By Dorsia Capital :
These sophisticated investors are using "cash EPS" as their preferred metric (Valeant has made 81 cents of GAAP EPS so far this year).
Cash EPS is an interesting metric that Valeant has done a great deal to popularize. Indeed, many people have started to use cash P/E interchangeably with regular GAAP P/E. For example, Morningstar gives a 4.6 P/E for Valeant in 2016 (That figure must be cash EPS because otherwise Morningstar is predicting that VRX will have GAAP earnings of about $7 billion, even though the company's now retracted 2016 guidance was only for $7.5 billion of EBITDA and it has about $2.8 billion a year in depreciation and $1.6 billion in interest).
Valeant's argument for cash EPS is relatively simple: "We are skilled operators who acquire durable pharmaceutical products like branded generics." Those products will have reliable revenue streams ad infinitum . Therefore, depreciation of these assets is not a real expense. Valeant is basically arguing that the assets it acquires are like Coca-Cola ( KO ).
This argument makes some sense when Valeant spends $8 billion to acquire Bausch & Lomb. Bausch & Lomb is a very well-known brand and does not primarily rely on patented products ( although B&L products make up most of the pipeline that Valeant is currently hyping ). It is reasonable to argue that the $8 billion in depreciation that VRX has to take for B&L distorts its GAAP EPS and the associated expense should be excluded when considering the company's actual earnings power.
There is one strong counterpoint. Valeant borrowed $6.5 billion (and issued $1.5 billion in stock) to buy Bausch & Lomb. The $8 billion in depreciation largely matches up with the debt that VRX presumably will eventually have to pay off. Using cash EPS makes debt repayment vanish from earnings per share. As long as the cash flow covers the interest, any acquisition that Valeant makes will be accretive to cash EPS, even if the acquired asset, over the long term, will not be able to generate the cash flow necessary to repay the debt. Now, of course, the company could just keep rolling over its debts. But that would mean that cash that could otherwise go to shareholders through buybacks and dividends would instead be going to repay interest. Valeant also has borrowed its money in a Fed-set zero bound interest rate environment. As the prime rate rises, Valeant's cost to roll over its debt will increase significantly.
Nonetheless, if Valeant is buying truly durable assets, there is a reasonable argument that cash EPS is an appropriate measurement.
But, in early 2015, the company spent $15 billion to acquire Salix (SLXP) ($1.5 billion in stock, $13.5 billion in debt). Salix is basically Xifaxan and a few other much smaller drugs.
Xifaxan is a patented drug. It consists of the antibiotic Rifaximin. Rifaximin was first marketed in Italy in 1987. It is now widely available throughout the world. Salix acquired patents on Rifaximin in 2009 and got orphan drug status for Xifaxan in 2010. Orphan drug status means that Salix's patents on Rifaximin can't be challenged at least until 2017. After 2017, a competing drug manufacturer could challenge the validity of Salix's patents, presumably based on prior art as Rifaximin was first synthesized in the 1980s. Salix's main patents on Xifaxan expire in 2024.
The point is that Xifaxan does not appear to be a durable asset. If it is very successful, then it is quite likely to at least face a patent challenge in 2017, and the patents expire in 2024. Therefore, it does not seem legitimate to exclude the cost of the $13.5 billion in debt that Valeant took on to acquire this finite-life asset from the cash EPS calculation (I'm not including the $1.5 billion in stock that Valeant also had to issue to finance the deal, although technically that should be included in a depreciation schedule).
Last quarter, VRX guided for cash EPS of $11.67-11.87 for 2015. Taking the midpoint of 11.77, we multiply that by the outstanding share count of 341 million to get a cash EPS figure of about $4 billion (I understand that Salix was acquired after the first quarter, but this depreciation exercise is more prospective than retrospective). Now, what would happen to the cash earnings per share if the debt cost of acquiring the finite life assets of Salix were included? Here are three potential depreciation schedules and their impact on Valeant's cash EPS for the year.
Valeant's own depreciation schedule
I looked in Valeant's quarterly reports and did not find a clear explanation of how it is depreciating the cost of acquiring Salix. Its depreciation did increase from $407 million to $635 million in the quarter after it acquired Salix. This implies that the company is depreciating at about $225 million a quarter. To finish depreciating the debt from the cost of acquisition would then take about 15 years, which implies a somewhat optimistic view of the durability of the asset but not necessarily unreasonable.
Using Valeant's own apparent depreciation of Salix in the cash EPS calculation would still reduce VRX's cash EPS from $11.77 to $9.09 for 2015.
A 10-year depreciation schedule
There's a reasonable argument that a 10-year depreciation schedule, running from 2015 to 2025, for Salix is appropriate because the Xifaxan patent expires in 2024 and, presumably, the profitability of Xifaxan will largely disappear once generic competitors enter the marketplace (The Glumetza patent, the other notable Salix product, expires in 2020).
