The central question for Valeant (NYSE:VRX) at the moment is if and how it can manage its debt load. Various SA contributors (most recently here and here) have claimed selling assets is essential if Valeant is to have a chance. On the other hand, CFO Paul Herendeen claimed on the latest earnings call that Valeant has a path for managing its debt without selling any assets at all.
Absent from the debate so far has been any detailed analysis of what the debt repayment path might look like. In this article, I provide such an analysis to show that Valeant does indeed have a clear pathway for managing its debt without selling assets, as CFO Herendeen claimed. Sales of non-core assets will make the pathway broader and easier, so this is a 'worst-case' analysis to show that Valeant is not reliant on asset sales to manage its debt.
The pathway I consider is deliberately conservative - it involves paying down the bank debt by $1.2B per year, which is significantly lower than the $1.6B of debt repayment Valeant has already made this year. (Justification for why this is a manageable amount will be given later.) I also assume that maturing bonds will be rolled over as new bond debt with significantly higher coupon interest rates than the old ones: 10%, 9%, 8.5% and 8% for the new bonds which replace the maturing bonds in 2018, 2020, 2021 and 2022, respectively.
The basis for assuming these coupon rates is the following. First, Valeant's bond yields are currently ~11%, after varying between 8% and 11% in recent months. Second, I am assuming that Valeant's core businesses stabilize at their current levels (no major further deterioration) and slowly return to industry growth rates in the coming years. Right now, the market is taking a "we'll believe it when we see it" approach to Valeant's stabilization and return to growth, hence the high current yields on the bonds. But by 2018, when the first bond needs to be rolled over, the stabilization and return to growth should be apparent. The assumed coupon rates above are conservatively high and could easily turn out to be lower in reality.
Note that long-term interest rates are lower now than during most of the period when Valeant issued its bond debt in 2011-2015. Considering that the global low growth environment is expected to continue for many years, I do not expect future interest rates to rise to an extent that they push up the coupon rates on Valeant's new bonds in a material way.
I assume that unpaid portions of maturing bank loans will be rolled over as new loans. It will not be a problem for Valeant to do this, since the value of its assets, used as collateral to secure the bank loans, far exceeds the current total bank debt of ~$11B.
Furthermore, I assume that the interest rates will remain the same on the rolled over loans. This assumption is justified by the fact that, as Valeant pays down these loans, the size of the bank debt compared to the value of the assets that secures it will rapidly decrease. Moreover, the bank debt will become a smaller part of Valeant's capital structure, dwarfed by the bond debt. In this situation, the credit rating of the secured bank debt will rise, implying lower interest rates are appropriate for the loans. In fact, it is quite possible Valeant will be able to receive lower interest rates on its rolled over bank loans in coming years. But to be conservative, I assume the rates remain the same.
Rising Fed rates will cause corresponding increases in the interest Valeant pays on its bank loans. However, rising Fed rates signal rising inflation, and this gives Valeant and other companies' justification for raising prices. Since Valeant's bank debt (~$11B) is of the same magnitude as its annual revenue (~$9.6B), an increase in the Fed rate of, say, 0.25% can be balanced out by a small price increase of ~0.3% across Valeant's product lines. Or by a price increase ~0.5% on its U.S. drugs and consumer products. Such small price increases will not affect Valeant's commitment to keeping its annual price increases in the single-digit range. In light of this, I ignore Fed rate increases in the analysis below, regarding them as balanced out by corresponding future price increases on Valeant's products.
For calculating Valeant's future debt leverage, I assume that EBITDA will be $4.3B in 2018 - the same as the current guidance for 2016 - and thereafter increase at the rate of 5% per year. This is a more conservative version of Valeant's own projections in the Q2 earnings presentation, modified to take $4.3B as the starting EBITDA (rather than the $4.8B in the Q2 guidance). It reflects management's indication that EBITDA will be lower in 2017 before recovering in subsequent years. In this model, EBITDA will be $5.2B by the beginning of 2023, which is the terminal point in my analysis.
For the benefit of readers who don't want to wade through the lengthy details below, I summarize the findings here:
Summary of the findings of the analysis:
(1) It turns out that the additional coupon costs for the new bonds neatly cancel against the savings of interest payments on the paid down bank debts through to 2023. Thus, Valeant will not drown in the higher interest payments on the new bond debt, contrary to what others have claimed.
(2) The pathway is long but leads to a happy-end: By 2023, Valeant's debt leverage will be reduced below 4.5x EBITDA, a normal leverage in the pharma industry, and further refinancing of its debt will be unproblematic.
