Sunday 27 September 2015

Outsiders in chemicals, the story of Rockwood Holdings

This piece is not about an investment idea, rather a case study of capital allocation and a story of a few Outsiders. When I first came across Rockwood I thought that this could be summed up neatly in a few paragraphs. Technically speaking it is possible but wouldn’t do justice to the story. The content is rather on the long side, however if you don’t feel like reading the whole thing the below quote, from Rockwood CEO Mr Seifi Ghasemi in early 2014, will give you the essence of what I’m about to describe.

This post is for investors who are (i) interested in stories of shareholder friendly managements and capital allocation, (ii) doing their homework on Air Products and want to learn further about Mr Ghasemi or (iii) doing their homework on Albemarle (eventual acquirer of Rockwood in 2014).

Onto the quote from Mr Ghasemi.

“In the last nine years as a public company we have on a consistent basis said that the only strategic goal we have at Rockwood is to maximize shareholder value. This is the guiding principle, which drives everything we do. Our job is to make money for our shareholders. We believe that in the long-term what matters is the increase in the per share value of our stock and not overall size or growth. 

I want to emphasize this point since it is the core of our strategic thinking; the act to increase the per share value of our stock and are not enamored with building an empire and running a bigger company. So it is this fundamental principle of focusing on increasing the per share value of the stock that led us to this strategic decision to sell 60% – yes, 60% – of our company and focus on our two core businesses. We publicly announced the key elements of our strategy a year-ago in January of 2013 and set the target of executing it in two years. Today, a year later, we have delivered on all elements of that strategy well ahead of plan.”

Rockwood was a specialty chemicals company ran by a great management team and ultimately acquired by Albemarle, returning 18% p.a. since the IPO in 2005 until the announcement of the sale in July 2014. The CEO – Mr Ghasemi - has since went on to transform the Bill Ackman backed Air Products.

Rockwood's history goes all the way back to the 1880s when the founder, Bernard LaPorte (back then known as LaPorte Chemicals) came up with a process to manufacture hydrogen peroxide, which was used for bleaching wool and straw hats, key products of the British economy during the industrial revolution. To conveniently skip 100+ years, LaPorte has fallen on hard times and certain parts of its business were acquired by KKR in 2000 for c. $1.2bn. In 2001 KKR brought in Mr Ghasemi, who has a background in chemicals and industrials and previously worked for GKN and BOC (now part of Linde). Mr Ghasemi, along with CFO Mr Robert Zatta and SVP of Law and Administration Mr Thomas Riordan took this business and over time converted into the largest lithium player in the world, handily enriching shareholders in the process.

Flying under the radar

The reason you probably haven’t heard about the company (unless you follow the lithium business) is partly because it’s a B2B company and partly because promotion was the last thing on management’s mind. Companies like Rockwood can hide in plain sight while compounding away. Management was always very low key but very much aligned with shareholders. Their mantra was decentralised operations and they ran Rockwood out of a small office in Princeton, small staff (25 at IPO, 35 at last count) with everybody on one floor. You could think of Rockwood as a holding company with a bunch of assets and hard-nosed approach for lean operations and value creation.

Comments from Mr Zatta (who previously worked at food giant Campbell Soup) in a 2014 interview paint a clear picture.

“Rockwood, with its small management team, definitely provided that [working in a non-bureaucratic environment]. “We didn’t have any bureaucracy,” Zatta recalls. “We didn’t create levels and layers of management for the sake of having process. We were very much focused on getting things done.”

The business is run on a decentralized basis, with the managers of individual units given the autonomy to run things how they needed to be done. “If they had a big capital project, or if they needed approval to shut something down, they only had to talk to myself and Seifi. And now that he’s not here, basically just myself.”

Rockwood’s leaders were crammed into close quarters. “We had our treasurer sitting in the kitchen. That was his office,” Zatta says. “Our controller sat at a secretary’s desk outside my office. When the CEO came in, the only room he could work out of was a conference room.” Zatta said the nimble management structure was a breath of fresh air compared to his big corporate background. “It was exhilarating,” he says.”

A slide from a 2014 investor meeting gives you an overview of their MO. Having read their filings, conference transcripts and presentations I’ve stopped counting the occurrence of “shareholder value creation” at about 347. At least I got the point.

