After what turned out to be a relatively strong year in 2017, the U.S. farm sector was poised to experience a pivotal year financially in 2018. Growth that reemerged for U.S. agriculture in 2017 was at risk of losing momentum as farmers endured a multiyear slump in crop and livestock prices. Interest rates and energy costs have been on the rise with the economy warming up, resulting in increased borrowing costs and tighter credit conditions for farmers. Many key indicators that measure debt repayment ability and liquidity reached unsustainable levels. All of these concerns emerged prior to the escalation of trade disputes with China and other major trading partners.
In late July 2018, the Trump Administration announced a $12 billion trade compensation package to make up for farmer’s market losses due to the ongoing trade war. Specific details of the program implementation will not be known until after Labor Day. For purposes of the 2018 Authoritative Analytics forecast, the assumption is that the $12 billion will be paid out in the 2018 calendar year. Since it is direct compensation, the impact of this highly criticized maneuver can be assessed on the U.S. farm sectors financial outlook.
The Authoritative Analytics August forecast for net cash income would have been below $100 billion in the absence of any trade compensation. Monthly values for net cash income forecasts have ranged from $98.2 billion to $107.8 billion since the first prediction in October of 2017. The initial impacts of trade disputes were estimated by Authoritative Analytics to be between $3.5 to $5 billion. With the $12 billion trade compensation being paid (assuming all payments go to production agriculture) in calendar year 2018, government payments would jump to $21.4 billion and net cash income reaches $111.7 billion. This would represent a 6 percent increase over what is estimated as the final 2017 value for net cash income and the second consecutive annual increase since falling 11 percent in 2016.
The dramatic difference in these alternative income trajectories is further amplified by examining how liquidity and debt repayment for the sector change. In each case, values of these critical measures would revert to 2016 levels. This would represent a challenging situation for maintaining cash reserves and timely repayment of debt.
Liquidity is vital to any business, but may be even more critical to agriculture depending on what is produced, the length of the production cycle, and how it’s marketed. Two measures that are often associated with liquidity are the amount of working capital and the current ratio. Working capital is calculated by subtracting total current liabilities from current farm assets. One rule of thumb, is that a business should have enough cash available in working capital to cover 5-8 months of operations expenses; although this will vary with the type of business. Liquidity as measured by working capital divided by cash expenses reached its lowest level at 0.21 (reserves of less than 3 months of expenses) in 2016 and would gradually rise to 0.26 in 2018 with trade compensation payments.
Leverage and Debt Repayment
Even though debt can become more challenging to acquire when low levels of income are sustained for extended periods of time, borrowing has increased noticeably since 2013. In addition to real estate transactions, much of the recent rise in debt can be attributed to short-term loan conversions and efforts to refinance and lock in fixed interest rates. As incomes have been declining or stagnant; operating costs have not fallen by the same amount, thus depleting cash reserves. Farm sector debt divided by earnings before interest and taxes (EBITDA) provides a leading indicator of leverage and repayment problems. The value of this measure exceeded 3.0 for the first time since the 1980’s in 2016. Much like liquidity, debt repayment would show gradual improvement through 2018 as trade compensation payments would help reduce debt divided by EBITDA to 2.79.
Uncertain Path Forward
Since some uncertainty remains regarding the specifics of trade compensation payments and the short- and long-term prospects for agriculture exports, 2018 remains a pivotal year for financial circumstances. Trade compensation payments represent, for the sector, an opportunity to sustain some of the economic growth that began in 2017 and ward off potential liquidity and debt repayment problems. Without trade compensation payments (or reinvigorated export potential), several key financial indicators will revert to 2016 levels and trigger some difficult financial adjustments for many participants in U.S. agriculture.
The monthly World Supply and Demand Estimates (WASDE) report released last week by USDA represents the first attempt to incorporate impacts from recently enacted tariffs by the U.S. and its trading partners on 2018 commodity markets. Authoritative Analytics uses these same numbers to update the financial outlook for U.S. agriculture. As a result, the initial impacts of reduced trade on U.S. Agriculture can be evaluated by comparing the June and July forecasts.
