Description

ARCHER DANIELS MIDLAND (ADM), SHORT                  

COMPANY BACKGROUNG

ADM owns assets that process soybeans, corn, wheat, cocoa and other agriculture commodities into a variety of food, biofuel and industrial end products.  It engages in these activities through out the world although greater than 55% of its assets are located in North America making the company less international than Bunge and the private agriculture firms such as Cargill and Louise Dreyfus.  ADM also owns an extensive array of storage and logistics assets that provide a significant return to the company and enable the company to profitably trade agriculture commodities to enhance its overall returns.

THESIS

Driven primarily by the dramatic increase in demand for biofuels, most of ADM’s assets are currently experiencing or have already past peak earnings conditions.  Thus ADM is likely to experience declining earnings in 2008, especially relative to wall street consensus expectations and thus, the 12x multiple the stock is trading for is rich.  Further, ADM invested its capital unwisely in the past few years (primarily in Bio Fuels) and is likely to achieve near term returns on incremental capital that are far less than expected and far less than cost.  In FY 2008 (june yr end), we expect ADM to make less than $2.30 per share versus consensus expectations of $2.70. 

Given the long term outlook for the agriculture sector and the expected reasonably fast demand catch up to this recent overbuild phase, ADM should trade based on it earning a 20% premium to its long term average ROCE of 8%.  Thus the stock should trade for $25 per share. This DCF value is based upon only one year of down earnings (2008) and then a resumption of 5% top line growth and operating margins equal to the historic 4% average.  Returns are higher than historic in our model because we have asset turns increasing.  That stock price is a theoretical value at which we would be inclined to cover our short, but, in reality, the short should be covered when we are satisfied that the market fully understands the likely near term earnings. 

THE BUSINESSES

SOY PROCESSING – this is ADM’s single largest business. The soy assets represent 26% of the total company and roughly 55% are in North America.  The business has been an integral part of the company’s recent eps growth.  Operating profits more than doubled from $301 million in 2004 to $705 million in 2007 on asset growth of just 7%.  Operating profits are expected to increase again in 2008 to $744 million.  However, this business is a commodity business that has tended to be very cyclical. 

Chart A

ADM is able to use its physical assets and market intelligence to make money in the soybean business that is often very different than what one would expect by looking at the near month soy crush on the CBOT.  In fact the correlation between the average quarterly near month soy crush is nearly non-existent.

Chart B

However, ADM is, as mentioned, able to exploit its position as a physical player.  If one uses the central Illinois spot price for soybeans against the CBOT meal and oil price the fit is much better.  What this tells us is that we can do a better job of predicting ADM’s soy margins by watching the basis differential as well as the crush.

Chart C

The bad news for the short position is that the basis differential blew out this past quarter and thus despite the depressed crush margin, ADM’s soy margins should be very good.  They should beat the ~ $155 million expected by up to $45 million ($.05 per share).  However, this basis differential appears to be being driven by the lack of storage space due to the huge corn crop.  Storage space and transportation capacity is at a premium, so farmers must give the owner’s of those assets a discount in order to have their crops stored or delivered.  This strength should also show up in ADM’s Ag Services division (transportation & storage among other things), but again, the strength is probably being understated in the numbers. 

However, this quarter or next is likely to be the peak for the soy business.  Even in the intermediate term, the soy crush business is likely to come under much pressure.  This will occur for the following reasons:

1)      ~ 6% domestic capacity expansion this ADM FY including a Greenfield Louise Dreyfus plant that just started up in Indiana that adds 3.5% new capacity.  The other capacity expansions are being done by Cargill in Kansas City and Bunge in Illinois and Purdue farms in North Carolina.

2)      A very small U.S. crop, a very small Chinese crop (12% smaller than expected due to drought) and perhaps a problem in Brazil due to an already dangerous drought situation is likely to pressure bean prices and hurt utilization.

3)      Meal is going to be hurt by dry corn mill capacity expansion for ethanol production.  The only material by-product of those mills is a primarily protein feed ingredient called Dried Distillers Grains (DDGs).  According to our estimates, the incremental DDGs produced in 2006 and 2007 will equal ~ 13% of total soy meal supply on a protein adjusted basis.  As per Chart D below, much of that production is just beginning to hit the market or will over the next few months. 

Chart D

4)      It appears the food processor lobby has had success in pushing forward legislation that will prevent the export of biodiesel.  This is significant because of the impact biodiesel has been having on the bean oil price. Ever since the $1.00 per gallon tax credit on bio diesel went into effect, bean oil has been priced at parity with diesel fuel minus the $1.00 credit (after operating costs, methanol costs and credit for glycerin is included).

