Description

?? Balfour's business model is highly dependent on leverage and off balance sheet vehicles. Its earnings and cash flows have been inflated by strategies similar to those used by Citigroup (SIVs), Macquarie / Babcock & Brown (infrastructure funds), and many other financial institutions during the credit boom.

- Balfour is believed to have a strong balance sheet, but its consolidated leverage is 8x net debt / EBITDA. It has generated >1/3 of its earnings through off balance sheet vehicles, which have a leverage of 36x debt / EBIT 

Balfour's opaque disclosure makes it hard to see the true leverage in the consolidated company. Its financials are setup so one focuses on the equity investments through a separate DCF valuation done by the company, equity income below the line on the income statement, and the limited debt on the balance sheet. However, in its notes Balfour discloses £1.4bn worth of JV and associates debt, and if one adds up all of its commitments, one reaches a whopping £1.9bn worth of senior debt (excluding subordinated debt since that is not disclosed). Even on a basic level one can see how its returns and valuation excluding off balance sheet entities portray a completely different company than the conventional view, as shown below.

While the off balance sheet debt is non-recourse, it is a prerequisite to generate these earnings, and that is why its infrastructure peers only talk about consolidated leverage-Ferrovial is net cash excluding nonrecourse debt, despite being widely considered a distressed company. Furthermore, given the political nature of many of these projects (hospitals, schools, etc), it is common for contractors to inject more cash into insolvent projects rather than endure reputational damages.

Furthermore, Balfour has made several acquisitions in the last few years. In the bubble period of 2006-1H08 it spent £829mm on acquisitions. It has only made minor writedowns to goodwill so far. In fact it increased the valuation on its 1H08 investments at its recent annual results. The goodwill has grown aggressively, eclipsing the book value such that Balfour's tangible equity is now negative £337mm!

- These off balance sheet vehicles, primarily Public Private Partnerships ("PPP's") and Private Finance Initiative ("PFI's"), also serve as captive customers for Balfour's core construction and support services businesses. PPP/PFI's generate an estimated ~30% of Balfour's core revenues, ~45% of operating profit, and an even larger proportion of its free cash flows-its margins are significantly higher than non-captive work and working capital terms are the most advantageous in the industry.

An example of a hospital project (used to) goes something like this. A typical £200mm project used to be funded with 90% debt / 10% equity, with the contractor taking 50% of equity, therefore putting in only £10mm (more on the financial investor's 50% of equity later). In return, the contractor gets a £200mm construction contract over 2-3 years from this SPV, with margins of ~3% that it doesn't have to bid for since the SPV is a "captive" customer-usually the bidding adds 1% of the cost. Then the contractor also gets a 25-30 year facilities management (FM) contract (also with no bidding) of £15mm per year with 4.5% margins (which they keep even if they flip their equity investment later). Therefore, the £10mm equity investment is a minor part of the overall economics of the project, and not that dissimilar from the costs of bidding for the construction/facilities management contracts of a traditional ("non-captive") project. In a non-captive contract, typical margins were ~1% for construction and 2.5-3% for FM (confirmed by plugging in these assumptions into Balfour's model). Therefore, while PFI's represent an estimated 30% of Balfour's revenues, they account for a significantly higher percentage of EBIT.  Lastly, the working capital terms for PFI's are by far the most advantageous of any construction project-according to global peers, and can be seen by Balfour's cash conversion cycle relative to peers. Therefore, as a percentage of Balfour's free cash flow, PFI's are >>50% of the core business (not to mention the bulk of the valuation of its investments, more later). As such, Balfour's business model is extremely dependent on the PFI pipeline.

- PFI/PPP's are part of an entire supply chain dependent on leverage (SPV's, customers, lenders, investors, subcontractors), which makes the pipeline unsustainable.

UK PFI's are the most highly leveraged infrastructure asset, with the debt mix of Balfour's portfolio at 87% of total capital. Private equity investors commonly leverage an additional ~2/3 of their capital at the fund level, meaning that the amount of equity put into many of these projects was ~3%. During the credit bubble, these investments yielded >20% IRR's due to multiple refinancing and flipping of assets to increasingly leveraged players. Currently the financing mix, debt terms (duration, spread, refinancing clauses), and fund level debt availability have reversed dramatically to more reasonable levels. As such, some large players are exiting the business, such as Trillium.

