Risk, uncertainty and Forge

The sage has long advised embracing risk. Machiavelli, for example, counselled that ‘nothing great was ever achieved without danger’. For the investor, this is equally true. If we agree that the future is unknowable (a sensible but not universal starting point), investing will always involve accepting a degree of risk. Yet accepting risk is not the same as accepting a gamble.

In his excellent book, The Signal and the Noise, Nate Silver makes a distinction between risk and uncertainty. Risk, says Silver, can be measured – it is a probability that one outcome from a range of foreseeable outcomes will occur. Uncertainty is present when probabilities are immeasurable and possible outcomes unknown.

In the famously obtuse parlance of Donald Rumsfeld, risk is a known unknown; uncertainty an unknown unknown. For the right price, investors should embrace risk to their comfort and abhor uncertainty. The problem with Forge Group is that the risk is unquantifiable and the uncertainty manifold.

Uncertain business

It isn’t just the stunning volatility that investors have endured in recent months – the company’s share price went from over $6 to 29 cents following contract losses before rebounding back to $1.50. This is a business that is inherently uncertain.

Forge makes money by completing contracts on which it makes a margin. Most of these contracts are competitively tendered and the company must compete to win. Plenty of competition and chronic cyclicality make it tough but Forge must also price contracts without information about their true cost. As the recent debacle illustrates (see Forge and the fine art of stock picking on 29 Nov 13 (Under Review - $0.68)), that can cause dire problems.

Forge’s profitability reflects aggregated profits on individual contracts. Yet investors (and often management themselves) can’t know how profitable each contract will be. Worse still, a failed contract can have a negative value as it is delivered regardless of cost. Potential liabilities are enormous if things go wrong, making the business not merely risky but uncertain.

Profits on individual contracts are extremely variable; competitors may sacrifice margins for work; true costs may vary from estimated costs; industry activity may slow, as it recently has. Few hazards are in the company’s control.

Forge can control the contracts it wins but, much like an insurance business, we must rely on management to write sensible contracts and forego poor ones. Some businesses in the industry – Monadelphous for example – have established a fine record of doing just that. Forge has not.

Damage done

Problems identified in the power division have shaken confidence in the company and we must approach it with more suspicion than peers with better records.

According to management, problems first arose in early September, and yet the company did not enter a trading halt until November, two months later. That’s a long time to keep investors in the dark. A class action lawsuit is already underway. We don’t know if other contracts will prove problematic but it’s hard to imagine how the company blundered two contracts and sailed breezily through the rest.

One explanation for the troubles may be in the company’s formation. For most of its history, Forge was merely a holding company for four different businesses: Cimeco (construction), Webb (concrete, pipes, tanks), Abesque Engineering (design, project management) and CTEC (power plants). Although each business was vetted and acquired by Forge, contracts were written and won by previous managers. That Forge management themselves erred when making acquisitions demonstrates that, even for insiders, the sector is a minefield.

Valuation woes

Valuing such a business is fraught with difficulty. Contractors are afforded discretion in how they recognise profits and problems in the power division suggest past profits have been overstated. We should treat earnings multiples with suspicion. Today’s PER of two is not necessarily a sign of value because it reflects overstated profits that won’t be replicated.

On the balance sheet, the company has an equity value of $213m. At first glance it appears Forge trades at a reasonable discount to book value. But assets include $40m of intangibles likely to be written off, $90m in cash that has already been depleted and $150m in work in progress, which deserves trimming. This is a poor quality balance sheet.

The recent confirmation of an $800m contract win at the long awaited Roy Hill iron ore project might be worth about 40 cents per share today if all goes well. Yet with so many unknown unknowns, we can’t reliably say how much the business can earn in aggregate. If it were a matter of mere risk we could weigh it against price but that is not what confronts us. This is a business beset by uncertainty. No one knows the probability of the Roy Hill project succeeding, whether costs will meet expectations or if any more contract problems will emerge.

If we were to buy such a business it would be for liquidation value only. That would demand a discount to net tangible assets. After making adjustments to cash, intangible assets and inventories, we estimate NTA for Forge to be around 40 cents per share compared with the current price of $1.37.

That is not to suggest the company is a basket case; it is not. It has been profitable for years by winning contracts building facilities and power plants. If industry conditions change so will the fortunes of Forge. Yet it is heavily reliant on a small number of large projects – it's top three customers last year accounted for 50% of revenue and the Roy Hill project alone will account for almost all activity for the next two years –and industry expenditure has collapsed.   

Perhaps there are no more mishaps and perhaps industry demand booms again. In that case today's share price might appear cheap and our caution overdone. We could say it’s not worth the risk but that would be a misrepresentation. We simply cannot estimate what the risk might be. Forge is not worth the uncertainty. With the share price up 101% since 29 Nov 13, AVOID.