Five steps to fewer forecasting follies

Those who have followed this blog long enough will know that I have often been critical of those that claim to be able to forecast commodity prices. My concern was initially spiked after it appeared that many investors, companies and governments had been fooled into believing that commodity prices would continue infinitum. That sparked me to write my second book, Crude Forecasts: Predictions, Pundits & Profits in the Commodity Casino which set out to provide the evidence, lay out the incentives and offer some ways that things could be improved for the better.

I’m certainly not the only one who is skeptical of the abilities and claims of Wall St forecasters. In a recent piece for Bloomberg, Barry Ritholtz of Ritholtz Wealth Management offers 5 suggestions as to how the forecasting business could be made a whole lot more transparent and potentially more successful.

No. 1. Share the underlying model’s past performance: As Ritholtz highlights, “if the forecaster has an audited track record showing how the prognostications stacked up versus reality during the past five years, and can demonstrate how these made clients some money, that might be worth notice.”

But he rightly caveats this by saying that past performance is no guarantee of future results since “a good track record may not be repeatable; that those winning outcomes could have been the result of luck or that specific era or some other random element.”

In my book I analysed the forecasting ability of investment banks and other institutions over a 10 year period to 2016, which of course covered commodity booms, busts and financial crises. But that was only one cycle, and just one commodity. It tells you nothing about whether that institution was just lucky, or unlucky. It may of course just be survivorship bias.

No. 2. Acknowledge the unknown variables: The article highlights the importance, and the rarity of the caveat, “Not locking oneself into any single outcome because things might change is simply common sense. Unfortunately, that is a rare characteristic in too many forecasters.”

The financial media being what it is (although there are welcome exceptions such as Real Vision) rewards the confident, no doubt about it, banging on the table kind of forecast. It just doesn’t have time (nor the appetite) for a forecaster saying, “on the one hand this or that could occur.”

No. 3. Acknowledge inherent biases: According to Bloomberg Business Week, economic forecasters are “more likely to miss recessions than to predict ones that never occur.” Ritholtz rightly highlights that there are other incentives at work, “it wasn’t because the economists were necessarily bad at economics, but rather, because of basic game theory. Career risk for being wrong is very real.”

To examine what the age and experience of the forecaster has on the degree of herding, research published in the RAND Journal of Economics examined over 8,000 forecasts by equity analysts between 1983 and 1996. They found that younger analysts tend to herd more than their more experienced colleagues do. Less experienced analysts, meanwhile, are more heavily punished for getting their forecasts wrong and so they have every incentive to stick with the herd. In contrast, older analysts, who have presumably built up their reputations, face less risk of termination. The researchers also found that, contrary to expectations, making bold and accurate predictions does not significantly improve a young analyst’s career prospects.

No. 4. Use errors to make better forecasts: Accurate predictions are held up as evidence of forecasting prowess, and a great marketing platform to sell clients something else. Poor forecasts are quietly swept under the carpet, where, hopefully no one will notice. But we can all learn from failure, and so maybe we shouldn’t be so quick to sweep the evidence away.

As Ritholtz explains, “All models do is take a series of data inputs, sprinkle a little fairy dust on them, and then generate an output. But even if the model does OK, how has the forecaster used its output? Can they make money for clients with it? Alternatively, do they anchor themselves to these predictions, regardless of subsequent data? There is a specific skill to adjusting to errors and failures in order to improve. It is a skill used too little by economists and financial analysts.”

As I explain in Crude Forecasts, “Cognitive dissonance is a psychological phenomenon that refers to the discomfort felt at a discrepancy between what you already believe to be true and new information that presents itself. This is an especially big risk for forecasters of all types, but essentially means that the forecaster will either discount new information that conflicts with the stated forecast or attempt to reframe the evidence to validate the success of the forecast.”

5. Learn from the pros: Finally, the article highlights the work that Philip Tetlock has done on why so many forecasts fail, “His 2006 book “Expert Political Judgment” studied thousands of forecasts, and came to the conclusion that people simply are not good at making predictions about much of anything. With one caveat: Buried within Tetlock’s huge dataset of failed forecasts was a surprising subset of superforecasters. Those in this group stood out for their ability to make more accurate predictions than others.”