The largest impacts were on soybean and dairy product receipts (Figure 1). Receipts for these two commodities were forecast to be $4.7 billion lower than in the previous month as a result of reduced exports. Most of the reduction in soybean exports was attributed to “The tariff that China recently imposed on U.S. soybeans…” Similarly, the export situation for dairy products was affected as indicated in the report, “China’s tariffs on certain U.S. dairy products hampers exports to some extent…” The overall implication for net cash income would be a $3.5 billion drop, representing a 3-percent decline from 2017. In contrast, I reported net cash income forecasts for 2018 in June as being essentially stagnant. The impacts on net cash income would have been potentially larger (minus $5 billion) were it not for the positive adjustment to 2018 fruit and tree nut receipts stemming from the recently released NASS report.
How Bad Could IT get?
The good news is, year-to-date, 2018 was running slightly ahead of 2017 in terms of the value U.S. agriculture exports (through May, the latest month for which statistics are available). The bad news is that the remaining months, on average, account for about 60 percent of the total value of exports during the calendar year. Of course, not all of this trade value will be adversely impacted.
For soybeans, China accounted for about 60 percent of calendar year export volume on average between 2014 and 2017 (Figure 3). Through May, China had already purchased one-fourth of the previous 4-year average volume of soybeans (7.3 million metric tons relative to 31.5 million metric tons). By my math, that leaves about $10 billion of soybean trade with China at risk for the remainder of the calendar year, of which 31 percent ($3.1 billion) has been discounted in the recent WASDE report. This suggests that some of the lost sales to China will be made up for elsewhere. Through May, soybean export volume to China was off more than 20 percent from the previous year, but total soybean exports were only down 2 percent.
Trade remains vital to the financial health of U.S. agriculture. Ramifications of the trade dispute with China will have a negative impact on 2018 earnings for U.S. agriculture. The magnitude is dampened somewhat by the export potential to other countries and unmet demand in China. Brazil, the world’s top soybean exporter since the 2012–13 season doesn’t have the capacity to meet China’s demand alone, and there are few, if any, other producing countries that could step in this fall. What happens for the remainder of 2018 depends on whether and to what degree other trading partners retaliate and additional commodities are implicated.
It seems that U.S. Agriculture has been overwhelmed with negative news in the last several months. Concerns about trade tariffs, renegotiation of NAFTA, and implementation of the new Farm Bill have dominated discussions in social media and news outlets. This, in addition to the usual concerns over weather and commodity prices. I wanted to revisit, the sentiment analysis that I performed regarding U.S. agriculture that demonstrated the extreme negative reaction when trade tariffs were first announced back in March. Scraping the recent news headlines (Yahoo News and Google News) for May 25 suggests that the same concerns remain (Figure 1).
When examined for polarity, this day’s headlines scored 0.149 suggesting neutral to slightly positive sentiment (0.33 is the threshold for strong positive). The polarity scores for news headlines during the last several months ranged from a low of 0.005 on June 8 to a high of 0.39 on June 30 and exhibited extreme volatility during this period (Figure 2). News headlines such as: “Agriculture caught in US-China trade dispute” , “Mexico Retaliates With Tariffs on US Agriculture” and “Canada to US: Explain that $30 billion farm spending war chest” are what drove the sentiment index to its lowest levels. In contrast, the prominent stories that help push the index to its highest level had to do with renewed efforts to push a Farm Bill through the Senate.
Sentiment expressed in social media regarding U.S. agriculture exhibits the same volatility (Figure 3). The longer-term view of twitter sentiment scores pertaining to U.S. agriculture show that for most days during the last several months viewpoints were neutral to moderately positive. There were three brief periods of extreme negativity. On or around April 2, 2018 having to do with trade concerns, on May 17, 2018 in relation to the failure to pass the Farm Bill in the House of Representatives, and on June 20, 2018 as there was heighten concerns regarding the escalation of U.S.-China trade disputes. Similarly, there were a few brief periods when sentiment moved towards strongly positive (above 0.50). The monthly trend in the hybrid sentiment score was 0.37 for March, 0.27 for April, 0.34 for May, and 0.45 for June. All four months were neutral-to-slightly positive and trending towards slightly positive. Later we will look at alternative sentiment measures for U.S. agriculture to see if this positive trend is further substantiated.
