You can’t just slap a 14x multiple on managed and franchised fees.
Continuing our review of the valuation of HOT, we focus this post on valuing the Fee business and once again use JPM as a proxy for “consensus”. At first glance, the “consensus” multiple of 14x times 2010 depressed Fees may look reasonable. However, there are a few problems with this simple valuation of Starwood’s fee business.
- 2010 Fee estimates include about $80MM of amortization of deferred gains which are non-cash and have ZERO value. They are remnants of the Host sale and will go away. Take a look at Starwood’s cash flow statement; “amortization of deferred gains” are plain as day deductions on HOT’s cash flow statement. That means zero cash.
- JPM’s Fee estimates also include roughly $40MM of termination fees and other one-time items. While termination fees are certainly part of the fee business, by definition they are not “recurring.” Once someone terminates a contract you get the cash – that’s it. We would value these fees as cash, perhaps even put a little multiple on them, but no one would pay 14x for them.
- JPM’s Fee number also includes $15-20MM of “miscellaneous other revenues” that used to be lumped in with Bliss before HOT sold that business. As a reminder, Bliss revenues for Sept 09 TTM were $85MM - all included in “fee revenues” - which under the JPM method would have been valued at 14x. Bliss Sept 09 TTM EBITDA was only $5.3MM, so the associated $80MM of expenses were captured in HOT’s SG&A, which gets a 50% reduction and a 12x multiple according to JPM… does that make sense to you? Anyway, we don’t know exactly what's in the $15-20MM of other fee revenues line, but we’d bet that it’s low margin and doesn’t deserve more than 8x multiple.
- Another "small" issue with the above methodology is reducing total SG&A by 50% and classifying that reduced amount as "unallocated". We know that roughly 75% or so of our estimated $325MM (ie $240MM) of SG&A in 2010 is related to the managed & franchised business. Another $15MM or so is related to overhead of running the owned business. The balance of roughly $70MM represents corporate overheard and unallocated expenses. So again, I'm not sure how the 50% reduction makes sense, but this brings us to issue #4 relating specifically to the fee valuation – this isn’t a start up; EBITDA, not revenue, should be valued with multiples. We don’t value Marriott or Choice by slapping a multiple on their revenue, so why would we value Starwood’s fee business this way? Management/ Franchise businesses have roughly 65-70% margins. We would actually apply a higher multiple on the management and franchise base fee EBITDA since it's still depressed in 2010 and the highest quality piece of HOT’s fee business. On a “revenue” basis, base fees are approximately 60% of HOT’s fee business.
- Incentive Fees – Sorry but this highly volatile revenue stream deserves a lower multiple than base fees since, as this cycle and every other cycle has illustrated, they are highly cyclical and cyclicals trade at lower multiples.
So where do we shake out on the value of HOT’s Fee business versus JPM and consensus. With the all-encompassing 14x multiple, JPM values the Fee business at $9.7BN or $52/share versus our valuation of the fee business at $5.2BN or $28/share - only a $24 difference. To make things a little more "apples to apples," we can apply all of JPM's Unallocated SG&A which they value at 12x as a reduction to the fee business valuation, and come up with a net value of HOT's fees at $7.8 billion, which still represents an $18 differential in value per share compared to where we shake out.