Hedge Fund Performance

Machine learning beats traditional forecasting methods in multi series forecasting. In one of the latest M forecasting competitions, the aim was to advance what we know about time series forecasting methods and strategies. Competitors had to forecast 40k+ time series representing sales for the largest retail company in the world by revenue: Walmart. These are the main findings: ▶️ Performance of ML Methods: Machine learning (ML) models demonstrate superior accuracy compared to simple statistical methods. Hybrid approaches that combine ML techniques with statistical functionalities often yield effective results. Advanced ML methods, such as LightGBM and deep learning techniques, have shown significant forecasting potential. ▶️ Value of Combining Forecasts: Combining forecasts from various methods enhances accuracy. Even simple, equal-weighted combinations of models can outperform more complex approaches, reaffirming the effectiveness of ensemble strategies. ▶️ Cross-Learning Benefits: Utilizing cross-learning from correlated, hierarchical data improves forecasting accuracy. In short, one model to forecast thousands of time series. This approach allows for more efficient training and reduces computational costs, making it a valuable strategy. ▶️ Differences in Performance: Winning methods often outperform traditional benchmarks significantly. However, many teams may not surpass the performance of simpler methods, indicating that straightforward approaches can still be effective. Impact of External Adjustments: Incorporating external adjustments (ie, data based insight) can enhance forecast accuracy. ▶️ Importance of Cross-Validation Strategies: Effective cross-validation (CV) strategies are crucial for accurately assessing forecasting methods. Many teams fail to select the best forecasts due to inadequate CV methods. Utilizing extensive validation techniques can ensure robustness. ▶️ Role of Exogenous Variables: Including exogenous/explanatory variables significantly improves forecasting accuracy. Additional data such as promotions and price changes can lead to substantial improvements over models that rely solely on historical data. Overall, these findings emphasize the effectiveness of ML methods, the value of combining forecasts, and the importance of incorporating external factors and robust validation strategies in forecasting. If you haven’t already, try using machine learning models to forecast your future challenge 🙂 Read the article 👉 https://buff.ly/3O95gQp

The real challenge in HR is not identifying talent. It's knowing who is ready for leadership. Great IC. High performer. Strong technical skills. But the moment they become a manager, everything shifts. Because being good at the work is not the same as leading the 𝙥𝙚𝙤𝙥𝙡𝙚 who do the work. And that is where we get into trouble. We often assume a strong IC will turn into a strong manager. But management is a completely different job. → Different pressures. → Different responsibilities. → Different skill set. Here are the core skills every manager needs: ✅ 𝗖𝗼𝗺𝗺𝘂𝗻𝗶𝗰𝗮𝘁𝗶𝗼𝗻 Managers spend a lot of time explaining what matters. Priorities. Expectations. Decisions. It is not about talking more. It is about giving people enough clarity to do their best work. ✅ 𝗖𝗼𝗮𝗰𝗵𝗶𝗻𝗴 𝗮𝗻𝗱 𝗗𝗲𝘃𝗲𝗹𝗼𝗽𝗺𝗲𝗻𝘁 ICs grow their own skills. Managers grow the skills of everyone on their team. That shift from doing the work to developing the people who do the work is a major jump. ✅ 𝗗𝗲𝗹𝗲𝗴𝗮𝘁𝗶𝗼𝗻 𝗮𝗻𝗱 𝗙𝗼𝗹𝗹𝗼𝘄 𝗨𝗽 It's not just handing off tasks and hoping for the best. It is assigning the right work, checking progress, and keeping accountability steady. ✅ 𝗘𝗺𝗼𝘁𝗶𝗼𝗻𝗮𝗹 𝗜𝗻𝘁𝗲𝗹𝗹𝗶𝗴𝗲𝗻𝗰𝗲 Managers deal with conflict, frustration, and change. Staying steady and reading the room is part of the job. Teams watch how you react long before they listen to what you say. ✅ 𝗦𝘁𝗿𝗮𝘁𝗲𝗴𝗶𝗰 𝗧𝗵𝗶𝗻𝗸𝗶𝗻𝗴 Managers cannot stay deep in the weeds forever. They must think ahead. Spot risks. Plan for the next move. Connect the team’s work to the larger goals. That is why promoting a great IC into management without support can set them up for a rough landing. 𝗪𝗵𝗮𝘁 𝘀𝗸𝗶𝗹𝗹𝘀 𝘄𝗼𝘂𝗹𝗱 𝘆𝗼𝘂 𝗮𝗱𝗱 𝘁𝗼 𝘁𝗵𝗶𝘀 𝗹𝗶𝘀𝘁? If this post helped you, share it with someone in your network who is moving into management. ♻️ I appreciate 𝘦𝘷𝘦𝘳𝘺 repost. 𝗪𝗮𝗻𝘁 𝗺𝗼𝗿𝗲 𝗛𝗥 𝗶𝗻𝘀𝗶𝗴𝗵𝘁𝘀? Click the "𝗩𝗶𝗲𝘄 𝗺𝘆 𝗡𝗲𝘄𝘀𝗹𝗲𝘁𝘁𝗲𝗿" link below my name for weekly tips to elevate your career! #HRLeadership #Management #CareerGrowth Adams HR Consulting Stephanie Adams, SPHR