Most of the tweets had a well defined emotional content., with the most dominant emotion being anticipation(Figure 4). Despite the high level of anxiety, positive emotions outnumbered negative emotions by nearly 2-to-1. Trust and joy were the most common positive emotions. The predominant negative emotions were fear, sadness, and disgust.
Alternative Agriculture Sentiment Measures
The Purdue/CME Group has, for several years, conducted monthly opinion surveys of agriculture producers to gauge economic sentiment. The results are used to calculate an index called “The Ag Economy Barometer.” It is designed to be a comprehensive measure of the health of the agricultural economy. In a recent report it was noted that “producers’ weakening perceptions of current conditions in the production agriculture sector, along with a decline in their expectations for future economic conditions” led to a sharp drop in the ag producer sentiment index in April 2018 and the second month in a row of declines. The report also postulates that “the undercurrent of concern expressed by producers in March became more pronounced in April as the trade dispute with key export customer China continued.” It was also noted that “the attitude shift identified in the survey extended beyond crops into animal agriculture.”
Looking at the recent history of monthly index values reveals that the largest monthly increase occurred in November 2016 (Figure 5). Coincidently, a Presidential election was settled during the same time frame. The May reading of the Purdue University/CME Group Ag Economy Barometer was 141, 16 points higher than in April 2018 and the highest barometer value since January 2017.
To get a sense of what is driving the sentiment indicated in the Ag Economy Barometer, I examined the correlation with monthly indices of major commodity profit developed for Iowa and the general prices received and paid indices from NASS (Figure 6). The Ag Economy Barometer is most strongly correlated with the profitability of corn and soybeans and had a strongly negative correlation with cattle profitability. No relationship was evident with the general prices revived index, while prices paid was negatively correlated.
Figure 6. Correlation Between Purdue/CME Ag Barometer and Commodity Indices
What about monthly changes? There are periods where the Ag Economy Barometer tracks closely with indices for several commodities (March 2016) and periods where they seem to be moving the opposite direction (Figure 7).
The Ag Economy Barometer was found to be most closely associated with corn and soybean profitability (Figure 8). Corn profitability was nearly four times more influential in explaining the Ag Economy Barometer. The results of the regression indicated that the model explained 66 percent of variance and that the model was a significant predictor of the Ag Economy Barometer, F(6,25) = 10.85.
Sentiment analysis is the measurement of positive and negative language. It is a way to assess written or spoken language to gauge if the expression is favorable, unfavorable, or neutral, and to what degree. It provides one approach to discerning opinions of individuals or groups. When done consistently, it provides a way to evaluate changes in topic views over time. At least as it pertains to U.S. agriculture, I have demonstrated that sentiment does reflect changes in the underlying economic fundamentals (commodity prices, expenses, etc…) to some degree and that news headlines seem to strongly influence sentiment. More importantly, sentiment about U.S. agriculture is extremely turbulent, which may not be surprising for an industry that is known for economic volatility.
Last month I revealed the Authoritative Analytics monthly farm financial forecasts (normally reserved for subscribing clients) in an article for the Farmer Mac, Feed magazine. Most of the questions I received had to do with “why my numbers were so much higher than USDA’s February release?” Since the last of the major commodity receipt components for 2017 was just recently released (Noncitrus Fruits and Nuts 2017 Summary, June 26, 2018), I thought it may be instructive to lay out the process of transitioning from a forecast to an estimate and how that relates to the timing of National Agricultural Statistics Service (NASS) publications.
USDA will officially report the 2017 calendar year estimates for farm income in it’s August release; nearly a full year after the fact. Some wonder why it takes so long to evaluate the preceding year. It all has to do with the timing of NASS surveys and reporting of statistics for commodity receipts and farm production expenditures. The array of reports used to compile commodity receipts are well defined in the ERS Farm Income and Wealth Statistics Documentation. These reports are released at different points between February and June of the calendar year following the year in which they are referencing. In addition to relying on the NASS reports for production and value, there is more information required (such as crop marketing patterns) to convert those commodities reported on a crop year basis.
