Three years on and he has succeeded in transforming the business. On 1 July this year, he will hand over to Jean-Sébastien Jacques, currently in charge of the copper and coal part of Rio. But Mr Walsh was also charged with preserving and increasing shareholder value and headwinds have made that a tough call. The share price has almost halved while he's been in office. That's hurt shareholders. And it's hit him in the pocket, too.
Tragedy of the commons
Demand for iron ore has continued to grow. Logically, producers would maximise their profits if they worked together to supply the market. But that's not going to happen, so instead they act individually. And their scale is staggering. Investments such as Rio's 15 technically sophisticated mines, 1,700 km rail network (equivalent to the length of Britain and then some) and four port terminals at Pilbara have taken years. The problem is that these are long-term investments and others, like BHP Billiton, had the same idea. The result is that both are now high-volume, low-cost producers in a world of excessive supply that has driven down prices. They are caught in a bind of their own making.
So what should Rio do? Having made the investments, its best game plan is to keep churning out the ore: 73 per cent of its underlying earnings still depend on it (down from 90 per cent when Mr Walsh took over). But should it invest further now to reduce production costs in a few years' time or mothball investments until it sees markets recovering? Mr Walsh's answer has been to focus on Rio's most profitable projects. He's cut back Rio's capital expenditure by three-quarters since 2012 but still plans to spend $4bn in 2016.
But iron ore suffers a double whammy - its end market, steel, faces similar issues of excessive supply. As producers with the cushion of state cash (as in China, which produces half the world's steel) drive others out, demand for iron ore is likely to falter. This is hardly likely to help Rio, where Mr Walsh's strategy is a balance between hunting for short-term value while laying down the groundwork for long-term returns in the hope that rivals will fold and the cycle will turn, even if it only gets back to a 'new normal'. Andrew Forrest of Fortescue Metals Group, a competitor, dubbed this the "last man standing" strategy.
Paying for performance
This is the context in which Mr Walsh managed down Rio. And it's fair to say that managing contraction presents more of a challenge than managing growth - it's all very well talking about cost efficiencies, but in practice they herald the behavioural consequences of wage freezes, business closures, redundancies and demotivated workforces, all of which need to be delicately managed. At Rio, how well executives do this counts towards 30 per cent of their annual bonus. The scorecard for these individual targets includes "business transformation, cost reduction, performance delivery and leadership and engagement". Another 20 per cent is linked to safety, in proportion to the number of injuries each year. In 2015, injuries were cut from almost three per million hours worked in 2014 to just over two and Mr Walsh's safety bonus was doubled. But four people died and so the safety bonus was halved back to the full 20 per cent. The other half of the bonus depends on financial targets: based on earnings and free cash flow for the year.
These performance conditions mostly depend on the discretionary judgements of Mr Walsh's fellow directors. The financial targets they've set expect earnings to slip and cash generation to fall - Mr Walsh is effectively being paid to make the best of a bad job. They must think he's done well - and in fairness, many shareholders would agree - for although his target annual bonus is about £1.2m, he's tended to earn closer to the maximum of £2m. (He's paid in Australian dollars; these figures assume 1.90 dollars to the pound.)
Making the best of a bad job also applies to his long-term performance, which is assessed over five years. The initial face value of this annual share award is about £4.5m and he's expected to end up with half of it. This depends on three things: how Rio's total shareholder return (share price including reinvested dividends) compares to that of other miners (the HSBC Global Mining Index); how it compares to other global companies (the Morgan Stanley Capital World Index (MSCI); and by how much Rio's margin improves relative to its global mining comparators. So two-thirds of this award depends on how Rio's margin or share price compares to other mining companies - in the slide to the bottom, as long as Rio's metrics fall by less than its peers', the award will be paid.
When he leaves, his outstanding share awards will continue but will be pro-rated. Curiously, though, he is also entitled to about £1.5m for, would you believe, long service leave and unused annual leave (Australian rules apparently).
But forget these carefully crafted performance conditions. Another, less obvious, one illustrates the financial perils of managing a cyclical organisation.
The reckoning
Share prices matter enormously in this sort of pay structure because share awards are defined in shares, not currency. Half Mr Walsh's bonus is paid in shares that are held back for three years and he receives a long-term share award. In 2015, he received 29,609 bonus shares plus 138,010 long-term shares, of which he is expected to receive half (69,005) when the performance condition is tested in 2020. The fall of Rio's share price since these were awarded has reduced their current value by about £1m. This is about equivalent to the cash that he took home last year (after deducting the cost of housing him, pension costs and tax from his gross £2.9m cash pay).
Another condition of being Rio's chief executive is that he has to own Rio's shares equivalent to four times his salary; on 1 January 2015, his 146,993 shares fell in value by about £1.5m. Worse, these were worth 5.2 times his salary; by the end of the year, maturing share plans had boosted his ownership to 194,780 but these were now only worth 4.2 times his salary.
More than a dent in his total expected pay package worth about £6m a year maybe, but it highlights how heavily geared executive pay can be to the share price. Mr Walsh's share awards have to be held for several years, of course, so there's time for the share price to recover. Just as well, since it will take some time yet for Rio to end up as one of the last men standing.
Sharing risks
Critics of high pay say that this gearing turns pay into a lottery, especially since many share price movements are beyond the executives' control. They dislike the windfall gains that higher share prices can bring but the same goes for windfall losses, for they can mean that directors have little control over what they are really paying their people. They also argue that paying too much in shares pushes up top pay levels because executives tend to migrate towards growing companies. This obliges other companies to pay more to recruit and retain people.
Many investors see this differently: they welcome skin in the game. Confidence in the top management influences the share price so it makes sense to hold these executives financially accountable by gearing their pay with shares. If executives and shareholders experience similar, meaningful, risks, that in itself is a safeguard. Mr Walsh and his team might have done a good job but, at the end of the day, long-term investors are worse off than they were a year ago and it's not unreasonable for executives' fortunes to be aligned with those of shareholders.
This symmetry of risk also applies to the chopping of Rio's dividend. Its 2015 dividend was $2.15; a minimum of $1.10 is promised for 2016 and there are no guarantees after that. Core dividends will in future be more sensibly linked to earnings per share (and presumably their cyclical low point if that can be identified); additional dividends are promised in upturns assuming the new normal allows it. One shareholder likely to receive about £150,000 less in dividends this year is Mr Walsh himself.
