Deere Short Write-Up

Deere & Company (“Deere” or the “Company”) is the largest global manufacturer of agricultural equipment with over 50% market share in North America and competitive positioning in international markets. Deere divides its business between equipment operations and financial services division (“JDFS”). Within the equipment operations Deere does the following:

• Ag & Turf (“A&T”): Approximately 75% of sales and operating income. Overwhelming majority of operating income is derived from large agriculture equipment such as 100+ horsepower (100+HP) tractors, combines, sprayers, etc. These large pieces of equipment are most exposed to the major crops such as corn, soybeans and wheat. Topline is roughly 70% North America, 30% overseas, though North America margins are more than 2x overseas margins (i.e. North America is the market that matters). Large overseas markets include Brazil, Europe and Russia. In this write-up we will refer extensively to AEM industry sales data for new large tractors in North America as this data correlates very closely with Deere A&T sales (0.92 R2 on a 14 year regression).

• Construction and Forestry (“C&F”): Approximately 25% of sales and operating income. 80% of this division is equipment for the North America construction industry. About 15% of this construction exposure is direct sales to oil & gas customers.

Deere sells substantially all of its products through independent dealers. Deere’s relationship with its dealer network is important as dealers typically take on a substantial amount of inventory (and thus risk from falling second hand values) in order to facilitate sales of new tractors to farmers. For example, the majority of new tractor sales are done on a “trade-in” basis (i.e. farmer sells a dealer a 5-year old tractor, plus cash to buy a new tractor). Through JDFS, Deere helps to support the dealer network and helps to facilitate new tractor sales by offering loans and leases on new and used tractor purchases. JDFS also provides floorplan financing for Deere dealers and provides unsecured revolving lines of credit to farmers.

Thesis

The US ag equipment market just went through the largest upcycle in history and is now in the early innings of a multi-year downturn. The implications are very negative for Deere. Evidence of this downturn is everywhere; second hand prices are falling, inventory levels are rising, dealers are struggling, and new equipment sales are collapsing. Analysis of historical data tells us long-periods of excess equipment spending are typically followed by long periods of below normal spending. Rarely (if ever) does the downturn last less than 5 years. Despite the dire near and long-term outlook, sell-side is projecting a shallow dip in sales next fiscal year (FY2016) and then a recovery starting the following year (FY2017). We think there is strong evidence the industry sees a sharp downturn in 2016 and that the low level of sales persists at least through 2020. This downturn will also impact JDFS as lower sales will result in a shrinking loan book, falling lending margins and possibly losses from residual values on leases. Lastly, falling sales, net income and free cash flow will impact Deere's ability to continue to return significant amounts of capital to shareholders in the form of share buybacks and dividends. We believe Deere will have to curtail its share buyback program in the next fiscal year. The combination of earnings misses, downward estimate revisions and decrease in shareholder return should catalyze Deere stock to fall to fair value, which we believe is almost 50% below current trading price.

Historical review of ag equipment

Before delving into Deere, it’s worthwhile to go through the history of ag equipment in North America. The ag equipment needs of this country have not grown in 30+ years. Perhaps surprising to some, the US has been able to steadily increase crop production for several decades while holding the amount of farmland flat (i.e. productivity has increased). Large ag equipment needs are a function of acres planted. The charts below show that both acres planted for major crops (corn, soybeans and wheat) and the size of the tractor fleet have essentially remained unchanged since the mid-1980s. (Note: the US tractor fleet shown below includes all sizes of tractors, whereas we only focus on large, 100HP+ tractors in this write-up; however, this is the only source of data on fleet size and directionally shows that the tractor fleet has not grown).

Similarly, new ag equipment spending / tractor spending has gone through long periods of stability. For example, in the 20 year period between 1986-2006 new large tractor sales averaged 21,700 units per year and in almost all years during this period, sales fell within 5,000 units of this average. However, historically there have also been several long periods where ag equipment spending / tractor sales have deviated significantly from these levels. The large deviations to the upside (i.e. more spending) relate to more rapid fleet renewal when farmer income is at unusually high levels (historically caused by exogenous demand shocks). These large deviations to the upside are then followed by lower than normal sales as fleet renewal has a logical limit (i.e. there are only so many new tractors you can buy, regardless of your income). In the past century we have seen several large deviations for normal levels including the period from the 1970s to the early-1980s and the period from 2007 through today.

Contrary to popular belief, ag equipment spending / new tractor sales are not usually correlated with farmer income. For example, during the 20 year period between 1986 and 2006, the correlation between farmer income and ag equipment spending was close to zero. The low relationship between tractor sales and farmer income over such a long period of time highlight the fact that farmer income is not the key determinant in ag equipment / tractor spending.