The expected commodity receipt revisions for 2017 as part of the process of transitioning from a forecast to an estimate should be in the neighborhood of $8 billion. Most of the revisions are coming from crop receipts ($7.4 billion) and in particular from specialty crops. The revision for fruit and tree nut cash receipts should be nearly $7 billion and vegetables and melons -$1.9 billion. It so happens that the commodities with the largest revisions have the least amount of survey activity and statistical reporting during the production cycle. Livestock cash receipt revisions are much smaller, totaling just over $800 million, with cattle and calves representing the largest share ($782 million). Most of the revisions were to increase the estimate from the forecast. The three exceptions were vegetables and melons, cotton, and poultry and eggs.
The extent of revisions clearly illustrates the importance of NASS commodity surveys (all commodities) and how they are critical to achieving an accurate portrayal of farm financial conditions. Whether these revisions translate directly into 2017 income depends on the results of the expense survey (Agricultural Resource Management Survey) reported in August. Moreover, 2017 is a census year with data collection winding down and reporting to occur next year. There will likely be additional revisions to come. Stay tuned.
David Ricardo in his theory of rent emphasized that rent is a reward for the services of land, which is fixed in supply. Secondly, “it arises due to original qualities of land which are indestructible.” As a result, the basic tenets of Ricardian rent and its empirical definition have been the subject of contention in the farmland valuation literature for many years.
So, while commodities produced on farmland are the primary determinant of its value, other returns associated with land can be capitalized into farmland prices. Some of the distinctive characteristics of farmland markets are its thinness and immobility . The fact that a very small portion of total farmland is sold in a given year defines a thin market. Land is immobile, so market interactions are local and the nature of personal relationships between landowners and farmers alters the outcomes of individual transactions. Furthermore, multiple year rent contracts may lead rental rates to react sluggishly to market shocks.
Farmland Real Estate Investment Trusts
A contemporary institutional arrangement that best illustrates Ricardian rent theory is the triple net lease. Under a triple net lease, a tenant pays all of the operating expenses associated with a property in addition to the payment of rent to the landlord. Suppose that an individual owns a piece of land and leases it to a tenant, with the tenant paying for all improvements (including the construction of buildings), maintenance, insurance, taxes and other costs. The amount that this tenant pays to the landlord could be considered rent in the Ricardian sense, for the only contribution of the landlord to the real estate project is permission to use the land. Farmland Real Estate Investment Trusts (REITS) almost exclusively utilize the triple net lease in their management of farm properties. There are two real estate investment trusts (REITs) involved in the ownership of agricultural land, Gladstone Land and Farmland Partners.
Gladstone Land is a publicly-traded (NYSE: LAND) real estate investment trust that invests in farmland located in major agricultural markets in the U.S., which it leases to farmers, and pays monthly distributions to its stockholders. The company reports the current fair value of its farmland on a quarterly basis; as of December 31, 2017, its estimated net asset value was $13.96 per share. Gladstone Land currently owns 73 farms, comprised of 63,014 acres in 9 different states across the U.S., valued at approximately $534 million. Its acreage is predominantly concentrated in locations where its tenants are able to grow fresh produce annual row crops, such as berries and vegetables, which are generally planted and harvested annually; as well as permanent crops, such as almonds, blueberries, and pistachios, which are planted every 10 to 20-plus years. The Company also may acquire property related to farming, such as cooling facilities, processing buildings, packaging facilities, and distribution centers. LAND invests heavily in specialty crops. As of the 2017 10-K, in terms of revenue LAND is 65 percent in produce, 15 percent in permanent crops such as nuts and only 10 percent in primary cash crops like grains. The remaining 10 percent rent is attributed to farm related facilities.
Since they both enjoy geographic diversity, one of the primary characteristics that distinguishes these farmland REITs is the composition of the commodities produced on the managed land. Farmland REIT values should reflect rental rates on managed land. More specifically, they represent the discounted value of expected future returns. These rents, in turn, reflect the expected returns from commodities produced on the land. In the case of Gladstone Land, the returns are predominately made up of specialty crops in contrast with Farmland Partners where a larger share of production on managed land is for traditional row crops.
