Let’s step through the process of turning 100% of R&D expense into a capitalized investment using the PharmaCo hypothetical company example in an incredible valuation book, Tim Koller’s McKinsey Valuation 7th Edition chapter 24, page 500 (of the softcover version).
I have put all the numbers and exhibits related to this from the book in screen grabs below. As usual, click on the grab to enlarge it. The first is Exhibit 24.1 which is a summary of key income statement items needed to get to NOPAT and the invested capital for the hypothetical company PharmaCo:
Your first response is likely that the ROIC looks great – 33%. One might say, these guys manufacture and sell little pills at about a 10% NOPAT margin, NOPAT is growing fairly nicely (3% CAGR) and the business employs a minimal amount of capital to get those sales out the door into what seems like a large and possibly growing market. As we have discussed, strong ROIC comes from either high margins or low capital intensity, and in this case, it looks like the latter. I’ve found See’s Candy, except it’s the medicine version!
But wait:
“…this ROIC does not represent the company’s true economic performance, because the invested capital includes only the purchased capital and not the intellectual capital created internally from R&D.” (page 501)
We need to fix this, or at least fix how we are thinking about it.
3 Adjustments
To correctly account for the R&D investment and get a better read on ROIC for this business, there are 3 adjustments. Mauboussin uses the same adjustments in One Job, which we will get to. But here are the adjustments as per McKinsey Valuation 7th Edition, which uses this simpler PharmaCo example and is therefore a better place to start understanding the details of these adjustments:
1. The PharmaCo R&D must be capitalized. This is done by summing all of the R&D expenditures beginning from some moment in the past when R&D investment is assumed to have started (“time 0”) to get to starting R&D asset value for each year and then burning down that number each of the years from then to the present using some a straight-line amortization schedule assumption (again, see below). So for each historical year, you have the R&D investment asset (the capitalized “expense” from the GAAP income statement) being added to the previous year’s capitalized R&D amount, and then the amortization is deducted from that sum to get to “Capitalized R&D, ending.”
That’s a mouthful – the numbers explain this better than I can. McKinsey started accumulating R&D expenses for capitalization purposes for Pharmaco in 1995 (year 0), and they use an 8-year amortization period (straight line).
Here is where the numbers fall out:
2. Adjust Invested Capital. Per Exhibit 24.2 we now have a value for R&D as an intangible asset for each year, which we can add to invested capital (see Invested capital, adjusted at the bottom of Exhibit 24.2 above). Note how much the invested capital has increased via the inclusion of capitalized R&D – a lot! If this is really a better estimate of the amount of capital investment required to continue to generate income and growth of that income in this business, the number we were using before was wildly inaccurate and misleading…
3. Adjust NOPAT. We now also have an amortization expense amount, which can be used to adjust NOPAT in the adjusted ROIC ratio. We decrease NOPAT by the R&D expense – because we have taken it out to capitalize it – and we replace that with the amortization of this intangible asset. These adjustments can be seen in Exhibit 24.3 below, which gets us to Adjusted NOPAT.
We can see here that on a net basis, NOPAT has gone up, because the R&D expense was greater than the amortization, at least based on the assumptions used.
The Big Reveal
The result of these adjustments is in the last two lines of Exhibit 24.3 which show the ROIC before and then after the changes to NOPAT and invested capital. The numerator has gone up, but the denominator has gone up by a much higher percentage, resulting in lower ROIC. The ROIC for 2020 has gone from an exciting 33% to a ho-hum 9%. Now it’s not even entirely clear that Pharmaco earns above its cost of capital. This likely changes our thinking about Pharmaco as an investment, at least a little.
Per Koller:
“As long as R&D investments needed to support earnings remain unchanged, PharmaCo’s adjusted ROIC is the better estimate of its true economic return and underlying performance.”
This makes intuitive sense: pharma companies invest a ton via R&D to develop drugs – that is their version of investment, which is equivalent to a ball-bearing manufacturer building a ball-bearing plant. But it’s actually worse, because often that R&D investment is worthless, while the ball-bearing manufacturer at least has the plant when it’s done paying for it. This probably means a higher discount rate/lower multiple for Pharmaco, to reflect this risk to its cash flows.
Free Cash Flow
Exhibit 24.4 below is a reminder that free cash flow does not change from these adjustments:
“The amortization is a noncash charge in NOPAT and is added back to calculate gross cash flow. This effectively moves R&D expenses from gross cash flow to investments, leaving free cash flow unchanged.“
We can and sometimes should apply the same thinking and methodology we applied to R&D for PharmaCo to other kinds of intangible investments that are expensed under GAAP. Examples include Sales & Marketing, and General & Admin. I think we have to use our judgment in doing this and not go overboard; it makes sense for only certain kinds of businesses and in certain situations. I will get into this in more detail in a later post.