Chart E

This is because there is, relative to edible oil demand, an infinite demand for diesel fuel.  Unlike with ethanol, there are no logistic, transportation, engine or other impediments to blending biodiesel into the diesel pool.  Thus every pound of bean oil produced would be converted to diesel if its price were below that of diesel.  At parity, refiners have no incentive to blend bio-diesel.  However, several European jurisdictions give subsidies to bio-diesel that can be collected by U.S. producers.  So, despite the lack of incentive from U.S. prices to produce bio-diesel, significant demand comes from exports.  The food processors who use oil, believe this additional demand is forcing up the price of bean oil. They have petitioned congress for help and it appears highly likely this congress will pass a law that prevents the export of bio-diesel.  We are skeptical of the food processors view on the impact of this legislation on bean oil because, it should trade at parity with diesel fuel regardless.  However, at the margin, perhaps it will loosen up the oil situation some what.

For the reasons stated above, we expect FY 2008 Oilseed processing margins to be equal to or below those of FY 2007 ($705 mm).

CORN PROCESSING

ADM is North America’s largest corn processor and has a significant presence in the business around the world.  Most of its corn is processed in “wet mills” that can produce a variety of end products including High Fructose Corn Syrup (HFCS), Food Starch, ethanol and others.  These mills also produce three by products; corn feed, corn meal and corn oil.  Corn feed and meal are used primarily for animal feed and corn oil is consumed by humans directly or in processed foods.  The corn processing assets represent 14% of the company total but in 2007 accounted for 31% of operating profit.  Like all commodity businesses, corn processing has been highly cyclical in ADM’s results.  However the business has generally been twice as profitable as the oil processing segment.

Chart F

INDUSTRY BACKGROUND

The domestic wet mill industry has consolidated dramatically in the past decade going from eight major players to four who control 95% of capacity.  Two other companies control the remaining 5%.  In terms of front end grind capacity, the industry breakdown is as follows:

ADM               45%

Cargill              27%

CPO                11%

Tate & Lyle      13%

Roquette          3%

Aventine           2%

Also, more than 5 % of industry capacity has been mothballed or permanently closed in the past 5 years. 

Of the companies in the industry, four produce ethanol, five produce high fructose corn syrup and 3 three produce both.  Of those that produce both, ADM controls 75% of switch able capacity.  For those mills that can produce both ethanol and HFCS, between 10-20% of corn grind can be diverted easily between the two.  In addition, those mills that do not produce one or the other can be reconfigured to do so with $.25 per bushel capital investment.

As mentioned above in the soy section, there has been massive capacity added to corn ethanol production over the past 1 ¾ years.  This capacity has been in the form of “dry” ethanol mills.  These mills are thought to have lower capital costs, although based on the cost of Tate & Lyle’s new facility the differential does not appear to be as large as some think (~ 5.50 per bushel of front end grind vs. $4.30).  However, a wet mill is only economic in the 125k bpd size, so one can build a dry mill for a much lower total cost and size is an issue because being small and proximate to corn can enhance operating efficiency.  However, generally the operating costs are higher and by-product credits are lower for dry mills.

Wet mill profitability has been very high, driven by high ethanol prices, which drove switching from HFCS finishing to Ethanol finishing, which coupled with capacity rationalization, tightened up the HFCS market substantially.  This occurred despite flat to down HFCS demand as health issues, the promotion of low calorie food and consumer preferences has weakened the market.  This weakening of demand appears to be accelerating.

Recently however, due to the gross overcapacity in ethanol capacity, prices of ethanol have plunged.

Chart G

Despite fewer problems blending ethanol in the winter which will lead to greater demand, the schedule of new plants (shown above in Chart D) means that supply will more than offset this greater demand.  Thus the pricing situation for ethanol is not likely to improve any time soon. 

Although not likely to make the ethanol situation worse at current price, the Brazilian Sugar situation will continue to exert downward pressure on ethanol because cane is available to be distilled into ethanol rather than converted to sugar.  Low prices in Brazil can impact U.S. prices if they are low enough to make it economic to pay the $.54/gallon tariff + ~ $.10/gallon of transportation cost to ship ethanol to New York.  This has happened in substantial quantities and depressed domestic ethanol prices recently.  However, as mentioned, prices are not likely to fall further from this impact as the arbitrage is no longer open.