Since 2020, the markets have given insane returns and a 4x hike in demat accounts! However, many new-age investors chase unrealistic expectations of 25-30% annual returns. The recent 10% dip in Nifty (from its all-time high in September 2024 to its lowest in November) has left many impatient and questioning their strategies.  So, how do you build a portfolio that can handle market volatility and grow sustainably? → Multi-asset funds combine debt, equities and assets like real estate or gold. These funds adjust allocations based on market trends like equity valuations, gold-silver ratios and broader economic factors.  → For high-net-worth individuals, AIFs like corporate credit or long-short funds offer uncorrelated returns, adding stability to portfolios. These reduce market dependency, have the opportunity for regular income and enhance risk-adjusted returns  → Financial goals can be planned or unplanned. Maintain an emergency fund in low-risk options like arbitrage funds for unplanned goals. For planned goals, invest based on timelines, risk tolerance and your vision for the future.  → Investor behavior often has loss aversion and recency bias. In 44 years, Nifty has seen 10% dips within a year 40 times but ended the year positively 35 times. So missing just the 10 best days in 20 years can cut your returns by 50%. → Periodic reviews ensure your portfolio aligns with goals and risk tolerance. Rebalancing helps restore asset allocation, reduce risk exposure and capitalise on market opportunities  Markets will have ups and downs, but wealth creation is about resilience. Focus on disciplined investing, managing emotions and building a strategy.  How are you preparing your portfolio to handle market volatility? #investment #portfolio

𝗠𝗲𝗮𝘀𝘂𝗿𝗶𝗻𝗴 𝗟𝗲𝗮𝗱𝗲𝗿𝘀𝗵𝗶𝗽 𝗗𝗲𝘃𝗲𝗹𝗼𝗽𝗺𝗲𝗻𝘁 I've been asked this at least 3 times in the last two months. "How do I know that my leaders are improving?" This is where we distinguish knowing from application. 10% of capability comes from learning from formal sources. 20% comes from networks and interactions. 70% comes from application to portfolios and projects. One thing that sets this all apart are data points. Even if I apply skills to my projects, how do I know I did it well? Most large companies have a 360-degree or leadership assessment process in place. So, I'll share my thought process for this in case you are attempting to develop this for your own organization. Step 1: Determine organizational strategy and business outcomes. This is necessary to align expectations of desired behaviors. This is where a Balanced Scorecard can come in handy. Step 2: Assess expectations of leaders. You'll then assess them across leadership behaviors for new, mid and even senior managers. Granularity of differences supports focus and clarity. Often, a list of pre-existing behaviors/competencies are used to make the exercise easier. Validated psychometric tools such as the 16PF help to anchor it to scientific rigor. Organizational psychologists like me conduct surveys to gather insights. Then, focus groups are used to drill down to details information. After that, we'll create categories basedon the information and produce working behavior-based definitions. Step 3: Prioritize the list Now, the leadership team decides which behaviors are more important by way of ratings. Step 4: Build the 360 We then build a 360-degree feedback survey questions. These questions are reviewed for validity. Step 5: Allocate the survey A system specializing in the 360 (there are many) can be used. Feedback Recipient selects 6 to 12 people to rate them. In organizations, to avoid selection bias, leaders of the feedback recipient can review and veto the people doing the rating. Then, the participant does the survey too (self-rating) Step 6: Debrief of survey Usually, participants need guidance from a trained coach who understands feedback requirements. This is to provide grounding and objective input. Often, 360 surveys tend to be met with resistance unless the coach is skilled in facilitating the reflection conversation. Step 7: Action Planning The participant then produces a set of actions for improvement. This plan and the priority of focus should be made known to the feedback givers. Step 8: Pulse Surveys After a designated time (within 6 to 12 month period) a validated pulse survey is set up for the observers to rate improvement in specific behaviors. Step 9: Continued Leadership Coaching, Mentoring and Peer Support A combination of these can be used to enhance development. Step 10: Final Comparison Survey Toward the end of the year, a comparison survey is done to see how the key areas have improved or not. ---

