I recently spoke with Rob Trippe, M&A, corporate finance advisor and analyst for Corporate Finance Consulting & Advisory, on the role that models play and the challenges that they present to FP&A departments.

Bryan Lapidus: Please introduce yourself and tell us what you do.

Rob Trippe: Rob Trippe, corporate finance professional. I help executives make decisions. It really boils down to our present value formula, which is in a steady growth scenario, cash flow, divided by discount rate. That's value.

That's what finance brings to the table, that nobody else does. Executives at the end of the day can make decisions and weigh them among different functional areas and different mindsets any way they want, but what finance brings to the table are simple tools and analysis. They can be employed and are often not.

Lapidus: In January 2019, we put out a survey about how companies go through investment projects and look at investment projects. And we gave a list of different evaluation techniques, NPV, IRR, ROI, and all the other mathematical ones came in three, four, five, six, seven. Number one was strategic and number two was payback/breakeven. Why are those number one and number two?

Trippe: I'm going to be a little cynical on the strategic. At the end of the day, everything is ‘strategic.’ I think what that really means is we're not quite able to express mathematically our decision-making, but we have tools for that and this is what finance brings to the table. We have real option theory, we have binomial assets and we can employ these.

These are methods of thinking through an uncertain future and quantifying our thought and decision. Marketing and HR don’t come to the table with that, and that is not their role. They bring different expertise. We think through a series of cash flows in the future, show the inflection points and what we believe are the options and outcomes. We instill that discipline, and we are also creating risk and control points. Little stop and start nodes where we can say, ‘Okay, where are we right now? And which direction do we now want to go?’

One example from my background would be developing the rolling balance sheet and cash flow forecast for Hertz. We did 12 refreshes throughout the year and each month, that was a soft way where we could assess where the balance sheet was and what changes might need to be made, and we could assess where we are now relative to where we want it to be year end. Management was dragging their feet on selling a $50 million fleet and we kept pointing out the cash flow impact that indecision was having. Management finally got to a point and said, ‘Okay, we really have got to do something about that.’ Phone calls were made, the fleet was gone.

During these month-end periods, where we could look at our statements and say, ‘Okay, where are we relative to where we thought we would be? And what action can we take to get there.’ Maybe you're going to tighten receivables, maybe you're going to try to loosen payables, maybe you didn't need all that cash on the balance sheet. Finance should be bringing the options, pointing them out.

Common Errors

Lapidus: What common problems have you seen in corporate finance modeling?

Trippe: These are the two that always come to mind. Number one, do we really know what we're solving for? And number two, do we have proper modeling practices? And quite frankly, many organizations don't.

Model builders and your users should be in contact with each other to make sure you know what you're building and why you're building it. Once I had a client who confused fair market value with fair value. Well, fair value is very specific to the impairment exercise where certain costs are excluded. Fair market value being willing buyer/willing seller. They will not yield the same answer.

Here’s another example about know what you are solving for. EBITDA is a great number, right? But EBITDA is not cash flow and you won't find it anywhere on a GAAP statement. What we're trying to do is develop a figure that is very real and tangible, because at the end of the day, we buy an asset and we pay cash, not with a theoretical income number that comes through accrual accounting. Finance has to stay true to modeling principles.

Thinking about best practices within the corporate finance world is to develop your discount rate within a framework such as CAPM, but not stick to this. For example, it is very common to use some type of company-wide, standardized discount rate that doesn't move over time and doesn't specifically reflect the embedded risks of any one particular project.

I've seen large companies, household names, where the discount rate was not refreshed, not quarterly, not even annually. I don't know when. And I’ve seen confusion on what a discount rate even is. Example: if you have a net margin on the income statement of 15%, that doesn't mean your discount rate on your cash flow is 15%. I really saw that.

One client did have a discount rate that we used on the cash flow; it was provided by their major (larger) supplier. And guess what? It was the same 15% at this other Fortune 10 company that I was at; they both used 15%. Coincidence? Two entirely different industries, entirely different profiles. By the way, we used 15% in grad school, just as a quick and dirty when we were learning how to develop discount rates. Defining your modeling practices is just essential for good work on individual projects and comparison across projects.

Advice to Leaders

Lapidus: You've been in the corporate world, an investment bank, private equity, and are now working for yourself for a number of years where you parachute in and out of companies. If your customer CFO asks you about best practices for model governance and model discipline, what advice do you offer?

Trippe: You need some form of model management in place. A base level is standardization. At the end of the day, I don't care if input cells are blue or green, but you should communicate a standard across all models in our department. We need to build models that have the flexibility to be recalibrated on a periodic basis.

After models are built, you need to understand: What are the models that we employ? How we store and secure them? How we can reduce key person dependence? And how we can monitor model output in the future, to ensure that it remains relevant?

Here is a negative example: Once, I worked in a large bullpen environment where you can just talk to the person in front of you and on either side. Yet that department had no standardization for its myriad Excel workbooks, absolutely none. I thought to myself, “You sit right next to each other, how are you not talking?” Because I'm going to design the way Rob wants it, and the next person is going to come up with a workbook that I think is appropriate. And it becomes remarkably difficult to even marry the two. That shouldn't happen.

Best case examples would be from an investment bank where I worked. We had an objective that anybody, at any time, could step into a model and populate it with very little risk, because we were all taught the same model standards and methodologies. So if Rob is sick today, Bryan can step in and continue the work.

Lapidus: That makes me think of a story from my own experience. I was head of FP&A for a company for a number of years and before I left, I trained my replacement. About three months later, my old boss called me, saying it didn't really work out with the person who replaced me, and there was nobody left who could operate the forecasting models. And the forecasts tied finance and the customer churn models, so without them, the company could not figure out who should receive marketing promotions over the next quarter. I ended up contracting with them and working nights to keep that company running.

Trippe: That is an example of why to build a model with component pieces. Instead of one huge model that is extremely difficult refresh, it is better to have “plug and play” component pieces so you can decide, say, to move the churn rate piece in, now the revenue, then the expense piece in. Great. Now, we've gotten that income.

Lapidus: What would be different on this conversation, for models that are built using those sort of SaaS-based, cloud-based EPM tools?

Trippe: Modeling is a process; it's not beholden to any language, to any application, or to any proprietary system. It is a wing-to-wing framework: initial inputs, calculation processes, final model output. Models weave in and out systems, languages, and applications. It might even end up on paper. Seeing the entirety as a process, and part of the decision-making process, puts us in far better shape and we're going into probably eliminate many difficulties, many risks, and we're going to identify numerous risking control points and end up with better answers, period.

Rob Trippe is M&A, corporate finance advisor and analyst for Corporate Finance Consulting & Advisory. Read the other parts of this series here.

---

PARTNER CONTENT: