Stalexport Autostrady SA: Polish toll road at a 20%+ FCF-yield

A reader whispered the name Stalexport Autostrady SA (WSE:STX) in my ear. This Polish company used to be an importer and exporter of steel products (hence the first part of the name), but in 1997 it obtained a 30-year concession to operate a 61KM stretch of the A4 as a toll road between Katowice and Kraków (respectively the 10th biggest and 2nd biggest cities of Poland). The steel business was sold more than a decade ago when Atlantia SpA (BIT:ATL) took a majority position in the company. Atlantia is an Italy-based operator of toll roads and airports in various countries with a €22.5 billion market cap while Stalexport Autostrady is worth just €225 million at the moment. Atlantia holds 61.2% of the outstanding shares, but it’s just a tiny tiny sliver of their overall business.

Toll road between Katowice and Kraków on Google Maps

Conceptually, understanding a toll road business is easy. The costs you have are mostly fixed: road maintenance and operating the toll booths. How much money you make is a direct function of how many cars use the road, and how much toll you ask. It’s usually a pretty good business to be in because most people have no real choice to use your road if they want to get quickly from point A to point B, and no-one is realistically going to build a competing  road in the same direction. In this case we know that the concession will end in March 2027, so this is the kind of situation that screams for a discounted cash-flow model to estimate intrinsic value.

Financials: the balance sheet

Before we start building a DCF-model it’s a good idea to first take a look at the historical financials, and especially the balance sheet. There are a couple of big items on there that are important to understand. I have color-coded them in one column:

Source: company data and author’s calculations

The assets colored red are basically the operating assets of the company, and the concession itself (the intangible assets). As you can see the value of the concession drops every year, and is expected to be zero in 10 years time. Maybe the physical operating assets will have some residual value, but this part of the balance sheet isn’t very important.

More important are the yellow colored cells. On the asset side these are mostly labeled “other non-current investments” while we have various provisions at the liability side. These two items cancel each-other almost out, and that is no coincidence. The investments are “cash reserved for capital expenditures” while the provisions are the estimated costs of mandatory capital expenditures (with the concession also came the obligation to maintain/upgrade the road). This is very handy for our DCF-model. We don’t have to worry about modelling capex: we can just assume that these balance sheet items will cancel each other out in 10 years time. To check that this is sort of reasonable we can compare the implied annual capex/year with historical capex levels. Total provisions are PLN412 million which would imply roughly PLN41 million of capex/year versus a seven year average of PLN46.5 million. This is a small gap to begin with, and it will be a bit smaller because the provisions are discounted while the investments will also generate a certain return (and we have some excess investments to begin with). I also think it makes sense to assume limited capex close to the end of the concession.

The green and blue cells are closely related, and basically represent how much excess cash/assets the company has at this point in time. They do have some debt, but they have been paying that down rapidly in the past years. Note that the “other non-current liabilities” also represent a very debt-like liability since they are concession payments that have to be made to the Polish state. This amount went down by a large amount compared to 2016, but that at the same time “trade and other payables” saw a huge increase. As a result net working capital turned negative, but this should be a temporary situation. They are basically paying down debt, and they now have roughly PLN60 million in excess cash while holding PLN60 million in positive working capital as well if we normalize the situation.

Financials: earnings and cash flow statement

Source: company data and author’s calculations

Looking at the income statement is useful, but we have to know what we are looking for. Earnings are sort of meaningless (except for the fact that they determine taxes) since we don’t care about the depreciation of the concession, or the variability in earnings because of lumpiness in roadworks. Stalexport Autostrady doesn’t really have a meaningful “cost of sales”. A large part of the cost of sales are the amortization of the concession and changes in the valuation of the provisions, and costs related to roadworks. Therefore, it doesn’t move based on revenue, and as a result higher revenues are resulting in higher operating margins. Because of that, the business is at the moment throwing off cash.

We mainly care about the amount of operating cash flow. The most important driver of the amount of operating cash flow has been the revenue, that has been growing at an average rate of 10% per year. If you dive in the companies annual reports you see that this growth is driven by a combination of higher traffic and higher tolls. The maximum amount of toll that the company can charge is capped in the concession agreement, but the rate for passenger cars can be increased by 40% before hitting it while the rate for trucks has 18.3% of upside left. Tolls from passenger cars represent roughly two-thirds of all toll revenue.


Now that we sort of know the current state of the business, and understand what is important, we can use that to make some predictions about the future. I have made a very simple cash flow model based on the trailing twelve month operating cashflow. I have adjusted this number for changes in working capital, and also deducted the dividends paid tot non-controlling interests (and assumed that this amount will remain the same in the future). The value that the models spits out depends on four major assumptions:

  1. Incremental revenue drops directly down to the bottom line (after 20% taxes are paid).
  2. Revenue continues to grow with an average rate  of 10% per annum (not so conservative).
  3. No additional cash is required for maintenance capex above the amount currently provisioned (with the PLN164 million in net tangible assets on the balance sheet providing an additional buffer).
  4. We use a 10% discount rate (I think this is pretty conservative).

The interaction of the first two are really the crucial ones. Because we don’t model increases in costs and overhead while revenues grow the amount of operating cash flow that the business generates the next ten year will just explode. If you look back at the historical results that doesn’t seem to too crazy: this is exactly what we see happen in the past results.

Without further ado, the model:

Source: company data and author’s calculations

If these assumptions are reasonable, we find that Stalexport Autostrady should be capable of generating roughly PLN3.2 billion of operating cashflow with a net present value of PLN1.9 billion using a 10% discount rate in the next 9.5 years. Since the current market cap is just PLN945 million this gives us a large margin of safety, and if we increase the discount rate to find the same present value as the current market cap we need to use a whopping 27.4%.

If we input zero revenue growth in this model we get undiscounted cash flows of PLN1.7 billion for a discounted value of PLN1.1 billion. So basically the market is pricing the toll road as if revenues will suddenly stop growing, something that I think is unlikely to happen with an asset like this that has so much pricing power. And even if that turns out to be the case, getting paid 10% per year is nothing to sniff at in today’s low interest rate world.


To cut directly to the chase: Stalexport Autostrady seems to be valued very attractively compared to its expected cash flows in the next 10 years. According to my, admittedly very crude model, the stock should be trading at roughly twice the current level. It’s quite possible that a 10% revenue growth rate will turn out to be a too aggressive assumption, but since expected returns are adequate even with a zero growth rate I think there isn’t a whole lot that can go really wrong here. A toll road (concession) must be one of the more stable assets one can own, so I’m pretty excited to be able to buy this at such a high yield.


Author is long Stalexport Autostrady

Tejoori finally liquidating

In the beginnen of this year I bought a small low-conviction position in Tejoori. The company had substantially sold all their assets at that point in time, and my guess was that it would soon fully liquidate. Tejoori took their sweet time though on making this decision, and if you would have asked me in the past couple of months if I would have wished I never invested in the company I would have said yes. “Luckily” trading stock on the SEAQ segment of the AIM market isn’t possible without incurring huge frictional costs, so I never bought or sold any shares after establishing my initial position. Last week the company announced that it would fully liquidate and return to shareholders its cash balance of approximately $17.6 million.

The company doesn’t provide an estimate of liquidation proceeds, and while knowing the cash balance is nice it doesn’t tell us how many liabilities are remaining on the balance sheet and how much costs will be incurred to finalize the liquidation. However, I suspect that there are no meaningful liabilities left. At the end of 2016 the company had $19.0 million in cash (and Wakala deposits) with total liabilities of $1.4 million for a net equity of $17.6 million. Since the company has a minimal amount of cash burn (less than $200,000 annually) I would conclude that most liabilities should have been extinguished given that the remaining cash balance is now close to the previous net equity number.

With the stock trading at $0.50/share the potential upside is significant. If we would budget half a million to wind down the company, net assets value per share would be $0.62 for an upside potential of 23.5%. For a liquidation that’s a pretty big spread, and I couldn’t resist increasing my position. At the same time I’m still not willing to bet too big on this. I deeply distrust the average AIM listed company, and so far Tejoori hasn’t done a lot to stand out positively.

I also wonder how the liquidation will be executed. The company is planning to hold a vote on delisting the company later this month, followed by a second vote at a date to be determined to authorize the liquidation. There should be no problem with passing the first vote. They need a 75% majority to approve it, but only from the votes present. If 51% of the shares vote that is sufficient, and if this threshold isn’t reached the meeting is delayed by one day and the threshold is lowered to 33%. Management owns 16.42% of the outstanding shares, and will vote in favor of the delisting, so this really shouldn’t be a problem.

After the delisting the company is planning to cancel their CREST facility (electronic depository), and I’m not sure what exactly the impact is of this move. I own my stock through CREST, so I’m wondering how liquidation distributions will be made if this facility is cancelled. Secondly I’m wondering if this will impact the second vote required for starting the liquidation. Shareholders who own shares in uncertificated form also vote through the CREST system.


Author is long Tejoori

A bit more on the Ocera Therapeutics merger

Yesterday Mallinckrodt officially launched the tender offer for Ocera Therapeutics with a deadline on the 8th of December while the merger should be concluded shortly after. Concurrently with the tender offer document Ocera also filed a “solicitation/recommendation statement” that has some interesting information that can be used to value the CVR. The document contains some tables with managements estimates of the probabilities of passing the various hurdles that need to be taken to commercialize OCR-002, and what kind of revenue it expect to generate. Management’s estimates of the probabilities are as follows:

The two most important milestones for the CVRs are those related to successfully starting Phase 3 trials. As you might remember from my previous post, I estimated a 80% probability for an intravenously version of the drug entering Phase 3 trials and a 20% probability for the oral version. It’s nice to see that as a total pharma nitwit I managed to guesstimate something in the right direction. Ocera management is going with 75% and 35.7% respectively. This would value the first two milestones at $0.43/share, pretty close to the number I was getting.

There is a big difference between my estimate of the probability of commercial success versus Ocera’s, but since my estimate was a super low 1%, that’s not really a surprise. They estimate a 34.6% probability of getting FDA approval combined with a 60% probability of commercial succes which implies a 20.8% probability of hitting the last CVR with just the intravenously version. How this probability increases with adding the oral version to the mix is a bit unclear. Presumably all outcomes here are highly correlated with each other, and adding the oral version to the mix will not meaningfully increase the probability of hitting the last milestone. Hitting it stand alone has a 16.5% times 80% probability of success which is a 13.2% probability, and presumably most of the outcomes overlap with those in the 20.8% pie of hitting commercial succes with the intravenously version. If we simply go for the 20.8% probability the third milestone is worth $0.34/share, although this number does need to be discounted significantly. Payment will not only be very far away in the future, a Mallinckrodt receivable might also carry a significant amount of credit risk

The tables with projected revenues also give some hints with regards to the timing of the various milestones. With the intravenously version possibly hitting the market in 2022 and the oral version possibly hitting the market in 2023 I guess we should expect the milestones related to entering Phase 3 trials to payout relatively soon. That good, not just because of the time value of money, but also so we don’t have to worry too much about credit risk since most of Mallinckrodt $5.9 billion in debt is due between 2022 and 2025.

Clear is that we don’t have to expect a quick payment on the last milestone. Using these projections cumulatieve revenue would hit the $500 million somewhere in 2027 while the oral formulation only hits it in 2030. And we shouldn’t forget the reason why this deal is partly financed using a CVR: presumably Mallinckrodt didn’t agree with all those projections and they certainly could be too optimistic. On the other hand a CVR can also simply be a way to share risk and borrow money. Given the highly leveraged nature of Mallinckrodt not spending too much hard cash on a deal must be attractive to them.


