A Short Thesis on Precision Castparts: Why We Believe Margins Will Decline

Jan.25.11 | About: Precision Castparts (PCP)


We are short Precision Castparts Corp. (NYSE:PCP) and have spent the better part of a year developing and debating our views. Our initial thesis focused on earnings quality and was profiled in last year’s May 28th edition of Grant’s Interest Rate Observer. PCP management spent a good portion of their June 2010 quarter call (see here) attempting to refute our thesis. It seems like most of Wall Street accepted management’s response; but to us it only raised more questions. We are writing this report to expand on and defend our initial earnings quality thesis, as well as to detail competitive pressures that have recently developed in PCP’s business that are being ignored by investors. Management, on the other hand, doesn’t seem to be ignoring these pressures: they have been selling PCP stock at record levels since the Grant’s article.

PCP is a leading manufacturer of complex metal forgings, fasteners, and investment castings for critical aerospace and industrial applications. It is a high quality business with dominant positions in some segments of its markets. The company has a history of superior earnings growth and was the fourth best performing stock in the S&P 500 during the last decade, up 1580%. This track record has created a cult-like following around the stock and its long-time CEO, Mark Donegan.[1]

More recently, PCP’s stock is up over 140% since the beginning of 2009, giving the company a market cap of $20 billion. Some of this outperformance is due to investor enthusiasm about the rebounding commercial aerospace market. But a large part of it can be attributed to PCP’s remarkable ability to attain record-high operating margins of 27% in the teeth of the recession.

Most companies experience margin deleveraging when their sales decline as their fixed-costs get spread over fewer units of production. The margins of PCP’s competitors reflect this phenomenon:

Figure 1: Operating Margins for Ladish Co. (NASDAQ:LDSH), Allegheny Technologies (NYSE:ATI), Precision Castparts (PCP), LISI SA (FII.FP), and Vallourec (VK.FP) [2]

PCP’s margins, in contrast, have expanded since the onset of the recession, and risen to a level far above that of their peers. Management claims that this performance in the face of a +20% collapse in sales has been due to sustainable productivity improvements, and that 2009-2010 margins are a new base to work from. Accordingly, analysts are forecasting record margins in the next two years, as sales rebound along with their end markets.

In our judgment, PCP’s operating margin performance the past two years is a mirage, achieved by aggressive accounting and one-time benefits. We estimate that various earnings quality issues have artificially and/or temporarily inflated margins by at least 450 bps. The primary items are:

  1. Revenue/cost mismatch from the collapse in metal prices
  2. LIFO accounting/bloated balance sheet, and
  3. Aggressive acquisition accounting that has lowered depreciation and amortization.

As these benefits run out, we expect margins to come under pressure, and to be further impacted by intensifying pricing pressure in their product markets. Meanwhile, management has papered over these issues with misleading statements about the true health of the business.

Consensus estimates have PCP increasing its earnings run rate by 20% over the next twelve months, and then at a 10-15% annual rate thereafter. We forecast earnings will decline by over 10% from the current run rate over the next two years despite a rebound in sales; our FY2013 (March 2013) earnings estimate is 35% below consensus. PCP trades at 22x trailing twelve-month earnings, which we think makes the stock grossly overvalued. We see downside of 45% if our earnings estimate is correct, even without much in the way of multiple compression.

Earnings Quality Issues

Earnings quality issue #1: Revenue/Cost mismatch (200 bps benefit)

Metal purchases are very important to PCP, representing 30-40% of their total costs in normal periods. In order to avoid getting squeezed, PCP attempts to adjust their finished goods prices by employing escalators that vary with increases or decreases in the relevant metal price compared to a preset reference price. PCP’s metal strategy has taken on greater importance in recent years owing to extreme price volatility, as seen below with nickel, PCP’s most important input.

Figure 2: Spot Price of Nickel ($/lb)

As metal prices collapsed throughout 2008 and 2009, PCP was buying metal at prices far lower than before. PCP contends that this reduction in their metal costs has had little impact on their profitability because they pass on the benefit of lower metal prices to their customers via reductions in their escalators and/or lower finished goods prices.[3] PCP has repeatedly assuaged investors about the effectiveness of their metal hedging strategy. In their June 2010 quarter earnings call, when PCP attempted to address some of our concerns, management said that “substantially all” of their contracts today have escalators. Shawn Hagel, the CFO concluded:

We are not a metal spectator—excuse me, speculator.

This characterization of their metal price sensitivity is correct in normal times. But in the context of the recent extremely volatile period—where nickel dropped 75% peak to trough—we think their characterization is wrong. Whereas in their public comments management uses the phrase “substantially all”, their latest 10-K says:

We have escalation clauses for nickel and other metals in certain of our long-term contracts with major customers, and we employ “price-in-effect” metal pricing in our alloy production businesses to lock-in the current cost of metal.

