Market Research -- China Wheat-Flour Supply-Chain

Recent Trends and Export Prospects

Prairie Grain Portal

Wheat Consumption Trends

Wheat Production Trends

Flour Milling Industry

Wheat Import Outlook


China became the world’s largest grain importer and increased its imports from Canada to become our largest grain export destination, with a 20% share. China’s grain imports increased to 130 MT/year, but mostly driven by feed rather than food needs, with soybean becoming the principal import-crop at close to 75% of the total. Wheat has been a minor share in China’s import-mix; having peaked at 7 MT in the early 2000s and again in the early 2010s, lately it has been in the 4-5 MT/year range, little more than canola but less than sorghum or barley.

China is the world’s 2nd largest wheat producer, also at about 130 MT/year, only 12% less than the largest, EU. Wheat and rice have been the country’s two food staples, consumed in equal quantities. Self-sufficiency has been the goal in both, though achieved in rice, not always in wheat, requiring an average of 10 MT of imports a year through the 1980s and 1990s. Lately China has been producing as much wheat as it consumes, but it continues to import to build up its stocks – largest wheat-reserves in the world, equivalent to one year of its consumption.

We on the other hand are a much smaller wheat-producer, 7th in the world with 30-35 MT/year but export two-thirds of what we produce, making us 5th largest exporter. But more important than our export volume is the quality of our grades or varieties, particularly durum. Going back 20 years, China used to source 60-70% of its wheat imports from us, lately that share has been below 10% – 2-3% of our wheat-exports of late, down from its peak at 16% in the mid-2000s.

Lately our trading relations with China have soured, threatening our grain exports altogether. We are not sure whether we can restore these relations to continue exporting grains to China, but we thought we would revisit some of the opportunities we had identified a few years ago, particularly with respect to containerized wheat exports to the milling industry, which has gone through a technological revival, and needs high grades of wheat to improve flour quality.

Wheat Consumption Trends

Wheat consumption in China has been quite steady of late but we should pay attention to a few finer details. It increased from 102 MT in 2006/07 to 122.5 MT in 2011/12, 20% in 5 years. If this trend had prevailed for another 5 years, now consumption would be close to 150 MT. However, over the next 4 years consumption actually declined to 112 MT; only due to a sharp pick up in 2016/17 that it recovered to 118.5 MT. This is still a fairly significant increase, 16% over 10 years, while the country’s population increased by only 5%. Thus, per capita wheat consumption increased at close to 1% per annum, from 77 kg/year to 86 kg/year.

This rate of increase pales in comparison to China’s GDP per capita, which doubled over the same period, or 7% per annum. It may not be an indication of a strong income effect on consumption, but cannot be dismissed if it were to be sustainable. In 10 years from now, China’s annual wheat consumption would be close to 140 MT, 20 MT more than today. The income effect on wheat consumption is controversial (which we will get into later), but given China’s size we would be amiss to ignore it. We acknowledge at the outset that income effect is not huge, and taken on its own may even be negative. But with development and concomitant rise in incomes, other effects come into play – such as rural-to-urban migration, shifts in culinary habits or traditions, or even the type or grade of wheat consumed.

Below we show wheat and rice consumption in China together with coarse-grains, which are mostly corn with some barley and sorghum in the mix. Wheat and rice are the main food staples (only 5% for animal-feed), which the Chinese consume in more or less same quantities on a per capita basis. Rice consumption has been quite steady over the years at 100 kg/year, while wheat is consumed somewhat less but on a per capita basis, increasing modestly over time. Per capita coarse grain consumption is higher with significant increases over the last 10 years, from 120 kg/year to 180 kg/year, but used primarily for animal-feed.

The above charts display per capita wheat and rice consumption for select countries around the world. The 8 countries shown on the left (all top-10 grain producers but India which is included in the Asian group to the right) are compared to China. Among them only Brazil is a significant rice consumer with almost 40 kg/year per capita; Australia, US, Canada, and Argentina consume in the 11-16 kg/year range, while EU, Russia and Ukraine even less, in the 2-8 kg/year range.