Author is long Ocera Therapeutics

Ocera Therapeutics: merger arbitrage with CVR

Earlier this month Ocera Therapeutics (NASDAQ:OCRX) announced that it would be acquired by Mallinckrodt (NYSE:MNK) for $1.52/share in cash plus a CVR that could be worth up to $2.58/share. In general I like mergers with CVRs because I think market participants are often conservative with valuing them. I especially like gettings CVRs for free, but unfortunately that is not the case here since the stock is trading at $1.70/share.

The CVR has three milestones on which it will payout:

  • $0.34/share if the first patient is enrolled in a Phase 3 trial for an intravenous formulation of OCR-002 (before 2029).
  • $0.52/share if the first patient is enrolled in a Phase 3 trial for an oral formulation of OCR-002 (also before 2029)
  • $1.72/share if cumulatieve sales of OCR-002 worldwide exceed $500 million before 2029

So to make this merger arbitrage a succes you basically need to hit the first CVR milestone. With a share price of $1.70 you are effectively paying $0.18 for the CVR, and if it returns $0.34/share at some point in the next couple of years it should generate a nice internal rate of return. The merger will be structured as a tender offer that is scheduled to be launched no later than November 16, 2017 and should be concluded before the end of the year. So with most of the cash being returned soon the only thing that really matters is how much are we paying for the CVR, and how much it’s expected to payout.

I know basically nothing about medicine, so take that in mind reading this post, but when reading the press release it’s clear that OCR-002 is not a sure thing to make it to a Phase 3 trial. It was unable to meet statistical significance in its primary endpoint in its Phase 2 trial, but it appears to be that this was caused by using a too low dosage of the drugs. That’s something that’s easy to correct, and Mallinckrodt is paying $42 million for Ocera Therapeutics so they must have a decent amount of confidence that they will not only be able to progress to Phase 3, but also get the product to the market. And I guess if you can make OCR-002 work intravenously there is also a decent probability of making it work orally.

I wouldn’t ascribe to much value to the sales milestone though. It not only requires OCR-002 to become a big success, you also have the problem that for a sales milestone your interests aren’t very well aligned with the company. If they are close to the deadline and close to hitting that $500 million milestone they will have a big incentive to stay below that number in order to avoid the $75 million milestone payment. You don’t have this problem with the two milestones connected to starting Phase 3 trials. Their whole purchase of Ocera Therapeutics will be a waste of money if they cannot start those trials.

To be honest I’m a bit surprised that there is no milestone related to successfully concluding a Phase 3 trial, but I guess that’s a good thing for investors here. Just starting Phase 3 trials is a milestone that is a lot easier to reach than successfully finishing them.

Attaching numbers to the various probabilities is a huge guess, but in my mind something like the following doesn’t sounds unreasonable. Maybe there is a 80% probability of starting a Phase 3 trial, maybe a 20% probability of also doing it for an oral formulation and maybe just a 1% probability of hitting the $500 million sales milestone. This would value the CVR at $0.39/share compared to a market price of $0.18/share, so as long that is roughly in the right direction it would be a good bet. I think it is.


Author is long OCRX

My take on TABS Holland

Last week a fellow Dutch value investing blogger wrote-up TABS Holland. The company supplies wood and building products to the Dutch market, and is traded on the obscure NPEX platform. The platform will actually cease all trading next month because its software doesn’t comply with the latest regulations. They have implemented a weird auction system without a central order book that makes it possible that different lots of shares are simultaneously sold and bought for different prices. An obscure illiquid stock on a weird exchange sounds like the rights ingredients for a cheap valuation, and on the surface that seems to be the case for TABS Holland as well. It’s trading at a 6.6x P/E-ratio while the other valuation metrics are as follows:

Last price: €28.76
Shares outstanding: 6,768,303
Market cap: €194.7 million
Net debt: €63.4 million
Enterprise Value: €258.0 million
P/E (ttm): 6.6x
P/B (mrq): 1.56x
EV/EBITDA (ttm): 5.23x
EV/EBIT (ttm): 6.12x

Based on market cap and enterprise value Timber and Building Supplies Holland (TABS Holland) is actually not that tiny and obscure. Most of the stocks I own are smaller, but in this case the float is ridiculously small. According to the free float is roughly 3% which translates to less than €6 million. Most of the shares are held by HAL Holding while management and some other “big” shareholder own the rest.