So far so good, but here is where the stories diverge. One year ago (May 9, 2017) these REITS had a similar valuation of about $11 per share. Since then, their values have moved in opposite directions. Have the expected future returns deviated that much between specialty commodities and traditional row crops? USDA’s Agricultural Resource Management Survey (ARMS) provides farm real estate values as a component or the reported balance sheets by production specialty. This allows some evidence of how real estate values have grown over time for specialty crops in comparison with row crops such as corn and soybeans (Figure 2).
The data do suggest that since 2013 the growth of farm real estate values have been quite different for specialty crops versus corn and soybeans. The value of farm real estate on farms that produce specialty crops increased by 40-percent from 2013-2016. In comparison, the value of farm real estate on farms that earn 50 percent or more of their total value of farm production from corn saw almost a 2-percent decline. Similarly, after peaking in 2014, soybean farm real estate value is down almost 6-percent. The 3-year moving average values represent a longer-term, expectation view and would suggest increasing values going forward for specialty crops and stagnant growth for corn and soybeans.
Farmland total returns consist not only of price appreciation, but also the earned income stream from the land. One estimate of the cash returns to farmland is cash rental rates less property taxes and other ownership costs. Using this approach, the rate return attributable to income can be computed by dividing the cash rental rate by the market value of land in the same year. The Hancock Agricultural Investment Group provides a regional breakout of the return components for their managed farmland portfolio (Figure 3).
The income returns (rent) to farmland within the Hancock Farmland Services portfolio show much less annual variability than does price appreciation. Those regions where permanent crops such as fruits and vegetables are predominant (Pacific West) have the highest percentage income returns, although declining since 2015. In fact, there is no region represented where income returns have grown since 2015. Furthermore, as is the theme of this article, the relative differences in the income returns between regions should reflect the configuration of commodities produced.
As in any investment situation, there are likely additional factors at play in explaining the divergent nature of these REIT valuations. These include, but are not limited to, revenue growth, dividend rates, dividend coverage, and operating costs. Many of these are discussed in detail at Seeking Alpha. No doubt that tariffs, rumors of tariffs, and potential trade retaliations have dampened the outlook for row crops and soybeans in particular. Beyond the sentiment, details of these tariffs are an important consideration. Fundamentally, though, farmland REIT values reflect the potential income stream from rents and not the appreciation of land itself.
Rents Versus Value
The relationship between farmland value and cash rents remains dubious. Whether rents follow or lead land valuation could be argued either way. Cash rents would follow land prices since rents are longer term and take longer to change. Earnings are capitalized directly into land values rather than cash rents. The cash rent as a mover of land value argument would view returns per acre as setting the cash rent. Land values would follow based on a capitalization of those rents. In either case, the relationship or ratio between cash rents and land values warrants monitoring.
There are a variety of ways to evaluate the relationships between rents and value. This relationship should remain relatively consistent over time regardless if cash rents or land values react first. The stability stems from the balancing of renting versus purchasing in the market. Simply dividing farmland value by rent gives a quasi price-to-earnings ratio (Figure 4). A 10-year moving average calculation allows for comparison with alternative investment returns ( Shiller PE-10 and reciprocal of the 10-year treasury yield ). The imbalances in the U.S. housing market during the late 1990s to early 2000s is easily seen along with the impact of historically low interest rates in recent years helping to push PE values above 30 for some states.
Another way to evaluate the relationship between cash rent and farmland value is to compare the amount of a 30-year fixed rate mortgage payment to rent (Figure 5). The extreme imbalance that occurred in the early part of the 1980s is readily apparent regardless of which State you measure. A sharp run-up in interest rates pushed the mortgage payment associated with farmland to a multiple of 3 in Illinois and higher in Kansas during the peak interest rates in 1981. Subsequent to the correction, rents and the cost of owning farmland track pretty closely in most states.
Effectively functioning land markets have an important role to play in the performance of U.S. agriculture. Proper development of land markets requires secure land tenure agreements and low transaction costs.
