A note - "They've invested a lot of capital" is not a sufficient explanation as to why Churchill Downs' return on that invested capital has been poor - sorry MEMO. The real reason is very obvious if you look at their financials for a few seconds - a massive goodwill account, principally assumed in 2022 when I'm guessing they made a major acquisition for a fairly high multiple. I'm sure I don't need to explain this, but if you have a shell company which pays 20x earnings to acquire a business, the ROIC will (initially) be 5% - even if the underlying business earns a 50% ROIC.
Thanks, Matt. You're right that CHDN made a large acquisition in 2022 - Peninsula Pacific - which added considerable goodwill to the balance sheet. However, CHDN also ramped up its capital expenditures over the same period. If you exclude goodwill from invested capital (which is not an uncommon practice), ROIC would indeed be higher, but management is still responsible for having allocated that capital, so I prefer to include goodwill in invested capital.
It depends somewhat on the purpose of the calculation, but broadly I do not think goodwill should be included in ROIC calculations. The calculation is generally intended to estimate (a) the return on incremental capital you can expect from the company and (b) more broadly, whether it’s a capital intensive or capital light business.
In the case of (a), the goodwill is only relevant if they’re a serial acquirer, and this is their primary method of growth, and industry multiples are roughly in line with historical levels. Constellation software is a great example. On the other hand, if two businesses which each earn 40% on invested capital (incremental and existing) come together, and the multiple paid is 20x, the ROIIC you can expect from the combined business doesn’t suddenly plummet.
In the case of (b), it’s a similar story - the capital intensity of the combined business is essentially just a weighted average of the capital intensity of the merging businesses, and is independent of the multiple paid.
ROIC is also often used to gauge management’s capital allocation ability. Here there is a stronger case to be made for keeping goodwill, since an acquisition is an allocation of capital the same as any other, however I’d still prefer to try to look at the major acquisitions on an individual basis, considering whether the price paid was fair, rather than trying to sum it all up with one number. An example of how this could go wrong is if they’ve recently purchased a fantastic high growth business for 15x earnings - this may be an absolute bargain, but those earnings will take time to come, and in the immediate term, this drags their ROIC (if you include goodwill) down towards the earnings yield of the acquisition of about 7%.
Thanks for the additional thoughts. I think there's good value in comparing ROIC with and without goodwill, as the results can be quite different, particularly when we're talking about asset-light companies. Ideally, we would go through each transaction and only include "overpayment" in goodwill. If a company paid up for synergies, those synergies should - eventually - show up in the numerator (NOPAT).
It's an important discussion. Mauboussin's Return on Invested Capital paper, for example, he shows how MSFT's 2022 ROIC could be 94% or 34% depending on how you account for intangibles and goodwill. Both are quite impressive, of course, but can have material differences in how you size up the business.
In the case of CHDN and its acquisition, the acquired company had operating losses prior to the acquisition, so it's not surprising that there hasn't been a lift in ROIC. Perhaps with time.
Thanks for the interesting ideas Todd.
A note - "They've invested a lot of capital" is not a sufficient explanation as to why Churchill Downs' return on that invested capital has been poor - sorry MEMO. The real reason is very obvious if you look at their financials for a few seconds - a massive goodwill account, principally assumed in 2022 when I'm guessing they made a major acquisition for a fairly high multiple. I'm sure I don't need to explain this, but if you have a shell company which pays 20x earnings to acquire a business, the ROIC will (initially) be 5% - even if the underlying business earns a 50% ROIC.
Thanks, Matt. You're right that CHDN made a large acquisition in 2022 - Peninsula Pacific - which added considerable goodwill to the balance sheet. However, CHDN also ramped up its capital expenditures over the same period. If you exclude goodwill from invested capital (which is not an uncommon practice), ROIC would indeed be higher, but management is still responsible for having allocated that capital, so I prefer to include goodwill in invested capital.
It depends somewhat on the purpose of the calculation, but broadly I do not think goodwill should be included in ROIC calculations. The calculation is generally intended to estimate (a) the return on incremental capital you can expect from the company and (b) more broadly, whether it’s a capital intensive or capital light business.
In the case of (a), the goodwill is only relevant if they’re a serial acquirer, and this is their primary method of growth, and industry multiples are roughly in line with historical levels. Constellation software is a great example. On the other hand, if two businesses which each earn 40% on invested capital (incremental and existing) come together, and the multiple paid is 20x, the ROIIC you can expect from the combined business doesn’t suddenly plummet.
In the case of (b), it’s a similar story - the capital intensity of the combined business is essentially just a weighted average of the capital intensity of the merging businesses, and is independent of the multiple paid.
ROIC is also often used to gauge management’s capital allocation ability. Here there is a stronger case to be made for keeping goodwill, since an acquisition is an allocation of capital the same as any other, however I’d still prefer to try to look at the major acquisitions on an individual basis, considering whether the price paid was fair, rather than trying to sum it all up with one number. An example of how this could go wrong is if they’ve recently purchased a fantastic high growth business for 15x earnings - this may be an absolute bargain, but those earnings will take time to come, and in the immediate term, this drags their ROIC (if you include goodwill) down towards the earnings yield of the acquisition of about 7%.
Let me know your thoughts.
Thanks for the additional thoughts. I think there's good value in comparing ROIC with and without goodwill, as the results can be quite different, particularly when we're talking about asset-light companies. Ideally, we would go through each transaction and only include "overpayment" in goodwill. If a company paid up for synergies, those synergies should - eventually - show up in the numerator (NOPAT).
It's an important discussion. Mauboussin's Return on Invested Capital paper, for example, he shows how MSFT's 2022 ROIC could be 94% or 34% depending on how you account for intangibles and goodwill. Both are quite impressive, of course, but can have material differences in how you size up the business.
In the case of CHDN and its acquisition, the acquired company had operating losses prior to the acquisition, so it's not surprising that there hasn't been a lift in ROIC. Perhaps with time.