Chart H

There is no doubt, due to the ethanol debacle; it is relatively more profitable for a wet mill to make HFCS at current prices than ethanol compared to most of last year.  The difference is striking compared to the first 9 months of 2006, when bullishness on ethanol was peaking. 

Chart I

In addition, ADM receives between 12% and 20% of its corn processing revenues per domestic bushel from by-products that compete with the by-products of the expanding dry mills that produce ethanol. As mentioned above, the main by-product of ethanol production from dry mills is DDG’s.  DDG’s can be substituted in certain animal feed for corn meal and feed.  There is going to be so much DDG’s around the price will likely fall to 0 after freight as the material cannot be stockpiled due to EPA regulations.  Since this low-cost feed can be substituted for corn meal and feed (and as previously noted, soy meal), the prices of these by-products are likely to suffer materially.  Our estimate is that nearly 10mm short tons of incremental DDG are going to be produced by mid 2008 vs. mid 2006.  This amount equals 400% of total corn gluten meal produced and 100% of gluten feed produced annually . . . a staggering quantity.  

Our thesis on this impact has not played out yet, as shown in Chart J, below.  Industry insiders have seen the wave of DDG’s coming and have likely done a good job in moving the first large production amounts.  However, we believe that the timing and size of the incremental DDG production makes a substantial negative impact on meal and feed inevitable and imminent. 

Chart J

DUE TO THE ETHANOL SITUATION, HFCS PRICING IS LIKELY TO COME UNDER PRESSURE

As noted above, ADM has ~ 75% of the switch able capacity.  Whether they actually switch capacity is almost irrelevant to the question of HFCS pricing because of the way pricing is set (see below “HFCS Industry Structure”). The issue is who has greater power and leverage in the price negotiations between the large buyers and sellers.  The large buyers are very sophisticated.  They know that ADM has 1.15 billion gallons of wet mill ethanol capacity and they are building an additional .55 billion gallons of dry mill capacity.  They know that 10 - 20% of the wet mill ethanol capacity can be diverted to making HFCS and this could add 1.5 – 3.0 billion lbs (dry) of HFCS or 7.5 – 15% of total demand.  They also know that the relative profitability of ethanol has dropped dramatically from where it was last year (see Chart I, above).  They also, we are sure, have heard the argument that ADM will not kill the HFCS business to benefit ethanol; however such a conclusion does not appear to be warranted. 

It appears that ADM, after its new dry mill capacity is on line, controls ~ 25% of domestic ethanol capacity.  From our calculations, at current prices, ADM should be making ~ $.25 per gallon less at spot prices than it averaged in 2006.  That equates to $.288 billion less operating profit yoy ($.32 per share) and $.425 billion ($.46 per share) less than they expected if the new capacity is included.  In comparison, if they were able to force HFCS prices higher by 10%, it appears they would only increase operating income $90 million ($.10 per share).  ADM controls ~ 30% of the HFCS market so its ability to impact both markets appears similar.  It seems that the incentive is for them to try to improve ethanol margins and to give up some HFCS profitability to do so. 

The buyers who are negotiating with them know all this and will insist on prices that reflect that reality, regardless of what ADM does or does not actually do.  However, ADM will likely actually switch.  At a recent conference, the CFO of ADM stated that the company would maximize its grind and produce whatever products maximized the value of each kernel.  That stood in contrast to previous statements by the company that they would not switch Ethanol back to HFCS.   It seems surprising that ADM would make such a statement which would obviously hurt its position in the ongoing HFCS negotiations.  Our only explanation is that they felt the market was becoming very nervous about the near term earning impact from ethanol and found it more important to assure the market that it might do something on ethanol as opposed to HFCS.

We are not swayed by the argument that the NAFTA ruling freeing HFCS to go into Mexico will help HFCS pricing leverage

There are many questions with respect to how much additional HFCS will be shipped to Mexico due to the above-mentioned ruling.  The best industry consultant expects 200 – 300k metric tons (2 – 3%) of additional domestic capacity to be shipped to Mexico in 2008, thus absorbing any loosening of the market up north.  However, many analysts and consultants, think for a variety of reasons, very little additional capacity will make it to Mexico and to the extent it does, it will be offset by Sugar displacement here.  Corn Products itself recently said that they expect no additional supplies to go to Mexico next year. 

