term considerations. The challenge for monetary policy is
to ensure that inflation is brought back to within the target range, and also
to convince the markets that we will are serious about this. In this way we
will ensure that inflation expectations remain anchored within the target
range.

I should note that we are not alone in missing our targets. During the past
year or so, a number of central banks have temporarily exceeded their targets,
the most recent notable example being the Bank of England. Pressures emanating
from oil and food prices are a global phenomenon and are posing a challenge to
many central banks.

Dealing with inflation shocks: International oil prices.

Rising international oil prices and food prices have been on the top of our
list of risk factors for some time. I am often asked what monetary policy can
do about oil or food prices? If we tighten monetary policy in the face of
higher international oil prices, this will not cause OPEC to raise output or
market prices to decline. Similarly with food prices- raising interest rates
will not make it rain. Yet these are two variables that are high on our list of
concerns.

Let us first consider oil prices. At the beginning of 2004, the price of
North Sea Brent crude oil was around US$30 per barrel. It is now almost US$80
per barrel. In the first five months of this year the domestic petrol price
increased by R1,34 per litre, although it has since declined by R0,23. Most of
this increase has been due to higher international product prices rather than
exchange rate effects. International oil prices are affected by a variety of
factors ranging from higher demand in a fast-growing world economy,
particularly from China; and risks to supply emanating from OPEC cutbacks,
supply disruptions in a number of oil-producing countries, geopolitical
tensions and adverse weather conditions. These factors have contributed to the
upside risk to the oil price.

How should monetary policy react to this? The impact of a change in the
international oil price is direct and relatively quick, given that domestic
petrol prices are adjusted each month in terms of the formula of the Department
of Minerals and Energy. Petrol has a weight of about five percent in CPIX, so a
10 percent increase in the petrol price would increase CPIX inflation by
approximately 0,5 percentage points. This is the impact effect or what we call
the first-round effect, and clearly there is nothing we can do about this.

Our concern is with the broader impact of this effect, that is, the
so-called second-round effects. We have to assess whether or not these
increases will be passed on over time in the form of higher transport costs
which will increase prices of other commodities, for example food, or through
higher production costs. We also have to assess whether there will be an impact
on wage and price setting, which will impact broadly on inflation. Over the
past two years these second-round effects have been relatively moderate. This
could be due to a number of reasons, including increasingly anchored inflation
expectations as a result of enhanced monetary policy credibility; increased
competitiveness in the economy; and the fact that some of these increases were
in fact reversed late last year. The longer the international oil prices are
subject to upward pressure, the more likely these second-round effects will
feed through to generalised inflation.

The challenge for the MPC therefore is predicting not only the future course
of volatile international oil prices, but also assessing how these increases
will feed through to further inflation pressures over time. In deciding on the
appropriate monetary policy response, it is not always easy to differentiate
between first and second round effects. Some analysts have argued that we
should exclude food and oil from the price index that we target. We do not
believe that targeting such an index would be credible. The Economist magazine
has appropriately referred to such an index as 'the cold-and-hungry index'.

Food prices

Food prices have also become a cause for concern. Food prices have a weight
of almost 26 percent in the CPIX basket, although this proportion is much
higher at around 51 percent in the consumption basket of the lowest category of
income earners.

As is the case with oil prices, there is an international dimension to the
increase in food prices. Higher food inflation is an international phenomenon.
This is partly weather induced, but perhaps more significantly, the increased
diversion of maize and sugar to biofuels production has resulted in significant
price increases in the international markets. Domestic maize and wheat prices
are directly affected by these higher international prices and combined with
domestic drought in some areas of the country, have caused the spot price of
maize in South Africa to increase almost four-fold over the past two years.

Higher maize prices have not only increased the cost of maize products but
have also affected meat prices, which last year were increasing at rates of
almost 20 percent. Meat has a weight of around seven percent in the CPIX
basket, and in 2006 was the single biggest contributor to overall CPIX
inflation. Food prices are currently increasing at rates in excess of nine
percent per annum, and we are still feeling the effects of the maize price
increases. One small consolation is that meat price increases have moderated
from their peaks last year as more cattle are brought to the market during
periods of drought. This trend may not persist as farmers are likely to restock
at some stage.

Household consumption expenditure

Much has been said about the strong growth in household consumption
expenditure and the associated increase in domestic credit extension. Household
domestic expenditure has been increasing at a year-on-year rate of between
seven and eight percent for the past three years, compared to rates of growth
of around three percent in the preceding three years. At the same time, credit
extension to the private sector has been growing at rates of around 26 percent
despite the tighter monetary policy stance. These are uncomfortably high levels
and I have expressed my concern in this regard on various occasions.

Excessively high rates of expenditure growth will ultimately impact on
domestic inflation, although the timing and extent of these effects are at
times uncertain. These developments will remain an important focus of our
monetary policy deliberations.

My concern, however, extends beyond simple monetary policy considerations.
As a result of the high levels of credit extension, household indebtedness has
increased to record levels, currently at around 76 percent of disposable
income. Although the debt service ratio reached a low of six percent of
disposable income in 2004 and 2005, this has now risen to nine percent.

Although part of this increased indebtedness is justified on the basis of
improved household balance sheets, my concern relates to the socio-economic
impact of excessive debt accumulation. As more and more people find employment
or better paid employment, the temptation is to rush out and buy a bigger car
and a bigger house. Very soon, we see people becoming overextended and
repossessions become commonplace. Part of this development is due to excessive
exuberance on the part of consumers. There is not a culture of saving in South
Africa, as demonstrated by our very low savings rate, and people are happy to
borrow excessively against future income in order to finance current
consumption.

