A Sharpfokus special report on palm oil
Jakarta
September 14th, 2019
September 14th, 2019
Palm oil decade
It’s been a tough decade for palm oil. After rising sharply in the mid to late 2000s, the price has fallen 50% from above $1,000 to just above $500. This has affected the growth & financials of the sector. As Indonesia is now the most important producer with 2X the production volume of Malaysia, we look at the trends in Indonesian plantings, production, spending & price performance to asses implications for the industry outlook.
It’s been a tough decade for palm oil. After rising sharply in the mid to late 2000s, the price has fallen 50% from above $1,000 to just above $500. This has affected the growth & financials of the sector. As Indonesia is now the most important producer with 2X the production volume of Malaysia, we look at the trends in Indonesian plantings, production, spending & price performance to asses implications for the industry outlook.
Planting
Government
For Indonesian plantings, we have Agriculture Ministry historical data going back to the late 1960s. The planted area has increased from 100k ha in 1967 to now 14m ha! The data has occasional large adjustments & we have removed these & smoothed with a 3 year moving average for growth. The conclusion is that planting growth peaked in the mid 1980s & late 1990s (not 2008 as popularly believed). Today’s growth is at 0%, the lowest ever.
Companies
To confirm the validity of the industry numbers, we have data for the last 7 years planting of 11 of the major Indonesian (& some Malaysian) plantings. This shows & confirms the sharp slowdown in growth after 2015. Not only has growth of the company plantings fallen to 0% it is now negative due to the beginnings of replanting by the larger, older plantations.
Government
For Indonesian plantings, we have Agriculture Ministry historical data going back to the late 1960s. The planted area has increased from 100k ha in 1967 to now 14m ha! The data has occasional large adjustments & we have removed these & smoothed with a 3 year moving average for growth. The conclusion is that planting growth peaked in the mid 1980s & late 1990s (not 2008 as popularly believed). Today’s growth is at 0%, the lowest ever.
Companies
To confirm the validity of the industry numbers, we have data for the last 7 years planting of 11 of the major Indonesian (& some Malaysian) plantings. This shows & confirms the sharp slowdown in growth after 2015. Not only has growth of the company plantings fallen to 0% it is now negative due to the beginnings of replanting by the larger, older plantations.
Response
We have monthly US$ palm oil price data from 1960 (World Bank). This appears to show that the investments in new plantings are direct responses to market prices. In 1985 plantation growth peaked at 22% a year after the price peaked in May 1984 at almost $1,000 (yes 30 years ago). In 1998 planting growth peaked at 21% with prices at $700. The price above $1,000 in 2008 was only able to encourage growth of 7-8%. This slower growth shows the cost of planting has increased & therefore a higher price is now required.
We have monthly US$ palm oil price data from 1960 (World Bank). This appears to show that the investments in new plantings are direct responses to market prices. In 1985 plantation growth peaked at 22% a year after the price peaked in May 1984 at almost $1,000 (yes 30 years ago). In 1998 planting growth peaked at 21% with prices at $700. The price above $1,000 in 2008 was only able to encourage growth of 7-8%. This slower growth shows the cost of planting has increased & therefore a higher price is now required.
Production
Long term
In the long term, production of palm fruit & oil follows the trend of planting. There has been no material increase in industry yields for many years. Production can also be affected in the shorter term by plantation spending & weather patterns. If we analyze the data of the individual plantation companies which provide quarterly production data, the average annual growth rate of fruit production has been 4%, matching the average planting growth since 2013.
Short term, rain
In the shorter term, total production growth measured from the companies quarterly fruit production data fell from 2015 up to 2017. As mentioned, because the fruit absorbs water to create oil, rainfall plays an important part & rainfall was below average during 2013 to 2015. This affects fruit 2 years later. With better rain after 2015, production growth then picked up to reach 20% in 2018. This all of the 4% average growth was captured in just 1 year.
Future
In 2019, production growth has turned negative & is -7% from a year ago in the second quarter. Production is now rising & is expected to be stronger in the third & fourth quarters. But we are unlikely to strongly exceed last years levels meaning the average 4% growth will decline. Over the next 3 years production growth should follow planting growth down to 0% & even negative. Production lags by 4 years from the maturity of the trees.