They only have escalation clauses in certain of their long-term contracts with major customers. This leaves open: minor customers, shorter-term contracts, and long-term major customer contracts where they don’t have escalation clauses. In our mind, this implies that a meaningful part of their business is impacted by a change in metal prices. Indeed, they seem to agree when they say later in the 10-K:

"We have escalation clauses for nickel, titanium and other metals in a number of our long-term contracts with major customers, but we are not usually able to fully offset the effects of changes in raw material costs"

Below we make two observations that might explain the above 10-K language.

  1. We contend that a significant portion of their sales still utilize fixed-price contracts that have no escalators. In other words, PCP takes the metal price risk for the duration of the contract, which can be up to five years in length. This was the norm in the industry before the commodity price boom and, based on our industry discussions as well as PCP’s 10-K disclosures, this approach is still common today. PCP’s Fastener division in particular has had minimal impact from escalators, meaning that prices are fixed.[4] When these contracts are renegotiated, both PCP and their customer consider changes in the relevant metal price and incorporate this into the new contractual price. These types of contracts obviously hold great potential for a revenue/cost mismatch. Since metal prices collapsed in 2008, this mismatch has been a big positive for PCP, as they have been buying metal at prices far lower than the metal price they charge their customers. Considering the long duration of some of these contracts, we think it is reasonable to believe that this mismatch was a significant margin tailwind from late 2008 to early 2010.
  2. PCP often uses price bands for triggering the escalators. The way this works is that PCP takes on the metal risk within a band around a previously agreed upon price. If there is a big price move above or below this band, then the escalator comes into play to start laying-off additional metal price risk to the customer. In the declining metal price environment of 2008-2009, PCP must have been benefiting from metal prices veering towards the low end or piercing these bands—in this scenario PCP is paying lower prices for metal than what they are effectively charging the customer.

Estimating the impact on margins from the use of price bands is impossible without more transparency from the company. However, estimating the impact from the use of fixed-price contracts is feasible. In our estimation, this revenue/cost mismatch has boosted operating profits by around $240 million in the past two years, helping margins by 200 bps during this period.[5] This estimate comes with a lot of imprecision and is based on multiple assumptions that could be incorrect. But we think it is a conservative estimate for the revenue/cost mismatch because we are ignoring the benefits of price bands.

It is important to note that these profits are real and have emerged regardless of the accounting policy (LIFO vs. FIFO) employed. But these profits are not sustainable, and investors do not realize this.

Earnings quality issue #2: LIFO accounting/bloated balance sheet (140 bps benefit)

In addition to the use of escalators, PCP cites their use of forward metal purchases as a key component of their hedging strategy to match up revenues and costs. From our perspective, however, these forward purchases have not served as hedges but have instead been used as levers with which to inflate earnings.

On behalf of their customers, and particularly with long lead-time, multi-year orders, PCP will lock-in future metal costs by purchasing it forward for delivery up to two years in the future.[6] The use of forward purchases means that PCP entered the recession with commitments to buy metal at high prices for many months after prices collapsed.

This should not necessarily lead to a mismatch for PCP since they can pass on the high-cost metal to their customers via escalators when they ship the finished good. But it is our contention that PCP has been able to fill these orders by buying low-cost nickel or revert (scrap metal) in the spot market and pocketing the difference.

We believe that much of this high-cost metal from the forward commitments never actually ran the through the income statement—rather it has been building up on the inventory line-item on their balance sheet. Instead of actually utilizing this forward purchased metal when they received it, PCP went out and bought more metal on the spot market, and it is that lower-cost metal that has flowed through the income statement. The stale, high-cost inventory they received has stayed on their balance sheet, as evidenced by the spike in inventory and DSI:

Figure 3: PCP Inventory ($ million) & DSI

While competitors maximized cash flows by reducing inventory and ordering less metal, PCP did the opposite—they built inventory in the face of a +20% sales decline thereby inflating reported earnings.

PCP values inventory on a LIFO (last-in, first-out) basis, which has ensured that the lower-cost raw materials they bought (after the bust) were “last-in” and therefore “first-out” on the income statement. Because PCP has kept purchasing and producing more each quarter than they have sold, they have been able to avoid dipping into and expensing inventory that is carried at much higher cost. If PCP used FIFO (first-in, first-out) accounting they would have been forced to expense the old high-cost inventory first, and their profitability would have been far lower.

We can confirm that PCP did indeed benefit from this accounting maneuver by looking at the growth in the LIFO reserve, which shows the cumulative benefit PCP got from LIFO accounting vs. FIFO accounting:

Figure 4: PCP LIFO Reserve ($ million)

Since September 2008, the LIFO reserve has risen by $166 million as PCP was afforded the opportunity to buy excess metal at low prices. This means that operating profits have been $166 million higher during this time due to the use of LIFO instead of FIFO. This has added 140 bps to operating margin during this period, which is in addition to the aforementioned 200 bps from the drop in metal prices.