(3) The pathway has enough slack to accommodate (exceedingly unlikely) $2B in fines from current state investigations and civil actions. This assumes the fines will be covered by issuing additional $2B bond debt in 2018-2019 at 10% coupon rate. The leverage by 2023 in this scenario is 4.7x, still reasonable for future refinancing.
Debt data for the pathway analysis:
The rest of the article consists of the analysis of the debt repayment pathway described above. We will need the list of bank and bond debts from Valeant's latest 10-Q:
The coupon interest rates of the bonds and their maturity dates are indicated in the list. We will also need the effective interest rates Valeant is paying on its bank loans, also taken from the 10-Q:
The interest Valeant pays on its bank debts is important when analyzing its debt pathway, since paying down bank debt generates savings on Valeant's future interest payments. During Q4 2016 - Q2 2018 most of the repayments will go to the Revolving Credit Facility loan and A-3 loan. I will use 3.5% p.a. as the averaged interest rate during this period. This corresponds to 0.86% per quarter. From Q3 2018 onwards, I will use 4.2% p.a. as the averaged interest rate. This corresponds to 1.0% per quarter.
As a warm-up, it is useful to start by looking at the debt repayment Valeant has already made this year. Endorama Global claims that Valeant has misrepresented the true debt it has repaid this year, so this is also a chance to check that.
Debt repayment during 2016 - has Valeant misrepresented it?
According to the latest 10-Q, during Q1-Q3 Valeant repaid $1.92B long-term debt and issued $1.22B new debt (drawn on its revolving credit facility), giving a net debt paydown of $0.70B. To this, we should add the $500M final payment for the Sprout acquisition in Q1. This is appropriate since Valeant's total debt is due to acquisitions, and no debt had previously been issued to cover this Sprout payment. The total debt paydown in Q1-Q3 is then $1.20B, averaging $400M per quarter.
With this paydown rate, Valeant would be expected to pay down a total of $1.6B debt in 2016. In fact, in the Q3 earnings presentation, management claimed that $1.61B debt had already been repaid by November 8, and that the total debt paydown target for 2016 is $1.7B. This is compatible with the paydown rate from the 10-Q mentioned above, so there is no reason to doubt it.
According to Endorama Global, the misleading nature of the claimed debt paydown is revealed by comparing the long-term debts at the end of Q3 2016 and end of 2015. So let's take a look at that.
Valeant's total long-term debt at the end of Q3 and at the end of 2015 are stated in the first figure above. At first sight, from the bottom line in the figure, it seems the total debt actually increased from $30.27B at the end of 2015 to $30.39B at the end of Q3 2016. However, those are the wrong numbers to use since they exclude the 'current portion', i.e. debt that will be maturing during the next 12 months. This is the "short-term part" of the long-term debt, and is normally excluded when stating the long-term debt in financial statements. It should not be excluded when considering debt repayment, though, since a major part of Valeant's debt repayments in 2016 has been to pay down the short-term part (current portion) of the debt as it stood at the end of 2015.
Including the current portions, the total debt decreased from $31.09B at the end of 2015 to $30.45B at the end of Q3 2016, a decrease of $0.64B. This is compatible with the net debt paydown of $0.70B claimed by Valeant in the 10-Q as mentioned above (before including the Sprout payment of $0.50B). The small discrepancy of $60M neatly matches the increase in cash on Valeant's balance sheet, which increased from $0.60B at the end of 2015 to $0.66B at the end of Q3 2016.
Thus, there is nothing misleading in management's description of the debt repayment in 2016, contrary to Endorama Global's mistaken claim. For the analysis in this article, the relevant part of the preceding is that Valeant has been averaging $400M in debt paydown per quarter during Q1-Q3.
Modeling the cash available for future debt repayment:
First, Valeant's free cash flow in Q4, and how much of it that will be available for debt repayment, is estimated as follows using Q1-Q3 data and the current guidance for 2016 EBITDA ($4.3B), which implies Q4 EBITDA of $1.05B.
|CF from ops||$558M||$448M||$568M||$505M (from opCF/EBITDA below)|
|opCF/EBITDA||55%||41%||49%||48% (avg. of Q1-Q3 margins)|
|Capex||$70M||$69M||$91M||$80M (rounded up from avg. $77M)|
The average free CF in Q1-Q3 was $448M per quarter. Since the average debt repayment per quarter was $400M, this indicates that the amount of free CF needed for things other than debt repayment (e.g. for lease payments) is $48M. Let's round this up to $50M. Then, if free CF remains at the Q4 estimate of $425M in the table above, the amount available for debt repayment per quarter is $425M - $50M = $375M.
