When FAS 157 took effect last November, many finance departments and business units faced a new accounting requirement but lacked any new technology with which to address it. To determine the fair value of a wide range of balance-sheet items, they turned to that old standby, the spreadsheet, to piece together models that they hoped would make sense to shareholders and auditors.
A year later, not much has changed. Spreadsheets are often still the starting point when trying to figure out what a portfolio of credit default swaps or a series of collateralized debt obligations would fetch on the market. That’s a worrisome thought for auditors and CFOs — especially because many users construct spreadsheets so poorly that the results may be impossible to verify.
FAS 157 builds on an older rule, FAS 133, which forced companies to divulge the fair value of derivative instruments. But FAS 157 goes a major step further, telling companies how to value the assets and liabilities on their balance sheets that they mark to market. It has affected financial companies in a big way. “If the subprime crisis hadn’t happened, FAS 157 would have been a ‘Who cares?’ kind of thing,” says Jiro Okochi, CEO of Reval, a New York–based vendor that recently added a 157 module to its software for managing derivatives. “Auditors wouldn’t have paid much attention.”
They do now. But while FAS 157 introduced a much stronger emphasis on the methodology behind fair-value calculations, that has not inspired software that can automate that process. “I haven’t seen a single solution in the marketplace that would address all the issues,” says Peter Marshall, a principal in Ernst & Young’s treasury advisory practice. “People are using existing systems and doing manual workarounds.”
Information Vs. Data
Most people, anyway. Wesley Walton, vice president of finance at CBC Federal Credit Union, has managed to tap newer technology, in this case an analytics function contained in software from Brick & Associates that allows Walton to perform fair-value calculations. But a credit union with just $300 million in assets and one primary software vendor has an easier time of it than larger companies.
Indeed, big companies face the irony of having too much software. A large bank will typically use different applications to handle accounting for bond transfers, commercial-lending decisions, and foreign-exchange transactions, to cite just three of many activities potentially affected by FAS 157. Therefore, multiple vendors must modify their systems to handle the dictates of FAS 157.
It also isn’t clear that FAS 157 can even be captured in computer code. The rule is partly intended to force companies to articulate how they arrive at valuations for illiquid securities — the so-called Level 2 assets, where there may be some observable inputs in the market; and Level 3 assets, where there is nothing comparable in the market and valuations are determined by models. Explaining valuation methods amounts to disclosure notes in financial statements, a form of information that doesn’t fit neatly into the fields of a database.
Still, a database, with its central controls and its knack for forcing information into a consistent format, is ultimately where all valuation information needs to end up, according to those who understand the pressures CFOs face in tracking and accounting for hard-to-value securities. The database can be the same one a company uses for other purposes — an Oracle database with programs written in Java, for example, or a SQL Server database with programs written for Microsoft’s .NET framework. “Those are all off-the-shelf,” says Duff & Phelps valuation expert Joseph Pimbley. “It’s not a case of a specialized firm inventing something new.”
Piece by Piece
In theory, the financial assets on a company’s balance sheet are ready-made for tabulation, sorting, and what-if analyses — work that software does well. So the earliest attempts to address FAS 157 have come as enhancements to asset/liability management (ALM) systems, the software used at banks and corporate treasuries.
“Many companies came out fairly promptly with a patch or upgrade” addressing FAS 157, says Denise Valentine, an analyst who covers asset-management systems for the Aite Group in Boston. British Columbia–based analytics provider FinCad and Reval are among the software firms that have added FAS 157 functionality to products that are live with clients. Bank of New York offers its clients a FAS 157 reporting package as part of its Workbench online portal of software tools.
The difficulty of addressing FAS 157 comprehensively hasn’t stopped software companies from tackling individual pieces. Reval was among the first to make it easy to see which financial instruments have moved from Level 3 to Level 2. That’s useful, because those assets could get a company, and its auditor, into trouble if their value later evaporated and the company had to justify their recategorization.
SAP is trying to sell banks on the idea of using its accounting for financial instruments (AFI) product to comply with fair-value requirements. Introduced three years ago for European banks, AFI functions as a financial-products subledger layer between the general-ledger and operational systems, taking data from such systems and putting it into a format that lets users value most financial instruments, while allowing some work to be done in decentralized systems.
Using AFI “doesn’t have to be a big-bang major paradigm shift,” says Mike Russo, a former bank CFO and regulator who is now an industry principal in SAP’s financial-services group. “It basically allows you to have a single point of control.”
And a single point of control is vital, says Duff & Phelps’s Pimbley, especially in an era when CFOs have to aggregate the fair values of many different things and, per the demands of Sarbanes-Oxley, attest to their accuracy. This is one reason that although Pimbley endorses the use of spreadsheets for initial model development in pricing hard-to-value securities, he shudders at the idea of spreadsheets as the sole fair-value calculator.
And yet, he says, with “no game-changer out there,” companies will continue to rely on spreadsheets for this delicate work. Let’s hope proper controls become part of the equation.
Robert Hertzberg is a freelance writer and editor based in Port Washington, New York.
Slow Going in the FAS 157 Lane
Four reasons why fair value lacks a technological silver bullet
1. NEWNESS OF RULE: Companies are still wrestling with how to apply it, auditors with how to enforce it. This makes FAS 157 hard to systematize. Word that federal regulators will move to deemphasize or suspend fair value to stem the credit crisis probably won’t help the cause.
2. NATURE OF RULE: FAS 157 involves many qualitative judgments. This doesn’t play to software’s strengths.
3. DECENTRALIZATION: Most banks use multiple IT systems to help value their financial assets, making a single-vendor solution unviable for most companies.
4. INGRAINED PROCESSES: The people responsible for calculating valuations often have a preferred method for addressing the problem and may not want to change, even if a new product promises a better result. — R.H.