Source: company filings

Act I: The beginnings - IPO (2005)

Rockwood started life as a portfolio company of KKR, when the PE fund acquired the pigments, additives, metal processing and other businesses, which constituted over 50% of LaPorte’s (UK chemicals co) sales, for $1.2bn (around 1.2x TTM sales and 8x EBITDA). Apparently, LaPorte’s share price drastically underperformed the FTSE at the time as the business had a bit of a rough patch due to a strong pound, rising oil and other input costs, hence they hoped that selling assets and focusing on their core operations (fine and performance chemicals etc) would help performance. LaPorte was also a bit of a hodgepodge of loosely related assets so selling things off seemed like the right thing to do. In general, the growth profile of chemicals and industrials businesses is modest, but PE firms like these assets given the stability i.e. they can be levered to the hills. In this deal KKR put about $0.3bn equity with the rest financed by debt.

As a side note, what makes a chemicals company a “specialty” business is that the products they sell make up a small percentage of customers’ operations but are critical to performance. These are in general lower volume, higher margin products as compared to their commodity cousins. Rockwood has built a portfolio of inorganic chemicals assets (i.e. no relation to carbon related minerals such as oil) in a diversified way.

Rockwood has grown via acquisitions of which the largest was the 2004 acquisition of Dynamit Nobel. The Dynamit acquisition included the businesses of Sachtleben (TiO2), Chemetall (lithium and surface treatment) and CeramTec (ceramics, hip implants etc) amongst others. Total consideration was $2.3bn (inc. assumed debt) and the seller was MG Technologies (now GEA Group) as the German conglomerate was separating its engineering and chemicals businesses. At the time Dynamit had sales of $1.6bn (compared to Rockwood’s $0.8bn) and EBITDA of c. $300m (paying just above 8x TTM).

With this acquisition Rockwood expanded into new platforms. While the actual operational overlap was limited, these businesses had one thing in common with the existing operations: speciality chemicals and the products represented a small portion of customers’ production costs.

From Mr Ghasemi at the time (the last point is key). “The four businesses we are buying are profitable stand-alone companies,” says Mr. Ghasemi. “There is no particular product overlap, but there is overlap in customers in the consumer products, construction, electronics and coatings markets. These are specialty chemical businesses where the definition is that the products they make are a small percentage of customers’ costs and essential to the way they make their products.”

The key here, as you’ll see below, was to get to the lithium business. Rockwood’s management was developing a thesis that an inflection point in the demand for electric vehicles in 10-15 years would cause a surge in lithium demand and a scramble for supply. But to get that business they had to buy the whole thing from Dynamit, with a few problem children.

Following this acquisition, as well as other smaller bolt-ons such as Groupe Novasep (subsequently sold in 2006 in an MBO) and the pigments business of Johnson Mathey, Rockwood would report $3bn in revenues and $570m in adj. EBITDA in 2005 (year end). The other side of this breakneck growth was the increase in leverage, on a gross basis Q1 2005 was close to $3.5bn with interest expense over $200m annually.

KKR decided that it was time to take Rockwood public and did so in August 2005 by selling 23.5m (inc. 3m greenshoe) shares at $20 apiece, raising $440m after discounts, which almost entirely went to pay down debt. Rockwood starting trading on the NYSE under the ticker symbol ROC (Jay Z must have been a buyer).

Prior to the IPO, the business was 75% owned by KKR, 22% DLJ (remember them?) while mgmt owned c. 1% of the business. Post IPO KKR’s stake was down close to 50% and has exited the business via a multiple of sell downs during November 2007, June 2008, December 2010, May 2011 and finally in October 2012.

The below slide gives you an overview as to the development Rockwood has gone from formation to post-IPO.

Source: company filings

Act II: The middle part – 2006 to 2012

As Mr Ghasemi elaborated during a 2011 investor meeting:

“We had two five-year plans that we have executed. The first five-year plan was to change the company culture, service the debt and live within the covenants, grow by acquisitions to more than $3 billion and take the company public. That was our main goal for the first five years. We accomplished that. 

