A farmland futures contract does not have a related instrument with a robust daily correlation. Nonetheless, farmland prices are influenced by many common marketplace factors. Farmland, though, maintains a longer-term economic focus and tends to ignore low-frequency marketplace noise. The resulting lower volatility makes replication modeling difficult and limits short-term usefulness.
Corn (NYBOT:JCK14) futures can react to an export or crop-related announcement. Interest rate futures answer to more mundane economic releases. But the long-term nature of farmland is not tuned to respond the same. It overlooks most short-term marketplace influences, focusing instead on events that bring about fundamental change. Interest rates can fluctuate up and down on their own, but a Fed policy change is pivotal and will influence both interest rates and farmland.
With identification of the major risk contributing factors embedded in farmland’s pricing, it is possible to determine the core drivers. These impact variables can be observed in the price action of other commodity future contracts. We can then replicate farmland values with an economically equivalent basket of specified futures.
There is documented evidence of long-term relationships between farmland prices and crops, particularly in states with a predominant crop. Kastens and Dhuyvetter (2011) evaluated Kansas farmland values vs. U.S. wheat prices from 1880 to 2010. They found a correlation coefficient between the two of 0.90. The discovery is not surprising, considering Kansas is such a wheat-dominant state.
Having identified such stout long-term land/crop correlations, replication of a farmland futures contract is feasible. While a model based on current futures contracts may be of limited value for daily trading, there is strong reason to suspect trading opportunities for much shorter periods than previously thought.
Current farmland values can simplistically be viewed from an Income Capitalization Model. Because farmland is a capital asset, its present value can be defined in a manner that exposes interest rates as a contributory factor to property values. Interest rates affect everything from loan payments to discounting future income streams. Low interest rates generally support higher land prices. The future income is less discounted while in the currency markets a weakening dollar encourages agricultural exports. Higher interest rates yield opposite effects.
But farmland pricing is often irrational, can temporarily disregard fundamentals and overreacts both on the upside and downside. Featherstone and Baker (1987) point out that “part of the explanation of change in farm asset values may lie in the time paths of adjustments to the fundamental variables, returns and interest rates.” Past studies have suggested other contributing factors in farmland pricing, including a linkage to returns, asset ratios, expected inflation and cash rents.
Moss, Schonkwiler and Reynolds (1989) attribute the 1980s price tumble to income, asset ratios and interest rate increases. Feldstein (1980) describes “a fundamental link between general price inflation and the relative price of land that deserves particular attention.” Moss (1997) found that inflation contributed the most to Florida land prices from 1960 to 1994.
Expected inflation must clearly be accounted for in a reproduction attempt. Historically, gold futures have provided a good proxy for inflation expectations. Recent inflation worries from the Fed’s quantitative easing were blatantly expressed in gold’s pricing.
Correlation analysis is subject to the period being viewed. Intervals can be found where almost any two items can misleadingly imply a strong correlation. However, specific commodities inherently share common data in their pricing. An economic synthesis of farmland values must use commodity futures that capture commonality. The recommended index has a strong economic connection with corn, soybeans (NYBOT:JSK14), gold (COMEX:GCJ14) and interest rate futures. These commodities can reasonably be used for replication purposes. Interest rate risk can be captured with either five- or 10-year Treasury note futures.
Data for the Peak Soil Iowa Cropland Value Index begins in 2005; correlation and regression analysis is limited to January 2005 to June 2013. With the understanding that a longer-term analysis might suggest otherwise, this observed 8.5-year period reveals an index that would present intriguing hedging possibilities as a futures contract.
Crop prices, expected inflation and interest rates have a commanding economic bearing on property values and are considered in the analysis. The analysis uses the month ending, spot futures contract price. The five-year T-note is used for interest rate risk, corn and soybeans for commodity exposure and gold for expected inflation. A wheat analysis was reviewed to examine the correlation to a crop not generally produced in Iowa.
Quarterly analysis used actual index levels while monthly analysis extrapolated prices from the previous and following quarter. A quarterly basis provides 34 observations while the monthly analysis offers 100 data points.