The September statement also included a modest strengthening of the pledge to do more if conditions warrant, as follows: "The Committee will continue to monitor the economic outlook and financial developments and is prepared to provide additional accommodation if needed to support the economic recovery and to return inflation, over time, to levels consistent with its mandate" (emphasis added). This marks the first change to this wording since it was introduced in March 2010. So, the FOMC is saying that it will do more if the recovery falters and inflation does not show signs of returning to desired levels "over time". This provides policymakers with the flexibility to implement asset purchases on a data-dependent basis going forward.
Growth data still take precedence. The change in the inflation language, while important, does not, in our view, signal an elevated emphasis on the incoming inflation data itself as a possible trigger for asset purchases. To be sure, the inflation data do matter, but the growth indicators matter more because, from the Fed's perspective, the pace of growth in economic activity is a leading indicator of inflation. Here is a key excerpt from Bernanke's Jackson Hole speech that helps explain the perceived link between growth and inflation:
"...the FOMC will do all that it can to ensure continuation of the economic recovery. Consistent with our mandate, the Federal Reserve is committed to promoting growth in employment and reducing resource slack more generally. Because a further significant weakening in the economic outlook would likely be associated with further disinflation, in the current environment there is little or no potential conflict between the goals of supporting growth and employment and of maintaining price stability" (emphasis added).
Renewal of asset purchases unlikely, but it's a close call. We believe that the incoming growth data will be sufficiently positive to prevent the reintroduction of asset purchases in coming months. Indeed, since September 1, our tracking estimate for 3Q GDP has risen from +1.7% to +2.6%. But it's a close call, and the FOMC could act at any time if the data disappoint. So, some analysis of an asset purchase program seems warranted. Here is a brief Q&A:
Q) What is the trigger for renewed asset purchases?
A) This is a tricky one to answer because it's the assessment of tail risk (not the baseline forecast) that really matters. If Fed officials perceive a significant probability that recovery is stalling out, then they have to do something. To be more specific, we suspect that a perception of a one-third or higher probability that GDP growth will slip below the +2% ‘escape velocity' pace for a few quarters would trigger another round of asset purchases.
Q) What is the timing?
A) Could happen at any time (on an FOMC meeting day or even between meetings) since it is data-dependent. In particular, we do not believe that the timing of the November FOMC announcement (the day after the midterm elections) has any bearing on the likelihood of a move at that time.
Q) If implemented, what would the program look like?
A) There is a wide range of options. The Fed could merely revive the LSAPs and announce a targeted amount of buying over a certain timeframe (e.g., $500 billion, $1 trillion, $2 trillion over 3, 6, 9 or 12 months). Alternatively, the Fed could adopt a more flexible approach - recognizing that in the first round of the LSAP, it had to scale back the amount of agency purchases and probably wound up buying too many MBS. Former Fed Governor Don Kohn has suggested a program that would entail a modest starting size that could then be scaled up depending on incoming economic data. Finally, we actually prefer a rate cap approach (along the lines outlined by Bernanke in his 2002 ‘Deflation' speech) whereby the Fed specifies a target for 10-year Treasury yields as opposed to a quantity of purchases. However, we admit that this approach does not seem to be getting much traction inside the Fed these days.
Q) What will be the impact on the markets and the economy?
A) Obviously, this is a critical issue but it is also highly uncertain. Suffice to say that we are skeptical that the economic effects would be large. The Fed might well be successful in driving Treasury yields 25 or 50bp lower from here, but the economic benefits of that outcome alone are minimal in the current environment.
Before analyzing the potential economic impact in greater detail, it's worth highlighting the difference between the Fed's past actions and the asset purchases that are now under consideration. For some reason, the markets seem to equate the term QE with central bank asset purchases. In fact, QE refers to any form of central bank balance sheet expansion. In the US, QE 1 actually started with the introduction of the Term Auction Facility (TAF) and foreign central bank swap lines - and all this occurred several months before the Fed started buying MBS and Treasuries. The initial round of quantitative easing had a clearly defined objective - to provide sufficient liquidity to rein in interbank funding pressures that were contributing to a tightening of credit availability. The Fed's actions were successful in reining in the spike in measures of liquidity stress, such as the Libor/OIS spread.
