Flawless Financials the Financial Forecasting Online Newsletter from Minotaur Financial and David Brode April, 2005 Sent monthly to over 500 subscribers. Please pass on Flawless Financials to those in your network. To leave Flawless Financials, follow instructions at bottom. * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * This month: Operationalizing A Model: Part 2 * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * Overview ---------------------------------------------------------- This newsletter is the third in the series on the evolution of the model after financing is complete. The first (http://www.brode.net/resources/Newsletter-2005- Feb.txt) was about uses of the model, the second about the problems in “operationalizing” a strategic model (http://www.brode.net/resources/Newsletter-2005-Mar.txt) , and this focuses on solutions to those problems. In general, the challenge is to keep model accurate, relevant, and manageable. In doing this, remember the 80-20 Rule for Models (http://www.brode.net/resources/Newsletter-2003-Jan.txt). First, a quick review of the problems:s 1. Tying to actuals. 2. Syncing to other forecasts. 3. Adding detail for definitional nuances. 4. Feeding assumption changes Solutions ---------------------------------------------------------- The solutions fall into four categories: process, automation, standards, and submodels. Process ---------------------------------------------------------- A simple example of process is agreeing on how to interface with data sources. For example, each month the model has to get actuals. Typically, the controller publishes this data, and typically I have to translate the line items into the model line items. Over time, I tend to push this activity back to the controller, so that her system can map into the line items I need, thus ensuring a quick and accurate turnaround for updating the model. Last year I saw a more complex case while worked with a client who had about twenty active deals, any one of which could have large cash flow impact. The deal list had regular adds/deletes/changes, and the impact of these changes was highly material over a month. Initially, the process we had was to call up one guy in the department and go through each deal with him to verify we had the proper forecast. But because other people ran many of the deals, we had to go back and forth and wait for him to gather data from his colleagues. All in all, it took about three days of effort to update this forecast and nobody felt confident in the results. Over time, we created a better system. First, we had high-level management resolve—the senior executives from each of our two departments agreed that all deal data would be kept in one database, which prior to that was in spotty use in their department. Later we found that the database was being continually updated, so we never knew when was a good time to take a snapshot of the data. From that, we evolved a process where there was a regular Monday meeting to discuss what changes were approved for the database; subsequently the database was updated Monday afternoon, it was frozen until the next update, and we were notified that the update was complete. Automation ---------------------------------------------------------- Creating automated links between systems is the key to rapid turnaround. Where process is about ensuring that the data coming from another source is accurate, automation is about making it easy to take their data into my system. So if process with the controller was about his special output format, automation is about getting it in a specific Excel file AND having hooks in my model where I can do just one copy and paste and then see the results in the model. In the more complex case with the outside database, I worked extensively with the database developer to get highly customized reports that gave me the information I needed. Of course, there’s a limit to how far you can push this before the activity becomes uneconomic. The database report takes about ten minutes of effort to get into the model, but it’s not worth spending $2,000 to turn the ten minutes into three minutes. Standards ---------------------------------------------------------- Strategic modeling isn’t accounting. For tying to actuals, 80-20 is all you get. Think about a forecast showing a $10M cash balance on 12/31/04 and a forecast showing you going to $2M on 12/31/05. If someone objects that you’re really at $11M, ok, fine, but the forecast still shows a severe decrease in cash. That’s the level at which I interpret the results of models. That is, I want to discuss the fact that cash is falling by $8M in a year and be able to describe the economic dynamics causing that and be able to give advice on alternative scenarios. Now in reality the tough things to tie out to aren’t really the financial ones, but the operational data, like subscribers or cell sites or properties. Often the model will say that 11/30/04 units were 62 while an operational report says 65. And the reality of tying things out is that it takes RIDICULOUS staff time to tie these out for no real gain in the quality of the forecast. Think of a graph with trend line going down: if you’re a bit off at the start, the end is just a bit off, too. And in modeling, the TREND and TENDENCY is what matters, not the absolute number, because it’s in the absolute numbers that people joke that “models are always wrong.” It’s simply the case that 95% accuracy can take an hour but 98% accuracy can take a week—so the return doesn’t justify the investment. If people want exact actuals, they can look at quarterly accounting reports. But if they want to understand the business, a 95% accurate model should be acceptable. I don’t think this is defeatism: I think it’s good time management. Submodels ---------------------------------------------------------- Part of success is a balancing act of extra detail vs. growth in size. I want the model to have good detail for credibility, but it needs to remain a MODEL where things can adjust. Sometimes the solution to this is tying in to linked submodels. Here the trick is often to be super-explicit how data moves between the main model and the submodel. In my most careful work on this, I’ll have InLink and OutLink sections for these data flows. The advantage of a submodel approach is that it allows complexity, but isolates it so the complexity doesn’t impact the entire model’s structure. Conclusion ---------------------------------------------------------- So there you have it—solutions to the problems of strategic modeling in an operating company. There’s no doubt that it’s easier to model the pie-in-the-sky ideas, but operating companies need these models too—I think the input they give on strategic direction and tactical maneuvers is invaluable. Until next month, all the best, David Brode -- Minotaur Financial Removing Financial Issues as a Deal Roadblock * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * NEWSLETTER ARCHIVE AVAILABLE Make sure to visit the Minotaur Financial website for the Newsletter Archive at http://www.brode.net/resources/ * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * http://www.brode.net mailto:David@Brode.net 1919 14th Street, Suite 510 Boulder, CO 80302 (303) 444-3300 * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * ABOUT DAVID BRODE I’m a financial modeling specialist. Over the last fifteen years I’ve completed dozens of models and certainly thousands of versions to support corporate development, M&A, strategic planning, and debt and equity transactions. These models have raised over $1B in debt and $100M in venture capital and private equity. Over time I’ve consistently revised software tools and work processes to get the job done quickly and well. If you have a financial forecasting issue, I’d love to help. * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * (c) 2004 Minotaur Financial, All rights reserved. You are free to use material from the Flawless Financials newsletter in whole or in part, as long as you include complete attribution, including a live web site link. Please also notify me where the material will appear. The attribution should read: "By David Brode of Minotaur Financial. More articles on financial forecasting can be found at http://www.brode.net " * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * Are you struggling to convince others to do a deal which you think is a no-brainer? To discuss how you can take numbers off the table as a deal roadblock, call (303) 444-3300. I'm very accessible and glad to help. * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * DO YOU LIKE THIS NEWSLETTER? You are welcome to share this email with colleagues who would benefit from better numbers. Your feedback is always welcome and appreciated. Write in to mailto:feedback@brode.net. * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * PRIVACY POLICY: I never rent, trade, or sell my email list to anyone for any reason whatsoever. You'll never get an unsolicited email from a stranger as a result of joining this list. To SUBSCRIBE FREE to this newsletter, send an email to mailto:subscribe@brode.net. And you'll get Minotaur's Financial Forecasting White Paper in the deal. To be REMOVED from this list, send an email to mailto:remove@brode.net. (Please note that this message needs to come from the email address that originally subscribed. If you need help determining this, please email mailto:David@Brode.net.)