𝗦𝗸𝗶𝗹𝗹‑𝗯𝗮𝘀𝗲𝗱 𝗮𝘀𝘀𝗲𝘀𝘀𝗺𝗲𝗻𝘁𝘀 𝗽𝗿𝗲𝗱𝗶𝗰𝘁 𝗠𝗮𝗻𝗮𝗴𝗲𝗿𝗶𝗮𝗹 𝗣𝗲𝗿𝗳𝗼𝗿𝗺𝗮𝗻𝗰𝗲 𝗳𝗮𝗿 𝗯𝗲𝘁𝘁𝗲𝗿 𝘁𝗵𝗮𝗻 𝗽𝗲𝗿𝘀𝗼𝗻𝗮𝗹𝗶𝘁𝘆 𝘁𝗿𝗮𝗶𝘁𝘀 𝗼𝗿 𝗮 𝘀𝗶𝗺𝗽𝗹𝗲 𝗱𝗲𝘀𝗶𝗿𝗲 𝘁𝗼 𝗟𝗲𝗮𝗱 ! 🧠 Great managers aren’t just good at leading people, they think differently. Great managers have higher fluid intelligence and make stronger economic decisions than their peers. 📈 And the impact is huge: A 1 SD increase in manager quality boosts sales 1.4x more than the same increase in the average quality of store employees. Management really matters. But here’s where it gets interesting… 🚫 Self‑promoted managers underperform; significantly. On average, they perform worse than managers who were randomly assigned to the role. 👀 Why? Because many “I want to be a manager” candidates overestimate their social skills, scoring lower on objective emotional‑perceptiveness tests like the Reading the Mind in the Eyes Test (RMET). And importantly: ⚖️ There is zero difference in managerial performance by gender, age, or ethnicity. Talent is universal. Opportunity is not, according to an interesting research published recently by a team of international researchers using data 📊 from publicly LinkedIn data limited to early-career workers. Researchers compared three ways of selecting managers: 1️⃣ Self‑promotion, 2️⃣ Random assignment (lottery) 3️⃣ Choosing based on four measurable skills The differences are huge. 🚀 When managers are chosen for their economic decision‑making skills, they perform 0.7 standard deviations better than those who rise through self‑promotion. ✅ 𝙈𝙮 𝙥𝙚𝙧𝙨𝙤𝙣𝙖𝙡 𝙫𝙞𝙚𝙬:  I keep coming back to one powerful insight: how we select managers shapes the entire organization. These findings make it clear that great management isn’t about personality, charisma, or who is the most eager to lead. It’s about measurable skills. And this research highlights something we often overlook and also researchers pointed out: in many workplaces, people who are more extraverted, more confident, or simply more vocal about wanting to lead are more likely to be promoted. But those traits don’t necessarily translate into better management. In fact, they can sometimes mask gaps in the very skills that matter most. That’s why these findings are so important. They suggest that organizations willing to look beyond self‑promotion, widen the pool of candidates, and screen for real skills, like economic decision‑making, can dramatically improve managerial quality and team performance. For me, this isn’t just research. It’s a roadmap. 🙏 Thank you international researchers team for these insightful findings: Achyuta Adhvaryu sonia bhalotra David Deming Anant Nyshadham Jorge Tamayo Ben Weidmann 🔑 Are we choosing leaders who want the role or leaders who can elevate others? #Leadership #Manager #SkillsBasedLeadership

India’s financial behavior is undergoing a silent transformation. In May 2025, mutual fund assets under management (AUM) reached 31% of total bank deposits in India, a significant jump from 16% in FY20, right before the pandemic. That’s a near doubling in just five years. This rise has come despite: > A global pandemic that froze economies > War in Europe and geopolitical instability > High global inflation and rate hikes > FIIs pulling capital out of emerging markets > Domestic market volatility and elections But this isn’t just a story about numbers. It’s a story about how Indians are beginning to think differently about money. Traditionally, Indian households parked most of their savings in fixed deposits, gold, or real estate, prioritizing safety over returns. But the post-COVID era has triggered a structural rebalancing: > Low interest rates nudged savers out of FDs. > Digital access & financial literacy brought mutual funds to the mainstream. > Rise of SIPs and retail equity participation changed the nature of capital formation. The result? A growing portion of household savings is now moving from guaranteed returns to market-linked risk capital. This shift has three major implications for India’s economy: 1. Rewiring financial intermediation: Banks are no longer the sole channel for household savings. Capital markets are stepping up. 2. Boosting domestic risk capital: A larger share of equity financing now comes from Indian investors, reducing dependence on volatile FII flows. 3. Sowing the seeds of an equity culture: Millions of Indians are learning to think like owners, not just depositors. This mindset shift is critical for long-term wealth creation and economic resilience. This is more than a cyclical boom. It’s a structural change in India’s financial DNA. And we’re just getting started. Despite the chaos, confidence is compounding.