Before trying to figure out how much this business is worth it’s a good idea to start with a look at the historical financials, even though that might not be that useful in this case. In 2015 PontMeyer NV acquired Deli Building Supplies and the combined company continued as TABS Holland. The acquisition basically doubled the size of the company. The transaction was completed the 1st of September, so in the financials for the year 2015 the last four months include the results of Deli Building Supplies while 2016 was the first full year of the combined company.

When looking at the historical results it is clear that TABS Holland isn’t a great business. Revenues didn’t grow at all in the 2011-2014 period while profitability was around breakeven. Only in 2015 and 2016 things started to turn around in a big way. While the acquisition of a big competitor can’t have hurt, I think the main reason is simply the fact that the Dutch housing market has rebounded after the crisis and is now running white hot again. The low for the Dutch housing market was in 2012 and 2013, while now things are pretty crazy again. I live in Groningen, which isn’t as hot as Amsterdam, but I  have never seen so many houses being renovated or just being torn down and rebuild in my neighbourhood as in this year. So it’s not a surprise that the more recent results are pretty good, and presumably the full year results for 2017 will be even better. The timing of the acquisition was certainly excellent!

It was a big acquisition, and to finance it TABS Holland issued a bunch of new shares while the amount of debt outstanding also exploded. They have been paying down the debt rapidly, and last year they even found some money to pay out a decent dividend.


How much is TABS Holland worth? it’s a good question. It’s a cyclical business that probably is near the top of the cycle. The housing market in the Netherlands isn’t showing signs of slowing down yet, but when you play a game of musical chairs the music always continues playing until it doesn’t. Based on current earnings the business is clearly cheap, but a lesson about cyclical businesses that stuck with me is that they are best bought when they appear expensive, not when they appear cheap. But maybe a 6.6x P/E ratio is cheap enough to overcome this risk.

A reasonable base-case scenario is perhaps to assume a couple of years like this year, followed by more average results (adjusted for the doubling in size of the company). This gives some credit to the current good results, while also building a bit of mean reversion into the valuation.

In the valuation I have doubled the operating profit numbers from 2011 till 2014 to account for the fact that the business is now roughly twice as big. Since 2015 already included 4 months of the combined company I multiplied those results by a factor 1.5. This gives us an average operating profit of €21.2 million. Deduct €1.4 million in interest costs (assuming they keep operating with the current debt load), deduct 25% in taxes and normalised net income is €14.9 million. Slap a 10x multiple on that and we have a €149 million terminal value. Additionally I will assume that the next 3 years will be great, and the company will earn €30 million in each of these years. Discount this with a 10% rate and we get a valuation of €182 million.

The current market cap is actually a bit higher than that, so to me the market price seems pretty reasonable. A P/E-ratio of 6.6x sounds pretty cheap, but at the same time I don’t want to be paying a 10x ratio if it’s indeed near a cyclical top.


When I read about TABS Holland I thought that it was a pretty interesting situation, and I already had opened a NPEX account and transferred some money. After doing a little bit of research I’m less convinced. The company doesn’t appear particularly expensive, especially compared to other publicly traded equities, but at the same time it isn’t obvious cheap either. I think a conservative valuation needs to take into account a bit of mean reversion at this point in time, and because of that it should be looking cheap from the outside. You could come up with some reasons why looking at the past results is too pessimistic or why a 10% discount rate is too high though. The combined company probably has realised some synergies while the market has become less competitive.

I have to admit that I would have loved to buy something on such an obscure platform as NPEX, but I think I’m going to pass on this opportunity… TABS Holland is probably a bit undervalued, but not enough for me.


Author has no position in TABS Holland