Paragon is leaking cash

One of the things that attracts me to Castings is that listen to the doomsayers, it oughtn't to exist. It's a British manufacturer of iron castings that are subsequently machined into parts, like differentials and steering knuckles, for trucks, tractors, and cars mostly. Roughly sixty per-cent of its sales are exported to Sweden (20%) and other European countries (40%) supplying companies like Scania, DAF and BMW from foundries (run by two companies, Castings, and Wm Lee) and factories (its subsidiary CNC Speedwell does machining) in England.

We're not supposed to be good at this kind of thing any more.

For a long-established industrial company, it is saddled with surprisingly few of the usual burdens. It owns its workshops and warehouses, so no operating leases are hidden off the balance sheet. And although it operates a defined benefit pension scheme (closed in 2009), it's ''6m in surplus (also not disclosed on the balance sheet). Castings has no debt, and finances its operations and capital expenditure from its own deep reserves under the guidance of a chairman who has been with the business since 1960 and owns a significant chunk of it.

I've held it in the Thrifty 30 for two years, and its been a good servant, earning a return of about 50% including dividends.

I thought the shares were cheap in 2009 because the automotive industry was in disarray. Since Castings was a financial fortress and had decades of unbroken profitability behind it, I figured it would emerge from the crisis stronger, and that is exactly what seems to have happened. It's half year results last week indicated that it is operating at or beyond pre-recession levels, although the company reports early indications of a downturn in the commercial vehicle market.

The shares now trade at around 1.5 times book value and although it's still possible the market underrates Castings, I'm not as confident it does.

My doubts centre on capital expenditure as well as price. One of the reasons Castings competes with foundries and machine shops in Eastern Europe is investment in the latest machinery. This investment is a cost that is depreciated, or charged, to the income statement, and therefore reduces profit over the period the machinery is useful. Whenever the company buys machinery or, as it did, in 2010 invests in a new foundry, the money comes out of its cash account in great lumps though.

This means that in any one year the company's fee cash flow (net cash from operations minus net capital expenditure) can be radically different from its net, or accounting, profit. In the long term, I reckon the two should even out (roughly).

Judging by the six years of data I have from the company's annual reports, this is not the case:

Even through the recession, Castings has earned a healthy median 13% return on equity, but free cash flow return on equity was just 4%.

Over the whole period, total depreciation was about ''20m, while total capital expenditure was about ''55m.

Maybe the company is spending more money than it needs to maintain production by replacing worn-out plant and machinery. Strictly speaking we should only count 'maintenance capital expenditure' in the free cash flow calculation but it's not possible to tell from the cash flow statement when the company is maintaining capacity and when its spending money to increase it. The chairman's statement is also vague although it seems quite likely the new foundry at William Lee (cost ''16m) and some of the heavy investment at CNC Speedwell, is new investment.

An alternative explanation is the company is not depreciating its plant and machinery by enough each year, but I'm going to give it the benefit of the doubt. Castings is run prudently in every other respect, and since the chairman is a major shareholder, he'd be kidding himself.

Another cash leakage during the period flowed as a series of exceptional payments into the company's pension fund. Since it now has a healthy surplus, I'm working on the assumption the leak is plugged.

I'm going to keep the shares in the portfolio because new investment should lead to higher returns in the future, but as a hard-bitten value investor this is a difficult thing for me to do because it increases the speculative element.

The alternative, though, is to discard a company that in every other respect is a paragon.