We disagree. To be sure, the major contributions from inventories to growth are over, because output has caught up to final sales. But we see four reasons for optimism: First, we think that a spring surge in imports temporarily boosted inventories; that is now unwinding. Second, inventories look lean in relation to sales. Third, prices at the wholesale level are accelerating, giving producers an incentive to hold inventories, if not add to them. Fourth, and most important, final demand now seems to be quickening again. As a result, we think that increased demand over time will translate into both production and inventory gains.
Pace of accumulation is only one metric. The bears look to the pace of accumulation as prima facie evidence that an inventory problem is surfacing. They contrast the spring and early summer inventory surge with the sales slowdown in the same period. For example, the book value of durable goods inventories advanced by 2.2% in the three months ended in August, twice as fast as durable goods shipments. Also, using a more economically meaningful measure, manufacturing and (retail and wholesale) trade (M&T) inventories in real terms accelerated to a 2.5% annual rate in the three months ended in June, while real M&T sales contracted at a 0.5% rate in the same period. This torrid pace of inventory accumulation is clearly unsustainable even as sales growth improves.
Import surge: It had to go somewhere. The inventory story is not just about domestic production; a spring surge in imports probably accounted for a major part of spring stockbuilding. In fact, the $8.5 billion (not annualized) surge in real merchandise imports over April-June just about matched the $8 billion increase in real inventories in the same period. Some of that import surge likely resulted from Chinese exporters accelerating shipments to beat the removal in July of export tax rebates on hundreds of items, including steel, fertilizers, glass, rubber and medicine. Those Chinese exports shipped here likely wound up in US inventories with a lag; although real imports fell by 2.2% in July, we estimate that real M&T inventories jumped by 0.7% in July.
How to measure ‘lean'. Beyond the import surge, moreover, analyzing growth rates for sales and inventories ignores the question of whether the stock of inventories was lean or top-heavy in the first place. Of course, how to measure whether stocks are lean in today's context is tricky. Inventory-sales (I-S) ratios are close to historical lows in real terms, and lower than in past cycles. For example, we estimate that the real I-S ratio in manufacturing and trade excluding motor vehicles stood at 1.288 in July, 14 months after the current recovery began. In contrast, it stood at 1.323 in November 2002, 14 months after that recovery began.
But the absolute I-S ratio level is not informative in a just-in-time (JIT) world of global supply chains. JIT inventory management techniques have fostered a secular downtrend in most I-S ratios, as companies have increased inventory turns and pushed back the holding of stocks onto suppliers. We estimate that this downtrend probably removes about three-quarters of a day's supply, or a little more than 2%, from desired inventory holdings each year. As we see it, I-S ratios early this year were well above the level indicated by that downtrend, suggesting that stocks were too high in relation to sales.
Three reasons to accumulate inventories. That was then. Judging by the real M&T inventory-sales ratio excluding motor vehicles in June, those ratios are now back to or below trend, suggesting that there is no inventory overhang. Other metrics support that view. For example, the ISM survey of customer inventories was well below its 16-year average of 45.9% through 1H10, and it remained below the average in August at 43.5%.
We think that three factors may drive firms to boost inventories slightly in today's environment. First, the ISM canvass of order policies for production materials, which maps the extent of ‘hand-to-mouth' inventory policies, sank to unsustainably low levels in the recession as credit dried up and uncertainty reigned. In the past year, that survey has rebounded sharply, hinting that inventories had become too low relative to any improvement in sales from low levels. Second, while talk of deflation permeates the air, wholesale prices are rising again, especially for materials and supplies, which gives companies an incentive to accumulate stocks. Prices of base metals like copper and aluminum are soaring; the CRB raw industrials price index has risen by 8.5% in the past 90 days.
Third, with the cost of financing inventories no higher than the prime rate, and for many companies at the 20-30bp one-month commercial paper rate, price increases mean that the real cost of carrying inventories is sharply negative. In effect, purchasing managers can now exploit the ‘carry trade' in stockbuilding by financing inventories. Negative carrying costs mean that I-S ratios even slightly above the JIT trend may currently be ‘optimal'. As evidence, non-financial commercial paper has jumped at a 41% annual rate this year, following a sharp slide in 2009. And even commercial and industrial loans held by banks have stopped declining.
