When Chris Hitchen, CEO of the UK’s Railpen, thinks about portfolio fossil-fuel exposure, there is no room for moral absolutes. The discussion is “quite different” now that oil trades at $45/bbl rather than $115/bbl, he insists. These are risks that must be measured in dollars and cents – and whose price can swing wildly from one year to the next.

Cheap oil removes our incentives to dial-down the central heating or trade in our gas-guzzlers, but it also removes the incentive for producers to extract many of their fossil-fuel assets. And while oil has halved in value, the world’s leading carbon credits, Certified Emission Reduction units (CERs), are now worth virtually nothing.

“It’s often said that carbon risk will only go in one direction – it’s not true,” as Ossian Ekdahl, head of ESG at AP1, puts it in our opening article (page 43). One might add that different markets price it so divergently that the information value for risk managers is zero.

Investors are looking for some certainty, and since Carbon Tracker first suggested it in 2011, the ‘stranded assets’ hypothesis has been the likeliest candidate to deliver it. The idea is that, if we are to limit global warming to 2°C, we set ourselves an emissions budget that implies that 80% of known carbon reserves owned by fossil-fuel interests would never be extracted. A more recent study from the UCL Institute for Sustainable Resources arrived at a similar number.

“Companies spent over $670bn last year searching for and developing new fossil fuel resources,” observes UCL professor Paul Ekins. “They will need to rethink such substantial budgets if policies are implemented to support the 2°C limit. Investors in these companies should also question spending such budgets.”

How should investors respond? The obvious thing to do is to re-allocate capital away from the companies – and sovereigns (page 59) – that are most exposed.

But some in our report, like PGGM’s Marcel Jeucken and PensionDanmark’s Torben Möger Pedersen, argue that selling a stock to someone else who doesn’t care about climate change has “zero impact on the climate agenda”. Jeucken urges positive engagement. Pedersen says investors should finance things like renewable energy infrastructure or energy-grid development (page 57) projects increasingly financed by the fast-growing market in ‘green bonds’ (page 58).

Others, such as Philippe Desfossés at ERAFP, argue that identifying the extent to which all companies exacerbate or mitigate the climate problem and then engaging and divesting where necessary, is the only way forward. These competing ideas are considered further by Gordon Noble and Matthew Kiernan of Inflection Point Capital Management (page 55).

Of course, divesting and engaging both require an initial assessment of an investment’s carbon exposure. This is why, in 2014, ERAFP collaborated with environmental data consultancy Trucost to publish a pioneering audit of its listed stocks’ carbon footprint. Mats Andersson, CEO at AP4, suggests it should be “mandatory” for pension funds to disclose their carbon footprints.

But Andersson also concedes that poor data and a lack of reporting standards make this a big challenge. Our report includes an article getting to grips with the 400 or so different emissions reporting rules identified by the Climate Disclosure Standard requirements Board, and efforts to standardise them (pages 56-57).

But let’s imagine we do eventually get the data that would enable informed divestment. Would this mitigate or exacerbate risk?

There would be at least two new risks to consider. The first is tracking error: naively divesting from, say, energy and utilities will leave you a long way from your benchmarks.

Like all such quantitative finance problems this has quantitative finance solutions. Trucost and index provider MSCI have worked on this for some time, and recent collaborations with Amundi, FRR and AP4 have resulted in low-carbon index strategies that explicitly target low tracking error: for Europe, an 81% reduction in carbon reserves intensity has been achieved in exchange for tracking error of just 0.72% (pages 51-52).

But the second risk is less tractable. What becomes of the ‘stranded assets’ thesis if we break the link between carbon reserves and greenhouse gas emissions?

This idea is pursued by Glashow University’s professor of energy engineering Paul Younger, who not only notes the importance of fossil fuel assets in the production of fertililsers, plastics, pharmaceuticals and other goods that would in fact represent a sequestration of carbon, but also the growing viability of carbon capture and storage (page 53).

Our report does not pretend to provide easy answers to any of this – but we hope that it starts to ask the right questions. It is critical that we do so, because there is no aspect of investment management where the stakes are so high, or the risks look so much like uncertainties.