Among these nine countries China’s per capita wheat consumption is only higher than Brazil’s; the US and Argentina consume more than China but by modest margins (14% and 34%). EU’s per capita wheat consumption is almost 3-times that of China’s; Ukraine consumes 2.7-times, Russia 3.2, Canada 3.4 and Australia 3.6. It is interesting to note how much more wheat EU consumes than the US; at 250 kg/year it consumes less than Russia (278 kg/year) but more than Ukraine (234 kg/year). Australia (308 kg/year) and Canada (289 kg/year) consume more than all three, and in this regard are more European than American in their dietary habits.

There is a stark contrast between the two sets of countries shown above. The group to the left (excluding China) on average consumes 184 kg/year of wheat and a mere 13 kg/year of rice, while the Asian group to the right (again excluding China) consumes 67 kg/year of wheat and 99 kg/year of rice. The combined wheat-rice total of those to the right is 197 kg/year compared to the other group at 166 kg/year (in both cases excluding China) – China’s total is 189 kg/year.

Among the Asian group China’s per capita wheat consumption (86 kg/year) is above all the others but South Korea, and even then only slightly (2 kg/year difference). India’s wheat consumption (76 kg/year) ranks behind China and the lowest in the group is Indonesia (39 kg/year) – the other 5 (Japan, Philippines, Thailand, Malaysia and Vietnam) all consume around 50 kg/year. Vietnam takes the top spot in rice consumption with 242 kg/year, followed by Thailand at 174 kg/year, Indonesia at 147 and Philippines at 126. The richest country in the group, Japan, has both the lowest rice (67 kg/year) and combined rice-wheat (111 kg/year) consumption, while the poorest, Vietnam, has the honor of being the highest on both counts.

There is a commonly held, but incorrect belief that there is a strong relationship between income levels and grain consumption – that people will eat more until they overcome their hunger, but as they become more prosperous, they reduce their grain intake by substituting calories from meat, vegetables, and fruits that they can now afford. Other than the comparison we drew above between Japan and Vietnam, there is little in wheat-rice consumption patterns to support this belief. In the chart below we rank the 17 countries we examined above by PPP adjusted GDP per capita. There is no evidence of an income effect on either wheat or rice, or their combined total, and the effect seems to be explained by culinary traditions.

In the highest income group ($37,000 to $57,000 per capita) Australia and Canada consume 3-times more wheat than the US, who they follow on the income-ladder. EU consumes 2.5 times more wheat than US, but like all three above it in the income ladder consumes little rice. Japan and South Korea that are at comparable income-levels to EU, consume considerably less wheat, and even wheat-rice combined. These two high-income Asian countries consume less wheat than the US, but more wheat-rice combined (particularly South Korea, 60% more).

In the middle-income group, Russia has the highest wheat consumption (even higher than EU) but consumes very little rice. Argentina is the 2nd highest wheat consumer in the group (but only little more than the US), but like Ukraine consumes very little rice. Thailand is the highest rice consumer and 2nd highest behind Russia with rice-wheat combined; rice constitutes 75% of its total. China is the 3rd largest in wheat as well wheat-rice combined; though its income level is about the same as Brazil, it consumes more of both wheat and rice. All three Asian countries in the group have more rice in the mix, but Brazil consumes quite a bit of rice as well.

In the lowest income group Vietnam (at the bottom of income ladder) is the highest consumer of wheat-rice combined (3rd highest among all 17). All 4 Asian countries in the group have more rice than wheat in the mix, but Vietnam is by far the highest rice consumer. In total Ukraine ranks 2nd in the group but almost all with wheat; India is the lowest with wheat-rice combined (in equal quantities). This lowest income group has a combined per capita consumption of 168 kg/year compared to 197 kg/year for the high and 185 kg/year for the middle income groups.

In the chart below, we compare GDP per capita (PPP) and combined wheat-rice consumption growth rates from 2006 to 2016, for each of the 17 countries we examined earlier. GDP growth rates are naturally more modest for the advanced economies, except for EU – but we should caution that growth may be overstated due to boundary and other structural changes as EU expands. Growth rates for the middle and low income countries in the western hemisphere (Russia, Ukraine, Brazil and Argentina) were not much better, but all the Asian countries experienced 1.5 to 2.0 times GDP per capita growth (including South Korea which can be regarded advanced). China was still above the rest with 2.7-times growth, followed by India.