Our interpretation of the increase in inventory is supported by the apparent evolution in the CEO’s thinking. Back on the March 2009 quarter call (see here) Donegan said that he was going to liquidate the company’s bloated inventory:

But I think there's opportunity that we are going to extract out of this business. I think we've got inventories that can come out and again we're going to do that [next quarter]. Some of the loss volume I'm talking about is we're going to try to keep extracting that inventory out.

He never did; as we know, inventory continued to rise. On the June 2010 quarter call (here), Donegan did an about-face and committed to keeping inventory flat as revenues rebound over the next few years. We question why he isn’t liquidating this bloated inventory as he said he would. Is he worried that this inventory is indeed old and high-cost and would hurt profitability?

Well, yes, he even told everyone this on the March 2010 quarter call (here):

Donegan: “Number one, I think we understand our LIFO layers abundantly clear, so I think we know at what point we can liquidate material at a certain [sic]. The value of the schedules going up, from a LIFO standpoint, as long as I hold my total inventory dollars flat, I'm good. I can, so again, deploying that same inventory level across a higher sales base lets me not take any negative LIFO hit and at the same time lets me get better leverages of working capital as a percent of sales. And then to that, as materials may rise, we know the strike point so we may choose at some point in time when they cross that threshold when there's no impact to make additional reductions in that environment. Did I say, is there anything there that you're un—“

Hagel (CFO): “No, that's exactly correct.”

Donegan: “So we have that type of clarity to it.”

In our mind these comments confirm that Donegan is afraid of liquidating inventory because he knows it will hurt their profits. We also think this is confirmation that PCP manages their earnings and exploits LIFO accounting.

In our judgment, Donegan has turned PCP into a metal speculator by managing LIFO layers. He is betting that at some point in the future metal prices will recover back to boom-time prices, and then he can start drawing down inventory without a “negative LIFO hit”. Donegan may indeed be successful. But that doesn’t negate the fact that margins have been unsustainably inflated. In the meantime, PCP operates with a bloated balance sheet as they add to inventory in the pools where avoidance of old high-cost inventory is most advantageous.

Earnings quality issue #3: Aggressive acquisition accounting (100 bps benefit)

We find PCP’s allocation of acquired intangible assets to be highly unusual, and estimate that it could be providing approximately 70 bps of margin benefit versus peers. Intangible assets in an acquisition result from the difference between the purchase price of an acquisition and the acquired company’s net tangible assets. The acquiring company must make assumptions regarding how much of the intangibles are amortizable, and these amortizable intangibles negatively impact the income statement in future periods. The remaining intangibles remain on the balance sheet as either non-amortizable intangibles or goodwill and do not negatively impact operating income. The lower the allocation to amortizable intangibles, the better future earnings appear to be.

PCP’s has shown a remarkable ability to avoid these amortizable intangibles, as they only represent 3% of PCP’s total gross intangibles. This compares to peers where amortizable intangibles range from 24% to 46% of total gross intangibles. We calculate that if PCP had used these peer’s assumptions regarding amortizable intangibles, margins would be around 70 bps lower.[7]

We also question whether PCP is being overly aggressive by under-valuing Property Plant & Equipment (PP&E) of acquired companies, thereby minimizing depreciation. Special Metals Corporations (SMC) is a useful case study in PCP’s acquisition accounting since it was a public company before entering bankruptcy and being purchased by PCP. We can therefore compare PCP’s assumptions regarding PP&E to the financials of the company prior to the purchase.

SMC’s PP&E turnover (Sales/PP&E) when acquired by PCP in mid-2006 was 8.8x, whereas it never rose above 3.1x in the three years leading up to SMC’s bankruptcy in late 2002. Additionally, 8.8x PP&E turnover is exceedingly high when compared to peers, who tend to be around the 3x level that SMC exhibited prior to bankruptcy. PCP valued SMC’s PP&E at $166 million when they acquired it in 2006, a decline from the $217 million it reported in its last quarter as a public company, despite the fact that sales nearly tripled since 2002.

We estimate that the number of SMC’s manufacturing sites more than doubled in 2007 to 21 from the 10 that SMC cited in their last 10-K in 2001, so a decline in PP&E is quite unusual.[8] Valuing SMC’s PP&E at this low level meant that SMC’s depreciation was 35% of the level it would have been if they had assumed PP&E turnover consistent with the 3.1x peak seen by SMC prior to bankruptcy. This helped SMC’s margins by an estimated 130 bps, not inconsequential considering SMC generated approximately 25-30% of PCP’s revenues the past two years

We estimate that the lower PP&E valuation on this acquisition alone was still helping PCP’s margins by at least 30 bps in the past two years. This is on top of the estimated 70 bps benefit from the intangible assumptions, making the total acquisition benefit around 100 bps.