What appears to be overlooked in the debate is the structure of the industry in Mexico and its relation to the U.S. and Canada.  The potential users of additional HFCS in Mexico are the same users as in the United States, i.e., KO, PEP and/or its bottlers.  FEMSA is by far the biggest player.  It controls 70% of the soft drink market in Mexico.  Although it appears that FEMSA has its own HFCS negotiating team, we find it hard to believe that KO would not be in close consultation with FEMSA on the negotiations and that KO wouldn’t dissuade them from making a move that might save a few cents on sweetener in Mexico at the expense of the parent and sister bottlers in the rest of North America. 

We are also not swayed by the argument that HFCS pricing can be forced higher because it is still substantially cheaper than sugar

There are material switching costs for carbonated soft drinks (“CSDs”) and other beverages.  The switching costs for dry HFCS and sugar are very low.  It appears that current domestic sugar prices plus refining costs are about $.27-.28 per lb.  The highest priced HFCS is being sold at ~ $.24 (on dry basis), so it is still ~ 15% cheaper than sugar.  Most HFCS is sold at an even greater discount.  

However, there are two important things to consider:

(1)   Canada has no sugar tariff and the price of sugar there, with refining is ~ $.17 per lb.  3.5% of North American HFCS capacity is located in Canada and, if displaced, can freely make its way south of the border.  At current prices, there is barely, if any discount on HFCS in Canada, so if prices rise, that HFCS may be displaced.  We are unaware of how much of the Canadian HFCS goes to beverage vs. food, but we assume it’s similar to the U.S./Canada total (i.e., 25%).  So there is likely to be a strong incentive to not cause food processors in Canada to substitute HFCS.

(2)   Price is not the only variable.  There is a growing health food backlash against HFCS.  We are told that the number of products marked “HFCS Free” increased from ~ 40 to over 150 in the past year.  Thus, food processors will argue in the price negotiations that they are willing to switch to sugar if the price premium is only 20% since they will be able to recoup that through greater demand from health conscious consumers.  

NOTE ON THE STRUCTURE OF THE HFCS INDUSTRY AND THE PRICE NEGOTIATIONS

The SELLERS

There are five producers.  Consultants often quote the capacity of each producer in HFCS 55 and 42; however we prefer to think in terms of front end grind.  Each wet mill has a finite capacity of front end grind but can produce a wide variety of end products.  It is a central part of our thesis that some mills can swing 20% of their corn grind between HFCS, ethanol and other products.  In rough terms, the U.S. and Canada industry is split the following way:

ADM               45%

Cargill               27%

CPO                 11%

Tate & Lyle      13%

Roquette           3%

In addition, often overlooked is a wet mill owned by Aventine that grinds 2.3% of the domestic total.

Mexico appears to have another 5.5% split between two companies, CPO and Almex, a jv of ADM and Tate & Lyle.

An additional 3% of capacity is being built by Tate & Lyle which should be operational in mid to late 2008.  Also, Cargill’s Blair Nebraska facility is apparently being run at 50% capacity and could easily be ramped up for an additional 1.5% of industry capacity. 

Of the 27 domestic mills, currently 8 have both ethanol and HFCS finishing capacity; however Aventine told us that an investment of less than $40mm would enable them to install capacity to add 2.3% to the supply of HFCS (they claim to be currently considering the possibility).  It appears all the existing wet mills can be configured similarly.  Tate & Lyle has stated that although its new mill currently has no plans for HFCS finishing capacity, it plans to move some specialty starch production from its Lafayette, Indiana mill to the new mill thus freeing up HFCS capacity in Lafayette. 

As noted above, demand appears to be declining 1% per year.

Of the switch able capacity, ADM appears to control ~ 75%.

The BUYERS

While focusing on the oligopolistic power of the sellers, often overlooked is the structure of the buyers.  70% of HFCS is sold into the CSD market.  55% to coke and Pepsi (actually, the price negotiation is done by a jv of the bottlers and the parents).  The other 15% is split between Cott, Cadbury and some smaller players.  Coke and Pepsi also buy another 5% of the market to go into non-CSD beverages.  It is difficult to figure out how significant sweetener pricing is to KO, PEP and their bottlers because in some instances the parent pays for the sweetener and sells the syrup to the bottler and in other instances the bottler buys only the concentrate and purchases the sweetener from the joint venture.  However, it is safe to say that it is a very material cost item (well over 30% of operating income for the bottlers and probably 20% for the parents). 