A deduction from this estimate needs to be made to take account of the expected revenue deterioration in the Diversified Products segment in the coming years. According to the Q2 earnings presentation, generic impact on Nitropress and Isuprel will drive an estimated 20% revenue decline in that segment over the next two years. The impact of this on free CF is estimated as follows:
Q3 revenue for the Diversified segment was $471M. A 20% decline would reduce this by $94M. I estimate that ~80% of this decline will go straight to the bottom line, i.e. will reduce the cash flow by $94M x 80% = $75M. This is because the generic-impacted drugs Isuprel and Nitropress have very high profit margins after their prices were jacked up by the former owners, Marathon Pharma, and again by Valeant.
Deducting this $75M from the above estimated amount $375M available for quarterly debt repayment gives $300M, or $1.2B per year. I will take this as the amount available for debt repayment in future years in the pathway analysis below. This is very conservative. First, the 20% revenue decline in the Diversified segment won't happen all at once, but my model assumes that it does. This introduces slack in the model.
Second, with the expected launches of major new drugs in 2H 2017, and Valeant's strong near-term pipeline, cash available for debt repayment should increase from 2018 to 2019 onwards towards the current average of $400M per quarter. Nevertheless, I stick to quarterly debt repayment of $300M from operations throughout the pathway in this analysis, and add only the additional cash freed up from no longer having to pay interest on repaid bank debt.
Valeant should leave its debt repayment at this level even if it can afford to pay more, and use the excess cash for investment (acquisitions/partnerships and R&D), since these have a higher ROC than paying down the debt. The excess cash can also be used in the unlikely event that Valeant receives large fines from current state investigations and civil actions. The impact of fines on the pathway will be discussed later.
Valeant's pathway for managing its debt - the details:
2016 Q4 - 2018 Q2 (seven quarters):
Bank debt repayment from cash flow [CF] during this period: $300M x 7 = $2.10B.
Cash saved from no longer needing to pay interest on this repaid debt: $300M x 0.86% x (6+5+4+3+2+1) = $54M. I envisage this being used for liquidity during the period and then used to pay down debt at the end of it. Then the total bank debt repaid during the period is $2.15B.
Amount freed up by annual savings on interest payments going forward: $2.15B x 3.5% = $75M per year.
Note that the $300M debt repayment in Q4 2016 is lower than Valeant's stated target, which is $1.7B - $1.2B = $500M. I assume this will be balanced by lower debt payments during 2017. The surplus $200M takes care of the $150M milestone payment Valeant has to pay after Brodalumab is approved by the FDA in 2017.
Refinancing the August 2018 6.75% bond ($1.59B):
I assume that a new five-year August 2023 bond is issued to cover this, with a coupon rate of 10%. The additional annual cost of the coupon payments is $1.59B x (10-6.75)% = $52M per year.
This additional cost is covered by the annual saving of $75M on interest payments mentioned above with a leftover surplus of $23M per year, or $6M per quarter. Adding this to the debt repayment of $300M per quarter from operations, the debt repayment per quarter from CF going forward increases to $306M.
2018 Q3 - 2020 Q2 (eight quarters):
Bank debt repayment during this period: $306M x 8 = $2.45B.
Cash saved on interest payments: $306M x 1% x (7+6+...+2+1) = $86M. This is not used for paying down bank debt this time since it will be used instead for the first coupon payments on the new bonds below.
Maturing bonds in 2020:
March 5.375% ($1.98B), October 7.00% ($0.69B), October 6.375% ($2.23B) - a total of $4.90B in bond debt.
I assume new five-year bonds are issued to cover each of these with a coupon rate of 9% each. This coupon rate is conservatively high and could easily turn out to be lower in reality. The total debt paid off at the end of Q2 2020 is $4.60B, and EBITDA according to my model is ~$4.6B, giving debt leverage of 5.6x. This is a significant deleveraging from Valeant's current leverage of 7.1x and is not extreme by pharma standards.
The weighted average of the coupon rates of the bonds maturing in 2020 is 6.06%. Hence, the additional annual cost of the coupon payments for the new bonds that replace them is $4.90B x (9-6)% = $147M per year. The cash saving of $86M mentioned above from interest payments during 2018 Q3 - 2020 Q2 more than covers the additional cost of the new bonds through to Q1 2021. The additional cost will thereafter be covered by savings on bank debt interest payments as shown in the following.
2020 Q3 - 2021 Q1 (three quarters):
Bank debt repayment from CF during this period: $306M x 3 = $0.92B.
Cash saved from no longer needing to pay interest on this repaid debt: $306M x 1.0% x (2+1) = $9M. Putting this towards the bank debt, the total debt repaid during this period is $0.93B.