REITS may be an effective hedge against inflation for current farmland owners. Moreover, farmland assets are a particularly attractive diversifier because they have a long documented history of generating returns that have low or negative correlation with traditional asset class returns.
Overall, farmland return expectations for 2018 remain muted; just outpacing inflation for the year. The main driver is the likelihood of a fourth consecutive year of commodity price declines, leading to declining on-farm profitability and tighter operating margins. Farmland markets are further susceptible to potential trade disruptions (NAFTA, trade relations with China).
One of Ricardo’s chief contributions (arrived at without mathematical tools) is his theory of rents, which has withstood the test of time.
Note that this article does not constitute investment advice nor do I own shares in any farmland REITS. The author also appreciates the helpful comments from Dr. Todd Kuethe.
The day was first celebrated on April 22, 1970. Today marks the forty-eighth observation of Earth Day. An opportunity to raise awareness for the need to protect the environment and promote circumstances for environmental conservation. I thought it appropriate to reflect on how farmers and ranchers have contributed to environmental stewardship and the role of USDA conservation programs.
Investment in agricultural conservation has come a long way since the 1985 Farm Bill first added a Conservation Title. The Farm Bill conservation programs, taken in total, are the largest single source of funding for land conservation. Farm bill programs create significant opportunities for land trusts to protect high-priority farm and ranch lands, grasslands, wetlands, and forests. Today, there is a portfolio of payment programs and other policy instruments designed to encourage better environmental performance and accountability on U.S. farms.
Perhaps one of the more well-known and controversial programs is the CRP. The primary purpose of the Conservation Reserve Program (CRP) is to conserve and improve soil, protect water quality, and provide wildlife habitat by establishing long-term cover on highly erodible land or land in need of conservation buffers that was previously in row crop production. The CRP program offers 10–15 year contracts for the retirement of land from crop production using submitted bids subject to field specific caps. In exchange for cost-share and rental payments, agriculture producers remove environmentally sensitive land from production and plant resource-conserving land cover to protect soil, water, and wildlife habitat.
Congress created CRP in the 1985 Farm Bill due to increased concern over unacceptably high levels of soil erosion. The 1985 Bill authorized USDA to enroll up to 45 million acres, though actual enrollment has never exceeded 37 million acres. Between 1985 and 2008, the enrollment cap was reduced to 36 million acres before being increased to 39 million and then reduced again to 32 million acres. The 2014 Farm Bill gradually lowered the CRP acreage cap from 32 million acres under the 2008 Farm Bill to 24 million acres in 2018.
Despite substantial changes in producer preferences and program design; fluctuations in total acres enrolled; and tremendous technological advancements in agricultural production (many of the arguments used for lower support in the 2014 Farm Bill), payments per acre enrolled have generally declined over time and remained remarkably steady in recent years. This suggest, that at least in relative terms, the CRP program is operating with greater economic efficiency today than when its was first developed.
I offer 2 alternative methods for inflation adjusting federal CRP outlays. I first use a more traditional chain-type GDP deflator (2018=100). I also deflate expenditures by the farm real estate value index (2018=100), since the primary focus of the program and it’s instruments are farmland. Although eligibility is centered on highly erodible land, most CRP land is selected from producer offers using the Environmental Benefits Index (EBI), a benefit–cost index that accounts for a broad range of environmental concerns and the cost of the contract to the government. Compensation is meant to capture the opportunity cost of foregone production on that specific land and its ownership and maintenance costs.
In addition to reducing the acreage cap to achieve cost savings, several other modifications were made as part of the 2014 Farm Bill. This included three transition options for expiring CRP land. First, within the 2 million-acre reservation for grassland enrollments, expiring CRP acres are prioritized; the land will remain in CRP but the economic use of the land for grazing and haying is greatly expanded. Second, it allowed producers with expiring CRP land to enroll in the Conservation Stewardship Program in the final year of their CRP contract. Third, it provided two years of extra rental payments to owners of expiring CRP land who rent or sell their land to a beginning, socially disadvantaged, or veteran producer who will practice conservation on the land through the Transition Incentives Program. These program refinements are potentially at odds with the direction of the current Farm Bill discussions.