At the same time, the leading providers of PFI debt during the credit boom are under extreme distress. The top 10 banks in 2007 included HBOS (#1), RBS (#2), Dexia (#3), Barclays, Depfa, Lloyds, and SocGen (see chart below). In 2H08 and especially 4Q08 most of them either dropped in the ranking significantly or were simply not present (only Barclays lent to a UK PFI in 4Q08). Even healthier banks such as HSBC have announced they are pulling out of the PFI debt market. In a screen for UK and Scandinavian PFI's, 4Q08 vs 4Q07 deal value dropped by 93%.  Japanese banks, such as Sumitomo and Mizuho, were expected to pick up some of the slack, but are now facing their own capital troubles. In fact, the upcoming M25 financial close date was pushed to April so it falls in the next Japanese fiscal year, highlighting the hope that Japanese banks would become major PFI players.

Moving to the other end of the chain, the customers are also getting squeezed by the credit crunch. Stepping back, the main benefit of the PFI scheme has been to transfer government liabilities off balance sheet, since it merely commits to paying a 30-year annual fee to the SPV rather than procuring the entire project directly with debt. The UK is by far the country with highest PFI/PPP penetration, increasingly relying on PFI's such that it represented 15% of government expenses since 1996-these off balance sheet liabilities now represent ~12% of GDP.

Even excluding this, the UK's public sector net debt / GDP has surpassed the 40% ceiling target, it is bailing out financial institutions whose balance sheets are multiples of GDP, and its unfunded pension liabilities are >2x GDP. Reflecting the UK's dire finances, 5 year CDS have increased 10x in the last six months. As a result, while the government is desperate to prevent the pipeline from collapsing before the elections in early 2010, there have been several delays recently, and it seems to be merely a matter of time before this happens. The February 2009 National Audit Office report summarizes the current environment:

Ultimately, all of these forces will result in a dramatic reduction in work for Balfour. As Skanska acknowledged on a recent call (2-6-09), the cost for the government has "increased quite substantially over the last 12 months" due to the cost of finance, cost for guarantees, etc, which is delaying PFI financing processes. This assumes that the PFI model survives the new credit reality, because otherwise the government would have to procure projects directly and put the debt on its balance sheet, which is practically unfeasible. Below, a table by Ernst & Young shows the drastic increase in PPP/PFI debt spreads--currently they are even higher at 250bps according to Infrastructure Journal.

Combined with the severe budgetary constraints of the UK government, this will ultimately result in significantly less deal flow, which will hurt the volume players (contractors) the most. Alternatively, one can interpret this to mean that the current backlog is severely overstated due to the indirect impact of the credit bubble.

- Furthermore, the UK/US private non-residential sector (offices, shopping malls, stadiums, hotels, etc) represents 20% of Balfour's revenues and Dubai accounts for ~7% of EBITA. Both of these end markets are now collapsing.

20% of Balfour's 2007/2008 revenues come from the UK/US private non-residential sector, building offices, shopping malls, stadiums, hotels, etc. Disney, for example, is a significant US customer that has announced significant capex cuts.  New projects in Dubai have come to a complete halt, including projects in Balfour's backlog such as the Waterfront Dubai and the Trump tower. Estimates as of February 2009 suggest that $582bn worth of UAE projects (52% of civil development projects) have been suspended.

Balfour's other "defensive" end customers include US states/proxies (company is focused on FL, CA, TX, and NC) and Dubai (~7% of 2007 revenues, but higher margin). FL and CA are two of the states in the US with the worse budget problems-Florida's CDS has gone from 50bps to >200bps in the last 6 months while California's CDS has gone from <100bps to >300bps.

?? Equity investment valuations in Balfour's PFI/PPP's are critical to the bull thesis.

PFI/PPP's alone represent ~25% of consensus valuations for Balfour. PFI/PPP valuations are highly sensitive to the cost and availability of credit, and had been inflated by private equity investors that added another layer of leverage at the fund level-they are now retreating from the asset class. The market for these levered assets has been reset, impairing the company's investment portfolio.

As described earlier, private equity and infrastructure investors were adding another layer of leverage to an already highly leveraged asset to get to their target returns. With that capital structure, deals in 2007/1H08 generally implied a ~5.5% discount rate. There haven't been many recent deals, but investors estimate the current discount rate is closer to 9%. With several forced sellers in the market (notably HBOS, which owns the largest portfolio of PFI's), it is likely that deal comps will be substantially below company's NAV assumptions. In its 2007 Annual Report, Balfour said that a 1% change in discount rates would impact its PFI portfolio valuation by £65mm (4% of total market cap)-even though the difference between the value of its 6% and 8.1% discount rate scenarios was £167mm.