Quantification of total benefits from earnings quality issues

In total, we estimate that the margins in the 2009-2010 period were inflated by around 450 bps, or 18% of operating income, due to the earnings boosts listed above:

The benefits listed in the table above are only those that we attempted to quantify. They therefore represent a minimum benefit from poor earnings quality. In addition to these benefits, we detail in Appendix A the following earnings quality issues:

  • Overproduction to lower fixed cost overhead per unit (level loading)
  • Discontinued operation selling to continuing operation
  • Pass-through restatements
  • Unnecessary pension contributions

These issues, while largely difficult to quantify, add to the overall picture of a company working hard to mask true operational performance.

Management’s rebuttal

As we mentioned in the introduction, PCP’s management spent a good portion of the June 2010 quarter call attempting to refute our thesis as summarized in Grant’s. We believe they only succeeded in raising more questions. For example, they introduced the term “inventory income” and quantified it at $20 million for the fiscal year ending March 2010. In the Q&A session an analyst, understandably, asked what this term actually means:

Joe Nadol (JPMorgan-Analyst): “So, Mark, and thanks for the additional info. Of course, on the inventory income, that was the words—those were the words you used, Shawn. How do you define that, the $20 million?”

Hagel (CFO): “That's really looking at the net benefit generated out of our inventory balances in the year that went to the P&L. So, it includes FCA [sic?], would include LCM adjustments, E&O, would include LIFO adjustments to that which would include all the productivity changes, material cost, that kind of thing. We looked at it all in. That's what we do when we put together the information for today’s quarters. We look at that total storyline and make sure that we're giving you any of the material movements.”

Does that clear things up? It doesn’t for us and apparently it doesn’t for the CFO either. Later in the call she said that they can’t actually quantify their inventory income on a quarterly basis—because it’s too complex! From the June 2010 quarter call:

Joe Nadol (JPMorgan-Analyst): “The way you're defining it, the way you're looking at it, what was the inventory income in Q1?”

Hagel: “We're still working on that. Because again, it's such a complex—I can't give you an exact number, but what I can tell you is, that there was not a lot of change in the quarter. The productivity improvements were pretty flat, as you can see, except for in our investment cast products group. Materials were relatively flat. So, really the storyline that we gave you which shows improvement going through investment cast is really where you are going to see that, what I would call inventory income generated.”

How can management be “giving [us] any material movements” in inventory income each quarter if they can’t calculate it themselves? Even if management could calculate their inventory income on a quarterly basis, we still don’t trust their “storyline”[9] because we don’t trust their definition of “material”.

We say this because of a dialogue the company had with the SEC back in 2007. At the time, the company was arguing that they didn’t need to give pro forma numbers for their SMC acquisition because it was not material—despite SMC accounting for one-third of PCP’s sales! From the SEC correspondence of March 20, 2007:

We read that you have not provided pro forma information concerning your acquisition of SMC because this acquisition did not have a material impact on your results of operations. We also assume that you concluded that this acquisition was not “significant” under Article 3-05 of Regulation S-X since you did not file a Form 8-K containing historical and pro forma financial statements related to your acquisition of SMC. However, your conclusions concerning this acquisition are unclear to us since your October 1, 2006 MD&A and your press release for your December 31, 2006 results indicate that the acquisition of SMC was one of the primary drivers for your improved results in fiscal 2007. In this regard, we also read on page 23 of your April 2, 2006 Form 10-K that SMC has annual sales in excess of $1.0 billion, which we note is approximately one-third of your annual sales.

One final note, the company wants us to trust them that they are calculating the income impact of reserves such as E&O correctly, yet they have manipulated these in the past in order to boost profit (see company restatement described in "Managing for the short term" section of this report).

Rather than disproving our initial thesis, the June 2010 quarter call proved to us that PCP’s cost accounting is subject to numerous, complex adjustments that management has discretion over. Management confirmed that they cannot quantify their cost drivers on a real-time basis. Yet investors should be comfortable that management is in control of “inventory income” and will tell us when it is material—even though their definition of material is highly questionable.

We stand by our assertion that PCP’s margins have benefited from unsustainable, low-quality earnings boosts and that today’s margins should not be relied upon as a base from which to grow.

Competitive Pressures

PCP’s product pricing reflects an assumption for the price of metal, as well as supply/demand dynamics. We have previously discussed how PCP benefited from metal price declines on fixed-price contracts. In this section we focus on how PCP’s pricing additionally benefited from strong supply/demand dynamics on old contracts in backlog that are now expiring. PCP is a long-cycle business where contracts can last several years.

This means that PCP’s prices post-recession to some extent still reflect pre-recession conditions—when capacity was tight and product margins were attractive. But demand has since decreased while competition has reacted to PCP’s blowout margins by attacking PCP’s businesses with more capacity. In our judgment, customers also must have recognized that PCP’s margins are unacceptably high and are demanding price concessions at renewal.

