By Bert Ely
As last month’s FCW reported, on December 1, the ABA sent a letter to Farm Credit Administration (FCA) chairman Dallas Tonsager registering ABA’s strong objection to CoBank’s participation in a $1.55 billion credit facility that had been extended to CyrusOne LP, the operating partnership of CyrusOne, Inc., a NASDAQ-listed Real Estate Investment Trust, or REIT.
CoBank’s portion of the loan participation was not disclosed, but could easily fall in the $50-$100 million range. CyrusOne is in the business of leasing buildings used to house computer data centers to a variety of businesses. It is highly unlikely that any of CyrusOne’s lessees is engaged in agriculture. Further, none of its data centers are located in rural areas. There is simply no way this loan participation is permissible under the Farm Credit Act. Although this complaint was filed almost two months ago, Tonsager has yet to respond.
FCA has failed to correct material loan-data misclassifications
In the November 2016 FCW I presented data demonstrating that the FCS’s on-farm lending as a percentage of its total loans outstanding had steadily declined from year-end 2011 to year-end 2015 due to the fact that the FCS increasingly lends off the farm, to agribusinesses, rural utility cooperatives and investor-owned utilities, and other non-farm borrowers. My analysis did not predate 2011 because, according to the FCS’s Annual Information Statement for 2015, “certain [FCS] institutions identified errors in loan type classifications that are used to report loan portfolio information” that were corrected only back to 2011.
That is, each institution correctly reported the total amount of loans outstanding, but individual loan categories were not correctly reported. The FCA later advised me that it would not restate FCS loan data by type of loan prior to 2011 even though these loan classification errors probably long predate 2011. These errors also mean that whatever reliance the FCA placed on data by type of loan was misplaced. For example, the accuracy of loan performance measures, such as non-performing loans as a percentage of total loans outstanding and the adequacy of loan-loss reserves by loan category has been questionable in recent years. These errors, in turn, are reflected in the FCS’s audited financial statements, which reported these inaccurate loan data in a footnote to those statements.
These classification errors are discussed at fca.gov, which the FCA quietly posted on its website on August 17, 2016. The title of this document, “CRS Disclosures,” provides no inkling that it discusses material loan-reporting errors by numerous FCS institutions. The document did state that in addition to correcting their December 31, 2015, call reports, “those institutions with material errors… were also required to submit corrected data [for the report’s schedule of loans outstanding by loan type]for December 31, 2014, and December 31, 2013.” However, the offending institutions were not required to file corrected call reports for those two year ends nor was any mention made of the 2012 and 2011 error corrections even though corrected loan-classification data for those years was published in the FCS’s 2015 Annual Information Statement.
Not only are those incorrect call reports still accessible through the FCA website, but there is no statement on the website alerting users of those reports to the errors in them. FCA justified this inaction by stating that “corrected data were not incorporated into FCA’s Consolidated Reporting System (CRS) reports because they would create inconsistencies between the schedules.” That is an absurd statement, for there is no reason corrected call reports could not have been filed for 2014 and 2013 and even for 2012 and 2011.
The FCA did try to help users of FCS data by providing a link in the CRS Disclosures document to a spreadsheet which showed reported and corrected figures for each FCS loan category for 2014 and 2013 for each institution that had made loan-classification errors. What is most interesting about these errors is the fact that they were concentrated in just 29 of the 81 FCS institutions that existed at the end of 2014. Further, the institutions with classification errors were concentrated in just two of the four FCS districts, the ones served by AgriBank and AgFirst Farm Credit Bank. No loan classification errors were reported by FCS associations served by the Farm Credit Bank of Texas and CoBank yet CoBank itself reported the greatest amount of loan-classification error of any FCS institution.
At the end of 2014, CoBank report owning $489 million of farm real estate mortgages; that number was revised to $1.75 billion. For production and intermediate term loans, CoBank initially reported $3.26 billion of loans; that number was revised down by almost half, to $1.64 billion. CoBank also misreported its loans to cooperatives by almost 10% and it initially underreported its processing and marketing loans by almost half. Overall, at December 31, 2014, CoBank had misclassified loans equally about 4.5% of its total loans outstanding on that date. Other notable reporting errors occurred at AgFirst , First South ACA, Cape Fear ACA, and AgCarolina ACA.
The question which has not been answered, but which should be addressed in the FCS’s Annual Information Statement for 2016, is how did these loan-classification errors occur, why were these errors concentrated in just a few FCS institutions, and perhaps most important of all, why did they go undetected so long? Why did both FCA examiners and the FCS’s auditors miss them for so many years? Answers, please.
CoBank CEO Bob Engel retired six months early
Long-time CoBank CEO Bob Engel retired on December 31, 2016, six months earlier than his retirement date first announced by CoBank on June 2, 2016. Engel was succeeded by Tom Halverson, previously CoBank’s chief lending officer.
Halvorson, who joined CoBank from Goldman Sachs, was the leading inside candidate to succeed Engel as he had been serving as CoBank’s “chief banking officer, with responsibility for all lending operations.” The most interesting question, which CoBank has not discussed publicly, is why Engel left six months earlier that he had planned to. Was his early departure possibly related to CoBank’s overly aggressive lending in recent years that led to sharp questioning at the most recent oversight hearings conducted by the Senate and House Agriculture Committees? It will be interesting to see if Halvorson reins in CoBank’s aggressive lending activity.
Other recent high-level turnover at CoBank suggests substantial changes at CoBank. Last August, CoBank announced that President Mary McBride and Chief Risk Officer Lori O’Flaherty would “leave the bank at the end of 2016 to pursue other opportunities.” Last month, CoBank announced that Dan Key, its current credit officer, would be retiring in March 2017. Key is being succeeded by Bill Davis, who had been serving as chief credit officer for Farm Credit Services of America, the FCS’s largest direct-lending association. It serves all of Nebraska, Iowa, South Dakota, and Wyoming. Other changes in CoBank’s executive line-up are expected as Halvorson settles in as CEO.