However, the rest of the market should not be overlooked.  The rest of the market consists of HFCS purchased to go into baked goods, candy and other processed foods.  The buyers of most of this HFCS are the large food companies such as GIS, HSY, CAG, K, etc.  These buyers are no slouches when it comes to procurement sophistication and they consume a material amount of HFCS.  Perhaps more important, these buyers can more easily switch to sugar than liquid manufacturers because there is capital equipment and transportation logistics necessary to get liquefied sugar in the correct formulation into soft drinks.  Apparently, it is much easier to substitute between the two dry forms. 

Another important point is that the same domestic buyers are the active buyers in Canada and Mexico.  This will be explained more fully below.

The NEGOTIATIONS

When analyzing the potential for HFCS price changes, we need to consider the important factors in the negotiations.  We are guided by Michael Porter’s 5 forces analysis of market power:

1. The threat of substitute products
2. The threat of the entry of new competitors
3. The intensity of competitive rivalry
4. The bargaining power of customers
5. The bargaining power of suppliers

We have concluded, based on the above factors, that this year these factors, on balance favor the buyers versus the sellers, relative to last year.  We have attempted to explain why above and would be happy to answer any additional questions on the subject.

TATE & LYLE’S ANNOUNCEMENT

Tate and Lyle recently gave an update of its business through the end of August.  They projected much worse than expected results and cited, among other things, weak HFCS volumes and poor ethanol margins in their U.S. operations. 

AGRICULTURE SERVICES

The Ag Services business represents 15% of company assets and has generally been the second most profitable behind corn milling.  This business consists of most of the company’s network of storage and transportation assets.  The business is very strong this year as USDA projections have bushels of row crop production up 16% year over year.  This is putting a tremendous strain on storage and transportation assets. 

Chart K

However, the business is likely to decline next year as substantial new capacity is being built to address the bio-fuels boom.  That capacity will be available for the next harvest.  For now, however, the business is strong as evidenced by barge rates in the chart below. 

Chart L

OTHER (INCLUDING WHEAT AND COCOA PROCESSING AND FINANCIAL)

This segment consists of wheat processing, cocoa processing, some other food processing assets and financial (which includes ADM’s own trading business as well as a futures broker).  The returns in the businesses other than financial have been awful. 

Chart K

Frankly, we don’t have a great feel for this business and the company does not like to break out detail on it.  As noted above, returns have been low (other than financial).  Wheat margins were very depressed as wheat prices ran up much faster than flower and other derivatives, but they appear to have now caught up. 

Chart L

Financial profits are not likely to be repeated however, as this past year’s high volatility was likely a driver of financial profits and the commodity boom has increased returns to the brokerage business. 

WASTED CAPITAL SPENDING

As the cycle has improved, ADM has reverted to its normal excessive capital spending. 

Chart M

More important than the total is what the capital has been spent on.  It appears that ADM’s maintenance capital is about $300 million per year.  In the past two years they have spent nearly $2b on capex, thus $1.4 billion has been spent on growth projects. It appears that $800 million of that was spent on Ethanol dry mills, whose first couple of years are likely to have negative full cost returns.  In addition, they spent $100 million on a rapeseed biodiesel plant in North Dakota that we are told is not being opened due to movements in rapeseed oil versus bean oil.  However, a look at the relative price of the two commodities in Chart O, does not indicate a problem.  We are trying to get more information on this plant, although it is not particularly material to the thesis.

Chart N

Regardless, at least 60% of the capital spend over the past two years has been, at best, ill timed.

EARNINGS EXPECATIONS AND VALUATION

There are a lot of moving parts in ADM, and some businesses appear to be doing well (albeit temporarily, such as soy processing and storage & transportation).   However, on balance, for the 2008 FY, operating earnings should decline by at least $100 million (equity earnings in affiliates by a like % amount).  This should bring eps in at ~ $2.30 vs. $2.40 in 2007 and $2.70 expected for 2008.  In 2009, eps should be down another $.10.  Our breakdown is as follows:

So, with EPS down slightly versus last year and down 15% vs. expectations, why do we expect the stock to go down so much? 

Since ADM is a cyclical stock, it generally trades at very low multiples when eps is peaking. 

Chart O

When the reality of peak eps hits the market, the stock will likely trade down to 11x peak eps like it did at the last peak ~ $26 per share.  However, if you don’t like peak, trough and normalized eps arguments, you should just use the keep it simple/stupid methodology: the stock price will follow eps estimates down as per chart chart P, below.

Chart P

Also, as mentioned at the beginning of the report, a dcf using the above projections and 1.2x the average normalized ROCE of the past 10 years and a 6.4% discount rate leads to the same conclusion . . . a $25 stock. 

Regards.

Catalyst