Description

Owens-Illinois Investment Thesis 

Company Background

Owen-Illinois (O-I) is the world's largest manufacturer of glass containers, producing in 22 countries and accounting for 31% of global production. The next largest global producer is Saint Gobain (SGO, ticker SGO FP) with a 20% share. The company is the number one player in each of its markets, North America, Europe, South America, and Asia Pacific.  The company sold its plastic container operations in 2007. At this time the company is segmented by geography. The following shows the breakdown is sales and operating profit by geography in 2008.

Industry Structure

The North American glass market has been highly consolidated since the mid 1990's when Saint Gobain and Ball Corporation entered into a JV to control 30% of the market behind O-I's 40%. The number three player, Anchor Glass, controls an additional 20%. Anchor is very much wedded to Budweiser with Bud accounting for almost 50% of its revenue. Anchor is a perpetual LBO. It has filed for bankruptcy protection three times in the past fifteen years (1996, 2002, and 2005) and was long considered the irrational player, chasing volume outside of its low profitability Bud contract. It last emerged from bankruptcy in 2007 with a much more manageable debt load. We will discuss the market dynamics in more depth below.

The glass packaging industry has undergone extensive consolidation over the past ten years in Europe. In the mid-1990's, the top three players in the European market controlled 50% of the market. Saint-Gobain (SGO, ticker SGO FP) had the largest share with 26% of the market. In 2002, the top three's share expanded to nearly 60%. With O-I's purchase of BSN (the combination of Danone and Gervais) in 2002, the top three now control 72% of the market. The following shows European market share by company:

The South American market is similarly consolidated with O-I and SGO combining to control 80% of the market. Asia is more fragmented, but O-I is by far and away the dominant player with a 50% market share. The number two player has a 20% share.

It is worth noting that SGO has been in the process of trying to sell its glass packaging unit for over a year now but was unable to do so before the financing environment turned.

End Markets

Glass container end markets differ by region. In the US the dominant use is for beer at 56%, and the buyers are mainly the large beer manufacturers. In Europe the end markets are more diverse, with wine constituting 31% of demand followed by beer at 21%. In turn, the European market has a more fragmented buyer base. This is an important distinction. However, O-I's mix is more towards beer in Europe than the overall breakdown. The following shows end market shares in Europe and the US.

 We do not have a good breakdown of end markets in South America and Asia. However, beer is certainly an important in South America. In that market Inbev is a growing force.

Cost Structure and Pricing

Typically energy costs (natural gas and electricity) constitute 15-25% of total production costs. In 2008 they were around 20%. Other raw materials (soda ash, sand, and lime) constitute another 25% of costs (with soda ash half that total). Transportation is 15%. Wages are 20%, SG&A just under 10%, and other the rest. O-I hedges 50% of their natural gas needs for each year and is hedged at $9 gas for 2009 (putting it at a disadvantage relative to competitors).

Pricing in North America and Asia is based on longer-term contracts with annual inflation adjustments. In Europe, pricing on 60% of volumes is renegotiated every year for roughly the start of the year. South America is hybrid of the two with more periodic inflation pass through. This contracting structure has given rise to volatility in annual earnings as rising costs in a year crimp margins with recovery in the next year.

Historically the industry chased volume at the expense of price. O-I was no exception to this trend. This dynamic was compounded in Europe by the addition of capacity by Quinn in 2003 and 2005 and BSN (bought by O-I in 2004) in 2001. Since that time, the industry has started focusing more on price. This has been led by new management at O-I. Albert Stoucken became CEO in December 2006. O-I began an extensive restructuring process in Europe where it set out to rationalize the capacity it acquired in the BSN acquisition. In 2008 they shuttered approximately 3% of capacity. SGO's desire to sell the business has also focused it on improving returns. At the same time, Anchor has a healthier balance sheet.

Current Dynamics

Through 2008, O-I shut considerable capacity in Europe. This served to tighten the market. At the same time raw material prices went north. Before the current crisis hit in full force, O-I was aiming for double digit price increases in Europe and South America to offset mid to high single digit cost increases. North American and Asian prices, as discussed before, will reset to capture most of the cost inflation. However, severe volume declines and high inventory levels have tempered those hopes. O-I's volumes were down 10% in Q4. We estimate 3% of this derived from the shutdowns they enacted last year. Another 3% was market volume declines. And 3% was inventory de-stocking. O-I ran at production down 2x its reported volume decline. The market has naturally been fixated on European pricing. It is clear that O-I is bearing the brunt of the volume declines and ceding share in Europe. However, they did work to protect share in some areas. As a result, we estimate that the double digit price increases they hoped to achieve will be closer to 4-5%. We estimate pricing in the US and Asia to be up similar amounts with double digit price increases in South America (due to higher cost inflation).

Investment Thesis

We believe an investment in O-I offers an excellent risk/reward. We expect O-I to be able to earn over $2.00 this year and into the future with significant potential for margin improvements longer term as they complete their restructuring program. The current volume environment has delayed anticipated improvement. But even so, we do not anticipate a return to sub-$2 earnings. Even if we see such degradation, the shares have little downside in our estimation. The following shows our earnings bridge from 2008 to 2009/2010.

Below are our arguments supporting our volume and pricing assumptions.

1) Volume assumptions are reasonable

For volumes, we assume Q1 down 15% with an improvement to down 8% for the rest of the year. In past recessions, volumes tend to be 3% and bottom six months before the economy. In this recession, we believe underlying volumes will be off more like 5-6% with O-I losing share in Europe to protect price. We would not be surprised to see additional closure announcements from O-I or SGO. We know SGO and Ardagh (the number three player in Europe) have both also temporarily shut some capacity. We would not be surprised to hear of some permanent closures.

