Farmland values went through boom and bust in the 1970s and 1980s, probably due to the volatile farm income situation at the time. Today, farm income is once again volatile as gyrating commodity prices, compensatory government policies, and unpredictable weather interact. How might all this affect future farmland values? Will we see a return to the devastating cycle of the early 1980s? To find out, let’s explore the historical relationship between crop returns, cropland rents, and farmland values in Minnesota.
Over the long run, logic suggests that
In this article, for the sake of simplicity, I’ll focus only on the returns from corn and soybean production and will ignore costs and returns from other farm enterprises—such as livestock production, which may subsidize or draw from the crop enterprises during the year.
The expected net return of tenants, however, involves making predictions about future crop prices and yields, government subsidies, and cash expenses. To calculate the expected rate of return for tenants who rent cropland, I gathered cost, income, and land-rental rates from the 1983–99 annual reports of the Southwestern Minnesota Farm Business Management Association (SWFBMA) and calculated a breakeven rental rate by subtracting out land costs from total costs. The results are shown in table 4.
| Table 4. Breakeven amount available to pay rent at different prices and yields, 50-50 corn-soybean rotation | ||||
| Corn/Soybean prices ($/bu.) | ||||
| 1.60/4.27 | 1.80/4.81 | 2.00/5.34 | 2.20/5.87 | |
| Corn/Soybean yields (bu.) | Rent per acre of cropland | |||
| 140/45 | 69 | 72 | 90 | 116 |
| 150/48 | 81 | 85 | 105 | 132 |
| 160/51 | 94 | 98 | 119 | 149 |
| 170/54 | 107 | 111 | 133 | 165 |
The data in table 4 were calculated using the following assumptions:
The data in table 4 show that when corn sells for $1.60/bu. and yields 140 bu./acre, the breakeven rent is only $69 per acre. For farmers to continue to pay rent at 1999 levels, however, corn must sell for $2.00/acre and yield about 140 bu./acre. In contrast, if corn sells for $2.20/bu. and yields 170 bu./acre, the breakeven rent increases to $165/acre. (These figures for corn, of course, are based upon the associated yield and price figures listed for soybeans in table 4.)
Figures 10 through 12 show the relationship between year-to-year changes in breakeven rents, rents paid, and land values during the period 1983–99. The numbers are adjusted to account for any required acreage set-aside that might have been in effect.
Figure 10 shows that the United States Department of Agriculture (USDA) land values seem to track the SWFBMA rents fairly well, but appear somewhat more volatile as values rose more than rents in both the early 1980s and the late 1990s. (Figure 10 uses data collected by the USDA’s National Agricultural Statistics Service, which defines land value as the value of land and buildings.) Because of the difference in geographic coverage, the statewide USDA rental rates and values are not directly comparable to the SWFBMA data—but they do give an indication of trends prior to 1983 before the Association rent data became available.
Figures 11 and 12 show the breakeven amounts that remained to pay rent. Figure 11 is based on estimated harvest-time corn and soybean prices, while figure 12 is based on the average prices received in the next calendar year. In both figures, the lower, light-shaded area is the (calculated) amount that remains to pay rent from market returns after subtracting cash operating expenses, depreciation, and the opportunity cost of operator labor and management. The upper, dark-shaded area shows the contribution made to breakeven amounts by government payments and small amounts of miscellaneous farm income.
Both “willingness to pay” rent based on next year’s expected returns, and “ability to pay” rent, based on proceeds from last year’s crop, probably figure into what tenants offer landlords each fall when they renegotiate rental rates.
Figure 11 is an attempt to get at tenants’ willingness to pay based on expected returns at the time of negotiation. Harvest-time prices are probably the most current price information available at that time. (I lagged the rental rates by one year for comparing to figure 11’s breakevens because the next year’s rates would have been negotiated around the time those breakevens were becoming apparent, around harvest time.)
Figure 11 shows that breakevens based on harvest-time prices never exceeded $120 in any year, but they did at least exceed actual rents in every year except for 1991–93. Over the past five years, Association corn-soybean breakevens averaged $14 per acre more than rents actually paid. Looking back further over the entire 16 years, breakevens averaged $9 per acre more than harvest-time breakevens.