The second phase of QE, announced in November 2008 and implemented starting in early 2009, involved an effort to drive mortgage rates lower in order to spur refinancing activity and improve housing affordability. Again, this goal was (at least partially) achieved as mortgage rates plummeted. I say ‘partially achieved' because the full impact of the drop in mortgage rates is not making its way through to the consumer due to blockages in the refi channel. We believe that future Fed action aimed at driving rates lower would be far more powerful if it were combined with something along the lines of our ‘Slam Dunk Stimulus' proposal (also, see the op-ed in last weekend's New York Times by Glen Hubbard and Chris Mayer that endorses a similar type of plan). In any case, the main point is that the Fed's easing measures conducted in late 2008/early 2009 were well defined and largely achieved its objectives. In our view, that is not the case with the next phase of asset purchases now under consideration.
Impact of new asset purchase program? Not much bang for the buck. Former Fed Governor Larry Meyer maintains a large-scale macro-econometric model of the US economy that is widely used in the private sector and in public policy-making circles. These types of models are good for running ‘what if?' simulations. Meyer estimates that a $2 trillion asset purchase program would: 1) lower Treasury yields by 50bp; 2) increase GDP growth by 0.3pp in 2011 and 0.4pp in 2012; and 3) lower the unemployment rate by 0.3pp by the end of 2011 and 0.5pp by the end of 2012. However, Meyer admits that these may be "high-end estimates". Some probability of a resumption of asset purchases is already priced in, and thus a full 50bp response in Treasuries is unlikely. Moreover, a model such as Meyer's is based on normal historical relationships and therefore assumes that the typical transmission mechanisms are working. For example, a drop in Treasury yields would lower borrowing costs for consumers and businesses, helping to stimulate consumption, business investment and housing. But there is good reason to believe that the transmission mechanism is at least partially broken at present, and thus the pass-through benefit to the economy associated with a small decline in Treasury yields (relative to current levels) would likely be infinitesimal.
Summary
The South African Reserve Bank released its September 2010 Quarterly Bulletin on September 22, with details on the balance of payments and the demand side of the economy. The current account balance surprised markets to the upside in 2Q10, coming in at just -2.5% of GDP, versus consensus expectations of -3.2%. Our forecast was for a -3.0% print. The improvement from -4.6% in 1Q10 was largely driven by a surplus on the visible trade balance and World Cup-related inflows - although the real surprise for us was the buoyancy of non-World Cup-related tourist activity. Smaller net transfer payments to the SACU free trade community trimmed the deficit too.
The current account deficit was more than adequately financed by a chunky R31.7 billion surplus on the financial account of the balance of payments, resulting in a widening of the basic balance to R16.2 billion. In the face of such a surplus, it is hardly surprising that the trade-weighted ZAR was some 4.5% stronger in the year to 2Q.
The Bulletin also provides a detailed breakdown of the demand side of the economy. Household consumption held up pretty well on account of World Cup-related spending, employment gains in 2Q10, significant above-inflation wage rewards, and a lower debt service cost. Notably, private fixed investment turned positive for the first time since 4Q08, supporting the already buoyant rate of public sector capital outlays.
Mining Exports Respond to Higher Commodity Prices...
In line with our forecast, total exports rose to R611.0 billion, as volumes of mining products accelerated by 5%Q (seasonally adjusted and annualised), mainly on account of strong demand from Europe (mostly base metals) and China (minerals). According to the Bulletin, gold export volumes were up 9.2%Q. Exports of motor vehicles and transport equipment were also boosted by a recovery in external demand. This was in line with our view that external demand conditions are the most important driver of South African manufactured exports (see South Africa: Gauging Susceptibility to the Vagaries of European Growth, June 21 2010), and underpins our view of a further recovery in manufactured exports in 2010. Export prices also rose some 5.5%Q, thanks to higher gold and platinum prices (up 7.7% and 4.4%, respectively).