Ag equipment from 2007-2014

The period from 2007 to 2014 marks the biggest agriculture equipment up-cycle in history. This wave of purchasing corresponded with a large upswing in crop prices and farmer income, largely driven by the implementation of the ethanol mandate in the US. This cycle was also aided by favorable tax policy (Section 179), that allowed for accelerated depreciation on ag equipment purchases (very helpful considering how much money farmers were making). During this period, farmers spent more than they ever had on farm equipment. This upturn was also notable for its length, as new equipment buying was significantly above normal levels for 7 years.

During this period, the size of the tractor fleet did not grow, thus the average age of the fleet was decreasing. A good measure of the age of the tractor fleet is the trailing 10-year new equipment sales (we define as the “10-year and younger fleet”). As a result of the enormous amount of buying that occurred from 2007-2014, the 10-year and younger fleet reached a multi-decade high in unit terms (and reached an all-time high in horsepower terms).

Ag equipment from 2015 and beyond

Ag equipment purchasing peaked in 2013, stayed high in 2014 and year-to-date 2015 is taking a sharp turn lower. The catalysts for the turnaround were a drop in crop prices, drop in farmer income, changes in tax policy (lapse of Section 179) and a very young tractor/combine fleet. The year to date drop in new equipment sales has been steep, at -23% year over year through September. Equally as concerning, the negative trends appear to be accelerating, with September sales down 40% y-o-y, a multi-decade low in terms of monthly sales declines.

How low will sales go and how long will they stay low for? We think the answer to this question lies in the fleet age. As pointed out earlier, the 10-year and younger fleet is now over 300,000 units. Based on historical tractor sales levels, and keeping in mind that planted acreage is not growing, we think “normal” age is closer to 200,000-250,000 units. Analysis of prior upcycles in equipment buying tells us that equipment age tends to mean revert. We don’t think this time is any different. The analysis below highlights how much new equipment sales would need to fall in order to bring the 10-year fleet size back to 260,000 units. Assuming a normalization process that lasts until 2020, we see sales falling to 20,000 units in 2016 and then 15,000 units in 2017-2020. Thus, we ultimately see sales falling another 50% from current levels and then staying low for another 5 years.

Excess dealer inventory, falling second hand prices

Further evidence of a pronounced and prolonged downturn can be found in the dealer channel. The dealer channel is important because it serves as Deere’s primarily distribution channel for new equipment. Dealers are also important because they facilitate new equipment sales by buying second hand equipment (i.e. a “trade-in”). Unlike for other types of machinery, almost every new piece of Deere equipment sold is done on a trade-in basis. For example, a buyer of a new tractor might turn in his 5-year old tractor to significantly reduce the cash component of the new tractor purchase. Without the trade-in, the upfront cash consideration would be too large for the farmer and thus the new equipment would not be sold.

Today, there is strong evidence of excessive inventory in the dealer channel, particularly in second hand equipment. Channel checks with dealers and consultants confirm that used equipment on dealer lots has been growing substantially over the past year, as do the balance sheets of public dealers. The table below highlights changes in inventory turns for Cervus, which is the only publically traded Deere dealer and the largest Deere dealer in Canada. Inventory turns on new and used equipment have fallen significantly over the past several years.

The excess inventory is putting pressure on second hand prices as dealers are become more desperate to offload their aged pieces of equipment. Our channel checks indicate that dealers are increasingly utilizing auction sales as a measure of last resort for clearing out equipment. Machinery Pete, a good industry data source for second hand pricing, suggests that prices for the highest HP used tractors (>200HP) have steadily trended downward since 2013. The table below tracks the average sale prices of used Deere 8360R and 8335R tractors. These models serve as good benchmarks for pricing trends within high HP tractors. While there are only a handful of data-points, pricing for newer tractors is down 10-23% from peak levels.

