There is no fixed form for a financial model, as it is always tailored for the type of business, and various models serve different purposes. However, a good design that is easy to follow and allows smooth data flow has a big impact on the model’s scalability and efficiency of use.

An integrated model that contains BS, PL and CFS is usually desirable for forecasting, as it reflects a more complete picture of the business. I’d like to share a few ideas in structuring this type of model. There is no right or wrong—the best structure is defined by the business and how good users feel about it.

Keep it neat and simple. Try to minimize the number of tabs, as long as the source data and inputs are categorized organically to support the final output. Navigating through too many tabs will make it hard to find your way back. Keeping similar information in the same tab can enhance efficiency and accuracy. I’ve seen models break output financials into BS, PL and CFS by month, quarter and year, which ends up creating nine tabs for similar information. Instead, data can be sorted into two tabs: a master sheet with BS, PL and CFS by month, which serves as a data pool; and a reporting tab customized to the layout you wish to see, such as financials by quarter, by year or both, with ratios and so on. You may also find that consolidating BS, PL and CFS into one tab will enhance visibility of data links.

Place assumptions in applicable worksheets. You may often see modelers put all assumptions in one tab in an effort to keep things looking neat and tidy. However, when the model grows larger and more complex, the artificially created inter-sheet links will require extra time to navigate and extra effort to scale or revise. For example, in most cases, you will have an operating expenses tab that contains numerous line items that don’t share the same assumption, input and projection method. Creating assumption columns next to these items will give you a straight look of how the projections are made. Be sure to shade the cells so you don’t miss the input areas.

Set up a warning system to trap errors. Ensure all data tie out and reconcile. It is critical to set up cross-check, total tie-out rules to ensure accuracy when flowing through tabs. Conditionally format  any unreconciled data or imbalance in red so you won’t miss it. Since BS often sees imbalance, placing these rules is essential to identify errors at each step and correct them so you don’t walk into a disaster.

Ultimately, the goal is to achieve simplicity in structure so people can follow it easily—to make it look like a book and read like a book. By doing so, you will help users—the CEO, CFO, COO and other top executives—see a much clearer future picture of the business and be able to make better decisions.

Should you ask your audit committee to evaluate your XBRL files for completeness, mapping, accuracy and structure under an agreed-upon procedures (AUP) engagement in accordance with the principles and criteria set by the AICPA? I get asked this question all the time, especially by companies whose limited liability is expiring. But everybody should consider AUP for their XBRL.

Why? In the absence of a mandatory audit assurance, an AUP engagement helps ensure that XBRL data brings meaningful value and transparency to the investment community.

Even if your audit committee has adopted a wait-and-see attitude, analysts and investors may be making investment decisions about your company that may be based on substandard and inconsistent data quality. For example, the SEC found several significant and recurring errors by large accelerated filers during the first two months of 2011. The most prevalent data-quality issues revolved around negative values, extended elements and tagging completeness.

Says XBRL US: “In the over 14,900 XBRL submissions to date, over 145,000 data issues have been identified related to the use of the XBRL US GAAP Taxonomy. These inconsistencies include incorrect signs, missing concepts and concepts used incorrectly.“

While a formal AUP is not required, it is best to have a mock AUP environment that ensures compliance of your XBRL-formatted information. A recent trend is for companies to leverage their internal audit function or professional service firm to implement a mock AUP environment to be better prepared for the formal AUP engagement.

What exactly is AUP?
First, an AUP engagement doesn’t deliver an audit opinion. The practitioner performs agreed-upon procedures and assessments, and then reports findings, alternatives and recommendations in a letter to management and the audit committee.

An AUP ensures the completeness, accuracy, proper mapping and structure of your XBRL files. These are the four important aspects of XBRL, according to the AICPA’s latest exposure draft for the XBRL process. Here is what an AUP engagement covers.

Completeness Do you have a procedure to ensure that all required source information is tagged in XBRL? For example, a few commonly missed tags are significant accounting policies embedded throughout footnotes, spelled out amounts and superscript footnotes.

Mapping Even though finding data-quality issues on proper mapping can be aided by software-assisted search and benchmarking analytical tools, at the end of the day, this core process can be subjective: choosing the narrowest tag and assessing materiality can be an art rather than a science. Likewise, the SEC considers mapping to be the most critical part of the XBRL quality control process, but there are no software tools that can detect this type of error prior to filing.

Accuracy Even if common data-quality issues, such as negative values, are flagged by software tools, you still need to assess their validity based on financial facts and the specific circumstances for comparative quarters and year-to-date periods.

Structure Technical validation errors of this type tend to be black-and-white and can be detected by third-party SEC and EDGAR validation tools prior to submission to the SEC.

AUP = quality assurance = market value
Whether you have a built-in versus a bolt-on XBRL solution, you need quality assurance over your XBRL data. Some AUP steps can be accomplished with software tools, while other procedures require professional judgment. Automated tools can only help you so much in highlighting inconsistencies and the usual suspects. Ultimately, you need to tell your company’s story by choosing the tag that best maps to the underlying transaction and translates that fact into meaningful information.

Because investors rely on your XBRL data to make investment decisions, it is ultimately your responsibility to avoid errors before they are disseminated to the public. Aside from compliance, the real benefit of XBRL is increased transparency and comparability, which can in turn increase the value of your stock when the analyst community gains more confidence in your XBRL data.

Learn more about RoseRyan’s XBRL expertise.