Empirical results support those three factors. As further evidence that those three factors are the critical ones explaining inventories, we estimated an empirical relationship that explains the M&T I-S ratio excluding motor vehicles as functions of all of them. The ‘desired' I-S ratio depends on a time trend (to explain the JIT decline) and the real cost of financing (which takes account of rising wholesale prices). The adjustment to that desired level varies with imports in relation to sales, a four-month distributed lag on the change in M&T sales (to capture surprises in demand), and a stock adjustment term that takes account of the fact that firms don't adjust stocks immediately. All these factors are statistically significant in this relationship, which explains 97% of the variance in inventories relative to sales.
Detroit the rule, no longer the exception. A leaner, meaner Detroit also fits the script. In recent years, two of the three legacy US vehicle makers felt compelled to produce vehicles to generate the billions in cash flow needed to pay for employee benefits. As a result, Detroit struggled to sustain high sales levels using financing and other incentives, and often found vehicle inventories high in relation to sales. More often than not, Detroit would put more money on the hood to clear dealers' lots.
No longer. Forced by competition to stop binging, and freed from the yoke of their pension and OPEB liabilities, Detroit's vehicle producers have gone on a diet and intend to stay lean. They have shuttered capacity and now gear production to demand. This new austerity has recently left vehicle inventories low in relation to the new, lower level of sales. Auto analysts Adam Jonas and Ravi Shanker reckon that light vehicle stocks represent only 47-49 days of sales at end-September, well below traditional industry norms. Clearly, if sales slide significantly, there will again be downside risk to vehicle output. But if sales now run at an 11.5-12 million pace as we expect, there is scope for output to increase. As a result, we think that motor vehicle output will add fractionally to 2H GDP, whereas such output subtracted slightly from overall growth in the spring.
It all comes down to final demand. In the end, the inventory cycle is a temporary cyclical phenomenon; sales and inventories will over time grow in tandem. Whether production declines or rises will all come down to the strength of final demand. If demand is weak, inventories will rise involuntarily and companies will be forced to cut production; indeed, the adjustments are quicker today than they were a decade ago, courtesy of the spread of JIT- and information-based supply and manufacturing chains.
And, as we've long noted, political gridlock in Washington could create major headwinds to demand: If Congress and the Administration allow the Bush-era tax cuts to expire on December 31, the economy will face fiscal drag amounting to more than one percent of GDP, trimming about a three-quarters of a percentage point from growth in 2011.
The good news: Apart from housing, demand in 3Q is picking up, with exports and capital spending - two key areas of weakness in June and July, respectively - rebounding sharply. We believe that the ISM export orders index (55.5% in August, to be updated this week for September), while well off its peak, is still consistent with 7-10% gains in real merchandise exports. Moreover, the sharp 4.1% rebound in August orders for non-defense capital goods excluding aircraft signals that overseas and domestic demand for capital goods is expanding again. In August, notes metals analyst Mark Liinamaa, the AIA Architectural Billings diffusion index for commercial and industrial projects remained in growth territory at 50.6, the fourth consecutive month above 50. That suggests we will see a pick-up in construction activity in mid-2011 and an associated pick-up in materials demand.
Rising Food Prices Push Headline Inflation Higher
We have been expecting a modest acceleration in headline inflation in the AXJ ex-India region into 3Q. However, over the last few months, headline inflation in the region ex-India has been surprising on the upside largely because of higher food prices. Inflation ex-food has remained largely stable. Headline inflation in the region ex India accelerated to 3.3%Y in August from 2.9% in June and 1.8% in Jan-10. The food inflation rate increased to 6.5%Y in August from 5.3% in June and 3.3% in Jan-10. India, on the other hand, saw a deceleration in headline inflation due to a high base effect and moderating food prices on good monsoons. While we do not have the breakdown of food inflation in the CPI as of July, headline CPI inflation in India decelerated to 11.3%Y in July compared with 13.7% in June and 16.2% in Jan-10.
Local and Global Weather Problems
In Asia ex-Japan, except for India, the countries have seen an increase in food inflation. India's food inflation is moderating on a high base. We expect further deceleration, as the summer crop output (harvesting in September-October) is likely to be very strong. In Korea, food inflation increased to 4.1%Y in August from 2.6% in May-10 and 2.4% in Jan-10, on abnormal weather conditions with the spring season unusually short and the summer unusually hot, coupled with heavy rainfall. Indonesia has seen an increase in food inflation from 6.3% in Jan-10 to 11.9%Y in August on unseasonal heavy rains during what is normally the dry season of July-September. Food inflation in Thailand accelerated from 3.2% to 7.5%Y during the same period as cultivation and harvests were disrupted by heavy rains and floods in some areas. Bad weather also affected food inflation in China, mainly in the form of higher fruit and vegetable prices. Food inflation rose to 7.5% in August in China from 3.7% in Jan-10.
