Latest news with #riskanalysis
Yahoo
2 days ago
- Business
- Yahoo
Amplify Launches Risk Assessment Tool
You can find original article here Wealthmanagement. Subscribe to our free daily Wealthmanagement newsletter. Amplify Platform, a wealth management tech provider, has launched a proprietary risk analysis tool to help advisors measure risk in client portfolios. The new tool, QuantumRisk is based on the work Dr. Ron Piccinini, director of investment research at Amplify, who has a background in fat tail risk modeling. He was responsible for developing PrairieSmarts' risk management system in 2017, which was subsequently acquired by Covisum, a software company that serves RIAs. 'Over the last decade, advisors became more and more focused on trying to explain portfolio risk to their clients,' Piccinini said. 'As an advisor, your goal is to maximize some kind of return for the risk you're taking.' Piccinini said Amplify had a risk scoring tool previously, but it provided a 'vague indication' of the actual risk of underlying investments. The firm considered using something off-the-shelf, but they didn't think any of the existing tools were sophisticated enough. 'The sad reality is, we think, most of these things out there are toys, at best,' he said. QuantumRisk uses a series of risk scores across a portfolio on a scale of 0 (cash) to 1,000 (penny stocks). The S&P 500 is scored at 100. 'So I know if I have a score of 100, I'm just as risky as the market,' he said. 'If I have a score of 200, I'm twice as risky as the market. If I have a score of 50, I'm half the risk of the market.' It uses real-world probabilities and market stress scenarios, rather than relying on backward-looking models. The tool will start with risk analysis of stocks, bonds, mutual funds and ETFs. Alternatives and options are on the roadmap for future inclusion. 'What we're trying to do is just inform people about the risk of their investment, and we're not cutting any corners on statistical distributions, calculations, those types of things,' he said. 'A lot of toys/tools out there oversimplify things and give people overconfidence.'
Forbes
21-07-2025
- Business
- Forbes
Opening The AI Black Box: What Financial Leaders Need To Know
Nosa Omoigui, CEO of neuro-symbolic GenAI and intelligent agents that transform alpha decision making and risk analysis. As artificial intelligence (AI) accelerates across the financial sector, a critical question remains: Can executives trust the answers AI provides? While Gartner reports that 58% of financial professionals already use AI tools—with growing adoption for forecasting and analytics—many remain wary. That's because some of the most popular systems, like ChatGPT and Gemini, offer confident responses despite a high error rate. In a world driven by evidence and risk management, this black-box approach is not acceptable. Financial professionals require AI that is designed to be transparent and defensible, and to enable more accurate reasoning. The answer lies in a new class of AI systems that combines the adaptive learning of neural networks with the clarity and structure of symbolic reasoning. The Limitations Of Traditional Generative AI In Finance Large language models (LLMs) excel at pattern recognition and next-word prediction, but this capability becomes a weakness in high-stakes financial analysis and forecasting. LLMs are inherently vulnerable to misinformation when data is flawed or misinterpreted, with recent research showing high and rising hallucination rates. Compounding this, their reasoning often lacks transparency, citations and clear logic. For financial leaders, these limitations are significant. When an LLM provides a confident yet unsupported answer without source attribution or clear reasoning, it undermines trust and introduces considerable risk. Defining AI Requirements For Financial Analysis Financial institutions operate in high-stakes, highly regulated environments. Whether assessing AML risk exposure, identifying compliance gaps or uncovering investment alpha opportunities, every decision must be water-tight. To be truly useful in financial decision-making, AI must deliver more than insights; it must offer traceability, auditability and control. That means every output must be linked to structured knowledge and clear rules, allowing executives to understand not just what the recommendation is, but why it was made. To earn trust, an AI platform for financial crime and AML compliance must not only flag control gaps but also benchmark them against peer disclosures and regulatory standards, tracing each gap to specific filing sections and rules for instant executive verification. This level of transparency streamlines fact-checking and enables confident, defensible decisions. For alpha discovery, the platform must benchmark portfolio growth and performance