In addition, HSBC Infrastructure's recent trading statement stated that several macro variables have reduced the valuation of its PFI portfolio, such as comparable debt spreads, collapsing inflation expectations, and deposit rates for cash.

?? Valuation: a typical bull argument is that the stock is very cheap, because excluding its investment portfolio and cash the P/E is <5x. With the conservative assumptions below, the stock is trading at a 14x adjusted P/E, or a roughly 100% premium to its peers.

- Investment portfolio: since the PFI market has reset as discussed earlier, and with most infrastructure stocks trading at fractions of their published NAV's, it is reasonable to assume at least a 25% discount to Balfour's estimate (several analysts still use significant premiums to Balfour's PFI NAV and book value for 2007/1H08 acquisitions).

- Earnings: assuming Balfour's PFI earnings converge to terms of outside work, assume a 15% volume decline for this segment, and a 50% decline for Dubai and UK/US private non-residential work (in line with market assumptions), earnings would be significantly below estimates.

Catalyst

?? Downturns inflict a vicious blow to cash flows in Balfour's cyclical core construction businesses (due to negative operating leverage, increased competition, and working capital deterioration)

- Balfour's cash position is weaker than it seems, even excluding off balance sheet liabilities, if one adjusts for its pension deficit, preferred shares, prepayments, and investment commitments.

It is widely believed that Balfour has a strong balance sheet. On a superficial level, it had an average of £250mm in cash during 2H08. However, after adjusting for its pension deficit (£309mm as of 1H08, with 42% in equities as of 2007), preferred shares, prepayments, and investment commitments, Balfour is actually in a weak position as it heads into a downturn. (This excludes the massive off balance sheet debt discussed earlier.)

- Sales will decline and margins will shrink due to increased competition and operating leverage

Although the book to bill as of 1H08 remains >1, it has continuously decreased to just over 1. Since Balfour is a late cyclical, it is hard to tell how much sales will decline in 2010, but given the earlier qualitative comments, it could be significant. Furthermore, as volumes fall, competition will increase, and margins should naturally contract. One example of this logic is how struggling regional contractors, currently locked out from PFI's, are lobbying the government to allow them to participate in these projects. This makes political sense since these companies employ a lot of people and changes may be announced in April.

Also, operating leverage is a major issue for contractors. Despite having relatively low fixed costs, contractors have thin margins so a decline in sales can have a huge impact on earnings. In the sample analysis below, I conservatively assume that only 10% of the construction division's costs and overhead are fixed. However, a 10% decline in sales leads to a 25% decline in net income.

- Severe working capital deterioration is common in downturns; especially likely in a credit crunch

Most construction companies operate with negative working capital, such that when they grow, their cash balance increases. That is mostly due to prepayments, since they get paid before doing each step of a given project. As such, when sales decline, prepayments will decline and so will the cash balance-that's why one excludes the prepayment cash balance to calculate a contractor's excess cash position. However, prepayments also increased as a percentage of construction sales during the market upturn. With customers now more focused on liquidity, this ratio should deteriorate back to the levels of the last downturn, burning through £125mm in my estimates.

Besides the natural decline in cash as prepayments and sales fall, the cash conversion cycle has also expanded significantly in the upturn and should come back to the levels seen earlier in the decade. This happens in a downturn as customers seek to delay payments (increasing accounts receivable days) and troubled subcontractors need cash (decreasing accounts payable days). During the credit bubble, large / prime contractors cheaply outsourced many risks to subcontractors, but the hidden leverage in the supply chain usually inverts in recessions leading to inferior working capital terms.  Credit insurance for UK subcontractors has been severely reduced recently, making this risk even more evident. Shrinking the spread between accounts receivable and accounts payable will also have a significant impact on Balfour's cash balance.

- Timing is opportune since sentiment and expectations are still overly optimistic with vocal bullish analyst coverage and low short interest. The stock price is within 10% of buy-rated analyst targets and is up 10% YTD after having rallied 47% from its October low through year end-partially on stimulus hype, despite peers in GS's US stimulus basket being down 25% YTD after rallying 57% from their November lows into year-end.

The construction industry is lobbying hard for a government PFI fund to bail out major projects that have been stalled due to financing issues, which may be announced soon. This is merely a vindication of this thesis, since the fund can only be a temporary buffer for a few projects until the government revamps the industry to adjust it to a sustainable form.

- Looking into 2010 things will get worse, with late cycle dynamics and UK macro issues kicking in.  This recent article discusses the issues well: