Market Comments – September 17, 2015

NY futures moved slightly lower this week, with December giving up 65 points to close at 62.40 cents/lb.

Neither the USDA supply/demand estimate nor the Fed’s decision on interest rates were able to inject new life into the cotton market, as prices remained confined to rather tight sideways range. December has now closed in a 95-point range, between 62.40 and 63.35 cents/lb, since August 26!

The USDA report came in more or less as expected, showing a slightly higher US crop and lower global mill use. However, despite these seemingly bearish adjustments, ROW ending stocks are still projected to drop by the end of the season, which would be supportive to prices.

The market’s initial reaction to the report was friendly, since the USDA raised the US crop by just 350’000 bales to 13.43 million bales, which was shy of traders’ expectations. The increase came mainly from a higher harvested acreage number, whereas the average yield was cut by another six pounds to just 789 lbs/acre, which would be the second lowest in the last ten seasons and over a hundred pounds below the record 892 lbs/acre of 2012. This is rather surprising, especially since the crop condition index is currently at the second highest reading in ten years!

The Texas numbers are even more confounding, since the USDA showed the highest boll count in Texas in the last five seasons, but has yield at just 615 lbs/acre, which would be the second lowest in ten years. Since 2005 Texas yields have ranged between 589 and 843 lbs/acre. Considering that we are in a “wet” year, it doesn’t make sense that yields would average less than in the previous three seasons, when Texas was experiencing a severe drought.

When we look at the USDA’s track record for estimating yields, we should definitely allow some room for error. In 2012 the USDA projected Upland yields at 774 lbs/acre in its September report, which proved to be exactly 100 lbs/acre shy of the final number. 2004 was another big miss, when the government was 96 lbs/acre off the final mark in September. There were three more occasions over the last ten seasons where the USDA underestimated the crop by 50-70 lbs/acre at this time of the season.

It is not our intention to put the blame on the USDA for these discrepancies, but rather to remind readers how difficult it is to get a good grip on the US crop at this juncture. In other words, we might still see significant swings in the crop estimate down the road. For every 10 lbs/acre in yield we get 170k bales, which means that a difference of 60 lbs/acre would amount to about a million bales.

Moving on to the global balance sheet, the USDA left world production basically unchanged at 108.74 million bales (down 0.25 million), but finally scaled back global mill use to 113.44 million bales (down 1.21 million). This resulted in an increase in ending stocks of slightly over a million bales to 106.26 million bales. While this looks bearish, we need to remind ourselves that most of these stocks are located in China (64.62 million bales) and while Chinese prices are in the process of closing the huge price gap that existed to international prices, they still remain slightly more expensive at this point in time.

Therefore, when we look at just the ROW balance sheet, we have a slight increase in the ROW production surplus from 2.3 to 3.3 million bales in the September report. However, since Chinese imports are still expected to amount to 5.75 million bales, they would not only absorb this ROW production surplus, but would also draw down ROW ending stocks.

In other words, based on the current set of numbers the statistics still paint a slightly friendly picture, although the question is whether we choose to believe these numbers or not? There seems to be a lot of disagreement on mill use due to the tough global macro environment we are currently in and most traders carry a 2-3 million bales lower mill use number than the USDA, including ourselves. Nevertheless, even if that were the case, it would only bring the ROW production surplus to the level of Chinese imports and therefore not really change the status quo next summer.

US export sales for the week that ended on September 10 were quite decent at 115’100 running bales of Upland and Pima for both marketing years, with 13 markets participating. Vietnam, Mexico and Turkey were the strongest buyers, while China is still mostly missing in action. Shipments were slow at 66’400 running bales, but that’s mainly due to low pipeline stocks. For the season we now have commitments at 3.1 million statistical bales, of which 0.7 million have so far been exported.

So where do we go from here? In the case of US cotton we still have a standoff between producers and mills, whose price ideas are too far apart to facilitate any sizeable business at this point. With the pipeline nearly empty and with the crop still mostly out in the field and vulnerable to adverse weather, growers are in no hurry to let go of their cotton and will likely play the loan game instead.

Mills on the other hand consider US cotton too expensive at the moment and therefore focus on foreign growths, which are being competitively offered due to weak exchange rates - be it Brazilian, Indian, West African or Greek. This has pushed the US once again into the role of residual supplier. However, unless the statistics are wrong and mill use is a lot weaker than currently anticipated, there is no way past the US since it accounts for 30% of global trade and is the main source for machine picked high grades. Whether the US is willing to play the waiting game or whether it eventually gets cold feet and follows other origins lower remains to be seen and probably depends to a large degree on how the crop turns out, both quantity and quality wise.

Speculators have been reducing their exposure in the cotton market in recent weeks, which is no surprise given the lack of momentum. The spec net long position amounted to just 2.9 million bales last week, but outright positions were still relatively large at 7.0 million bales long and 4.1 million bales short and they have the potential to move the market in case the long-term trading range of 61-68 cents is breached. In this regard we view the lower end of this range as the more vulnerable side, since we closed today just 2% above that level. For now the market is still confined to its 12-month trading range, but we view a potential breakout to the downside as the more likely scenario than vice versa!


This email has been scanned by the Symantec Email service. For more information please visit