Supporting the net zero agenda as technology leaders
Wednesday, November 3, 2021
Paul Frost, Chief Architect, Wholesale Banking Technology, HSBC
As COP26 gets into full swing, I’ve started reflecting on what we as technology leaders can do to support the sustainability and the net-zero agenda. Since the turn of the 21st century we as a society have become more and more dependent on technology both in our personal lives and at work, and the push to digitise continues at pace, especially in the financial sector.
Let’s start by saying from a carbon footprint perspective this is a good thing, for almost all processes in a digitised journey will always be more carbon-efficient than the equivalent manual process. Using paper and visiting a branch by example are evidently more carbon-intensive. Having said that we must also recognise that for us at HSBC and other banks, technology is a sizeable part of our carbon footprint. So what are the levers we have to influence the size of this footprint?
There are a number of choices we can make that have a positive impact — from large decisions around hosting, through to how we manage our day-to-day activities across technology. One of the most impactful things we can do is to use the cloud. Whilst we are moving to more sustainable internal data centres, cloud providers are generally ahead of this curve as they have had to drive a high degree of efficiency to support their hyperscale demands. That, and their advances in use of green energy, puts them ahead in the goals of net-zero carbon. But cloud isn’t the only answer — we need to think about how we use the technology more broadly. Here are just some of the things we as technology leaders should consider:
Are we rightsizing our applications? – Typically utilization figures of servers are low, so would efficient use of containers and virtualization help?
We often “gold plate” solutions – does the solution warrant the size of infrastructure deployed to meet this business need?
Do systems need to be up and running 100% of the time? We typically leave infrastructure running when not in use. Do those development and test environments need to be running 100% of the time? Can we scale on demand rather than pre-provision to meet spikes in capacity demand? This is another area where cloud helps.
What about data? It too has a carbon footprint. Do we always consider the data lifecycle and have efficient processes for data archiving and deletion, or do we keep the data around – just in case?
Duplication – do we have duplicate solutions that meet a business need? Do we really need this, or would consolidation help reduce the carbon footprint?
The desktop – are we thinking about carbon footprint in our desktop? There is significant variation across desktop devices that needs to be considered. Are we considering the whole life of a device through manufacture, transport, use, and eventually decommission?
These are just some of the things we need to consider, and I am sure many people are already aware and doing this. There are more options out there, but the most important thing we can do is start thinking about the various options we have at hand, make sure we are having these conversations and challenge the carbon efficiency of our technology use. At HSBC, we are committed to achieving net zero in our own operations and supply chain by 2030 or sooner – read more here.
Accelerating sustainable finance with technology
Wednesday, November 3, 2021
Jeff Sternberg, Director, Office of the CTO, Google Cloud
As global leaders gather this week at COP26 in Glasgow, one of their top goals is to mobilize finance to secure net zero emissions and keep 1.5 degrees warming within reach. To do this, COP26 leaders are challenging all financial stakeholders to take climate into account in every financial decision.
Technology is playing a vital role in this effort. I have a background in financial services and fintech, and work in Google Cloud’s CTO Office with technology leaders across financial services and other industries to transform sustainability in their organizations. Here are three technology areas that I think are vital for sustainable finance, both to manage climate-related risks and to enable new climate finance opportunities.
1. Understanding localized climate risks
COP26 leaders are calling on companies to be transparent about the risks inherent in climate change. Every business will be affected by a warming planet, and it’s critical to understand precisely how these events will affect company operations, employees’ lives, and entire value chains. As BlackRock’s CEO Larry Fink has stated, climate risk is investment risk.
However, physical risks from climate change are not evenly distributed around the world. Some areas will see increased floods, and others will face droughts or wildfires. Rising sea levels will affect coastal areas and severe winter storms will affect places with historically milder winters.
Geospatial analytics and modeling has the potential to give us a much more accurate view of these localized climate-related physical risks. The Task Force on Climate-Related Financial Disclosures (TCFD) recommends factoring the geographic location of the organization’s value chain (both upstream and downstream), as well as the organization’s own assets, into any climate change scenario analysis efforts. Using tools like Google Earth Engine and BigQuery, it’s possible to create a localized, precise understanding of climate-related physical risks for each location in the value chain.
2. Using AI to improve climate disclosure
Another key finance priority for COP26 is improving the quantity, quality and comparability of climate-related disclosures by companies, so that investors and regulators can better understand climate risks and opportunities. To date, most of this disclosure has been published as human-readable documents. One example is Google’s Environmental Report. These documents are very helpful if you’re in a position to deeply study an individual company, but what if you want a broad view of sustainability progress across markets?
This is where AI comes in. Specifically, natural language machine learning approaches can be applied to classifying and extracting structured information from sustainability disclosures. The TCFD recently used an AI approach to quantify the level of sustainability disclosure across 1,701 companies from 69 countries. A team of experts manually labeled text passages in a sample of 150 documents, assigning a “yes” or “no” value to each review question, such as “does the company disclose Scope 1, Scope 2, and, if appropriate, Scope 3 greenhouse gas emissions?”. Then they trained a model with these labels and used it to predict yes/no responses for the entire corpus. The resulting analytics were used by the task force to assess the state of adoption and level of disclosures.
Investors, regulators, and companies can all benefit from adopting natural language AI techniques to better understand sustainability documents, whether those are from suppliers, investors, customers, or even competitors. Tools like Google Document AI and Vertex AutoML can accelerate AI adoption in these organizations.
3. Collaborative sustainability data sharing
COP26 is demonstrating the power of collaboration to solve climate challenges. It’s clear that individual organizations cannot go at this alone; everyone must work together! In the financial world, companies need to share information with investors, suppliers need to share data with companies, and everyone needs to share data with regulators and central banks.
In an enterprise setting, sharing data can be difficult, involving legacy tools, slow cycle times, and one-way processes. What if sharing climate and sustainability data were as simple as sharing photos or documents in the cloud? Cloud-native tools like Analytics Hub, built on BigQuery, can definitely help here.
As a bonus, thanks to Google’s renewable energy matching, building these solutions on Google Cloud today means the net operational greenhouse gas emissions associated with your application is zero. And over time, the gross carbon footprint of these workloads will decrease as we make progress towards our goal of running on carbon-free energy around the clock by 2030.