Across the board combined wheat-rice consumption per capita increased but only modestly, 5% per year. There were actual declines in Australia (15%), Argentina (11%), Ukraine (6%) and (US (6%). The only significant consumption growth in the western hemisphere was observed in Russia, 9.5% -- exclusively from wheat. Consumption growth was more widespread in Asia – Thailand as much as 31% and Vietnam 25%. Malaysia’s consumption grew 11%, Philippines 9%, India 6.5%, South Korea 6% and even Japan 5.5%. Slowest growth rate among Asian countries was in Indonesia (1.3%), but wheat consumption grew faster while rice declined somewhat. Across Asia, though rice remained the dominant crop, most of the incremental growth was in wheat. As we saw earlier China’s rice consumption did not change much over this period, but overall per capita consumption growth was still close to 8%, naturally all from wheat.

As evident from both cross-country comparisons and last 10 year’s growth trends, there is little to support the belief that grain consumption (in particular wheat and rice) would decline with rising incomes. Consumption is growing albeit much more modestly than GDP per capita growth. Geography plays a key role in the wheat versus rice split, the latter more dominant in countries more conducive to rice production. But wheat consumption is growing faster in those countries where rice is the dominant crop. Also, we must acknowledge that local crop availability has shaped culinary traditions over centuries; still, wheat seems to be making inroads across Southeast Asia where rice has traditionally been the primary food staple.

There are a couple of other points to consider in assessing the future prospects for wheat-rice consumption in China. The country is divided along north-south lines; in terms of both production and consumption, wheat is the more dominant crop in the north and rice in the south. Per capita wheat consumption is 150 kg/year or more in some northern parts (national average 86) while rice consumption more than 200 kg/year in the southeast (national average 102). Also, while westerners generalize over Chinese-food, there are very different regional cuisines – with different local ingredients, some using more rice while others more wheat.

There is also a rural-urban divide; after decades of migration to the cities, now the urban population is larger than the rural. There are fairly sharp income differentials which give rise to different eating habits, be it due to disposal income or city-lifestyles. Per capita wheat-rice consumption is higher in rural areas, and there is a positive income effect, the higher quartiles eating more, but household surveys suggest a decline in per capita consumption. In urban areas consumption rates seem fairly flat over time with no appreciable difference between high or low income groups. Still, overall trends hold: flat per capita rice consumption, but increasing wheat consumption (10% in the last decade). The same modest trend could mean 10-15 MT of more wheat being consumed annually 10 years from now.

We do not put much stock in the direct income effect, but growth is still a critical factor; prosperity shapes culinary habits and opens entrenched traditions to external influences. In PPP terms China is already the largest economy, and with higher growth rates than most, in the coming years it is expected to overtake the US economy in nominal terms as well. China’s nominal GDP per capita still ranks around the 70th in the country rankings – $8,500 per capita compared to $57,000 for the US, $44,000 for Canada. In PPP-adjusted terms, China’s per capita GDP is higher, $15,500, but ranking even lower. We see no reason for China’s development model to falter in the next decade or two; we believe China is going to develop and prosper, at growth rates perhaps not as fast as 10 years ago but more in line with the last 5 years.

This prosperous future paves the way for changes in culinary traditions, what and how the Chinese eat – even more radical than what we have observed in the last 20 years. Firstly, every part of China is opening to the culinary traditions of other parts of the country. Secondly, we see the foreign influences – Japanese, Korean and Thai restaurants, bakeries and pastry-shops at every corner. Fast-food may be the ugly-face of Western influence, but KFC and McDonalds pave the way to fancier eateries where their derivatives are served at many times the price, while pizzerias popularize fancier pasta-varieties. There is a lot of wheat content behind these trends, different kind of wheat than what China grows. What interest us most is high-grade wheat and durum that we hold strong export positions in. This is where the import prospects lie, as we will turn our attention to later in this report, after production and milling trends.

Wheat Production Trends

In the last crop-year China produced 129 MT of wheat. There have been modest but steady increases over the last 10 years – generally ranging from 1% to 3% per year. More recently 3% per year appeared to have become the norm but in the last crop year production dipped slightly (by less than 1%), first decline in recent years. With the exception of 2011/12 when production fell 4% short of consumption, there have been modest surpluses, in the range of 3-7% but as much as 8%, 16% and 9% in the last 3 years (even with last year’s dip). In fact, last year’s production exceeded what OECD-FAO had projected to 2025 based on past trends.

