AmazonScience/document-haystack
Document Haystack Dataset This repository contains the dataset for the paper “Document Haystack: A Long Context Multimodal Image/Document Understanding Vision LLM Benchmark”. 📑 Abstract Paper The proliferation of multimodal Large Language Models has significantly advanced the ability to analyze and understand complex data inputs from different modalities. However, the processing of long documents remains under-explored, largely due to a lack of suitable… See the full description on the dataset page: https://huggingface.co/datasets/AmazonScience/document-haystack.
2089k
1Task Force on Climate-related Financial Disclosures (‘TCFD’) continued2Recommendation Response3Disclosure 4location5c) De6scribe the resilience of the organisation’s strategy, taking into consideration different climate-related scenarios, including a 2°C or lower 7scenario (Companies Act 2006 – Sections 414CA and 414CB 2A (f))8Embedding climate into 9scenario analysis10 – Sc11enario analysis supports our strategy by assessing our potential exposures to risks and vulnerabilities under 12a range of climate scenarios. It helps to build our awareness of climate change, plan for the future and meet 13our growing regulatory requirements.14 – In o15ur 2023 climate scenario analysis exercises, we explored five scenarios that were created to examine the 16potential impacts from climate change for the Group and its entities.17 Page 22518 Page 22519Key drivers of performance 20and how these have been 21taken into account22 – Cl23imate scenario analysis allows us to model how different potential climate pathways may affect and impact 24the resilience of our customers and our portfolios, particularly in respect of credit losses. Under the Current 25Commitments scenario, we expect lower levels of losses relating to transition risks, although we would 26expect an increase in the effects of climate-related physical risks over the longer term. 27 – Sc28enario analysis results have been used to support the Group’s ICAAP. This is an internal assessment of 29the capital the Group needs to hold to meet the risks identified on a current and projected basis, including 30climate risk.31 – In a32ddition, scenario analysis informs our risk appetite statement metrics. As an example, it supports the 33calibration of physical risk metrics for our retail mortgage portfolios and it is used to consider climate impact 34in our IFRS 9 assessment.35 Page 227 36 37 Page 22938 Page 22939Scenarios used and 40how they factored in 41government policies42 – Ou43r scenarios are: the Net Zero scenario, the Current Commitments scenario, the Delayed Transition Risk 44scenario, the Downside Physical Risk scenario and the Near Term scenario.45 – Ou46r scenarios reflect different levels of physical and transition risks over a variety of time periods. The 47scenario assumptions include varying levels of governmental climate policy changes, macroeconomic factors 48and technological developments. However, these scenarios rely on the development of technologies that are 49still unproven, such as global hydrogen production to decarbonise aviation and shipping.50Page 22551Page 22552How our strategies may 53change and adapt54 – Th55e nature of the scenarios, our developing capabilities, and limitations of the analysis lead to outcomes that 56are indicative of climate change headwinds, although they are not a direct forecast. 57 – De58velopments in climate science, data, methodology and scenario analysis techniques will help us shape our 59approach further. We therefore expect this view to change over time.60 – Cl61imate scenario analysis plays a crucial role helping us to identify and understand the impact of climate-62related risks and potential opportunities as we navigate the transition to net zero.63 – Ou64r target-setting approach to date, for on-balance sheet financed emissions and facilitated emissions, has 65been to utilise a single net zero reference scenario (IEA NZE 2021) to underpin both energy supply-related 66sectors (oil and gas, power and utilities, and thermal coal mining) and our published targets for demand-side 67sectors in transport and heavy industry.68 – We r69ecognise that the so-called ‘hard-to-abate’ sectors, such as cement, iron, steel and aluminium, and 70aviation have a large dependence on nascent technologies and the presence (or not) of enabling policies and 71regulations. We may consider tracking progress relative to 1.5°C-aligned ambition ranges for these sectors in 72the future, which could include industry-specific scenarios alongside the IEA NZE scenario. 73 – We d74o not currently fully disclose the impacts of transition and physical risk quantitatively, due to transitional 75challenges including data limitations and evolving science and methodologies. In 2023, we have disclosed 76the impairment impacts for our wholesale, retail and commercial real estate portfolios in different climate 77scenarios. In addition, we have disclosed losses on our retail mortgage book under three scenarios and 78flood depths for specific markets. For our wholesale book, we have disclosed potential implications on our 79expected credit losses for 11 sectors under two scenarios. We have also disclosed a heat map showing how 80we expect the risks to evolve over time.81 Page 225 82Page 225 83Page 22984Page 5385Page 4886Page 44087Risk management88a) De89scribe the organisation’s processes for identifying and assessing climate-related risks (Companies Act 2006 – Sections 414CA and 414CB 2A (b))90Process – We c91ontinue to integrate climate risk into policies, processes and controls across many areas of our organisation, 92and we will continue to update these as our climate risk management capabilities mature over time. 93 – We u94pdated our climate risk management approach to incorporate net zero alignment risk and developed 95guidance on how climate risk should be managed for non-financial risk types. While we have made progress 96in enhancing our climate risk framework, further work remains. This includes the need to develop additional 97metrics and tools to measure our exposure to climate-related risks, and to incorporate these tools within decision 98making. 99 – In 2100023, we enhanced our internal climate scenario analysis exercise by focusing our efforts on generating more 101granular insights for key sectors and regions to support core decision-making processes, and to respond to our 102regulatory requirements. In climate scenario analysis, we consider, jointly, both physical risks and transition risks.103 – We c104ontinue to review policy implementation as we apply our policies in practice, and our operationalisation 105of such policies continues to be enhanced. We take a risk-based approach when identifying transactions 106and clients to which our energy and thermal coal phase-out policies apply, and when reporting on relevant 107exposures, adopting approaches proportionate to risk and materiality. 108 Page 222 109 Page 222110 111 Page 225112 113Page 66114Integration into policies 115and procedures116 – We c117ontinue to integrate climate risk into policies, processes and controls across many areas of our 118organisation, and we will continue to update these as our climate risk management capabilities mature 119over time.120Page 222121Consider climate-related risks 122in traditional banking industry 123risk categories (supplementary 124guidance for banks)125 – We p126rovide further details of how we have embedded the management of climate risk across key risk 127types, including wholesale credit risk, retail credit risk, treasury risk, traded risk, reputational risk, regulatory 128compliance risk, resilience risk, model risk, and financial reporting risk. 129 Page 223 13072 HSBC Holdings plc Annual Report and Accounts 2023131ESG review | Environmental 