According to Tetlock pundits typically fall into categories – “foxes” and “hedgehogs”. Foxes pursue many different ends, often unrelated and even contradictory; they entertain ideas using divergent thinking (ie, looking at many possible outcomes), rather than convergent thinking, and they also don’t seek to fit these ideas into, or exclude them from, any one all-embracing inner vision.

However, many of the pundits courting the financial limelight are hedgehogs. You can easily spot a hedgehog – they are characterised by an attitude of relating everything back to a single vision, and they over simplify and come across as much more confident in their outlook on the world in order to produce a compelling narrative.

Overall, I sense that the process employed by forecasters has improved over recent years. Nevertheless, as the economic forecasting world debates the possibility of a global recession over the next few years these questions are unlikely to stray too far.

Unlock commodity market insight now and subscribe to our email updates or

Platinum boom on the horizon? Not so fast!

Palladium has been the stand out performer in commodity markets over the past three years – up 160%. Palladium, mainly used in catalytic converters to control emissions from gasoline engines has been supported by a switch away from diesel vehicles and by the introduction of tighter vehicle emissions standards in China and elsewhere in the world.

Investors look to play the price differential

Investors reluctant to follow the bull market in palladium have been looking at alternative ways to bet on increasing emissions standards. The other main metal used in catalytic converters is platinum. Platinum demand has been under pressure since 2015 following the Volkswagen diesel-engine emissions cheating scandal in the US.

Platinum is extremely cheap  compared with palladium. Platinum prices have declined by 17% over the past three years while the platinum/palladium ratio has declined to its lowest level since 2001.

Chart 1: Platinum/palladium ratio

However, that of course doesn’t mean the ratio has to correct through higher platinum prices. Lower palladium prices and/or a combination of the two works just as well.

Switch to platinum will take time

Some investors it seems are backing the former. Holdings in platinum-backed ETFs have surged by 15% this year to reach a 4 year high as investors bet that carmakers will start to use the metal in petrol car catalysts.

But according to the car industry a substitution towards diesel is far from being a done deal. Rahul Mital, global technical specialist at General Motors suggested at a London Bullion Market Association meeting late last year that it would take some time to switch and even then the price benefits don’t necessarily warrant it:

“Any time you want to make a substitution like that, it is at least 18 months to a two-year cycle if we’re going to switch. We have to be careful that by the time we do all that.”

According to the World Platinum Investment Council (WPIC) a 5% substitution would mean a 14% increase in platinum demand from the automotive demand segment. However, given that catalyst demand is roughly half as important to platinum demand as it is for palladium (40% vs 79%) it may take a large switch to have a material impact on overall platinum demand.

Chart 2: End market demand

Fog of uncertainty

The car industry is battling declining sales in China and other markets. If auto-sales growth really does fall off a cliff (perhaps if a global recession hits) then demand for both palladium and platinum decline.

Chart 3: Annual growth in car sales in China

The car industry is facing increasing demands to invest in electric vehicle capability to head off the perceived threat from Tesla and others who have a head start. Is now really the time that they will be tinkering with the amount of metal in their catalytic converters?

History is also not on the side of price booms in metals thought essential continuing indefinitely. Just look at the price of cobalt over the past year and rhodium a decade ago.

Related article: What lessons does rhodium have for commodity investors?

Constrained supply after one-off boost

A new report by the WPIC predicts that platinum mine-level output will rise by 6% increase in 2019 to 6.46 million ounces on top of a 3% boost to recycling. The WPIC said supply growth would come mainly from the release of material stockpiled by mines and smelters in South Africa (responsible for 70% of global mined output) during upgrades and maintenance over 2017 and 2018. If this occurs it will clearly be a one-off boost to supply.

More generally platinum output from South Africa is frequently disrupted by power cuts, strikes and unrest. There seems little chance of that changing. More than 60% of South Africa’s platinum mining industry is loss-making or marginal according to Minerals Council South Africa. Labour costs have been increasing faster than inflation, whilst revenue from metal sales has been declining due to low platinum prices.

Chart 4: Forecast change in platinum supply

An industrial metal but a precious one too

Both platinum and palladium have their uses in applications other than catalytic converters. Jewellery for one but they are also held by investors as a ‘beta’ version of more commonly held precious metals like gold and silver. If gold prices come under pressure (perhaps because of reduced geopolitical tensions, lower inflation expectations and / or higher bond yields) then investor demand for palladium and platinum will also suffer.