Spending
Investment
We can take a look at the behavior of the companies through their spending to see their growth & potential. We use the cash flow statements where available. There was a sharp slowdown is spending for investment (new plantings) which fell 50% from 2015 to 2018 following the slowdown in planting. Investment spend has since increased almost 50% for the replanting programs which have now begun.
Operational
The companies operational spending has generally increased as the focus has shifted from new planting & investment in immature plantations, to the increased share of mature plantations. Annualized spending has increased 36% from 2015 as investment fell, after initially declining from 2015 to 2016. The annualized spend matches well with the production growth which shows while weather is one factor, spending is probably the more important for fruit production.
Long term
In the long term, production of palm fruit & oil follows the trend of planting. There has been no material increase in industry yields for many years. Production can also be affected in the shorter term by plantation spending & weather patterns. If we analyze the data of the individual plantation companies which provide quarterly production data, the average annual growth rate of fruit production has been 4%, matching the average planting growth since 2013.
Short term, rain
In the shorter term, total production growth measured from the companies quarterly fruit production data fell from 2015 up to 2017. As mentioned, because the fruit absorbs water to create oil, rainfall plays an important part & rainfall was below average during 2013 to 2015. This affects fruit 2 years later. With better rain after 2015, production growth then picked up to reach 20% in 2018. This all of the 4% average growth was captured in just 1 year.
Future
In 2019, production growth has turned negative & is -7% from a year ago in the second quarter. Production is now rising & is expected to be stronger in the third & fourth quarters. But we are unlikely to strongly exceed last years levels meaning the average 4% growth will decline. Over the next 3 years production growth should follow planting growth down to 0% & even negative. Production lags by 4 years from the maturity of the trees.
Spending
Investment
We can take a look at the behavior of the companies through their spending to see their growth & potential. We use the cash flow statements where available. There was a sharp slowdown is spending for investment (new plantings) which fell 50% from 2015 to 2018 following the slowdown in planting. Investment spend has since increased almost 50% for the replanting programs which have now begun.
Operational
The companies operational spending has generally increased as the focus has shifted from new planting & investment in immature plantations, to the increased share of mature plantations. Annualized spending has increased 36% from 2015 as investment fell, after initially declining from 2015 to 2016. The annualized spend matches well with the production growth which shows while weather is one factor, spending is probably the more important for fruit production.
Receipts
But the spending has stalled & fallen back since peaking late last year. We can also see that spending on staff (from those companies which provide the data) which increased up to 2017 has been flat since then. We can see one possible reason why spending has stalled. Using cash flow receipts & subtracting spending, the companies are losing money even before interest & tax. This is a return to the losses in 2015 & probably means spending will remain low until prices rise further. This will hold back production.
Demand
Edible
A few words on demand. While we don’t have our own data for demand, we can draw some conclusions from other sources. We know for example that edible oil demand grows close to global GDP growth which is 3-4% per annum. This is because edible oil is used in about 50% of all supermarket products. From cooking oils to processed foods to personal care products including medicines.
Palm
We also know that demand for palm oil will be faster than for other edible oils as palm is now by far the largest of the 8 major oils. Palm is now more than 30% larger than the second largest, Soy & both account for about 50% of the total. Palm oil has become larger than Soy since 2006, which matches with Indonesia overtaking Malaysia as largest producer. The increasing demand for edible oil therefore relies heavily on Indonesia which is why we have focused here.
But the spending has stalled & fallen back since peaking late last year. We can also see that spending on staff (from those companies which provide the data) which increased up to 2017 has been flat since then. We can see one possible reason why spending has stalled. Using cash flow receipts & subtracting spending, the companies are losing money even before interest & tax. This is a return to the losses in 2015 & probably means spending will remain low until prices rise further. This will hold back production.
Demand
Edible
A few words on demand. While we don’t have our own data for demand, we can draw some conclusions from other sources. We know for example that edible oil demand grows close to global GDP growth which is 3-4% per annum. This is because edible oil is used in about 50% of all supermarket products. From cooking oils to processed foods to personal care products including medicines.
Palm
We also know that demand for palm oil will be faster than for other edible oils as palm is now by far the largest of the 8 major oils. Palm is now more than 30% larger than the second largest, Soy & both account for about 50% of the total. Palm oil has become larger than Soy since 2006, which matches with Indonesia overtaking Malaysia as largest producer. The increasing demand for edible oil therefore relies heavily on Indonesia which is why we have focused here.