America’s farmers and ranchers have a proud tradition of conservation stewardship working in concert with the federal government to develop innovative approaches and sustainable practices that benefit the land, water, and wildlife for future generations. The next farm legislation will have an important impact on whether this stewardship legacy is maintained. The signals, so far, are not encouraging.
The House Agriculture Committee’s proposed Farm Bill reauthorizes the conservation reserve program increasing acreage from 24 million acres to 29 million acres. It limits CRP rental payments to 80 percent of established local rental rates, with a 15 percent reduction for first re-enrollment and 10 percent thereafter. But working lands conservation is cut by 25 percent over the next 5 years with the Conservation Stewardship Program (CSP) being rolled into the Environmental Quality Incentives Program. Overall, conservation takes a nearly $1 billion cut over 10 years.
Often cited but rarely defined, the law of unintended consequences suggests that actions of people—and especially of government—always have unanticipated or unintended effects. There has been a lot of discussion in the farm press this week regarding the impacts of recently enacted steel and aluminum tariffs and the pending NAFTA negotiations on future agricultural exports. The general characterization of the situation was negative, citing expressions of fear and worry on the part of U.S. farmers.
I have been examining the potential of twitter analytics for evaluating current events in U.S. farming and tracking how the consensus mood changes over time. The heightened commentary and awareness of agricultural trade concerns seemed like a perfect opportunity to see what social media opinion has to offer. Twitter is an online social networking service on which users worldwide publish their opinions on a variety of topics, discuss current issues, complain, and express many kinds of emotions. I pulled 2,400 tweets from the past 10 days (March 10-20) from all over the U.S. containing the keywords: agriculture exports, NAFTA, tariffs, trade, ethanol, soybean exports.
Words contained in the tweets were evaluated using NRC emotion lexicon which holds 14,182 different words with their emotion and sentiment classes. Most of the tweets had a well defined emotional content. Negative emotions outnumbered positive by nearly 2-to-1. Emotion words referring to fear were over-represented and accounted for nearly half of all emotion words in the sampled tweets. Anger, however, did not seem to be as strong as what might have been depicted in many of the media outlets. Trust was the dominant positive emotion, although I am not sure how to interpret its meaning in this context.
Sentiment analysis revealed some interesting results. In particular, the extreme negative rating for one day (March 16th) in comparison with the earlier and later parts of the 10-day period. The R Syuzhet package. comes with four alternative sentiment dictionaries and provides a convenient method for accessing the robust, but computationally expensive, sentiment extraction tool developed in the NLP group at Stanford University. Since these alternative measures are scaled differently, I transformed all of the outcomes to the 0-1 range for comparison purposes. Negative sentiment would be towards zero, while values approaching 1 would be positive.
During March 10-14 the sentiment values were running slightly negative prior to reaching extreme negativity for 3 days and returning to near neutral (0.50) on March 18. This suggests that aside from some of the known concerns about twitter analysis, the large drop-off could have been media driven or reflect media twitter activity.
Exports As A Source of Income
Just how important are agricultural exports to farm income? To answer this question, you have to look at calendar year export amounts in relation to gross farm earnings. Since 2008, agricultural exports have accounted for more than 30 percent of gross cash income and averaged just over 32 percent. The largest increase in the importance of exports in the last 20 years occurred between 2005 and 2008 when the percentage share climbed from 22.6 percent to 32.8 percent. That event and subsequent high levels of exports relative to income are in large part attributable to a dramatic reduction in the value of the U.S. dollar against currencies of key trading partners and competing countries .
The dollar exchange rate compares its value to the currencies of other countries. When the dollar strengthens, it makes American-made goods more expensive and less competitive compared to foreign-produced goods. The Economic Research Service provide a specific dollar index for agricultural trade, Real Annual Commodity Trade Weighted Exchange Rates. Exchange rates are one of the many factors that determine trade patterns and U.S. exports of agricultural products, but their influence has been powerful in the last decade. Their impact will be tested in the current environment as the Federal Reserve accelerates the path for US rate rises and continues to scale back its balance sheet.