...While Imports Undershoot our Expectations
With regards to imports, the Quarterly Bulletin suggests that we may have been overly aggressive in forecasting a rather sharp recovery in 2Q10. The published reading of R597.8 billion undershot our R607.2 billion estimate by R9.4 billion. The Bulletin mentions that oil import volumes fell, partly explaining this undershoot. Overall, merchandise import volumes rose a modest 3.1%Q, of which mining imports were flat for the most part, while manufactured import volumes rose by 5%Q. Thus, although we expected the trade balance to swing from deficit to surplus in 2Q10, the actual out-turn of R13.2 billion was much higher than our forecast of R5 billion, and accounts for most of the upside surprise in the current account balance.
Looking forward, we expect the import bill to rise in 2H10, as capital imports relating to construction projects that were put on-hold just before the World Cup resume, and as the stronger ZAR supports a recovery in consumer imports. This should limit the beneficial impact of the 2Q10 trade surplus on our 3.8%-of-GDP CAD estimate for the year as a whole.
Tourism Inflows Exceed Our Expectation
On the invisible services account, transport receipts rose from R10.25 billion in 1Q10 to R13 billion in 2Q10, thanks largely to payments by foreigners for air travel services provided by South African airlines to the first port of entry into the country. This reading was marginally lower than our forecast of R15 billion - presumably because the SARB's preliminary estimate of 200,000 soccer tourists turned out to be lower than our 350,000 estimate. Travel and other receipts (including accommodation, transportation of foreign tourists within the country, food & beverages, etc.) came in at R100.5 billion - higher than our forecast of R95 billion.
According to the SARB, total spending by tourists amounted to some R15 billion non-annualised, of which R3.5 billion could be associated with the sporting event. The latter figure is less than our 2Q10 estimate of World Cup-related inflows of some R6.2 billion, although broadly in line with our estimate once we adjust for the 200,000 tourists the SARB mentions (for a detailed discussion, see page 7 of South Africa Chartbook: Gauging the Consumption Impact of the 2010 FIFA World Cup, January 26, 2010). Thus, while total travel and other inflows of R100.5 billion came in higher than our R95 billion estimate, the overshoot appears to have been in the non-World Cup-related component, and could be due to an over-adjustment by the SARB for seasonality (given that travel and other receipts are usually weak in the second quarter of each year).
Dividends Still Playing a Central Role
At R53.2 billion, income payments to the rest of the world were broadly in line with our estimate. A modest downside surprise in dividend receipts on non-direct investments and weaker-than-expected dividend payments on direct investments to foreigners explain the relatively modest R2.4 billion deviation from our forecast. Interestingly, dividend payments on non-direct investments maintained their strong 1Q10 reading with another buoyant outflow of R11.8 billion - the most since 1Q09. After these payments literally collapsed in the wake of the recession, investors were always set to be rewarded as corporate profitability improved. We continue to expect total dividend payments to re-emerge as a significant driver of future deficits on the current account as the business cycle progresses.
Finally, net transfer payments registered a rather low R10.7 billion, thanks largely to a sharp fall in central government payments to the SACU region. These payments fell short of our already conservative R20 billion assumption, as trade volumes within the free trade zone declined.
Basic Balance Swells Even Further
The financial account registered a large surplus of R31.6 billion, being the sum of net inward direct, portfolio and other capital transactions. This huge sum was more than enough to comfortably finance a R15.4 billion current account deficit, plug the R9.9 billion haemorrhage in unrecorded transactions, and still allow the SARB the luxury of being able to accumulate a further R6.3 billion in foreign exchange reserves!
The breakdown of the financial account continues to reflect an undesirable funding mix for South Africa, however: net portfolio inflows were some 1.7x the current account deficit, while net direct investments were actually negative on the quarter.
The Bulletin mentions that non-residents acquired an equity stake in a South African mining company during 2Q10. We suspect that this refers to the purchase of a 51% stake in Wesizwe Platinum Ltd by China's Jinchuan Group Ltd and the China-Africa Development Fund, for US$227 million. Looking forward, we expect the ratio of FDI to portfolio capital to improve, thanks largely to two large FDI transactions. These deals could be large enough fund the most part of the 2H10 CAD, and keep the rand overvalued for longer.