PCP’s backlog as of March 2010[10] was down 28% year-on-year despite sales being down only 10%, which indicates that they have been eating through their old high-margin backlog. As a result, pressures on the business are finally starting to impact results. Margins have declined sequentially in each of the past five quarters, yet management is misleading investors about the true cause of the weakness. We note two examples.

Competitive pressure #1: Seamless Pipe

Seamless Pipe is large-diameter pipe designed primarily for coal power plants. In FY2009 and FY2010 it was around 9% of sales and 18% of EBIT. Recently, it has been the main factor behind PCP’s disappointing earnings, as pipe sales dropped 50% sequentially in the June 2010 quarter and stayed around that depressed level (50% down year-on-year in the September 2010 quarter). On the June and September 2010 quarter calls (see here), Donegan claimed the Seamless Pipe weakness was due to a shift from less efficient 300 Megawatt (MW) and 600MW power plants to 1,000MW plants.

We have researched this issue and have not found any corroboration of PCP’s story. We couldn’t find anyone else in the industry who is talking about a dramatic shift to 1,000MW plants—it has been very gradual over multiple years. Vallourec, PCP’s primary competitor in Seamless Pipe, cites a more credible reason: increased competition from Chinese players.

From Vallourec’s 2009 Annual Report (see here - pdf):

The Group noted increased competition in this [Power Generation] sector in 2009, in particular in the Chinese market, linked to some customers’ decisions to reduce their technical requirements and give preference to some local manufacturers that have upgraded their range.

This competition took a while to manifest in Vallourec’s sales because of long lead-times; they said in their June 2010 quarter earnings call that sales were still benefiting from high-margin orders struck in 2008 at the peak. They warned on that call that they had worked through these old orders, and indeed in the September 2010 quarter their Seamless Pipe sales were down 45% year-on-year, similar to PCP’s 50% decline. Vallourec again blamed the decline on the “shift in project activity to the competitive Asian markets” and said that “price and product mix of sales for the conventional power market is expected to reach a low point by the end of the year and then stabilize”.

We think these competitive pressures are the best explanation for the weakness in PCP’s Seamless Pipe business. This is a clear example of PCP’s high margins (roughly 50% operating margins in Seamless Pipe) attracting competition and encouraging customers to switch over to lower priced competitors. It also shows how long-term contracts can shield PCP from feeling the impact of this increased competition for multiple years—it took until mid-2010 for economic weakness to hit this business—but that eventually it does impact results.

On the September 2010 quarter call Donegan said that they were seeing “solid growth” in the Seamless Pipe backlog as this issue resolves itself, and this is one reason why investors anticipate 20% overall earnings growth next year. Fueled by Donegan’s misleading characterization of the decline, PCP investors are awaiting an upturn that we believe will not materialize.

Competitive pressure #2: Fasteners

Fasteners are 25% of PCP’s sales and 29% of segment income. In the last two quarters, management has cited this business unit as the other temporary factor (in addition to Seamless Pipe) driving the margin weakness at the company. However, management’s explanation makes no sense.

On the June 2010 quarter call, Donegan was asked about the weakness in aerospace fasteners. He responded by blaming the weakness on an unnamed distributor that was consolidating the market within the past year:

If you look at a year ago, we roughly had three different distributors that we were supplying the same product into. They've been consolidated into one and that consolidated distributor now is working through the excess inventory.

Again on the September 2010 quarter call he cited distributor consolidation as the source of weakness for the Fastener business, but this time he named the distributor—B/E Aerospace, the largest aerospace fastener distributor.

From the September 2010 quarter call:

Donegan: “What's lagging and continuing to lag, is that consolidation where we we're supplying the same product to an M&M, a B/E, and a Honeywell, that was all out there in the chain and that all has to be consolidated before the new combined group under B/E starts ordering the same product again.”

Donegan’s statement is false. B/E did not buy any fastener businesses in the year leading up to this statement. They bought Honeywell’s fastener business in July 2008, and the M&M deal closed in September 2001! We find it incredible that Donegan would say that an acquisition that occurred nine years ago actually happened within the past year. And it is ludicrous to think that B/E is destocking today because of a deal that happened in 2001.

B/E’s finished goods inventory trend is further confirmation that Donegan’s explanation makes no sense. The table below shows that inventory understandably dropped in the December 2009 quarter as B/E fastener sales weakened, but this was a one-time drop and since then B/E has been restocking fasteners, not destocking them. Yet Donegan’s excuse has been accepted by unquestioning analysts.

