Stocks that have it all – including risk

When something looks too good to be true, it usually is. This is the chief concern I have long harboured with my Have-It-All stock screen. While the criteria used by the screen is not all that demanding, the screen itself tries to cover a lot of bases. The screen’s shopping list of stock characteristics is broad but unimaginative, which means for me it has been a surprise just how well it has performed over the seven years I’ve run it, with a total return of 148 per cent (131 per cent after applying a notional 1 per cent annual charge) compared with 75 per cent from the FTSE All-Share.

 

Last year, however, the screen had a bit of a comeuppance. The negative total return from the screen’s picks stood at a gnarly 12.9 per cent compared with a negative 6.2 per cent from the index. Often the reasons shares tend to look cheap at a time that fundamentals look strong (the type of share this screen attempts to highlight) is that risks are higher than they appear on the surface. The negative consequences of betting the market is being overly cautious played out for several of the stocks highlighted by the screen 12 months ago, and particularly the three housebuilders.

2018 performance

Source: Thomson Datastream

As well as the potential to highlight risky stocks, another drawback of a screen with such broad criteria is that it is often difficult to find shares that score a full house. That’s the case this year, with no constituent of the FTSE All-Share passing all the tests. If I allow the shares to fail the dividend yield test (arguably this is the weaker valuation test as it is based on historic dividends) then four companies qualify. In addition to these four, I’ve allowed Sirius Real Estate through the net, as the shares are priced below book value, which is a better valuation gauge for a property company than an earnings-based multiple. Five companies still represents thin pickings for a risky screen, though, so I’ve extended the results to 18 shares by allowing shares that have failed up to two tests on to the list, as long as one of the tests failed is not the low forward price/earnings (PE) ratio test. The full screening criteria are:

■ Forecast next 12-month price/earnings (PE) ratio among the lowest third of all stocks screened.

■ Historic dividend yield in the highest third of all stocks screened.

■ Average forecast earnings per share (EPS) growth in the next two financial years of 5 per cent or more.

■ Three-year EPS compound annual growth rate (CAGR) of 10 per cent or more.

■ Three-year free cash flow CAGR of 10 per cent or more.

■ Three-year dividend CAGR of 5 per cent or more.

■ Return on equity (RoE) of 10 per cent or more.

■ Three-year average RoE of 15 per cent or more.

Once again, a number of the shares highlighted this year face risks. I’ve looked at one of the picks below, Costain, which I feel makes an interesting example of the type of risks that can help explain what superficially appears to be an extremely attractive valuation. That said, Costain has a strategy and a trading backdrop that could provide the foundations for a rerating.

Have-It-All – weakened criteria

Have-It-All – criteria weakened further

Source: S&P Capital IQ

Costain

It’s not difficult to provide reasons for the apparent cheapness of shares in infrastructure engineering and construction group Costain (COST). If investors need a reminder of the risks lurking in the balance sheets of such companies, they only need to recall the collapse of Costain's rival Carillion a year ago, or for a more recent example, Interserve’s rescue rights issue. However, the fact that there is no short interest in Costain’s shares, based on data from Castellain Capital’s short tracker, suggests the market views Costain’s business in a very different light to either Carillion or Interserve, both of which were heavily shorted before things came unstuck. That said, there are noteworthy common risks associated with any company in this sector.

For one thing, Costain’s underlying operating margins are very low, coming in at just 2.9 per cent last year (see chart). This means a relatively minor upset could prove very painful for profits and cash flows. The risks associated with wafer-thin margins are increased by the complexity of the large projects Costain works on and their high working-capital requirements. While the group has negative working capital (it owes suppliers and clients more than customers owe it), the size of working capital items are large. Indeed, big movements in working capital have contributed to very lumpy historical cash flows (see chart).

What’s more, it is necessary for Costain to base some of its working capital numbers on management judgment, which is something investors tend be less comfortable about than numbers based on actual bills. For example, only about two-fifths of Costain’s £289m 2017 year-end receivables (income not yet received as cash) represented trade receivables: work that has been invoiced for. That said, there is reassurance to be had from the fact the company has not announced any major impact from two big recent accounting changes related to revenue recognition and operating leases. What’s more, over recent years the group has maintained a fairly clean record on reported 'exceptional' items, which is in stark contrast to some of its troubled peers.