Second five years, we wanted to optimize our portfolio, therefore we sold a lot of businesses. We executed bolt-on acquisitions to strengthen our core businesses. We improved our EBITDA margin to 20%. And we paid down debt. Our goal was three times, which we have – we are very ahead of that. So that was our second target. 

But I'm sure you are very interested in terms of what we are going to do for you in the future. I have some more details on that. But moving forward in the next five years, what are we going to do? First of all, why are we going to do what we are going to do? We want to create value for our shareholders. We measure ourselves strictly by the price of our shares. That is what – we exist to create value for our shareholders […]”

Rockwood closed the 2005 financials with $3.1bn in sales and $570m in adj. EBITDA. By 2012 the surviving businesses in the pursuit of higher returns on capital reported sales of $3.5bn and adj. EBITDA of $780m. While this might not seem like a lot of increase consider the divestments in the process. The core businesses such as lithium and surface treatment (previously part of specialty chemicals) or advanced chemicals have all performed well.

Source: company filings. Pie charts represent breakdown of revenues

During the 2005-12 period EBITDA margins increased from 18% to 22% given changes in the portfolio and operational efficiencies. More importantly net leverage decreased from about $2.7bn (almost 5x net leverage) to $1.5bn (about 2x), which is pretty substantial. In June 2012 management even instituted a dividend policy of paying out $35c per share per quarter ($27m per quarter). In the same period ROIC averaged 12%, FCF margin approx. 5% (positive every year, even in 2008/09) and FCF yield around 8%. This is pretty good in the chemicals space.

I’d mention a few key developments here and deal with a few housekeeping items. To start with the divestments, management sold a few businesses that you see in the 2005 breakdown but not in the 2012 financials, such as Group Novasep in an MBO (it produced pharma ingredients, for EUR425m or $540m EV at approx. 8x EBITDA), specialty compounds business to Mexichem ($300m EV at 1.3x sales and c. 9x EBITDA) or the electronics business to OM Group ($315m EV at c. 8x EBITDA). In addition, they bundled their TiO2 business (Sachtleben) with Kemira, Finnish chemicals co to form a JV where Rockwood owned 61%.

Management also made a run for Talison, a Canadian listed company with the world’s largest ore based lithium mine in Australia. Ultimately they lost out to a Chinese co named Tianqi, but as life works in mysterious ways Rockwood got another shot at it.

Management has been very disciplined in capital allocation. On a 2008 conference call, just before the crisis they noted the following: “Our priority we have said that our priorities number one is organic growth, number two is bolt-on acquisitions. We still have a lot of opportunities. We think actually that with the economic downturn, asset prices have gone down. We are not facing as much competition from private equity as before. As a result we think it's an opportunity to put our cash in use and to do some additional bolt-on acquisitions, which would be helpful.“

They had clear guidelines for acquisitions: the target had to be or have (i) global market position, (ii) adj. EBITDA margin of 25%+, (iii) global industry technology leader and (iv) limited exposure to oil-based raw materials (Rockwood’s bread and butter is in inorganic chemicals).

A word on the 2008/09 crisis. This had a pretty significant impact on Rockwood – sales and EBITDA declined 20% vs 2008. This was coupled with a pretty strong exposure to Europe as well as FX fluctuations. The business was helped by the diversified portfolio and a very proactive management – they’ve started cost cutting already in 2007 to reach $150m by 2009. As leverage was still high, they’ve renegotiated covenants and ultimately came out OK from the crisis. They also had to let people go (about 900 or 9% of the workforce) but mgmt. also froze their base pay, which didn’t increase $1 until 2013 (even then Mr Ghasemi’s stayed flat). Despite the solid results the share price dropped from above $40 at the peak in June 2008 to as low as $4 in early 2009.

A few housekeeping items before we move on.

You’ve noticed that in places I used adj. EBITDA as a metric. This was management favoured reporting method (and what they used internally too). This is basically reported EBITDA adjusted for one-off items. Now EBITDA is one thing, but when I see adjusted EBITDA all sorts of red flags are raised. To their credit they’ve at least been consistent and covenants were also tied to adj. EBITDA metrics. More importantly the company has been FCF positive all through the years (even in 2008/2009), while mgmt. comp was also tied to cash generation.