We think expiring contracts are the real reason for the weakness in Fasteners. PCP benefited from a fastener shortage in 2007[11] and it is likely that they struck favorable long-term contracts at that time. In February 2008, Donegan said the following[12]:

The one thing we've made sure we've done, or we focused on, is taking our base business and locking it up. So if you look at what we have, we have long-term positions established through 2010, in most cases beyond that now.

These lucrative multi-year contracts are rolling off and being replaced by contracts where pricing reflects a more normal balance of supply and demand. We think this will continue to be a headwind for PCP for quarters to come.

Competitive pressures summary

The parallels between Fasteners and Seamless Pipe are striking. We believe both businesses are finally showing the impact from increased competition as long-term contracts struck during boom times expire. And in both instances management’s explanation for the weakness is wrong and gives unquestioning analysts and investors false hope that the business will rebound. We expect further margin weakness as PCP’s favorable pricing on long-term contracts expire. These are the first two examples to crop up—we also expect the same pattern (more competition and pricing pressure upon contract renewal) to eventually show up in the rest of their businesses as contracts roll off.

Managing for the short term

PCP’s short-term focus explains a lot, and makes the recent record insider selling interesting. We know that PCP focuses on the short term—CEO Mark Donegan is famous for it. In a glowing 2008 article, titled “Zero Tolerance”, Forbes wrote[13],

[Donegan’s] system, evolved over years, starts with the fundamental requirement that a plant must cut its costs by 2% every quarter.

Imagine the pressure on a manager who is forced to constantly show 2% savings every quarter. It’s asking the impossible. The short-term focus doesn’t even stop there—Donegan has instituted daily P&Ls[14] throughout the organization to make sure that the pressure to reduce costs is felt every minute. Donegan is well known for castigating plant managers who don’t meet his short-term performance targets. He calls this “in-your-face” management.

Here is Donegan speaking at a Credit Suisse conference in December 2009:[15]

I think we're a different breed. I don't know if that's good or bad. So I'm not about to say one's good or bad, but we're kind of a blue-collar, in-your-face, slug-it-out, down-in-the-trenches type of company. And I take great pride in that. The fact that we put the energy and the effort—you've heard us talk about the quarterly review process—it ain't easy… it's not a pleasant thing. I wish I could tell you, oh, yes, it's great, we walk in and we shake hands, everybody tries hard, and away we go. It is a gut-wrenching process.

This extraordinary pressure to produce short-term results has led to problems in the past. In late 2005, PCP was forced to delay their 10-Q for the September 2005 quarter due to accounting issues. We quote from the December 2005 10-Q (emphasis added):

The investigation at J&L Fiber Services, Inc. [a subsidiary of PCP] identified errors in the Company’s fiscal 2003 through fiscal 2006 financial statements. These errors, which totaled $5.6 million, represented a net understatement of expenses through the second quarter of fiscal 2006. The errors primarily related to the overstatement of inventories resulting from the failure to properly record excess and obsolete reserves and inadequate cycle counting procedures, the failure to properly reconcile accounts, and the misapplication of GAAP and Company accounting policies, which included improper deferral of expenses. The identified errors resulted from both deliberate and inadvertent acts involving non-executive employees.

It’s not surprising that employees would deliberately overstate inventories to understate expenses—we think it is symptomatic of the culture of short-termism and excessive pressure to deliver results at PCP. Note also the reference to “the failure to properly record excess and obsolete reserves”. These are the same “E&O” reserves that the CFO said go into the mystical “inventory income” number that they can’t quantify on a quarterly basis but claim is not material.[16]

Donegan also had another incentive to manage for the short term and mislead about long-term pressures on the business. Since the beginning of May 2009 he has sold over 540,000 shares for nearly $62 million of gross proceeds. Fifteen different executives sold shares in the five weeks after PCP’s latest September quarter earnings report, the most insider-selling since our records began in 1986. We think this may be a sign that insiders realize there are no more earnings quality tricks left to counter the competitive pressures PCP is seeing.

Conclusion: We see 40%+ downside to the stock

PCP margins in the June 2009 quarter reached a phenomenal 27%. Donegan credited their internal cost-cutting efforts for the spike in margins, and said that they were a base from which to grow. From the June 2009 quarter call (here):

We will—I think that it was nothing abnormal in the margins this quarter. They were solid operating performances. So, yes, are we going to face a little headwind from outages and downages in some volume? Sure, we are. And there is just no overcoming that. But are these margins to build off of as we look beyond that? Yes, I think they are margins to build off of.

PCP has missed margin estimates for five straight quarters[17] since that statement (though they are still well above pre-recession levels):

Yet management has convinced analysts that these misses are due to short-term factors that will reverse; indeed the stock is up 50% since they started missing estimates. Consensus still believes that margins will only go up from here to attain a record level. Revenues are also expected to be buoyed by the prospect for aerospace-growth fueled by the Boeing 787, though we note that large commercial OE aerospace is only approximately 25-30% of PCP’s business.