In the last 10 years domestic wheat production exceeded consumption by 65 MT, but an extra 15 MT was still imported (after 10 MT of exports). The surplus, roughly 80 MT, went to boost the country’s reserves. China’s wheat stocks now stand at 110 MT, almost 3-fold increase over the last decade – more than 40% of the world total with only 20% of the global consumption. Some blame this stockpiling on China’s hoarding instinct and fear China would one day dump these stocks to flood global markets. In reality China imports certain types of wheat that it is short of, and the low-grade wheat in its stocks is of little value to global consumers at large.

On the surface, recent trends suggest self-sufficiency in wheat, as important a food-staple as rice – what China’s food-security policies aim to achieve. But not all “wheat” is the same; different kinds or grades are required for different purposes to meet market demand. The grade of wheat that traditionally grew in China was sufficient for steamed buns, noodles and dumplings that constitute the principal calorie sources in rural areas, and also consumed widely in urban areas. Wheat flour for other uses like bread or pastries requires higher grades (high or low in gluten). To satisfy these new market needs, China has to revert to imports, and naturally from Canada’s export perspective this is where we focus – what China needs to import.

China’s achievements on the wheat production front look even more impressive from a longer historical perspective. Going back to the beginning of the reform era, China’s grain production in 1966/67 was merely 25 MT. The early reforms (collective farms with more freedom to act on their own) managed to double the production in the first decade, to 50 MT in 1976/77. In two more decades, with more inputs and mechanization, the output doubled again to 100 MT in the early 1990s. Today’s production is more than 5-times what it was back in 1966/67.

During the first decade (1966/67 to 1976/77) China’s wheat production fell 24 MT short; imports during this period were 46 MT, covering the shortfall and starting to build up the stocks (a mere 4 MT at the beginning). Into the second decade (1976/77 to 1986/87) the production deficit was a massive 84 MT, which was more than offset by 98 MT of imports, boosting the stocks by another 12 MT. Into the third decade (1986/87 to 1996/97) the production deficit was somewhat smaller, 62 MT, but even more were imported, 103 MT, doubling the stocks to 79 MT. Into the 4th decade (1996/97 to 2006/07) the production deficit shrunk further but still remained significant, 46 MT; this time China drew on its stocks, reducing them to 36 MT. In the final decade (2006/07 to 2016/17) a sizeable production surplus was achieved but a further 15 MT were imported, bringing the stocks to 108 MT – 104 MT stock build-up in 50 years.

In the first decade China’s wheat imports fluctuated in the 3-6 MT range and briefly fell down to 2 MT in the mid-1970s. In the next two decades imports increased substantially, averaging at more than 10 MT per year from the mid-1970s to the mid-1990s, with interim peaks at 15-16 MT. In the next 15 years imports moderated considerably, to an average of 1.5 MT per year – despite an interim peak close to 7 MT in 2004/05 necessitated by a production slowdown (25% drop in output in the early 2000s). Into the 1990s (as we will come back to later) the grade-mix of China’s wheat output had started to change, generating surpluses in certain types of wheat that could be exported while importing higher grades that were still in short supply. In the last 8 years annual wheat imports averaged above 3 MT, while exports remained below 1 MT.

As the reform era got underway China had 24,000 ha of land allocated to wheat harvesting. This increased by about 25% during the first decade and stayed around 30,000 ha. Into the 2000s arable land allocated to wheat started to shrink and is now at about the original level, 24,000 ha. Thus, over these 5 decades the “production miracle” was achieved primarily through yield increases, from little more than 1 MT per ha to its current level of 5.33 MT per ha.

Needless to say, arable land is a scarce resource in China. With 20% of the world population China has less than 8% of the world’s arable land – about 0.08 ha per capita, compared to the world average of about 0.22 ha per capita. Though population growth has run its course, urbanization continues to encroach into agricultural lands. Total cultivated land is estimated below 120 Mha but continues to come under further pressure for development – grain production takes up more than 90% of this land allocation. Wheat takes up 24 Mha, corn 37 Mha, rice 30 Mha, soybean 7 Mha, and the rest is taken up by other crops (coarse grains and oilseeds). Over the 50 years land allocated to growing corn has doubled, soybean shrunk marginally, while wheat and rice remained the same – other grains shrunk by about half.