If you look back historically, the company for a long time traded at a discount to peers mostly due to the complexity of the portfolio (7 segments and 17 business units at IPO). The thing about specialty chemicals is that analysts are looking for a comp and if they cannot find it they’ll assign a discount, which of course creates opportunities for investors. Rockwood’s operations were also mostly based in Germany and other parts of Europe and issues such as FX (cost side) or macro risks (demand) hurt the company vs their peers oftentimes.

The good news was that management was always focused on shareholders and never engaged in empire building. They were always more concerned about the long term, never giving quarterly EPS guidance or forecasts but always laid out a framework of how they thought the business could compound, margins, how they controlled costs and capex, leverage and so on. Reading their transcripts of conference calls going back 7-10 years was very refreshing and full of common sense.

Rockwood positioned itself to be number 1 or 2 by market share in their respective businesses. About 70-80% of their sales came from businesses in such positions and management focused on getting out of the non-core ones. While changes were happening all along in the pursuit of a more rationalised portfolio, prompted by the undervaluation management decided to take drastic actions just as 2012 was coming to an end.

Act III: Management goes activist (2013 to 2014)

On a cold day as the world was getting over a NYE hangover management laid out something radical on a January 2013 investor day: shrinking the business by selling over 60% of revenue generating units and focusing solely on lithium and surface treatments. They certainly didn’t believe in sacred cows. The thinking was that the surface treatment business would be the cash cow: 25% EBITDA margins, low capex requirements (3% of sales on average vs 15% for lithium) and relatively good stability, fuelling the growth that was to come from lithium.

Management made a bold bet and starting selling assets off aggressively. In June 2013 they agreed to sell the advanced ceramics business (CeramTec) to Cinven (European PE fund) for $2bn (3.5x TTM sales and 11x TTM EBITDA). It was a high quality business with margins in the mid 30%s and given its presence in healthcare ceramics (think hip implants etc) a good exposure to the theme of ageing population.

Since they bought it as a package deal when acquiring Dynamit the like for like return comparison is not straightforward but assuming the $2.3bn price paid (inc. assumed debt) for then $1.6bn of sales (1.4x TTM sales), which then allocated equally to the segments (CeramTec was 18% then) would result in a 5x gross return (the multiple allocation is pretty subjective). If you assume constant multiples i.e. 3.5x TTM sales it’s a 2x gross return. The important point is that the return above is on an asset basis, in reality Rockwood used about $425m equity to buy the entire Dynamit business (as noted above CeramTec was less than 20% of the valuation) compared to selling just the CeramTec business for $2bn (gross proceeds – about 10% went to taxes, other fees etc - and a large portion went to pay down debt). It’s a pretty good achievement on an asset basis, amazing on equity. Why did they sell such a high quality business? On a conference call after the close Mr Ghasemi admitted that the exposure to a potentially huge medical liability didn’t let him sleep well at nights.

Next on the line was the Clay-based Additives business in July of the same year for $625m to Altana, a German specialty chemicals business for 3.3x TTM sales. This business was part of the performance additives segment. As a fun fact Altana (a company owned by one of the children of BMW founder Herbert Quandt) was considering buying Rockwood or parts of it in 2008, but talks didn’t materialise in the end.

Then to conclude the divestiture spree Rockwood neared the year-end by selling its TiO2 and the remainder of the performance additives business for $1,275bn (inc. pension) to Huntsman. The business generated $1.5bn TTM sales and $105m TTM EBITDA at the time of the announcement. Huntsman’s acquisition price was $1bn (exc. pension) so a multiple of about 0.7x TTM sales and 10x TTM EBITDA, however 2014 adj. EBITDA was expected to be over $200m (improvement of about $100m yoy, on volume, price recovery, raw material reduction etc) so the multiple is in effect was about 5.5x EBITDA at announcement (before synergies). The multiple seems low, but it’s roughly in line with peers. Some saw multiples fetching 6-7x given the speciality element in Rockwood’s business but that ultimately didn’t materialise. The deal happened amid a volatile period for the TiO2 industry with an industry restructuring so probably mgmt. preferred not spending any more resources on this business, which resulted in a decent sale in the end. This business was always a problem child for Rockwood, but as noted above the only way to get to the lithium business. The business was probably worth more in the hands of Huntsman. After a bit of back and forth on regulatory matters the deal finally closed in October 2014.

Lest you think we are done, Rockwood announced that after a failed attempt about a year ago it agreed to acquire a 49% stake in a JV with Tianqi (the parent company of Talison, which owns the Greenbushes lithium mine in Australia) for a total consideration of $516m ultimately (c. $1bn valuation). The saying of "if you cannot beat them join them" comes to mind. Mgmt. guided c. $50m proportional EBITDA from this JV (c. $100m on 100% basis) so would imply a multiple of around 10x, which is along the lines of peers. On a TTM basis with $60m EBITDA (100% basis) the implied multiple was 16.5x. In their 2012 run for Talison, Rockwood offered a valuation over $720m for $30m EBITDA, just as earnings were ramping up.

Talison operates the largest lithium producer anywhere in the world with about 100ktpa lithium carbonate capacity (global production capacity is about 250ktpa). The idea was that the eventual growth from batteries, cars etc would in the next 3-5 years more than compensate for the introduction of this monstrosity of capacity. The Australian mine doesn’t actually supply lithium carbonate but hard rock lithium, which other players (such as Tianqi, a Chinese hard rock converter and the largest in the world) process further. The mine is not the lowest cost but very close to China, while Tianqi is their single largest customer. In commodities you either make money with a cost or location advantage, if you’ve both - think nitrogen fertiliser producers in the US Cornbelt - you have found yourself a gold mine. Ultimately this move strengthened Rockwood’s presence in lithium globally. Additionally, Rockwood entered into an option agreement with Tianqi through 2016 which stipulated that Tianqi can take a 20-30% stake in Rockwood Lithium (which controls the EU and Asian arm of Rockwood’s Li business) at 14x TTM EBITDA less net debt if they wished so.

With the sale of approximately 2/3 of Rockwood’s business and over $3.5bn of gross proceeds received or on their way (vs 2013 ending market cap of over $5bn) you could say that 2013 has been a rather busy year, but all of these moves resulted in a clearer equity story and a better positioning for a large M&A. You could think that the story ends here but it didn’t take long for Rockwood management to shake things up and cap this remarkable story off.

Capital returns

Before we come to the end it’s worth making a quick detour to address capital returns.

Given the leverage and focus on growth, dividends and buybacks haven’t been too high on the agenda until 2012. The first dividend payment came in June 2012 when the company announced a $0.35c per share quarterly dividend ($27m per quarter), which was then subsequently raised to $0.40c per share in February 2013 and then to $0.45c per share (c. $35m per quarter) in August 2013. The dividend payout ratio was in the low to mid 30%s with a view to maintain a 2.8-3.2% yield (higher than industry average). In addition, Rockwood announced a $400m buyback in January 2013 just as it was embarking on its strategic simplification. This was completed in whole during Q3’13 and in November management announced an additional $500m, two-year buyback (this ultimately wasn’t completed). All in, shareholders received about $600m in buybacks and $300m in dividends from June 2012 through the third quarter of 2014. Furthermore, Rockwood paid down debt aggressively - from gross $3.5bn leverage in 2005 it was in net cash (pro forma for the last of asset sales) by late 2014. At the beginning Rockwood was essentially a public LBO.

A word about compensation and management shareholding. Compensation included a mix of base and performance based remuneration, however for the three key executives base compensation hasn’t gone anywhere between 2008-2013, while their performance based compensation was entirely tied to both short and long term financial - mostly cash flow - and relative/absolute share price performance goals. In a 2011 conference Mr Ghasemi commented that almost all of his net worth was tied up in Rockwood shares (both directly acquired and via options). At peak management beneficially owned close to 3% of the company, by early 2014 this was around 2% worth over $100m.

Act IV: Sale of Rockwood (2014)

Just as management was completing their strategic plan, they were on the hunt for their next move. In August 2013 Mr Ghasemi approached the CEO of Albemarle (US chemicals co) to acquire ALB, which didn’t go anywhere and Rockwood went on to focus on the acquisition of Talison. In February 2014 Mr Ghasemi approached again the CEO of ALB, now with a merger of equals plan but again this did not materialise. Following on from the two rejections Rockwood approached other PE and strategic investors. Some were interested in pursuing a transaction but not at the right price, while some PE shops only wanted the cash cow surface treatment business. In an interesting turn of events ALB approached Rockwood about an acquisition in early 2014. Mr Ghasemi indicated that shareholders would need a premium and a significant cash portion to even consider a deal.

Eventually in July 2014 the two companies announced a $6.2bn acquisition of Rockwood by Albemarle. The acquisition rationale was analogous to when Rockwood bought Dynamit Nobel’s businesses - increasing the number of platforms but not necessarily in already existing categories.

The per share consideration at the time was $85.53, paid $50.65 in cash and 0.4803 ALB stock (roughly 60/40 split) and ROC shareholders would become 30% of the combined company. The price meant a 13% premium to prior closing price. While this technically speaking is not a lot, the multiple paid was already on the high end. ALB paid a 14x forward multiple (around 11x including synergies), which is still pretty fair considering that only half of the Rockwood business is in high multiple lithium. Furthermore, as you’ll see below Rockwood’s share price significantly outperformed the market in 2013/14 given the delivery on the strategic plan hence one could argue that the then valuation already implied a high multiple. It’s actually a peculiar turn of events where Rockwood went from being a suitor to getting taken out at a premium. Management played it well.

An interesting twist before the acquisition (though probably on the cards for a while) was the resignation of Mr Ghasemi and his subsequent appointment as CEO of Air Products (Bill Ackman backed industrial gases co) in June 2014. To handle the acquisition, Mr Zatta and Mr Riordan stayed on.

As Pershing was wading through the corporate slog that APD was, they appointed Mr Ghasemi to the board in September 2013 given his experience in the industry and after a 10 month long search process the board decided that Mr Ghasemi was indeed the best candidate and elected him as chairman and CEO effective of July 2014. The industrial gases sector is a lot more concentrated than the specialty chemicals (owning to a previous industry consolidation) hence his remit is to basically restructure the business internally and focus on capital allocation. When his appointment was announced APD share price jumped 8% on the day and reached all time highs. Here is a recent video from Delivering Alpha where Bill Ackman describes how the board came to appoint Mr Ghasemi (watch from the 6:20 mark).

In Conclusion

I know that this is highly theoretical but since the listing in August 2005 through the announcement of the acquisition in July 2014 meant a compounded total return of 18% p.a., performing in line with the wider specialty chemicals sector but outperforming the S&P’s 8% p.a. return. In line performance with the sector comes from a larger drawdown during the 2008/2009 crisis. If you look at the second chart, since the end of 2008 (not even counting from the lows of March 2009) Rockwood returned 46% p.a., ahead of peers and the S&P. Since the end of 2012 (just before the announcement of the strategic dismantling of the company) the stock returned 44% p.a., again ahead of peers and the S&P.

Source: Bloomberg

If you made it this far you probably wonder why I took the time to put all of this on paper. I think that the fundamental performance of Rockwood is remarkable, considering the time period and the sectors they operated in and shows a few key lessons.

Management matters - “Outsiders” are by definition rare. While there is certainly a hindsight bias here, once you find them just sit back and hope they have a long runway ahead of them. Running Rockwood from small HQs, tucked away in Princeton allowed management to make decisions without the burden of bureaucracy and analysis-paralysis.

Disciplined capital allocation is paramount - Having clear guidelines, conviction and financial discipline for growth and/or capital returns can make or break the company. Management essentially bet the farm on lithium but remained disciplined. For instance, instead of getting into bidding war over Talison, Rockwood walked away from a deal in 2012 only to come back in 2013 when the prospects were better.

No sacred cows - While somewhat related to the above, Mr Ghasemi and team were willing to let go approx. 2/3 of their business because that was the right thing to do, without fear as to how that would impact their compensation or social standing (most CEOs like to run bigger, not better companies). It’s not about growth, but value creating growth.

Leverage works both ways - Leverage was a big part of the story and prudent financial management ultimately led to great returns on capital. From $3.5bn gross to net cash in less than ten years, without capital raise is remarkable considering the cards they were dealt at the beginning.

I’d close with two slides Mr Ghasemi presented on his first ever investor call with APD in July 2014 and at a recent investor meeting just a few weeks ago this September. The message is clear.

Source: Air Products filings