The problem has also been compounded on the supply side with banks and
others, including retailers, falling over themselves to extend credit,
resulting in a number of questionable lending practices. Perhaps the situation
is not as bad as in other countries such as Australia, where it was reported a
few months back that a woman successfully applied for a credit card for her
cat. But some of the stories we have heard locally are not much better.

Fortunately some sanity appears to be returning to the market with the
recent adoption by the banks of a voluntary code of good conduct with respect
to the marketing of credit. At the same time, further constraints on lending
have become evident with the introduction of the National Credit Act in June,
when stricter requirements were imposed on lenders to determine the
creditworthiness of borrowers. Although this may have a dampening affect on
certain categories of credit extension and introduce more careful risk
assessment, we do not believe that this will replace the need for monetary
policy restraint.

Economic growth and inflation

Excessively high economic growth brings with it potential inflationary
pressures to which monetary policy has to be sensitive. In each of the past
three years, the economy has been growing at a rate of five percent, which is
high compared to our experience of the past 25 years or so. Capacity
utilisation in the manufacturing sector has also reached record high levels.
While we in the Bank are pleased by this strong growth performance, we have to
ensure that this growth does not affect inflation adversely.

In economic terminology this refers to the concept of potential output,
which at a simplified level is the rate at which the economy can grow without
generating inflationary pressures. From a monetary policy perspective we have
to consider whether our current growth experience is inflationary. In other
words, are we growing in excess of our potential output?

Unfortunately, estimating potential output is not a simple task and the
estimates are subject to a high degree of uncertainty. Current research in the
Bank estimates that the potential output of the economy is now at least around
4,5 percent per annum. This can be compared to estimates of around three
percent or less during the 1990s. If this estimate is correct, it implies we
are currently growing at a rate above potential. This is not necessarily a
threat to inflation if this growth entails the necessary microeconomic reforms
which will allow for a higher potential growth rate. However if the growth rate
does not improve the productive capacity of the economy, there could be
inflationary consequences. These considerations are not easily observable,
which makes our job that much more difficult.

Wage settlements

In recent weeks, following the publicity emanating from the public servants
strike, much attention has been focused on the risk to inflation from higher
wage settlements. From an inflation perspective, higher real wage settlements
may increase domestic demand pressures. There also may be a cost-push effect as
producers pass on the higher wage costs in terms of higher prices. The latest
consolidated data for the economy as a whole are not yet available but there
does appear to be a worrying trend towards higher nominal wage settlements.
However we should also bear in mind that at this stage real wage settlements do
not appear to be significantly higher than in the recent past. In other words,
higher nominal wage settlements are simply compensating workers for the higher
trend in inflation. It is therefore important that we reverse this inflation
trend.

Furthermore, we also need to focus on the increases in unit labour costs,
that is, the nominal wage increases adjusted for increases in labour
productivity. The latest data indicate that in the first quarter of this year
unit labour costs in the non-agricultural sector increased by 2,2 percent
compared to the first quarter of last year of 2006. So although wage growth
poses an increasing threat to the inflation outlook, it does not appear to be
out of control.

Administered Prices

When we initially adopted the inflation targeting framework, one of the
constant laments we had at that time was the challenge posed to us by the range
of administered or regulated prices in the economy which remained stubbornly
high. Administered price setting by the regulators eventually became more
market-related, and from around 2003 till recently this category did not
feature among the list of our concerns. Unfortunately this situation is
changing. The administered price index (excluding petrol) has been increasing
steadily in recent months. The API excluding petrol increased at a year-on-year
rate of 5,6 percent in June this year, compared to 4,3 percent a year ago. Top
of our list of concerns are the proposed electricity price increases requested
by Eskom. These increases are of the order of 18 percent for the next two
years.

While we fully appreciate the need to expand electricity generating capacity
in the country as a matter of urgency, it is essential that alternative
financing options be carefully considered as these will have significant
implications for inflation and monetary policy. A similar argument can be made
for other areas where infrastructure expansion will take place. It is accepted
that increases in tariffs or indeed excise or fuel taxes should not be regarded
as 'true' inflation. However if they are a regular annual feature of the
economic landscape they become an integral part of the inflation process which
we have to take account of. We are mandated to achieve a particular inflation
outcome, but to achieve this, government and other regulators need to take
account of this in their price setting.

Conclusion

I have attempted to outline some of the major concerns currently facing the
MPC. This does not mean that other issues are not concerning us, or that we see
no positive developments. I should also emphasise that we do not have a target
for any particular variable; our focus is on the overall inflation outcome,
which is the outcome of the interaction of a number of variables. In other
words, we cannot say what the appropriate growth of household consumption
spending should be, as it would need to be seen in the context of the behaviour
of the other factors that determine the inflation rate.

Monetary policy-making is not a simple matter of adjusting an interest rate
lever in an automatic way. Because monetary policy has to be forward-looking,
and because the future is inherently uncertain, we have to make decisions on
the basis of our best estimates of the most likely outcomes. Despite these
difficulties, we will continue to strive to achieve our mandate and bring
inflation back to within the inflation target range. This will be the true test
of the success of monetary policy.

Thank you.

Issued by: South African Reserve Bank
2 August 2007
Source: South African Reserve Bank (http://www.resbank.gov.za)