In addition, the CRB commodity and food indices have continued to trend up on bad weather conditions internationally (a heat wave in Russia, dry weather in Kazakhstan, the Ukraine and the European Union, and flooding in Canada). Indeed, the CRB commodity and CRB food index increased by 5.7% and 8.5%, respectively, in September, after increasing 21% and 28% in the previous 12 months. The CRB commodity index is already near the pre-crisis high, while the CRB food index is just 5% below it. Even though most of the large countries in the AXJ region are largely self-sufficient for staple food items, they tend to import a few selective farm products. For example, while India and China are net food exporters, both are dependent on crude palm oil imports. Hence, in the past we have seen that a rise in international food prices does tend to pull up food inflation in the region with a lag.
Food Weighting in Emerging Asia's CPI Is High
Emerging Asian countries with low per capita income including China, India, Indonesia and Thailand have CPI weightings of 33-46%. Hence, persistently higher food prices pose a bigger risk of a rise in inflation expectations and wages in these countries as compared with higher per capita income economies on a relative basis. While job growth was affected by the latest global financial crisis, with GDP growth back to trend line and employment levels having recovered sharply, the risk of a rise in inflation expectations is significant. While employment statistics in the region are not very transparent, given the GDP growth trend, it appears that employment growth should have been strong.
Relying Too Much on Imports Not Really an Option for Large Population Countries
China, India and Indonesia together account for 40% of the global population. Any small increase in demand from these countries in the form of imports tends to push up global prices. The recent crop failure in India and the country's attempt to import sugar are a case in point. Moreover, there are some crops that are peculiar to local markets with very little global supply. For instance, in the case of India last year, the country fell short of pulses (lentils), and it was not really possible to import the crop even if the government had wanted to. Indeed, the top four (in terms of population) countries in the region (China, India, Indonesia and Thailand) are all net exporters of food items. All four countries tend to maintain inventories for staple items like rice and wheat, and have public distribution systems to ensure availability of these essential items at a reasonable price. Most countries in the region subsidize food for the poor.
Any Rise in Oil Prices Could Push Food Prices Higher
The sharp increase in the CRB food index in Aug-Sep-10 came despite the decline in crude oil prices due to concerns about a slowdown in the US (crude oil prices have declined to US$77 per barrel compared with US$82 at the beginning of August). We think that any quick acceleration in oil prices could pose a significant risk in light of already rising food prices. According to our commodity strategist, Hussein Allidina: "Crude oil prices have been surprisingly range-bound and are likely to remain so into mid-October, before ultimately trading higher as refiners exit maintenance, bolstering crude demand and tightening both inventories and spreads" (see The Commodity Call: The Time to Buy Crude, September 6, 2010). Before then, however, Hussein expects "crude oil prices to continue to trade in the upper end of a roughly US$70-80 per barrel range, moving higher as the market's appetite for risk increases".
Strong Growth and Slow Pace of Policy Exit Increases the Risk of Higher Pass-Through of Food Inflation
A combination of higher oil prices, strong job growth, narrowing output gap and delayed policy exit could increase the risk of pass-through of high food inflation into non-food inflation. India has already seen some pass-through of the sharp rise in food prices over the last 12 months due to crop failure, as domestic demand was strong. As discussed earlier, some of the other countries in AXJ have witnessed a meaningful rise in food inflation only recently. Although there is no clear evidence of major pass-through to wages in the region except in India, this is clearly a risk, if food inflation remains high for longer. We expect AXJ GDP growth to grow at a high rate of 8.2% in 2H10 and 2011. Moreover, concerns about a deeper slowdown in G3 and its extremely accommodative policies have meant that the AXJ policymakers are indirectly compelled to maintain policy rates at close to all-time lows. Even after the recent pick-up in the pace of policy exit, policy rates are near all-time lows and much below the neutral levels.
Who in the AXJ Region Faces Maximum Food Inflation Risks?
