After the most recent Weekly Export Sales and Shipments report, DTN Grains Analyst Todd Hultman provided the observation, "Something doesn't add up. There must be something going on behind the curtain in regards to China's soybean demand." I agree. Taking the thought further, the market's view of supply and demand has not agreed with USDA's numbers since the 2013-14 marketing year began. Last Friday's January reports just added more fodder for disagreement.
The math behind USDA's soybean numbers doesn't seem to add up. (DTN illustration by Nick Scalise.)
The elephant in the room is Brazilian production. Recently, the Brazilian state agency CONAB pegged soybean production at a record 90.3 million metric tons. In its January WASDE report, USDA upped its own Brazilian production estimate from November's 88 mmt to 89 mmt. An increase from the previous month in global beginning stocks (2012-13 ending stocks) to 60.55 mmt resulted in total supplies of 347.4 mmt. Total demand increased only slightly to 270.9 mmt, putting global ending stocks at an all-time high of 72.3 mmt. For the OCD readers out there, don't try to do the math. The numbers don't add up. USDA acknowledges as much in the fine print at the bottom of its official release.
The bottom line is this: Projections for new records in global production, global supplies, total demand, and ending stocks result in an ending stocks-to-use figure of 26.7%. This is the third-highest on record, coming in behind only the 27.8% estimated at the end of the 2006-2007 marketing year and the 27.3% at the conclusion of 2010-2011. Keeping in mind the easiest read on the bullishness or bearish of any government report is the ending stocks-to-use calculation (ending stocks divided by total demand), and the 2013-2014 marketing year is projected to be the third most bearish on record. Why does the soybean market remain strong?
Those familiar with my analysis know the importance I put on futures spreads for reading the market's fundamental view. In the case of soybeans, the 2013-14 forward curve (series of futures spreads that began with the November 2013 contract, but now starts at the March 2014 and runs through the September 2014 contract) continues to show a strong inverse. This means nearby contracts are higher priced than deferred contracts, reflecting a bullish marketing year outlook by those who actually trade the cash commodity.
Taking this analysis a step further, the trend of the 2013-14 spreads (my technical analysis of fundamentals) has turned up again. Recently, the inverse in the March-to-May had dipped to 13 cents (March higher priced than May) while the May-to-July slipped to 10 cents (both weekly closes only). However, the last two weeks have seen these spreads strengthen to inverses of 19 cents and 16 cents. In other words, the market's view of supply and demand is growing more bullish DESPITE continued increases in global ending stocks projections.
Why? Undoubtedly the U.S. numbers remain flawed. USDA has yet to acknowledge its own reported 2012-2013 exports of 1.331 billion bushels, leaving its demand estimate at 1.320 bb, creating ending stocks-to-use of 141 mb, and putting ending stocks to use at 4.5%. While 11 mb doesn't seem like much to quibble over, what would happen if it ever showed up on USDA's official ledger? Domestic demand would inch up to 3.11 bb while ending stocks fell to 129 mb, resulting in an ending stocks-to-use figure of 4.1% -- the tightest on record. USDA doesn't want that. Remember, it has unofficial floors that can't be breached when it comes to ending stocks-to-use. With corn, it is the 5.0% level. For U.S. soybeans, it appears 4.5% is the magic number because the beginning of the latest bull run in U.S. grains that began in 2005-06. Since then, this level (4.5%) has been estimated at the end of the marketing year four times: 2008-09, 2009-10, 2012-13, and now 2013-14.
The math behind USDA's soybean numbers doesn't seem to add up. (DTN illustration by Nick Scalise.)
As for this marketing year, soybean ending stocks were left unchanged at 150 mb with a 30 mb increase in total demand (20 mb in exports, 10 mb in crush) offset by a like increase in production. National average yield was increased to 43.3 bpa, the third-highest on record, while corn yield decreased from the November estimate of 160.4 bpa to 158.8 bpa. Why does this matter? The government's own crop condition numbers at the end of the 2013 growing season showed the soybean crop lagging the five-year average (2007-2011) by 20%, where corn lagged only 18%. Furthermore, if 2013 crop condition numbers are compared to those from the record yields of 2009-10 (43.5 bpa) and 2010-11 (44 bpa), a lag of approximately 21% is seen. In other words there is a strong possibility U.S. production for 2013 did not actually come in at USDA's number that conveniently kept ending stocks-to-use above 4.5%.
The quarterly stocks report showed unprecedented Q1 demand for U.S. soybeans of 37.8% of total supplies. Much of this was due to the extraordinary export pace -- 10% ahead of projections -- that if continued could easily add another 10 mb to total demand. If so, and if both beginning stocks and 2013 production are actually smaller than reported, then what might domestic ending stocks and ending stocks-to-use really be? It seems the futures spreads already know.