Adding this to the bank debt repayment of $2.45B in the 2018 Q3 - 2020 Q2 period gives $3.07B. Hence the annual saving on interest payments from repaid bank debt during 2018 Q3 - 2021 Q1 is $3.07B x 4.2% = $129M per year from Q2 2021 onwards.
Balancing this against the additional annual cost of $147M for the new bonds mentioned above leaves a deficit of $129M - $147M = -$18M per year, or -$5M per quarter. This needs to be deducted from the quarterly debt repayments of $306M, so the debt repayment per quarter from CF is reduced to $301M going forward.
2021 Q2 - 2022 Q1 (four quarters):
Bank debt repayment from CF during this period: $301M x 4 = $1.20B.
Cash saved on interest payments: $301M x 1.0% x (3+2+1) = $18M. This will be used for bond coupon payments below, so the total debt repaid in this period is $1.20B. The annual saving on interest payments from this going forward is $1.20B x 4.2% = $50M per year.
Maturing bonds in 2021:
July 7.50% ($1.61B), August 6.75% ($0.65B), December 5.625% ($0.88B) - a total of $3.15B in bond debt.
I assume new bonds are issued to cover each of these with a coupon rate of 8.5% each. The weighted average coupon rate of the original bonds is 6.81%. Hence the additional annual cost of the new coupon payments is $3.15B x (8.5-6.8)% = $54M per year.
This additional cost is covered (roughly) through to the end of Q1 2022 by the $18M cash saved from interest payments on repaid debt in the 2021 Q2 - 2022 Q1 period discussed above. From Q2 2022 onwards, it is covered by the $50M annual savings on interest payments after that period. There is a deficit of $50M - $54M = -$4M per year, or -$1M per quarter.
Deducting this from the quarterly debt repayments of $301M gives repayments of $300M per quarter going forward - the same as what it was at the beginning back in Q4 2016. Thus, the savings from interest payments on repaid debt neatly balances against the additional cost of the new bond coupon payments at this point.
2022 Q2 - 2023 Q1 (four quarters):
Bank debt repayment from CF during this period: $300M x 4 = $1.20B.
Cash saved on interest payments: $300M x 1.0% x (3+2+1) = $18M. Putting this towards the bank debt, the total debt repaid in this period is $1.22B.
There is a small bond maturing in 2022: the July 7.25% bond with a principal of $0.54B. I assume this is covered by issuing a new five-year bond for the same amount with a coupon rate of 8%. This is very close to the 7.25% of the original bond, and, considering its small size, the additional annual coupon payments for this are negligible.
The total debt paid off by the end of Q1 2023 is $7.95B, leaving a total remaining debt of $30.45B - $7.95B = $22.50B. According to my conservative model, EBITDA at this point will be more than $5B. Hence the debt leverage is now less than 4.5x. This is a normal and unproblematic leverage in the pharma industry, and Valeant will, therefore, be able to cover all future maturing bonds with new bonds at reasonable coupon rates similar to the original ones. (The model gives EBITDA of $5.2B, resulting in leverage of 4.3x.)
Valeant's debt crisis is over at this point. It just needs to pay down the debt a bit more so that the leverage is 4x, then it can - and should - keep the leverage at that level and start using most of its cash flow for things that generate a higher ROC than paying down debt, namely, acquisitions/partnerships with small biotechs and R&D.
It is worth noting, however, that Valeant has a mountain of bond debt coming due in 2023: $7.5B in total, with a weighted average coupon rate of 6.4%. (This includes the new bond issued in 2018 in the pathway above.) Therefore, it is very important that the debt leverage is sufficiently reduced by this time so that these bonds can be replaced with new ones at coupon rates of 7% or less. Any further deterioration in Valeant's businesses that causes the leverage to be above 5x at this point would be problematic.
Potential impact of fines:
There has been talks from the usual VRX bears - Wells Fargo analyst David Maris etc. - of Valeant being at risk of fines/payments up to $2B from the various state investigations and civil actions. I consider this exceedingly unlikely (although it would take a separate article to explain why). But to indulge the bears, let's consider the impact of a $2B fine on the pathway above.
In this case, I envisage Valeant issuing a new bond for $2B some time in 2018-2019 with a coupon rate of 10%. The required additional annual payment for this is $200M per year, or $50M per quarter. Adding this to the $300M debt repayment per quarter in the pathway above gives $350M. This is still far below the $400M per quarter debt repayment that Valeant has been averaging in Q1-Q3 2016. It should be very manageable, as discussed earlier, although Valeant's long-term growth will be negatively impacted due to the reduction in cash for investment.
The total debt at the end of Q1 2023 will then be $24.5B, so with $5.2B EBITDA, the leverage is 4.7x, which is still reasonable for future refinancing.
Disclosure: I am/we are long VRX.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.