We believe that both Indonesia and Korea's central banks are delaying the policy rate hikes even as growth remains strong and headline inflation is rising. However, we think that a further rise in inflation for three months or so could push these two central banks to accelerate policy rate hikes. We believe that India is likely to see moderation in food inflation and the central bank has already hiked policy rates meaningfully. We expect the RBI to continue to lift policy rates. We see low risks of a food inflation spike in China and policymakers have already taken measures other than interest rate hikes to manage the moderation in domestic demand. Malaysia and Taiwan have relatively low food inflation and even if food inflation were to rise further from current levels, we do not see a risk of non-food inflation pressures warranting a monetary policy response. We expect policymakers in both countries to take administrative measures to respond to food price inflation. Below is the summary of food and overall inflation outlook for the key economies in the region:
China (Qing Wang/Steven Zhang): Over the past few years, China has demonstrated a good supply response to match the structural rise in demand. China last saw a major rise in food and headline inflation in 2008 due to a sharp spike in global food and fuel prices at a time when domestic demand was strong. However, since the global credit crisis, both food and headline inflation has been low until recently. Food inflation accelerated to 7.5%Y in August from 3.7%Y in Jan-10 and headline inflation accelerated to 3.5%Y from 1.5%Y during the same period. However, we are less concerned about a major rise in food inflation going forward, as the government holds adequate inventory of essential food grains. Moreover, policymakers in China have already implemented a significant amount of tightening of domestic demand, reducing the risk of a major rise in inflation expectations. The tepid global recovery also constitutes substantial headwinds containing inflationary pressures, especially those stemming from commodity prices. We expect food inflation and headline inflation to remain moderate. We expect a fine-tuning of the policy stance to effect a modest easing in early 4Q10.
India (Chetan Ahya/Tanvee Gupta): India has been witnessing a steady rise in food inflation since 2005 due to a structural rise in demand and a less-than-optimal domestic supply-side response. Last year, unusually bad weather resulted in a sharp decline in food output, resulting in a major spike in food inflation. However, the summer crop is expected to see strong growth this year as the weather has been good. Headline WPI moderated to 8.5%Y in August from the peak of 11% in Apr-10. Food inflation (primary plus manufactured), as per WPI, decelerated to 10.6% in August from 20.2% in Feb-10 on a base effect and a moderation in food prices. We believe that food inflation will decelerate to close to 2% by end-December 2010. However, we expect non-food inflation to remain above the central bank's comfort zone over the next six months as strong domestic demand at a time when trailing investments and capacity creation have been weak has resulted in tighter capacity utilization. A spike in food inflation in the last 12 months also appears to have been reflected in non-food inflation. While the RBI has lifted the policy rate by 125bp from the trough, it remains a bit lower than neutral levels. In our base case, we expect inflation (WPI) to moderate to below 6% by March 2011 as long as the central bank stays on course to lift policy rates.
Korea (Sharon Lam/Jason Liu): Unlike India and China, where food has a weighting of 46% and 34% in CPI, respectively, Korea has relatively lower food weighting of 27%. However, Korea's self-sufficiency for grains is only around 30% and therefore it is more vulnerable to higher international commodity prices. Food inflation rose recently, to 4.1%Y in August from 2.6% in May and 2.4% in Jan-10. Headline inflation, however, remained stable at 2.3-3.0% in the same period. In addition, Korea has seen strong recovery from the crisis, driven by its export competitiveness, currency depreciation and policy stimulus. Moreover, Korea has one of the highest inflation sensitivities to international crude oil and food prices. With global food prices on the way up, we see increased risk of a rise in Korea's food inflation. Moreover, while the Bank of Korea has already begun its policy rate normalization cycle, we believe that the central bank has been slower than warranted, increasing the upside risk to our inflation forecast of 3.3% by 1Q11.
Indonesia (Chetan Ahya/ Deyi Tan): Indonesia has had a history of high food inflation. In the last five years, food inflation has averaged 11%, as the supply-side response tends to be weak. From a structural perspective, high food inflation in Indonesia can be attributed to the government's policy management. With a weighting of 33% in CPI, food inflation has tended to translate into higher non-food inflation at times when domestic demand growth is strong. Food inflation accelerated to 11.9% in August from 6.3% in Jan-10 due to bad weather affecting crop output. This drove headline inflation up to 6.4% from 3.7% during the same period, pushing up inflation expectations. Bank Indonesia (BI) has left the policy rate unchanged at 6.5% after cutting it aggressively during the global credit crisis. In its September monetary policy meeting, BI decided to move by controlling the ‘quantity' of liquidity rather than its ‘price', raising the primary reserve requirement ratio to 8% from 5%. BI appears to be taking comfort from the fact that core inflation is low at 4.2% and the government has promised administrative actions to bring down food inflation. Moreover, BI is concerned about rate hikes attracting more capital inflows. (Note that Indonesia has one of the highest interest rates in the region and one of the more open capital accounts.)