Yield improvements were not limited to wheat; they were experienced in all crop groups, but least of all soybean. With only 80% increase soybean was behind all other major crops; when coupled with 15% reduction in land, soybean output increased by only 55%. This was a deliberate policy choice since soybean is a low yield crop (1.8 MT per ha even with the 80% increase), held limited prospects for yield-improvement, and there were plenty of import choices – as discussed earlier, low cost sources from Brazil, Argentina and even US.

Rice experienced greater yield improvements, more than double from 3.1 to 6.9 MT per ha; with about the same land area, output also more than doubled. Yield improvements in corn were 3.5-times; with the land allocated to corn more than doubling, output increased 8-times. Still, of all the grains the rate of yield improvement (5-fold) was the highest for wheat.

Soybean yield in China, 1.8 MT/ha, is fairly low compared to other major producers – US at 3.5, Brazil 3.4 and Argentina 3.2. But this is not a fare comparison as China grows high-grade soybean mainly for food-products, while relying on imports for feed-grade soybean. This was a conscious decision on the part of the government, mainly because there was no way for China to produce anywhere near what was going to be needed for feed, thus the decision to import.

China rice yields doubled and at 6.9 MT/ha already the highest among major producers – Vietnam 5.7, Indonesia 4.8, Bangladesh 4.4, India 3.8 and Thailand 2.8. In the last 10 years, China’s yield improved by 10%, less than India (19%), Vietnam (18%) and Bangladesh (15%), and dramatic increases in the foreseeable future are unlikely. With per capita consumption fairly steady, the status quo will probably prevail. Given the crop’s “status”, China will undoubtedly protect rice-fields from any encroachment, unless there are unexpected yield increases.

Corn production in China is an interesting case; given twice as much land, yields have increased 3.5-fold and output 8-fold. At 6 MT/ha (only 5.3 according to USDA) China’s yield is still behind US at 10, Argentina 8 and EU 6.3, though higher than Brazil at 3.6. China’s output of 220 MT is still far behind the US at 385 MT but more than double that of Brazil. If yields could improve to the US level, China could produce 350 MT – even more than the US if USDA yield is correct. If last year is an indication, there will not be any more land allocated to corn, but even a 10% yield increase would eliminate the current import. Corn is not a food-staple in China but an important source of animal feed – more so than soybean and China is more self-sufficient in it.

In wheat China achieves the same yield level as EU, 5.3 MT per ha, but is above all other major producers – Ukraine is the closest at 4.1, followed by Canada at 3.6 and US at 3.5. The only countries above China and EU are New Zealand (9.1), Namibia (6.5) and Egypt (6.4) and Mexico (5.3). Argentina stands at 3.3, Brazil 3.2, India 2.9, Russia 2.7 and Australia with 2.6 at the bottom among the top-10. It is not clear how much room there is to increase the average yield; anyway, China’s priority would be to increase quality rather than quantity. Land is too precious when there are import options – especially since China has already achieved “food-security”.

In tandem with its efforts to drive up yields, for all grains but particularly wheat, China has also been trying to improve crop quality standards. A big effort was made in the late 1990s to streamline and upgrade quality standards for different types of wheat grown in different parts of the country, both strong and weak gluten varieties. The three principal grain-zones were targeted – strong gluten in the northeast (along Kinghan Mountain Range) and the central-grain-zone (Shandong-Henan-Shaanxi provinces along Yellow, Huai and Hai Rivers), and weak gluten further south (corridor that stretches east-west, Jiangsu-Anhui-Hubei provinces). Now the focus is to cross-breed new wheat varieties suitable for growth in these regions.

It was estimated that into the mid-2000s areas sowed with quality seeds had increased from less than 5% of all land allocated to wheat to more than 30%. The efforts continued with another drive to implement modified standards in the latter part of 2000s. There is no doubt that overall grain quality has improved, but unlike yields that are easy to measure, it is difficult to quantify and verify quality improvements. As impressive as the achievements might be, however, it is certain that quality standards are still lagging behind what the market demands.