Excess inventory and the drop in second hand pricing is having several negative effects on the industry. First, aged inventory is putting strain on dealer balance sheets. As many dealers are more focused on right-sizing existing inventory levels, their ability to take-on new inventory is limited or non-existent. Since most new sales are done on a trade-in basis, this impacts a dealer’s ability to facilitate new unit sales. Second, even if a dealer’s inventory is under control, the increasing differential between new and used pieces of equipment (new prices have not fallen) means that the cash component of the purchase is growing (i.e. the trade-in economics are worsening). As the cash component of a new equipment purchase grows, it further dissuades farmers from purchasing new. This second dynamic is important as in recent years, as a result of a strong second hand market, large sales incentives from Deere and tax depreciation for equipment purchases, a number of large farmers were buying new equipment every year (i.e. buy new, trade in 1-year old equipment, repeat). We believe a meaningful portion of new equipment demand in recent years came from these types of annual buyers. However, conditions that made these types of trades favorable are now evaporating, potentially erasing an important source of demand. First, the softening second hand market is having a large impact on young, second hand equipment. Whereas the difference between a brand new combine and a 1-year old combine may have been around $40,000 in 2012, today the difference might be more than 3x that amount at $125,000. The increase in the “trade-in difference” materially impacts the economics of buying a new piece of equipment ever year. Second, there is increasing uncertainty around the future of accelerated tax depreciation for farm equipment as Congress has yet to renew the law that allowed for such tax savings. Even if accelerated tax depreciation were to be renewed, the impact will be less than in prior years as farmers have less income to shield from taxes (some farmers likely have no income to tax shield).

Our channel checks indicate that dealers expect it to take 2-5 years to right-size second hand inventory levels. Of particular concern, dealers now estimate that it takes 3-5 used transactions to “wash out” a new equipment sale. This means that it takes a dealer 3-5 second hand transactions in order to fully cash out a new equipment sale. As used equipment inventory turns are now around 1.5x, this would imply it takes at least 2 years for a dealer to receive all the cash from a new equipment sale today. We think it could be hard to incentivize dealers to sell new equipment, when it might take 2+ years for them to receive the cash for such a transaction. Despite what we are hearing from dealers, sell-side (and to an extent Deere itself) seem to think that used inventory can be cleared up within a year, resulting in conditions for a rebound in new sales in 2017. We think this is far too optimistic and dealer inventories would seem to suggest a much more prolonged downturn.

Deere is now its own biggest customer

Through leases offered by JDFS, we believe Deere itself is now the largest end buyer for its ag equipment. The proportion of Deere sales that are done on lease have sharply accelerated this year. In a typical lease transaction, Deere will agree to buyback the piece of equipment after 3-years at a pre-set residual value. Leases are advantageous for the farmer/dealer because they require minimal upfront capital, are not dependent on a trade-in and move the second hand value risk from the farmer to Deere. In the past 10-years, leases on Deere balance sheet have grown from $1.3 billion in 2005 to $4.4 billion as of July 2015. Leases have continued to rise despite sales falling, up 40% between 2013 and Q3 2015 vs. a 32% drop in sales over the same period.

Similarly, the proportion of transactions financed with leases is increasing rapidly. According to UCC data, from 2003-2011, 16% of Deere’s new 100HP+ tractors were leased, vs. 29% for the past 12 months through July. In some of the most recent months, 40%+ of new tractor sales have been financed through leases.

As farmer demand for new equipment purchases is slowing, JDFS is effectively stepping in to provide demand by taking leases on its balance sheet. While Deere has historically seen minimal losses from residual values, going forward we think there is downside risk on residual values as a wave of leases hit the second hand market in three years time. By engaging in such elevated levels of leasing, Deere (as well as CNH and AGCO) is creating substantial headwinds to its business several years out as this large quantity of leased equipment has the potential to depress the second hand and new equipment market once off lease. One also has to wonder what demand levels for new ag equipment would be if Deere had not stepped up and massively increased its lease programs.

Challenges in Deere’s other business lines

North American large ag equipment sales represent a large portion of Deere’s sales and an even larger percentage of profits. However, Deere does have four other business lines worth mentioning.

JDFS has grown its finance book substantially from 2005 to 2014, from $20 billion to $40 billion, which has led net income to grow from $275 million to $544 million. In recent years, JDFS has benefited from historically low loss provisions, expanding net interest margins, and a growing amount of Deere new equipment sales (which allows JDFS to grow its book). Roughly 80% of JDFS’ portfolio comes from North American ag. As farmer profitability remains lower as a result of commodity prices, dealer cash flow remains stressed with excess used inventory, and demand for new equipment remains compressed for reasons stated above, JDFS will likely experience a shrinking loan book and lower margins. Assuming the loan portfolio returns to a more normalized $25 billion and loss provisions return to the 15-year average of 0.4%, JDFS net income should return to about $350 million. While this is a steady-state net income that will likely not occur in 2016 due to the lagging effect of sales on the loan book, it also does not account for the potential impact of rising interest rates on JFDS’ funding costs or for the increasing pressure on financing rates from banks (such as Wells Fargo) and ag-specific lenders (such as AgDirect).

