Sector-level portfolio temperature scores

[MichaelTiemannOSC]

Today we gross up the company's projections to the sector's budget and then in one step treat that as a globally comparable value. But we could look at how portfolios feeding into sectors do, at a sector level, add up (or not) to the global temperature. In this way, we could see how well the sectors--as filled by the companies in the portfolio--overshoot or undershoot the global number, thereby providing users with a sector-alignment analysis, not only a portfolio-alignment analysis.

[MichaelTiemannOSC]

Re-upping this: if we look at this report from ENI for their Plenitude and Power business (https://www.eni.com/assets/documents/eng/reports/2021/Fact-Book-2021-eng.pdf) we see that they split out S1 emissions for their electricity utilities business (40.1 Mt CO2e) and their gas utilities business (1.01 Mt CO2e). But they combine S2 and S3 reporting across both.

If we really want to make the ITR tool a sector-alignment tool, we could make the primary index for all companies an ISIN-Sector pair, rather than just an ISIN. In this way, Eni's Plenitude business would have two records from a single report, IT0003132476-Electricity and IT0003132476-Gas, each scored against their respective benchmarks.

An idea for fairly allocating S2 and S3 data would be to use benchmark-aligned weighting, meaning that for their respective budgets based on production, we could derive benchmark-aligned emissions amounts: S2-OECM-Electricity, S2-OECM-Gas, S3-OECM-Electricity, and S3-OECM-Gas. We could then estimate:

S2-IT0003132476-Electricity = S2-IT0003132476 * S2-OECM-Electricity / (S2-OECM-Electricity+S2-OECM-Gas)
S2-IT0003132476-Gas = S2-IT0003132476 * S2-OECM-Gas / (S2-OECM-Electricity+S2-OECM-Gas)
S3-IT0003132476-Electricity = S3-IT0003132476 * S3-OECM-Electricity / (S3-OECM-Electricity+S3-OECM-Gas)
S3-IT0003132476-Gas = S3-IT0003132476 * S3-OECM-Gas / (S3-OECM-Electricity+S3-OECM-Gas)

So now we have two questions: (1) do we want to more heavily emphasize the sector-alignment nature of the tool, and (2) does this estimation look like a good starting point for disclosures that do not separate based on sectors?

@ImkeHorten

[ImkeHorten]

Hi Michael,
Thank you, this is a very important point you make there.
1): Yes, I think we should score the different segments against their respective benchmark. I like the idea of an ISIN sector pair.
Did you already come across similar examples from other sectors?
2) Yes, the estimation looks good. Could we do a test on some companies?

@LeylaJavadova

[MichaelTiemannOSC]

Good news...I've implemented the first half of the task: an input template with multiple rows of production per company can have combined scope emissions aligned and allocated according to benchmark data. The second half requires working through the logic of deriving implied targets using the same logic. That will take some more days.

There are also some interesting edge cases, such as when companies do report scope emissions based on production. Right now the tool does its best to find the best and biggest number as the one metric to then align and allocate. When such work already exists in the report, finding and dealing with that is a whole new thing. Eni S.P.A. reports are a great example of both global scope emissions (for the group as a whole), sector-level reporting (O&G, Chemicals, Electricity Utilities, Gas Utilities), and how our Global->Sector mapping can both be validated (where exact data can be examined), or applied (to fill in gaps where data not otherwise available).

There's also the question of how the notional investment in an instrument should be allocated across the sectors represented therein. If an integrated utility has a refining operation (Oil and Gas), power generation (Electricity Utilities), and gas distribution (Gas Utilities), how should portfolio aggregation treat $1M of notional value in that instrument for the purposes of WATS, TETS, MOTS, AOTS, etc.? In one sense, this question of handling notional value across these sectors in a single instrument is itself a kind of portfolio alignment exercise, so maybe the problem (and solution) closes back on itself (if we think about it correctly).