Surprisingly, there has not been a lot of annual volatility in the importance of exports to farm income since 2008. This suggests that despite lower dollar valuations, other factors were countervailing. Changes in economic growth rates and incomes of importing countries can impact trade. Witness, the 2014 Ukraine Crisis and the debt crisis in Greece that has been ongoing since 2010. Changes in trade policies (often related to health or environmental concerns) can cause unexpected disruptions to trade patterns. For example, hog exports in 2013 were constrained by the PED virus and in 2015 they were limited due to concerns by China about feed ingredients. In 2017, the Chinese market to U.S. beef was reopened for the first time in nearly 14 years. Finally, U.S. and competing producers are subject to the whims of mother nature that influence world commodity supplies.
How Concerned Should We Be About Agricultural Exports
U.S. farmers and ranchers produce some of the most competitive, high-quality farm products in the world. They also consistently produce more product than can be consumed domestically. As a result, agricultural exports are important to both farmers and the U.S. economy.
FIVE Reasons We should Be Worried
Concentration of trading partners. The top destinations for U.S. agricultural exports have been limited to a few key countries. China, Canada, Mexico, and Japan represented more than half of FY2016 exports.
Strengthening of the U.S. Dollar. Any unanticipated moves by the Federal Reserve to further tighten monetary policy could result in a stronger U.S. dollar. For example, if the Fed decides to remove excess reserves by shrinking the balance sheet through asset sales rather than gradually reducing the balance sheet by ceasing to roll over securities as they mature, which began in September 2017.
Increasing competition across key commodities. The U.S. is the world’s top exporter of agricultural products, but Americas global export share has been steadily declining. Rival countries have adopted more modern farming and agricultural practices and improved their transportation infrastructure in recent years.
Changing consumer tastes and preferences. Food safety has emerged as an important global issue with international trade and public health implications that will continue to cause disruptions in trade. The consumer movement to reject modern agricultural technology, such as genetic modification, synthetic fertilizers, and irradiation has not faded.
Growing uncertainty surrounding trade agreements. Renegotiations have started on the North American Free Trade Agreement, a 23-year-old agreement that removed tariffs and significantly increased commerce between Mexico, the United States and Canada. The outcome of this trade deal will have far reaching implications.
Three Reasons to be Optimistic
Domestic and world economic growth. Steady global economic growth supports world food demand, global agricultural trade, and U.S. agricultural exports. Strong growth in developing countries is especially important because feed use and food consumption tend to be very responsive to income growth in those countries.
Greater access to new and existing international markets. USDA and other government agencies have, for many years, invested in a variety of trade promotion programs that have been successful at opening new markets and reducing trade barriers.
Continued weak U.S. dollar. Improved per capita income growth in the Eurozone coupled with a more restrictive European Central Bank Policy, and similar improvements in the economic outlook of other key U.S. trading partners in Asia should help dampen the agricultural export-weighted value of the U.S. dollar.
As Things Stand Now
Despite the legitimate concerns about U.S. agricultural export prospects in 2018 and beyond, the most recent Outlook for U.S. Agricultural Trade from USDA remains generally positive. Exports in FY2018 are projected to fall slightly below FY2017 levels and come in about $13 billion below the recent high established in FY2014. They regard robust per capita world GDP growth and continued weakening of the U.S. dollar against the currencies of many key trading partners as the two most important factors affecting U.S. agricultural exports. Growth in foreign incomes impact is mostly centered on developing countries since they have a much higher income elasticity for agricultural imports than developed countries.
The export situation for U.S. agriculture remains volatile, with the renegotiation of the North American Free Trade Agreement (NAFTA) and potential tariff wars creating the most uncertainty. This is in stark contrast to recent years where export unsureness was more depend on world growing conditions (weather) and U.S. and trading partner monetary policy changes. A complicated formulation just got more complicated. The return of Twitter sentiment to neutral territory suggests a wait and see attitude on agriculture export prospects. Stay tuned.
😀 Thanks to Mathew Shane for his constructive comments and suggestions.