Payables (money owed to suppliers and work paid for by clients but not yet done) are also large, considerably outstripping receivables at £403m in 2017. However, payables as a percentage of sales have fallen considerably over the past five years from 28 per cent to 20 per cent. It’s of some note that broker Liberum, using data from trade publication Construction News, ranks the group as second out of a nine-strong peer group for the slowness with which it pays suppliers and for not paying invoices to agreed term. That said, the broker does not view this as a concern.

A large defined-benefit pension with liabilities of £760m is another risk that Costain has in common with many other companies in its sector. It is encouraging that Costain’s scheme is currently in surplus (assets to cover future pension payments are currently higher than what actuaries estimate is needed) but the pension obligations are strikingly high at £760m (two times market cap). The scheme is also a noteworthy drain on cash with payments of £9.6m plus dividend-linked increases are scheduled until 2031. A new triennial review is imminent.

All in all, the cheapness of Costain’s shares relative to its earnings can, at least partly, be attributed to the fact that there are reasons to perceive the quality of those earnings as low. What’s more, the high returns the screen has highlighted are largely a result of the negative working capital which, as discussed above, increases risks.

However, for those not put off by the risks of companies in this sector, the good news is that Costain’s management is focused on improving earnings quality. What’s more, industry trends look as though they should support this effort. What’s more, a 13 per cent share price jump in response to an in-line fourth quarter update earlier this month suggests pessimism about prospects may be excessive. The update also provided encouragement by reporting an increase in the year-end order book from £3.9bn to £4.1bn; a rise in potential work at preferred bidder stage from £400m to £600m; and a continued move by clients to consolidate suppliers, which is benefiting the group.

Costain expects customer demand to move away from major capital projects towards technology-led asset optimisation. The view has been supported by the announced spending plans of its two biggest customers: Highways England and National Rail. Both customers are served by Costain’s infrastructure division, which accounted for 80 per cent of 2017 sales. Encouragement about the direction of travel can also be taken from recent contract wins in tech-focused areas such as connected autonomous vehicles.

To satisfy the growth in demand for 'smart' infrastructure, Costain now has about a third of its 4,000 staff in technology and consultancy roles, as well as 40 PHDs. It is hoped the move away from big capital projects to tech-focused work will both raise margins and make them more predictable. The predictability of cash flow should also be improved. While fewer major capital projects led to a 12 per cent fall in first-half revenue, underlying operating profit rose by 8 per cent.

On top of this progress at the infrastructure division, first-half profitability was helped by the ongoing margin recovery at the natural resources division. The division does work for the power, water and oil and gas industries and accounted for a fifth of 2017 turnover. The interim margin rose from a meagre 0.1 per cent to 2.6 per cent. The margin target for both divisions is 4 to 5 per cent, which infrastructure is expected to achieve this year.

*Forecast margins for 2018, 2019 and 2020

Given the working capital and pension position, comfort can be taken from Costain’s an average end-of-month 2018 net cash balance of £77m. While the average net cash figure is well down on the £97m reported for 2017, this was swollen by the timing of some large payments and average net cash in 2016 was £69m. Finances could be aided by non-core asset sales: Costain owns two golf courses and a marina in a Spanish resort valued at £26m.

Reassurance can also be taken from the group’s relative insulation from political and economic risk. As a supplier to infrastructure companies, a Corbyn government may actually be good for Costain due to the potential for higher infrastructure spending. It also does not own any private finance initiative (PFI) assets. What’s more, customer demand should be supported by the start of new five-year regulatory spending periods for roads, rail and water, all of which should underpin sustained or increased spending. These are due to get under way both this year and next.

There are grounds to hope that the changes being made at Costain will cause investors to reassess the quality of its earnings and consequently rerate the shares, especially from the depressed levels they’ve fallen to over the past year. That said, there are plenty of risks that are likely to keep the rating at a relatively low altitude.