The inventory correction seems to have run its course in the US in Q4. It appears volumes from brewers were down 4-5% in Q4 but sales at the retail level were only down 2-3%. Glass is losing some share to cans (as is typical in a downturn), implying underlying beer volumes are probably down 4-5% for glass bottles. Total glass container shipments were down 5% in Q4 as reported by the Census Bureau with production down over 7%. In turn, inventories were down at year end by 4%.

Europe has a more significant inventory overhang. It seems Western European volumes are suffering double digit declines. O-I reduced production by two times the underlying volume trend to reduce inventories. We don't have great insight into when inventories will be normal, but O-I claimed that it feels comfortable enough post-Q4 to run their factories in line with demand. Any stabilization here would be taken well by the market.

Inbev's Q4 report suggested that Latin American volumes are down low single digits. Asia is a mixed bag but does not seem to be as bad as Europe.

For 2010 we assume a further 3% decline in volumes. We consider this reasonable as well but acknowledge that trends could be worse if the recession deepened and extended into 2010. This outcome would far exceed the worst volume decline on record.

2) Fixed cost leverage

Our estimates imply about 50% of costs are fixed. The negative incremental margins are offset partially by cost savings from shutdowns. The low estimate on the Street for 2009 is $1.80, driven by an 8% volume decline. This estimate implies 60% fixed cost but does not seem to provide as much price as our model and does not provide the positive impact from cost savings. We believe the most likely positive offset to our assumptions would be a less dramatic volume decline through the second half of the year.

3) Pricing will not fall apart

Price for 2009 is almost completely set. We estimate that it is up 4.5% for the whole company. Market chatter about pricing pressure seems to reflect two points. First, O-I was looking for double digit increases in Europe. Weakness from that has been misinterpreted as real pricing pressure. Second, at the margin O-I did take below inflation pricing at some customers in Europe to protect share. This was offset by double digit increases elsewhere. In our view, the only remaining wildcard on earnings for this year is volume and the company's ability to deliver on cost savings. The latter are largely locked in from factory shutdowns.

The key question, then, is pricing for 2010. The market's concern is that with weak volumes we will enter into 2010 contract negotiations with expectations of price cuts. We have two responses to this. First, note that we do not assume any cost deflation in 2010 despite the fact that natural gas costs will be much lower for O-I (given 50% of 2009 gas is hedged at $9). In turn, even if prices were cut, O-I has room before they experience a real price decline. We also expect soda ash prices to roll over this year. Second, at year end 2009 O-I's contract with Bud and Miller come due. These represent significant North American volume (upwards of 50%). The current contracts with these companies are at well below market rates having been set in the 1990s. O-I is very confident it can improve margins on these contracts and in turn in North America. Note that O-I's North American operating margins were 8.4% in 2008 versus 13.7% in Europe, 16.9% in Asia, and 29.1% in South America.

Risk/Reward

We deem the risk/reward to be invested in O-I to be excellent. We base our probability weighted expected value on the following set of assumptions:

 Our probability weighted expected value is based on a multiple of 2010 earnings discounted back one year at 10%. The following shows our risk/reward.

We consider the 2010 bear case on top of the 2009 bear case to be a real bear case. It is difficult to get O-I's earnings to go to zero. Indeed, O-I averaged $1.30 in earnings from 2000-2006 when it suffered asbestos related problems and Europe had serious issues when new capacity put the market into oversupply. Since then O-I has grown and improved its operations. Furthermore, the bear case price target is putting a very trough multiple on a trough earnings scenario. We would only anticipate this to be fair if there were liquidity issues. We do not believe liquidity issues will arise as discussed below.

Other Issues

Pensions

In recent days investors have become more concerned with O-I's pension liability. In 2008 O-I's pension swung from being fully funded to being underfunded by $729 million. According to the 10-K, the company will need to inject $75 million into the plan in 2009. They do not need to make a contribution to the US plan. The plans assets are 48% in equities, 43% in debt, 6% in real estate, and 3% in other. In the case that equities and other are down 25%, real estate is down 35%, and fixed returns are up 3%, that would constitute an approximate $500 million swing in the funding status. We would not anticipate the company would need to inject the full amount into the plan. Rules in Europe are looser than for the US, centering more on the ability to pay the annual obligation out of plan assets than to have a fully funded plan. O-I's annual obligation is around $250 million. Even the US rules (which we anticipate would be under review in very dire scenarios) require 80% funding by year 2010. Indeed, at an underfunded status of $728 million at 2008 year end the company only needs to inject $75 million in 2009. Even if they needed to make a significant contribution, we don't anticipate any liquidity issues. While we do not see pension issues as debilitating to the thesis, a $500 million increase in the unfunded liability would is about a half turn on EV/EBITDAP.

Liquidity and Cash Flow

O-I does not have any significant near term maturities. The company has a $250 million 2010 debenture coming due. It has ample cash on the balance sheet to handle this. In addition, O-I has $768 million in unused liquidity under a revolver maturing in 2012. 

For the past several years O-I has been a cash flow story. During the Q4 earnings call, management disappointed investors by announcing a $120 increases in capex for 2009 to $480 million, or 1.2x D&A. Capex will include $200 million for restructuring (including severance and shutdown costs and some plant modernizations) and $40 million for expansion of their New Zealand plant. At the level of earnings we laid out above, the company will generate very little free cash flow in 2009 if they pursue these plans. In the past the company has taken capex to $280 million, and they could do it again if the need arose (as they indicated on the Q4 call). We see another weak year for free cash flow in 2010, but still anticipate roughly $250 million in free cash. With asbestos costs declining every year, we see normalized free cash flow to be greater than $400 million. On a $1.7 billion market cap, this is a significant free cash flow yield. 

Catalyst