Moreover, as domestic demand remains strong, we see high risk of food inflation passing through into non-food inflation unless there is major deceleration in food inflation soon. Inflation expectations have been on a general uptrend. In addition, Indonesia has relatively lower stockpiles and the ability of policymakers to manage food price inflation through administrative measures is weak. The government is indicating that food inflation will moderate soon. However, if food inflation stays higher for the next three to four months, we see a high risk of core inflation following food inflation, forcing the BI to lift policy rate quickly.
Thailand: (Deyi Tan/Shweta Singh): Thailand is one of the major foods exporters in the AXJ region, particularly rice. Yet, food inflation in Thailand rose from 3.2% in Jan-10 to 7.5% in August due to dry weather conditions. The increase was largely in vegetables and fruits and, to some extent, in rice, flour and cereal products. The concurrent increase in international food prices also matters for Thailand even though Thailand is a net food exporter, as it affects producers' incentives on whether to cater to domestic or export markets. However, given the higher reserve stockpile, we think that Thailand's policymakers are better placed to mitigate price increases as compared with other ASEAN economies such as Indonesia. We see only a moderate risk of pass-through of food inflation into core inflation. In general, inflation expectations in Thailand remain low. Core inflation also stayed relatively flat, at 1.1-1.2%, in the past four months. In light of the healthy growth momentum, we expect the Bank of Thailand to continue with policy rate normalization, which has just started. The policy rate has been raised from 1.25% to 1.75% in the past two meetings. However, for now, we do not expect food inflation to lead to an accelerated pace of rate tightening.
Malaysia (Deyi Tan/Shweta Singh): Headline inflation in Malaysia increased from 1.3% in Jan-10 to 2.2% in August, in line with our expectations. While food inflation increased from 1.2% in Jan-10 to 2.9% in August, it is still much lower than that in other ASEAN economies such as Indonesia and Thailand. Prices of some of the essential food items - such as sugar, flour, bread and cooking oil - are administered by the government. Hence, we believe that Malaysia is less likely to see any a major rise in food inflation. We also see a low probability of Bank Negara Malaysia (BNM) responding to food inflation with further monetary policy tightening. BNM was one of the first central banks in the region to embark on policy normalization, raising rates by a cumulative 75bp to 2.75%. With the last rate hike in July, we believe that the normalization process appears to have come to an end. Moreover, current core and headline inflation do not point to any significant inflation pressures in Malaysia, in our view.
Taiwan (Sharon Lam/Jason Liu): Although Taiwan's food inflation has increased in recent months, due to supply shocks caused by the hot and rainy weather, it has been much slower compared to other countries in the region. Headline inflation in Taiwan has increased from 0.3%Y in Jan-10 to 1.3% in July and -0.5% in August (mainly due to an abnormal base effect caused by typhoons last year). Food inflation increased from close to zero in Jan-10 to 2.3% in July and -3.4% (again because of an abnormal base last year) in August. Moreover, Taiwan has a relatively lower weight (26%) of food in its overall inflation compared with other economies in the region. Taiwan is nearly 90% self-sufficient in rice products, and thus the grain price in the international market has a relatively smaller impact on domestic grain prices. Like Malaysia, even in Taiwan, we do not see any acceleration in the policy rate on account of a rise in food inflation higher than that reflected in our forecast. If food prices continue to rise, we believe that the government may consider administrative measures and increase the food subsidy.
Will Food Inflation Become a Structural Issue for Developing Asia?
Will food prices in the world rise as China, India and Indonesia reduce poverty with a steady increase in food demand? We believe that the trend in China so far has established that the right supply response can ensure that food inflation does not become a hurdle in the process of development. China has achieved a large increase in farm productivity over the last three decades to ensure that food inflation stays within a manageable range. The risk is that India and Indonesia fail to improve productivity to match the structural rise in demand. Indeed, India and Indonesia's food inflation has started to move higher over the last few years compared to that in China.
Productivity levels in India and Indonesia are currently significantly lower than that in China. For instance, yield per acre for rice in India and Indonesia was 51% and 75%, respectively, of that in China as of 2008. Moreover, both India and Indonesia recently suffered from weak public distribution systems, which lack full effectiveness to ensure availability of essential food items to the poor during major crop failures. We believe that food price inflation could see a structural challenge if the supply response in India and Indonesia lags the steady rise in demand. The good news is that in India, the government is already working on measures to improve productivity and the public distribution system.
Read Full Article »
