The Illinois State Board of Investment has fired its longstanding farmland investment manager and plans to sell its $46-million farmland portfolio, citing high management fees and low returns. Though the holding is relatively small for an institutional investment, the exit is notable since the Board, which oversees $15 billion in pension assets for legislators, judges and other state employees, was among a handful of early public retirement funds to venture into agricultural real estate. The Board originally hired specialty manager Cozad/Westchester Agricultural Asset Management¹ in 1990 to invest $50 million in U.S. row-crop land.
The Board’s current “permanent” cropland portfolio includes almond, wine grape and citrus land in California, and wine grape, apple and cherry property in Washington. The portfolio is highly concentrated—by both geography and crop—with 90% of the assets in California and 75% of the assets held in wine grape and almond properties. Through June 30, the portfolio has generated a 5.6% annual net return since 1998, according to an August memo by Marquette Associates, an investment consultancy hired by the Board that recommended terminating Cozad/Westchester. The Board’s farmland return is in sharp contrast to the average 15% annual total return before fees reported over a comparable period for 69 permanent crop properties worth $388 million in the NCREIF Farmland Index. The Index is compiled by the National Council of Real Estate Investment Fiduciaries, an investment manager trade group.
Part of the performance lag comes from Cozad/Westchester’s fee structure, which includes an annual 0.80% investment management fee against the properties’ fair market value, and an annual 1.8% property management fee on gross income from crop sales. Cozad/Westchester also negotiated an incentive fee of 5% to 15% of the two-year average portfolio net income in excess of 5%.
Still, those fees don’t explain the yawning 9.4 percentage point return gap between the NCREIF Index for permanent crops and the Board’s portfolio. Stuart Meacham, Chief Operating Officer at Cozad Asset Management, didn’t respond to multiple requests for comment.
There is a dearth of practical performance data when it comes to tracking farmland investments. Investment promoters are often quick to highlight the compelling returns reported in the NCREIF Farmland Index. For the 15 years through June 30, the Index reports a 13.8% annualized return before fees. At mid-year, the Farmland Index included 539 properties valued at $4.6 billion. That’s less than a 2% slice of the estimated $2.5 trillion U.S. farm real estate sector. In addition, permanent crop properties, which typically generate higher income yields, comprise 34% of the value of the Index while such properties are estimated to make up no more than 20% of the overall U.S. farmland market. And because the Index is calculated at the property level before management fees, it doesn’t reflect the results generated by farmland investment managers.
Farmland Index Overstates Returns
In addition, the NCREIF Farmland Index excludes the typical three-to-seven-year development period required for new permanent crop properties to reach commercial production levels. As a result, NCREIF overstates the Index return and misrepresents the returns that investors can expect to achieve, contends Marquette Associates in its August Board memo summarizing farmland investment return characteristics and structural challenges.
Scott Richards, a senior portfolio manager at the Illinois Board, notes that private equity indexes do not exclude development periods from their return calculations. “These [NCREIF Farmland Index] returns look so great, but you are losing money those first three to five years, and it doesn’t go into the index,” he says. “To me, it is gaming an index.”
“The index is designed to understand how operating properties are performing,” says Dan Dierking, NCREIF’s chief information officer. “It’s not necessarily an investment benchmark.” NCREIF previously published investment returns for development properties in its detailed report, but ended that practice around 2005 to mask the results of Hancock Agricultural Investment Group, which was then the only manager contributing data on development properties. Though NCREIF continues to collect development property return data and now has multiple managers reporting development returns, it hasn’t resumed publishing the performance results.
The Illinois Board’s initial entry into farmland consisted of a portfolio of 25,456 acres of annual row-crop land in Illinois, Indiana, Ohio, Colorado, Mississippi and California, which was leased to local farmers for rental income. In 1998, the Board turned its attention to opportunistic real estate investments and moved to liquidate the row-crop holdings. The investment wound up in 2002 and generated $28 million in net earnings—a 7% internal rate of return or 4.2% after adjusting for inflation, according a Cozad/Westchester document.
The Board then re-invested half the proceeds in 1999 with Cozad/Westchester to develop a new permanent cropland portfolio dubbed Premiere Partners V. Westchester bought 1,700 acres in California, Florida and Washington and developed the land into the groves, vineyards and orchards that make up the Board’s current portfolio. (A 531-acre DeSoto County, Fla. pasture parcel purchased in 2001 for $730,600 and developed into a 409-acre orange grove was sold in 2006 for $4.85 million.)
The lengthy gap between the Board’s initial investment and when the new tree and vine plantings began producing a commercial crop made Premiere Partners V a net loser through at least its first six years, according to a 2005 Board response to an inquiry by Mercator Research on the portfolio. ■