Figure 12 attempts to get at tenants’ ability to pay by calculating the breakevens based on average prices received for corn and soybeans in the next calendar year. Many farmers store grain because they believe they can do better than selling at harvest time. Storage involves additional costs such as shrink- age, interest payments to the government on outstanding loans on the crop, and facility costs.
Farmers who store crops probably expect prices to increase in order to cover these added costs—but this is a bet that has not paid off over the past couple of years. Comparing annual average corn prices over the 16-year period 1983–98 with year-earlier harvest-time prices, annual prices averaged $2.22 compared to average harvest-time prices of $2.18. Cash corn prices were higher than at harvest in seven of the 16 years. Soybean annual prices averaged $5.87 compared to the harvest-time average of $5.77.
Calendar-year average prices have been more volatile than harvest-time estimates. As a result, figure 12’s calculated breakeven amounts based on the calendar-year prices are also more volatile. (Rental rates are lagged two years in figure 12 because the marketing year would not be completed in time for the first year.)
Returns were sufficient to pay the reported paid rent in 12 of the 15 years for which data are available, but there were shortfalls in 1988, 1993, and 1997. In years where a shortfall occurred, breakeven rents (after all costs were deducted) declined by 59 percent (1987–1988), 69 percent (1992–1993), and 56 percent (1996–1997).
The two main sources of economic returns that a landlord stands to gain from owning cropland are current-year rents and expected future capital gains when the land is sold. Future capital gains, in turn, may be related to expected future rent increases.
The relationship between land rents and values was a popular research topic during the run-up and later collapse of land prices during the 1970s and 1980s. A central question of this research was whether a simple capitalization formula (value=[annual rent / capitalization rate]) could predict land prices in the current year based on current rental rates, or whether a more sophisticated model is needed. We usually don’t observe a capitalization rate; instead, we infer it from the ratio of land price and annual rental rates.
In the mid-1970s, farmland increased in value by over 20 percent per year. Rental rates (after paying property taxes) increased even faster, and the calculated capitalization rate peaked at 8.3 percent in 1975. Expectations of future price increases may have been a factor in land prices being bid up during the early 1980s—even though rental rates were leveling off. The result was that capitalization rates bottomed out at 5 percent in 1981. When land prices declined in the mid-1980s, rates rose for a few years, then declined once again.
Another estimate of the capitalization rate can be calculated using the ending-market balance sheets of farms in the SWFBMA. In 1998 the average farmer owned 219 acres of cropland valued at $348,007. This translates into an implied cropland value of $1,589 per acre. The corresponding after-tax average rental rate was $76 per acre, which means that the inferred capitalization rate for farms in the SWFBMA was 4.8 percent in 1998.
Other trends are also apparent in figure 10. For example, in Minnesota as a whole, land price movements appear to have overshot the upward trend in rents in the early 1980s and “overshot” the downward trend in 1987—and might be overshooting rents at the present time. This suggests that a simple capitalization model probably does not predict land prices with any useful level of accuracy.
How much could land values change if they responded in direct proportion to the range of breakeven rents shown in table 4? It seems clear from figures 10–12 that rents and values do not respond very quickly to changes in farm income, but a series of good or bad income years eventually do bring about a response. Under a pessimistic production scenario of 140 bushels of $1.60 corn (using an after-tax cap rate of 6.2 percent and property taxes that vary in proportion to rent), the calculated capitalized land value is $930 per acre. At the other extreme, 170 bushels of $2.20 corn and a 4.8 percent cap rate yield a land value of $2,883 per acre.
The range of prices and yields discussed in this article suggest that in the next several years, land values might be as much as 40 percent lower or 80 percent higher than the current SWFBMA average value of $1,589 per acre. It depends on where prices and yields end up. By comparison, in the last boom-and-bust cycle, land values peaked in 1981, declined by over 50 percent during the next six years, leveled off in 1987, and climbed back to today’s levels.
So what does the future hold for land values in Minnesota? That depends on how farmers react to future changes in the prices and yields of the crops they grow. It also depends on whether or not the federal government maintains the recent very high levels of subsidies paid to Minnesota crop farmers.
Bill Lazarus is an associate professor and extension economist with the Department of Applied Economics at the University of Minnesota.
Back to Minnesota Ag Economist Newsletter Index Page
University of Minnesota Extension Home
Page