This optimism has resulted in analysts modeling annualized earnings growth in the high-teens over the next ten quarters. We think we have provided enough evidence in the previous pages to show that margins over the last two years should not be believed, and neither should management commentary about a rebound. We have analyzed and attempted to quantify four earnings boosts that we estimate have benefited recent margins by at least 450 bps. The unwinding of these earnings boosts, plus new competitive pressures, means that PCP’s margins are set to decline rather than expand as consensus expects.

We look back to PCP’s operating margins of 18.2% posted in the FY2007 as the last true and representative gauge of PCPs operating margin potential, and note this is still higher than PCP’s long-term operating margin target of 15-17% communicated in 2005. Since 2005 they did buy SMC and assuming that they were able to raise SMC’s margins to corporate average (a big assumption) then there should have been perhaps 180 bps of margin benefit from vertical integration. Eighteen-percent margins make sense to us as management’s adjusted target for peak margins. It also foots with where competitors averaged out pre-recession.

Since the recession, each of PCP’s competitors has predictably seen margin declines from 2007 levels, and PCP is experiencing increased competition and pricing pressure, which suggests PCP’s margins should be below the 18% level posted in FY2007. But we will give PCP the benefit of the doubt for driving some productivity improvements. We settle at an expected margin level of 18% in FY2013 when all the contracts struck in good times roll off and one-time earnings boosts recede.

We see FY2012 margins at 21% as PCP transitions from expected FY2011 margins of 24.3% to 18% in FY2013. This yields an EPS estimate of $6.70 for FY2012 and $6.35 for FY2013 versus consensus estimates of $8.53 and $9.84, respectively. Over the past five years the stock has traded at 14x next twelve-months earnings, so we generously assume a 13x multiple on FY2013 EPS, which gets us to a price target of $82, or 45% downside. We think the stock reaches this target as repeated earnings misses cause investors to question the true sustainable margins of the company.

Appendix A - Other Earnings Quality Issues

Earnings quality issue #4: Level loading

PCP also benefited during the downturn from “level loading” their plants, meaning they kept producing more than demand warranted in order to improve margins. Overproduction meant that they could allocate fixed-cost overhead over more units. On the March 2009 quarter call, Donegan said the following:

I'm not going to build inventory to keep people busy. I think we need to respond to that. So I would expect to see cash generation still holding up strong in Q1 and Q2.

Then a year later on the March 2010 quarter call (here) Donegan admitted to level loading when explaining why inventory continued to rise:

We had an opportunity to buy revert [scrap metal] when it was low, so we made sure we went out and maxed-out on that as much as we could. That was kind of across nickel and titanium. Those are probably the primary drivers. We level loaded some of our operations, we level loaded them kind of figuring out that we couldn't handle some of the slopes that were going to come.

This means they kept people busy, didn’t cut production, and created excess inventory. This had the effect of temporarily boosting short-term profitability. Note also the reference in the quote above to the maxing-out on as much low-cost revert [scrap metal] purchases as possibleit’s another clear example of PCP benefiting in the short term from binging on metal.

Earnings quality issue #5: Discontinued operation selling to continuing operation

In the March 2008 quarter PCP classified Kladno, a money-losing alloy manufacturing business located in the Czech Republic, to discontinued operations. This operation “primarily supplied steel ingots” to Wyman-Gordon, another PCP forging operation.[18]

Essentially, this means that a discontinued operation (which is ignored by investors) is selling to a continuing operation of PCP. In the FY2008 10-K, PCP noted they had decided to “close” this business. However, beginning with the FY2009 10-K, PCP began to note that they only planned to “sell” the business, implying that Kladno remained open. Indeed, as of the March 2010 10-K (two-years later), this business remains in discontinued operations. It is still listed as an active location on a PCP website and, therefore, we believe it continues to supply raw material to Wyman-Gordon.

We question why this business has remained in discontinued operations for so long—could it be that it is selling materials at a loss or at a low margin to the continuing operations of PCP, thereby flattering operating margins? We don’t know how significant this could be because PCP doesn’t disclose inter-company sales for discontinued operations like Kladno.

Earnings quality issue #6: Pass-through restatements

PCP says that they have contracts with customers where they can charge “pass-throughs” of raw material costs—these are the “escalators” we discussed previously. Management refers to these as “zero margin” because they are intended to simply pass through higher raw material costs to customers. They cite pass-throughs as a key way that they avoid exposure to commodity price fluctuations.

For the last three quarters, PCP has quietly restated historical pass-through numbers in their Forged Products business. For example, in the September 2009 10-Q they said Forged Products pass-throughs were $57.3 million. In the September 2010 10-Q they say that September 2009 quarter pass-throughs were approximately $42 million. Yet sales, costs, and operating profit for the September 2009 quarter are the same in both the September 2010 and September 2009 10-Qs.