Deere’s international business has sales in over 100 countries but Europe and Brazil are its key markets. Lower global commodity prices are hurting farmer profitability everywhere, but a stronger dollar has also pressured Deere’s international results. Specifically, in Europe the dairy sector has struggled and Deere has guided to a 10% decline in Europe ag sales. In Brazil, higher financing rates, lower farmer confidence, and economic uncertainty are leading to lower equipment sales. On its FQ3 2015 earnings call, Deere lowered its 2015 forecast for South American ag from a 15-20% decline to a 20-25% decline. In other machinery sectors, Brazil is down 70% from its peak, while ag is down only 40-50%, so there may be significantly more downside.

Deere’s small ag equipment business sells tractors and other equipment to livestock farmers, as well as lawn and garden tractors, compact utility tractors, residential and commercial mowers, utility vehicles, and golf and turf equipment to commercial and retail customers. In recent quarters, sales to livestock farmers were a bright spot of Deere’s A&T division. However, since August, cattle prices are down more than 10%, which will reduce demand for new equipment.

Deere’s C&F business sells machines used in construction, earthmoving, material handling and timber harvesting. Management has guided C&F revenue down 17% in FQ4 2015 as a result of slowing demand in the energy sector and the decline in rental company utilization rates impacting new equipment demand. Deere has also commented that competitors are getting more aggressive on pricing. The combination of continued end market pressure, soft demand driven by excess rental company capacity, pricing pressure and high decrementals set up for a surprise to the downside given consensus’ expectation for a 1% increase in sales in 2016.

Large divergence from consensus expectations, significant downside to stock price

As discussed earlier, we believe industry sales will need to fall to about 15,000 units per year through 2020 in order to normalize the fleet age of ag equipment in North America. Deere Ag & Turf sales have historically correlated very closely with industry unit sales, thus we can forecast with a high degree of confidence, Deere Ag & Turf sales for this level of industry activity.

Using the regression analysis above, we estimate that industry unit sales of 20,000 per year would result in A&T sales of $14.6 billion and 15,000 would result in sales of $11.1 billion per year. This compares to consensus estimates of $18.1bn in 2016 and $18.7bn in 2017. While sell-side expects this downturn to end in 2016, we are less sanguine about the market and think that the downturn could easily last through 2020.

Recent channel checks with dealers confirm not just the long-term view (2-5 years for inventory to normalize), but also the short-term view that consensus estimates are far too high. According to dealers we have recently spoken too, most see sales from early order programs (“EOP”) down 30%-60% versus early order programs from last year. While Deere is still in the early stages of the program and order books today do not definitively predict 2016 sales, the weak results to date are striking for two reasons. First, Deere provides larger discounts for earlier orders, so by delaying orders to later in the EOP, farmers will be paying more for their combines. We therefore think the fact that farmers did not place orders early in the EOP means it is more likely they simply skip out on this year’s EOP rather than pay more for their equipment. Second, the magnitude decline of this year’s EOP is shocking. Dealers that sold dozens of combines in 2014 EOP have sold low single digit combines in 2015 EOP. We view the level of weakness in this year’s EOP as strong evidence that FY2016 A&T sales will come in significantly weaker than the market expects.

Drops in sales are magnified in the operating income and net income lines as Deere has a significant amount of operating leverage. The table below highlights our operating income and net income / EPS estimates for Deere, given our view of 20,000 industry units in 2016 and 15,000 units thereafter. Our model also assumes a 5% decline in C&F (assumes current weakness carries on into part of 2016) and a gradual return to normalized net income in JDFS (for the reasons discussed earlier). Lastly, our model conservatively assumes decremental margins fall from 31% in FY2015 to 21% going forward as we assume Deere can continue to take out costs to manage the topline decline. We would note that managing the decremental will likely get more difficult as sales fall and that in historical cycles operating margins have gone negative as sales have fallen. While we think the downcycle will likely last 5+ years, we think that it’s appropriate to also think about Deere’s earnings power on a mid-cycle basis. Starting with long-term historical tractor sales average of 21,500 units, we see Deere A&T sales at $15.7bn. We apply the long-term average operating income margin of 7.5% to the equipment operations and assume JDFS falls to $350m of net income. Lastly we assume some share count reduction as excess cash is used for buybacks. Putting this all together we see Deere mid-cycle earnings at about $3.59 per share. Similar to topline, our EPS estimates for Deere are far below consensus estimates. For FY2017 we are more than 80% below consensus.

Deere’s long-term forward PE multiple has been 12.9x (1990-2015). Applying a DCF to the earnings stream between 2016-2020 and then applying a 15x multiple on mid-cycle/2020 EPS of $3.59 per share, discounting back at a 7.5% cost of equity, we arrive at a $43.00 stock price target, which is 46% below current price.