It took exactly five months but the strike by the Association of Mineworkers and Construction Union (Amcu) in Rustenburg against the main mining companies appears to finally be over, with workers set to return today.

We caution, however, that the level of increase promotes further the chances of restructuring by the companies in the next three years while securing some unfavourable precedents in the labour space.

Shorter term, we also caution that it may not be until into the start of the fourth quarter before production fully manages to ramp up again to “normal” levels.

In the past five months, the 70 000 employees on strike have lost R10.6 billion, while the companies have lost about R24bn, according to figures provided by the joint company task force.

Overall, we estimate about 0.7 percent of gross domestic product (GDP) in nominal terms from full-year GDP was lost by the strike, though in real terms this would have been slightly less at 0.5 percent.

The final wage agreement, however, in headline terms, does not seem dramatically dissimilar to what we understand was on the table in March. Zooming into the details, however, there were clearly significant climbdowns required on both sides to reach this final settlement.

The key details of the deal are still emerging, but appear as follows:

- R1 000 a month, or an 8 percent increase (whichever is greater), for all workers backdated to July 2013. This equates to around a 20 percent increase for the lowest-paid workers in the first year. A further R1 000 a month increase in the second year and R950 a month in the third year;

- A back-to-work package will include food and medical care, as well as a certain amount of backpay. This will be the difference between new and old salaries between July 2013 and January 2014. It is not clear at this stage if any small additional repayment for the time on strike will be made;

- About 236 laid-off staff will be reinstated;

- There will apparently be a no-strike clause for the three years of the current wage agreement by Amcu, which is less than the original five years asked by the companies; and

- The main climbdown for Amcu was to shift from the R12 500 monthly basic pay in two years (then three) requirement to now broadly members reaching the R12 500 monthly salary in three years, but only in terms of overall compensation, including other benefits.

The backpay may be particularly expensive, depending on its exact terms. We should also remember that the producers will not return to normal operations for maybe three months from now given the time it takes to restart operations that are dormant and then produce and process platinum ready for market.

As such, the inherent costs of the strike, on top of the back-to-work package which will be payable upfront, will continue for some time. Equally the macro data will only slowly reflect a turnaround, especially considering the lags in such data.

Overall, we see an industry that while not happy to accept such increases, is now in a position to more actively consider restructuring as the settlement is so far about productivity levels.

This varies from company to company, but is a mixture, in our view, of shifting operations to other sites (Zimbabwe or the Eastern Limb), selling off the most unprofitable shafts and making some productivity investments to cut the labour-intensive nature of operations (though there may be attrition or even some lay-offs). The involvement of the minister will likely embolden firms to make such moves.

Why so long?

Why it took so long to reach this point and why the resolution was seemingly agreed upon so quickly in the past week are complex and something of a mystery. Our own take on it is the following:

- Strong inventories and a higher platinum price sustained the companies for a long period, but into this month their cash flow situation started to look tighter;

- Significant levels of intimidation by a minority of Amcu members combined with many workers’ return home to their feeder provinces, as well as charity and Amcu assistance funds (together with a health-care provision from the mining firms) kept workers on strike despite no income. Equally, the normally very high proportion of their salaries given up to loan repayments meant the difference in disposable income was smaller than at first sight. Sentiment against the National Union of Mineworkers also helped, while the previous precedent (repeated in some way with the current agreement) of backpay for the strike time also helped swing the incentives to sustain the strike;

- Overall, a lack of meaningful communications and relatively few meetings between Amcu and the companies during the early and middle period of the strike sustained it;

- We dismiss some of the commentary that there have been other, sinister, forces sustaining the strike. However, other factors, such as animosity between ANC and Amcu leadership, certainly didn’t help, nor did the previous Department of Mineral Resources leadership’s total lack of engagement with the companies or Amcu;

- Engagement with the new mining minister earlier this month produced some positive sentiment on both sides to get an agreement in principal on the outline of a wage settlement but no fundamental shift in negotiation position on the details, in our view. However, Amcu made it clear that the ANC could not take any credit, hence the minister’s withdrawal; and

- Equally, the power structures within Amcu shifted to allow its leader, Joseph Mathunjwa, to call a rally and gain support for a return to work despite significant climbdowns. This then accelerated the ability of both sides to meet, after accounting for additional feedback from Amcu shop stewards over the past week.

Arguably, all sides should learn that the precedents of this strike can mean a shorter cycle of Amcu strikes in the future and less disruption to the economy.

Amcu appears to have come out of this strike with its support base intact, though our understanding is that the disquiet that was growing about the length meant it lasting much longer could have caused greater internal frictions.

Amcu has clearly failed on its original (totally unachievable) R12 500 monthly basic pay aim, in our view, but what they have managed to do is show sustained strike action in South Africa and preserve some element of backpay compensation. They have also achieved a massive reduction in inequality of pay of workers, with the lowest pay benefiting the most.

By our calculations, over the three years of the agreement, the workers will still be better off, even given the lack of pay during the strike, by on average R510 a month, or R120 a month for the highest-paid workers who were on strike (though again, the full terms still need to be seen).

This means a small increase in real terms, including lost wages of the strike, over the full period but maybe flat in real terms for the highest-paid workers. This dynamic again partly explains why the strike could be sustained.

The stability of any such deal will likely be easier to enforce than normal year-by-year increases, but we should still expect substantial demands at the end of the three years, as well as around corporate restructuring potential (not to mention political noise on that front).

We should not forget though that there is significant potential for further strikes across the economy in the third quarter – especially the likely strike by the National Union of Metalworkers of SA in two weeks. This will involve double the number of workers on strike, but will likely be a shorter one of two weeks, as well as ones by public workers later in the third quarter.

The end of this strike should have been broadly priced in by markets after the events of the past two weeks, but should not be read as an improvement in the underlying structural issues in the local labour market and labour framework.

* Peter Attard Montalto is an emerging markets economist at Nomura International.