Using a 10-year depreciation schedule for the $13.5 billion in debt would reduce Valeant's 2015 cash EPS from $11.77 to $7.77.
The depreciation schedule used by the senior secured loan Valeant took out to acquire Salix
The $13.5 billion that Valeant took out to acquire Salix was a mix of $8.5 billion in unsecured debt (i.e. junk) and $5.15 billion in increased term loan facilities.
The amortization schedule of the senior "Series A-4 Tranche A Term Loan Facility" is interesting (1 billion of the overall debt load).
It is perhaps not a coincidence that the amortization schedule for this loan increases dramatically once Xifaxan loses its orphan drug status in 2017, and finishes in 2020, shortly after the automatic 30-month extension that Valeant would get for responding to an ANDA challenge. It is unlikely but possible that Xifaxan could lose exclusivity in 2020 (and of course, the more of a blockbuster that it is, the more likely that it will receive a patent challenge that can rely on prior art dating back to the 1980s).
Using a five-year depreciation schedule would reduce Valeant's 2015 cash EPS to $3.81, leaving it with a hardly great P/E of 24.4. To be clear, I think it is unlikely that Valeant would lose Xifaxan exclusivity in 2020, but it is possible. I also admit that there are other more likely explanations for why this amortization schedule was chosen (like the senior loan with the lowest interest rate wanted to be paid first).
I think there are a few takeaways from this.
First, assuming Xifaxan and the rest of the Salix portfolio behaves like most other patented drugs, Valeant is overstating its cash EPS by between roughly 25% (using VRX's own depreciation schedule) to 65% (using the amortization schedule that the company's bankers demanded for the senior loan).
Second, I am surprised that Valeant's sophisticated investors have continued to uncritically accept the cash EPS metric, justified on the basis that Valeant was purchasing durable assets, even after it spent a very large sum of money on what appears to be a non-durable asset. I would think that if these investors do want to rely on the cash EPS metric, they should either adjust for depreciation for Salix, or ask Valeant why Xifaxan will be a durable asset, rather than a normal patented drug (I am not surprised that Wall Street investment banks are happy to tout cash EPS, because it encourages a great deal of M&A activity).
In the Reuters article, the investors said that they predicted cash EPS of around $12-14 in 2016 and a 9-11 multiple. If we take a $13 midpoint, and deduct by the 10-year depreciation schedule for the Salix debt that I think is reasonable, you would be left with about $9 in adjusted cash EPS. If you give that the suggested 10 multiple, that would suggest Valeant is fairly valued at about $90 per share right now under the metrics of its investors.
Third, I think Valeant's choice of a cash EPS metric raises questions about how it approaches deal making. As long as Valeant chooses to measure itself by this metric, any deal will be accretive to cash EPS as long as the operational cash flow is greater than the interest (and in this low interest rate environment, that's not a high bar). Valeant has touted a metric that heavily incentivizes its management to pursue deal making that can raise the cash EPS metric. But not every asset that can be bought should be bought. Private equity LBOs do go bankrupt with some frequency. Valeant's two biggest acquisitions, Bausch & Lomb and Salix, were both very publicly for sale and were passed over by every other player in the pharmaceutical industry. Perhaps Michael Pearson is a once-in-a-lifetime efficiency machine (although last I checked, he's an ex-McKinsey consultant, and other pharmaceutical companies can hire McKinsey consultants to give cost-cutting suggestions very easily) or perhaps there's a reason that every other pharmaceutical company did not purchase these assets. People do say there's a certain wisdom of crowds. And, Sprout, for example, does not raise confidence about Valeant's deal-making acumen. Time will tell.
Disclosure: I am still short Valeant through put options, but I have significantly trimmed my position. As this prior article shows,I am horrible at predicting short-term price movements(I expect and deserve to be mocked in the comments. To be fair, I did not count on Ackman choosing to engage in a martingale game). I think that Valeant will tell investors exactly what they want to hear at the upcoming meeting, and that investors will likely react very positively (although VRX's decision to hold the meeting on the day of the fed decision will add some fun to everything). Because I do not place faith in the statements of Valeant's management, I don't think that will alter anything about the fundamentals and will not take the company's unsupported public statements as evidence that I've been wrong. If Valeant can make $3-4 billion in debt repayments next year, then that would alter the fundamentals and I would certainly reevaluate my statements. But it will take us a while before we see whether that happens.
See also Is There Any Value In SodaStream After Its 8% Drop? on seekingalpha.com
The views and opinions expressed herein are the views and opinions of the author and do not necessarily reflect those of Nasdaq, Inc.