Inventories
To asses the demand for palm oil we use inventory levels. Unfortunately, we don’t have Indonesian inventory so we use Malaysian inventory as a proxy. With supply from the company data we have growing an average 4% since 2015, what has happened to inventory levels. The current 3 month average inventory is 2.4m tons & August was below this 2.3m. In August 2015 the 3 month average inventory was 2.5m & actual 2.6m. This appears to indicate the supply growth of 4% is not enough to satisfy demand. Worrying as we now expect supply to fall to 0%.
To asses the demand for palm oil we use inventory levels. Unfortunately, we don’t have Indonesian inventory so we use Malaysian inventory as a proxy. With supply from the company data we have growing an average 4% since 2015, what has happened to inventory levels. The current 3 month average inventory is 2.4m tons & August was below this 2.3m. In August 2015 the 3 month average inventory was 2.5m & actual 2.6m. This appears to indicate the supply growth of 4% is not enough to satisfy demand. Worrying as we now expect supply to fall to 0%.
Pricing
Inventory
The information on plantings, production, spending leads us to believe prices should rise. The question is by how much. We have 3 suggestions based on available information. First using Malaysian inventory. Inventories are now declining from a year ago. In recent years, inventory has declined on 2 previous occasions. On the first occasion in early 2014 the price was a $800. The second time early 2017, the price was above $700. This would mean prices rising 50% from current levels.
Returns
Second we can use a simple IRR model with some basic assumptions on fruit production yields & spending. The base assumptions are a maximum yield of 21 tons (the average yields are below this level now) & spending of Rp20m per hectare (hard to calculated as spending is mixed with 3rd party fruit purchases). Using this model, to achieve an IRR of 15% would require the palm oil price to be Rp16m per ton or $1,140 at current exchange rates.
Inflation
Lastly, while this is a controversial topic... we can look at what the theoretical price would be in real terms. As the price is in $s, we can use average US CPI inflation to calculate this theoretical value. Of course the controversy is over whether commodity prices should follow average prices & what start date you use. As we have prices back to 1960, we use that as our base. Palm oil was then $233. If it followed inflation, the price should now be $1,972.
Conclusion
To conclude, after a decade of falling prices, which is really 3 decades since the price first reached $1,000, the palm oil industry growth has slowed to 0%. The growth has in reality been slowing all the way from the 1980s matching prices. Now without higher prices, spending will lag as plantations lose money, meaning production growth will be slow. Demand is likely to continue at above the 4% level it appears to have been growing at. Prices must therefore rise & could increased substantially higher than any current estimates allows.
Inventory
The information on plantings, production, spending leads us to believe prices should rise. The question is by how much. We have 3 suggestions based on available information. First using Malaysian inventory. Inventories are now declining from a year ago. In recent years, inventory has declined on 2 previous occasions. On the first occasion in early 2014 the price was a $800. The second time early 2017, the price was above $700. This would mean prices rising 50% from current levels.
Returns
Second we can use a simple IRR model with some basic assumptions on fruit production yields & spending. The base assumptions are a maximum yield of 21 tons (the average yields are below this level now) & spending of Rp20m per hectare (hard to calculated as spending is mixed with 3rd party fruit purchases). Using this model, to achieve an IRR of 15% would require the palm oil price to be Rp16m per ton or $1,140 at current exchange rates.
Inflation
Lastly, while this is a controversial topic... we can look at what the theoretical price would be in real terms. As the price is in $s, we can use average US CPI inflation to calculate this theoretical value. Of course the controversy is over whether commodity prices should follow average prices & what start date you use. As we have prices back to 1960, we use that as our base. Palm oil was then $233. If it followed inflation, the price should now be $1,972.
Conclusion
To conclude, after a decade of falling prices, which is really 3 decades since the price first reached $1,000, the palm oil industry growth has slowed to 0%. The growth has in reality been slowing all the way from the 1980s matching prices. Now without higher prices, spending will lag as plantations lose money, meaning production growth will be slow. Demand is likely to continue at above the 4% level it appears to have been growing at. Prices must therefore rise & could increased substantially higher than any current estimates allows.