If pass-throughs are charged to the customer, how did they change retroactively? And how did this reduction in pass-throughs not impact sales and profit if indeed pass-throughs recover the increased commodity costs they incurred? We conclude that pass-throughs are not in fact “zero margin”, otherwise costs would have also gone down by an equal amount in the restatement.

Earnings quality issue #7: Unnecessary pension contributions

PCP made large and unnecessary pension contributions in FY2010 and FY2011 that ‘earn’ (purely in a GAAP sense) 7.5-8.0% pre-tax rates of return. The less underfunded a pension plan is, the more it helps earnings, especially given the current low interest rate environment. In the FY2009 10-K, they freely admit a reason for the FY2010 contribution was a desire to improve accounting profit (emphasis added):

In the first month of fiscal 2010, we contributed $190.6 million to our defined benefit pension plans, including a voluntary contribution of $188.9 million, to increase plan funding and to mitigate the impact of these pension investment losses on our income in future years.

Appendix B – Calculation of Earnings Quality Issues #1 and #3a

Quantification of earnings quality issue #1 (Revenue/cost mismatch)

The following table (click to enlarge) details calculations made to arrive at PCP’s cumulative earnings benefit from revenue/cost mismatch. We make the following assumptions:

  • 30% of PCP’s business employs fixed-price contracts
  • When nickel is at $10/pound, metals represent 25% of the total cost of production on this business, which is lower than company average because this revenue is skewed towards lower metal content products
  • Metal costs for this 30% slice of revenues vary with spot nickel prices with a four-month lag
  • Average contract duration is four years

Click to enlarge

Quantification of earnings quality issue #3a (benefiting from intangible allocations):

The following table (click to enlarge) shows what PCP’s amortization would be using allocation assumptions of various peers:

Click to enlarge


[1] Of the 21 analysts covering PCP there are 16 buy ratings and no sell ratings. The average U.S. stock has about one buy rating for every four analysts that cover it.

[2]PCP’s fiscal year ends in March, but for ease of comparison all of our charts are based on calendar year-end.

[3] One technical note: If PCP is able to pass-through 100% of metal price changes, a drop in metal prices would increase margins since sales decline even as profit stays the same. By our calculations, fixed-cost deleveraging during the sales decline should easily overwhelm this impact, as seen by peer results.

[4] PCP reports by division the revenue impact from escalators in the line items “pass-throughs”. In the latest two quarters, pass-throughs were only $5 million for Fasteners, or less than 1% of Fasteners sales. Fastener metal content (at around 20% of costs), while still significant, is lower than that of the other divisions. Fasteners are 28% of PCP’s overall profit.

[5] See Appendix B for details.

[6] Note that these long-term forward hedges on behalf of customers represent the minority of their sales. For the most part, PCP forward purchases on its own and for relatively short durations (typically 3-6 months).

[7] See Appendix B for detail. Using the assumptions of PCP’s international peers would result in an even greater margin difference, but we are hesitant to include them as a reference point due to differences in international accounting standards.

[8] Twenty-one sites from the following links: here and page 24 here (pdf)

[9] We find it interesting that in both the preceding quotes from the CFO, she uses the word “storyline”, a word that is usually reserved for “the plot of a book or play or film” according to TheFreeDictionary.com.

[10] They only disclose this at fiscal year-ends, so we don’t know what has happened in the six months to September 2010. Additionally, the gap between backlog growth and sales growth would be even higher adjusting for the Carlton Forge acquisition impact.

[11] See here

[12] February 6, 2008: Precision Castparts at Cowen and Company Aerospace/Defense Conference

[13] Forbes, January 28, 2008 (here)

[14] Daily P&L by department mentioned in RBC analyst report “Cleveland Rocks!: Casting the Upcycle” September 14, 2010. Also implied in Forbes article above.

[15] Credit Suisse Group Aerospace & Defense Conference, New York, December 2, 2009

[16] See quote including E&O in "Management's rebuttal" section of this report.

[17] Margin estimates based on average of published models

[18] PCP’s FY2008 10-K, page 25. Note also that Kladno was one of two businesses at the center of the 2005 accounting investigation detailed in the "Managing for the short term" section of this report.

Disclosure: Temujin is short Precision Castparts common stock (PCP) and stands to profit if the stock price falls.

Disclaimer: This report is intended for discussion purposes only. It is neither a recommendation nor a solicitation to purchase or sell any security or private fund offered by Temujin. The analysis and opinions of Temujin presented in this report are based on publicly available information sourced from SEC filings, earnings call transcripts, and other public sources that we believe to be reliable. Judgments have been made based on information sourced on or before the issuance of this report. There may be information concerning Precision Castparts that is not publicly available, which could lead the company to disagree with our analysis and opinions. Temujin expressly disclaims all liability for errors or omissions in, or the misuse or misinterpretation of, any information contained in this report.