Land supply is a severe constraint on agriculture in China, which we emphasize throughout, but water shortages and greenhouse emissions are equally worrisome. With 20% of the world’s population, China has only 8% of arable land, but even less fresh water, 7% and some claim as low as 5%. Rice production takes up 40% of all irrigation water usage worldwide; wheat ranks second, requiring 4-times as much as corn and close to 10-times as much as soybean, next two largest water-consuming crops – as we examined, all four have significant land-prints in China.

On the greenhouse front, concerns over coal and other fossil fuels overshadow emissions from agriculture, but the magnitude is more significant than generally recognized or acknowledged. In this regard threats posed by rice-fields (mainly methane emissions) are even more glaring than in the case of water. Worldwide, carbon emissions from rice (1.3 Mt CO2-equivalent) are double the next two biggest offenders, soybean and corn, combined. Wheat accounts for less than half of either of the latter two, but all grains combined constitute a significant source.

China gives utmost priority to food-security, and to that end pursues a policy of self-sufficiency in at least rice and wheat, and to a lesser extent corn. But facing other challenges like limited land supply, water scarcity and climate change, China is open to compromises. As long as a certain degree of food-security is achieved, it is open to import options, as it has proven over the years with soybeans. Wheat ranks far ahead in policy priorities, but China will not flinch on importing not just 10% but perhaps 20% or more of its wheat requirements, particularly when there are both price and quality advantages – as there are in high-grade wheat varieties. With huge trade-surpluses and foreign-reserves, China has ample ability to import plenty more.

Flour Milling Industry

The next link in the wheat supply chain after production is milling. Large milling companies have their own procurement channels, working through intermediate traders or buying direct from producers. Some of the largest among them are in the production end of the business as well, with large corporate farms of their own or producer co-ops under contract to them. Small producers tend to deliver to local mills nearby, which there are plenty of across China – leftovers from the collective era or the Town and Village Enterprises that followed. One way or the other 100 MT of wheat end up at flour mills – residual going to seed-supply for next year’s harvest, into the animal-feed-chain, or to food-processing plants that take in raw grain.

China’s vast flour milling industry is highly proliferated with more than 4,000 mills and ridden with overcapacity – annual capacity at more than 250 MT but actual grind only 85 MT. More than half (perhaps as many as two-thirds) are small mills with less than 100 T daily capacity. There are 5,000-1,000 mills in the medium-range with daily capacities varying in the 100-400 T range. There are about 350 above 400 T/day but even that threshold is small in today’s standards; modern ones that are being built can grind more than 4,000-5,000 T per day. The larger mills achieve better utilization but most if not all still have some capacity to spare.

With so much excess capacity and inefficiency, it is not surprising to see the milling industry consolidate, to the chagrin of small operators and delight of industry giants. The excess capacity is largely the industry’s heritage, a plethora of small scale mills built over time to grind the locally grown grains, initially owned and operated collectively and in time taken over by budding rural entrepreneurs. A few had the vision to scale up and modernize but most got left behind due to limited know-how or resources to adapt to the new ways – scale and modern technology. More than 2,000 mills across China are idle or at best poorly utilized, barely covering their costs, with their salvage value limited to the land value they sit on.

Material costs (wheat intake) make up 80-85% of the value of the flour; production cost margins are 15-20%, including capital (4-6%), operating (3-6%) and distribution (5-7%) as well as a narrow profit margin (3-6%). The scale-factor makes technology adaptation and automation possible, at lower cost with huge quality advantages. First, higher extraction rates can be achieved, thus more flour output from a given wheat intake. Second, grading, mixing, cleaning, and purification become easier and more effective. Third, flour can be produced with the desired particle-size, ash-content and other attributes most suitable for the end purpose.

Perhaps some of the advantages that come with technology (grading, cleaning, purifying and mixing) would not be so important if wheat supply was more reliable and consistent, but that is not the case in China. There are limited controls to ensure quality standards up the supply-chains; local grains are not subjected to the same scrutiny as imports, both at origin and inbound customs. Large wheat-processing companies have an advantage in that they have the market power to control suppliers, including domestic ones. Smaller, independent mills have to accept what is delivered to their mill-gate, but they have the advantage that the uses they produce flour for are more forgiving in quality – an issue we will come back to later.