(Sept 30): What you don’t know can hurt you, so it’s understandable that analysts have raised questions about Glencore’s opaque commodity-trading business. Without the ability to work out how it makes money, the unit is a “black box,” according to Morningstar’s David Wang.

But there’s more to fear from what’s in plain sight. Capital spending has been the biggest single drain on Glencore’s cash flow in every year since it started the takeover of Xstrata in 2012:

Even after announcing a US$10 billion ($14.3 billion) debt-reduction plan earlier this month, the company still expects to spend US$10 billion to US$10.5 billion by 2017 on capital works at its mines, farms, oil wells and processing facilities.

These aren’t the best tier-1 assets favored by BHP Billiton and Rio Tinto, Bernstein analysts led by Paul Gait wrote in a note to clients Tuesday, even if they’re still profitable right now.

That means that they’re more vulnerable to any further downturn in commodity prices, and in the meantime are throwing off less cash to help Chief Executive Officer Ivan Glasenberg reduce debt.

Glencore and, in particular, Xstrata rode the commodities boom by buying unloved assets just as a turn in the market during the 2000s was about to send them surging in value.

These takeovers didn’t always come cheap. While Glencore spent a median 1.4 times book value on the 20 deals for which valuation data are available, the 10 comparable pre-takeover transactions made by Xstrata came at 4.2 times book, according to data compiled by Bloomberg. That’s more than double the 2 times multiple in more than 3,500 deals across the mining, energy and agriculture sectors, separate data show.

The wave of deals also turned the new Glencore into a quite different business from the highly liquid, asset-light operation founded in 1974 by Marc Rich. Rich borrowed money from his father and father-in-law and got a partner to sell his vintage car to cobble together his 2 million Swiss francs of seed capital, according to the 2010 biography “The King of Oil.”

Using letters of credit secured against cargoes of oil and materials, the business could always settle its debts in a hurry. Moreover, by depending on volatility in commodity prices rather than the quality of any underlying assets, it could prosper in bear as well as bull markets.

Selling a tanker full of oil is easy; getting rid of a colliery is harder. Rio Tinto tried to sell some of its Australian and New Zealand aluminum operations back in 2011. Almost four years later, they’re still for sale, the Financial Times reported in May.

People think Glencore’s problem is its mysterious trading business. The more boring but pressing issue is its mediocre mining business. The company took on too much debt to buy assets which the market is now revaluing downwards.

That issue won’t go away in a hurry. Even after its shares rallied 17% in London yesterday on the back of a five-sentence statement from the company reaffirming its solvency and liquidity, debt investors drove credit-default swaps to their highest level since 2009:

Listing on the public markets has been a wild ride for Glencore, delivering Glasenberg a $9.3 billion paper fortune in the company’s 2011 initial public offering and reversing 37 years of Swiss secrecy.

Citigroup analyst Heath Jansen suggests that Glasenberg may now be better off taking the company private again. While he’s at it, Glasenberg should sell those illiquid industrial assets, and keep his hands on the black box.