against market leaders, highlighting gaps with direct links to source data. This empowers investment teams to act on reliable insights and refine their theses with greater conviction. These systems are built on verifiable evidence, provide automatic audit trails and adapt to changing internal policies, regulations and market conditions. Essentially, AI must be human-led for financial executives to trust it. An AI tool should empower its user with new, "weaved" insights—integrating data from diverse sources—while also providing the "breadcrumb trail" that engenders trust and defensibility. How Neuro-Symbolic AI Opens The Box New AI tools, such as neuro-symbolic AI, address key limitations of LLMs in high-stakes analysis and decision-making. This hybrid approach combines neural networks, which learn from data and detect complex patterns, with symbolic reasoning, which applies explicit rules and logic for transparent, explainable decisions. A neuro-symbolic AI system can map a bank's AML controls to precise Financial Action Task Force (FATF) requirements, benchmark them against peers and pinpoint gaps, issuing targeted alerts when regulatory risks are rising and linking every recommendation directly to the relevant regulatory clause. This is achieved by connecting dots across filings, regulatory texts and enforcement actions to deliver actionable, fully explainable guidance. Similarly, the system can assess a portfolio's innovation leadership by benchmarking R&D and patent activity against top peers, detecting hidden risks or opportunities. Critically, it understands context and materiality, so it's not misled by superficial "innovation-washing." Alerts are grounded in concrete data points, providing clear, actionable recommendations for alpha generation. LLMs alone cannot deliver this level of structured, multisource, auditable support. In high-stakes workflows, even minor inaccuracies or hallucinations can quickly compound, creating significant risk and eroding trust. With a hybrid approach, neuro-symbolic AI combines deep learning from large datasets with transparent, rule-based reasoning. For financial services, this enables AI to deliver rigorous analysis with clear, auditable justifications for every output. Equipping Financial Executives For Confident Decision-Making Neuro-symbolic AI empowers financial executives with advanced decision support, enhancing oversight and regulatory alignment. By analyzing vast datasets and benchmarking performance and compliance against peers, these systems help institutions demonstrate best practice parity and reduce regulatory risk. By integrating global frameworks such as FATF, FinCEN and Basel III, neuro-symbolic AI can identify specific compliance gaps and recommend next-best actions tailored to each institution's operational reality. Neuro-symbolic AI delivers analyses that compliance teams and regulators can easily validate. Unlike traditional AI, it produces decision-ready outputs that support a shift from reactive, fragmented oversight to proactive, unified governance. These platforms empower compliance, risk, finance and investment leaders to benchmark external risk posture and identify opportunities, without internal IT delays. Chief investment officers can target specific theses or portfolios, gaining actionable insights by comparing performance to peers and market signals. This 'smoke detector' approach provides rapid wins and builds internal momentum. Adoption can begin with high-priority risk vectors such as FinCrime, cyber or market risk, and scale to others like credit, operational, conduct, model, strategic and third-party risk. Firms can expand incrementally, targeting specific regulations, risk domains or investment strategies as value is demonstrated. This phased strategy helps overcome legacy barriers, accelerate results and justify broader investment. For example, one global bank I worked with rapidly closed cyber and AML gaps through external benchmarking, while an asset manager's CIO identified and addressed underperformance in key portfolios. Both achieved rapid impact, scaling adoption to strengthen risk and investment outcomes across the enterprise. From Black Box To Command Center The message to financial leaders is clear: Trust will determine the trajectory of AI adoption. Neuro-symbolic AI offers a way forward that combines human judgment with machine precision, and innovation with oversight. As the pressure mounts to modernize and compete, adopting transparent, controllable AI won't just be a differentiator. It will be a requirement. Now is the time for leaders to evaluate where these capabilities can deliver the greatest impact within their organizations and to take the first steps toward building a more future-ready command center for risk and opportunity. Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