Finally, the Bulletin mentions that other investments held by the domestic banking sector declined by R10.5 billion as domestic banks repatriated some of their foreign exchange holdings, and scaled back loans and advances to foreign counterparts. Similarly, foreign banks withdrew R5.2 billion of capital from the local banking system. On the whole, net other investments registered a positive inflow of R8.6 billion in 2Q10.
GDP Breakdown Portends an Improving Macro Backdrop
As alluded to in the September MPC statement, household consumption expenditure moderated to 4.8%Q compared to the sharp 5.7%Q rebound in 1Q10. Notable growth was registered across all consumption categories, particularly in the durable goods component (up a staggering 37.8%Q) where it appears consumers stepped up their purchases of transport equipment, furniture, household appliances and other recreational goods. Although locals may have scaled back on the pace of semi-durable and non-durable consumption (particularly clothing & footwear, vehicle components, and entertainment goods), the data suggest that total expenditure on these categories by foreigners, in South Africa for the World Cup, must have been pretty strong.
In line with international best practice, the SARB subscribes to the guidelines of the 2008 United Nations System of National Accounts (SNA) for National Account reporting. According to paragraphs 9.79-80 of the SNA, in order to calculate total household final consumption expenditure, it is important to adjust total household expenditures within the economic territory by adding expenditures by residents abroad and subtracting that by non-residents within the economic territory. Due to the paucity of data, most statistical agencies are unable to make these adjustments to each type of consumer spend (i.e., durables, semi-durables, etc.). Instead, the miscellaneous component of the services basket is adjusted by the full amount, and an equal and opposite adjustment is made to net exports, thereby keeping the overall impact on GDP unchanged. We believe this corrective measure was the chief driver behind the 7.7%Q contraction in services spend over the quarter.
On the whole, the consumer appears to be recovering gradually. Employment statistics released by Statistics South Africa earlier this week show that formal employment rose by a headcount of 41,000 in 2Q10, not much in light of the near 1 million job losses, but a turnaround nonetheless. At the same time, the Bulletin shows real disposable income rose 4.8%Q on account of above-inflation wage rewards and a decline in the debt service cost (to 8.0% of disposable income from 8.2% previously). In fact, private sector wage settlements ranged between 10.4% and 26.4% in the year to 1Q10. In the public sector, the average settlement was as high as 16.6%. As we showed in our 3Q10 Chartbook, however, lower and middle-income consumers are the weakest link, and certainly not at the same stage of the recovery as their upper-income counterparts (see South Africa Chartbook: Consumer Spending Blues, August 25 2010). We therefore remain cautious on household consumption expenditure in the short term.
Elsewhere, the Bulletin showed that government consumption steamed ahead in 2Q10 - not only on account of World Cup-related spending, but also due to the acquisition of four military aircraft and the payment of higher salaries and wages. Of course, with the current round of public sector wage settlements not completely finalised, we believe that the full impact of the state's 7.5% wage offer will only be reflected in 3Q10. We therefore remain comfortable with our out-of-consensus forecast of a 6.1% increase in government consumption in 2010.
Regarding fixed capital formation, private fixed investment turned the corner for the first time since the end of 2008, rising 0.4%Q. This is an admittedly modest - but nonetheless encouraging - development, considering its large share in total investment. Capital outlays in the mining sector (coal mines in particular) and further enhancements to the country's communications network received explicit mention in the Bulletin. Private investment remains negative on a year-over-year basis for the time being, but the momentum has clearly turned the corner.
The share of private investment's share in total GDFI has fallen from 67% at the start of 2008 to 59% presently - mainly as a result of the rapid pace of public sector capital formation (now 26% of total GDFI from just 17% previously), which rose 3.2%Q in the second quarter. In addition to airports, roads and railways, most of which is already at advanced stages - if not already completed - public investment is also being supported by capital expenditure on the New Multiproduct Petroleum Pipeline from Durban to Gauteng.
Finally, the pace of inventory drawdown stabilised to R7.1 billion, adding 0.3 percentage points to the 2.3%Q increase in gross domestic expenditure in 2Q10.
On a sectoral basis, the manufacturing, trade and transport sectors were responsible for the continued drawdown in inventories, suggesting that existing stockpiles remain the preferred method of meeting current demand.
Read Full Article »