In 2013, Moody’s indicated that water scarcity could increase rating pressure on global mining companies, particularly if the accompanying operational and political risk is poorly managed. There was a big ‘if’ associated with the warning, related to assumptions of regulatory pressure may increase project times and result in procurement of more complex water management systems. Consider that it takes over 700 cubic meters of water to produce a ton of gold in the extraction and processing phases, and it is clear that regulatory processes will limit the amount of water that can be allocated for mining processes.
The mining industry is aware of these issues, because they have faced water-related pressures for a long time, and have had the opportunity to gradually build in efficiencies in response to regulatory or operational risks. Are these risks likely to increase? Certainly. The amplitude of water risk impacts on business operations over the last two decades has increased significantly, as we have demonstrated for electric utilities. One of the prime opportunities for mining operations is that low quality or water unfit for human consumption (e.g. seawater) can be used. Moreover, to date, it is unusual for a mine not to have the potential to recycle process water within the mine’s closed cycle and store it either in a tailings facility or a dedicated water storage facility. Nevertheless, where there is risk, there is operational uncertainty with impact on cost and productivity, and thus potential for water to impact stock volatility.
This blog sought the quantify the ripple effect of water risk impacts on stock volatility using the waterVaR risk model (see earlier blogs). We looked at twelve large mining companies, focused in the precious metals industry, across three continents. Using input data such as residual stock price volatility, corrected for systemic market and sector risks, revenue risk exposure, and fixed asset risk (measured in terms of fixed asset productivity per unit enterprise value), both waterBeta’s and waterVaR values were calculated. The waterVaR model focuses on the physical risks, and does not explicitly take into account reputational or regulatory risks. Hence, it is a proxy valuation of total water risk exposure.
The values are based on water risk volatility during the reporting quarter with highest idiosyncratic impact (often the drought or summer/early Fall period), or when allocations face the highest regulatory pressure. Based on our analysis, waterVaRs are approximately one order of magnitude lower for mining companies (upper quantile of $10-45MM) than for electric utilities (upper quantile of $25MM-$300 MM). The waterVaR estimate is generally driven by two economic and financial risk drivers: fixed asset productivity risk (economic output) exposures to local water conditions (‘asset risk’), and the excess market risk embedded in the operational value of the firm (‘volatility risk’).
In the case of mining companies, 20-60% of fixed assets may experience some level of water risk impacts on productivity. In comparison, the asset risk of electric utilities impacts 33-65% of facilities. Hence, the embedded volatility risk is more significant for electric utilities than for mining companies. It is plausible that in the commodities sector, water risk management costs (as COGS or Capex) are already to some extent priced in the stock, whereas in the regulated portion of the energy business (often with greatest water risk exposures) Capex or COGS cannot be immediately passed on to the consumer.
Equarius Risk Analytics has promoted the concept that waterVaR data can be used as smart beta’s for portfolio allocation strategies, provided they represent financial risk drivers to the company’s performance. We have generally adopted the rule that waterVaR has to represent at least 3% of VaR to be financially material. Hence, Anglo-Eagle Mines was left out. In addition, companies with a low market capitalization relative to the portfolio (e.g. Kinross Gold) are excluded because their waterVaR is an outlier relative to the remaining companies. The figure below assumes a $1 bn. portfolio, and shows the allocation of stocks based on three strategies: market capitalization-weighted (blue), price-weighted (red), and waterVaR-weighted (green).
The allocation is highly variable between the three strategies, particularly as high waterVaR companies such as Newmont and Barrick Gold are significantly underweighted, and low waterVaR stocks such as Koza Altin Isletmeri, Northam, and Zhaojin Mining are significantly overweighted in the portfolio. The performance of the smart beta allocation indicates that annual returns are significantly higher than for market- or price-weighted portfolios, by a premium of 4-20%, while maintaining a similar volatility risk. As we have argued before, the waterVaR smart beta metric indicates that there are relevant market pricing signals for water risk that provide feedback to the companies regarding stock volatility risk impacts from water.
A particularly useful perspective on business water risk impacts is the sensitivity of the stock to the water risk, as expressed by the waterBeta (shown below). Stocks such as Kinross, Barrick Gold, Newmont Mining and Northam are very sensitive to propagation of revenue and asset risk into stock volatility. On the other hand, Anglo-Eagle, Zijin Mining and Newcrest are the least sensitive, largely because neither revenue risk, nor asset risk exposures are significant, and the stock risk relative to the market (financial beta) is low.
Hence, the message to the risk sensitive firms is: water risk management pays off to reduce stock volatility, and asset managers with smart beta allocation strategies may be less inclined to include these stocks in their portfolios.
Dr. Peter Adriaens is CEO of Equarius Risk Analytics LLC, a financial IT firm focused on quantitative risk analytics related to equities and portfolios exposed to natural resource constraints. He can be reached at firstname.lastname@example.org or at www.linkedin.com/in/peteradriaens/