Unlike gold and to a lesser extent silver, the PGMs suffer from poor liquidity and an opaque market. The PGM group of metals (palladium, platinum, rhodium and others) are very small markets relative to gold or indeed even silver. This means that prices can go from being an investors best friend to their worst enemy in the blink of an eye.

Related article: Platinum prices: The top 10 most important drivers

Platinum prices have been on a downward trend since 2011 frequently coming up against trend line resistance along the way. Since August prices have attempted to push up towards the trend line again (rising $100 per oz to reach $875 per oz) before failing to push past in in late February and then retracing half of that run-up in just a few days.

Chart 5: Platinum prices

Although speculators have recently turned net long it is going to take a break of this trend-line combined with confirmation of car-makers intent to seriously switch towards platinum for investors to run platinum higher.

Chart 6: Managed money positioning

Unlock commodity market insight now and subscribe to our email updates or

The negative cascade from the Brazilian mining disaster may have only just begun

On 25th January 2019 it happened again. Just three years and two months after Samarco’s tailings dam failed, the world’s largest iron ore producer, Vale, saw one of its tailings dams collapse in Brazil. The death toll is expected to exceed 300.

Aside from the obvious human tragedy the episode holds import lessons for commodity investments and by implication the outlook for commodity prices.

Just as climate concerns and safety failings (e.g. Deepwater Horizon) have arguably deterred investment in the oil industry, the same may happen for the mining industry as poor safety and community relations take centre stage. As Clyde Russel outlines in a recent article for Reuters the recent disaster risks a cascade of negative outcomes for the mining industry:

“If investors don’t want to commit funds to the industry because of fear over damage to their own image from poor mining practices, then the industry will be starved of capital. Lack of capital means it becomes harder and harder to get new mines built or exploration drilling undertaken.”

As international investors scale back funding (perhaps indiscriminately in the short term) then the share price of mining companies will fall while also pushing up the cost of funding for new ventures. Note that the share price of Vale’s competitors are only rallying right now because of the potential iron ore supply deficit that Vale’s difficulties leaves behind.

Governments that have long outsourced the provision of social services to the mining companies, especially in remote regions not easily accessible may be under pressure to take over where the mining companies have failed. Indeed, there may be renewed pressure for the state to have a bigger say in the mining sector and a bigger share of the spoils – a return to resource nationalism. That being said mining accounts for about 5% of Brazil’s GDP and Vale’s biggest shareholders are some of Brazil’s largest pension funds, so in Brazil at least they will be careful not to kill the golden goose. The attitude elsewhere in the world may be different.

According to Amsterdam-based Responsible Mining Foundation many of the world’s largest mining companies are not able to ‘know and show’ how effectively they are addressing the risks of tailings dam failure and seepage. The foundation’s research shows that failure risks are greatest for large, steep and old tailings dams in tropical zones where seismic activity and extreme weather events can precipitate dam collapses.

Although you could argue that iron ore and coal mining are the dirty heaving lifting of the mining sector there is a risk that all mining activities get tarnished with the same brush – that safety isn’t as high up the priority list and not quite as good as the mining companies say it is. The risk of future disasters remains high.

Resource companies are trying to position themselves as the suppliers of the ingredients necessary for an electric future – whether that is cobalt, vanadium or copper. Many people are not aware that commodities like cobalt and vanadium are typically by-products of much larger activities involving the extraction of more conventional commodities. Any negative impact to production of the latter will ultimately affect the former.

There are some parallels between this latest disaster and the BP Deepwater Horizon accident in 2010. Just as with that corporate disaster, rivals are quick to twist the knife but at least then tighter regulations and a long and arduous approval process were  introduced. Although activity dropped sharply in the aftermath, oil production in the US Gulf has since rebounded to hit record levels.

There is no sign of a similar response in the mining industry following this latest disaster. But that is arguably what it must get unless it wants to face the cascade of negative outcomes outlined in this post.

Related article: Shining a light on extractive industries: An interview with Åsa Borssén from Raw Talks


Unlock commodity market insight now and subscribe to our email updates or