Steady Growth Amid Long-Term Challenges Hedge fund assets under management (AUM) grew by an impressive $349 billion in 2024, reaching $4.88 trillion—a 7.7% year-to-date increase . However, this growth masks the underlying challenge of consistent net outflows over the past decade. Institutional investors, including public pension funds, have trimmed their hedge fund allocations, with the average allocation dropping to 7% from 8% five years ago. Despite this, performance has remained robust. Hedge funds collectively delivered a 10.1% return in 2024, outperforming public debt and demonstrating their value as risk mitigators. Strategies like niche and equity-focused funds led the pack with returns of 41.7% and 11%, respectively . Key Insight: Hedge funds remain a crucial component of diversified portfolios, offering downside protection and risk-adjusted returns despite allocation headwinds.

Are we really delivering the best possible forecasts with state-of-the-art foundation models if our models stop at historical patterns and ignore the external signals shaping the future? In the last few years, we've all seen how foundation models started transforming time series forecasting — unlocking strong zero-shot performance and making high-quality predictions possible without task-specific tuning. But the problem is that most of these models are univariate: they treat time series as isolated signals, leaving out exogenous factors that are often critical for accurate prediction. And that's not how forecasting works outside of a benchmark. Promotions, holidays, weather, pricing — these external influences often explain as much of the future as the past itself. Ignoring them leads to wider prediction intervals and forecasts that are harder to translate into real business decisions. So the real challenge now is: how do we bring that missing context into foundation models? That's the problem Chronos-2 was designed to solve. We built Chronos-2 to handle covariates and multivariate data in a zero-shot manner, and on benchmarks focused on these tasks, it achieves significant reductions in forecast error. But building a foundation model that can handle such diverse, context-dependent signals is not straightforward. Each forecasting task is unique — the number of features, their semantic meaning, and their interactions differ. The solution is a model that can adapt with in-context learning (ICL). Chronos-2 tackles this with two key components: 1. Architecture. In addition to standard temporal attention, we introduce group attention layers that enable information mixing across dimensions, allowing the model to learn from exogenous signals. 2. Training data. Multivariate and covariate time series data are extremely scarce, so we use synthetic data augmentation, adding multivariate structure on top of the univariate series commonly used for pretraining. The result is strong empirical performance across domains. In retail, Chronos-2 captures the impact of promotions on sales. In energy, it learns how weather influences energy consumption. In both cases, incorporating covariates significantly improves forecast accuracy and narrows prediction intervals — making forecasts more actionable. Chronos-2 is available under the Apache 2.0 license and ready to use. Give it a try and let us know what you think! 📄 Technical report: https://lnkd.in/d4RZG8Rq 💻 GitHub: https://lnkd.in/d9mvFT5B 📓 Example notebook: https://lnkd.in/dz69pCyu Abdul Fatir Ansari, Jaris Küken, Andreas Auer, Yuyang (Bernie) Wang, George Karypis, Huzefa Rangwala, Michael Bohlke-Schneider, Nick Erickson, Boran Han, Pedro Mercado, Syama Sundar Rangapuram, Huibin Shen, Lorenzo Stella, Amazon Science

We are happy to share the results of our exhaustive benchmarking study on forecasting models, where we assessed 87 models across 24 varied datasets. This project aimed to evaluate the performance of univariate forecasting models ranging from naive baselines to sophisticated neural networks, using a comprehensive set of metrics such as RMSE, RMSSE, MAE, MASE, sMAPE, WAPE, and R-squared. The 24 datasets contained a wide range of frequencies, including hourly (4 datasets), daily (5), weekly (2), and monthly (4), quarterly (2), yearly (3). Additionally, there are 4 synthetic datasets without a specific frequency. Some of the datasets also contain covariates (exogenous features) of static, past, and/or future nature. For each model, we aimed to identify hyperparameters that were effective on a global level, across all datasets. Dataset specific hyperpameter tuning for each model was not performed due to budget constraints on this project. We use a simple train/test split along the temporal dimension, ensuring models are trained on historical data and assessed on unseen future data. The attached chart shows a heatmap of the average RMSSE scores for each model, grouped by dataset frequency. The results are filtered to 43 models for brevity, excluding noticeably inferior models and redundant implementations. RMSSE is a scaled version of RMSE, where a model's RMSE score is divided by the RMSE of a naive model. With RMSSE, the lower the score, the better the model's performance. A score of 1.0 indicates performance on par with the naive baseline. Key Findings: - Machine-Learning Dominance: Extra trees and random forest models demonstrate the best overall performance. - Neural Network Success: Variational Encoder, PatchTST, and MLP emerged as top neural network models, with Variational Encoder showing the best results, notably including pretraining on synthetic data. - Efficacy of Simplicity: DLinear and Ridge regression models show strong performance, highlighting efficiency in specific contexts. - Statistical Models' Relevance: TBATS stands out among statistical models for its forecasting accuracy. - Yearly Datasets Insight: On yearly datasets, none of the advanced models surpassed the performance of the naive mean model, highlighting the difficulty of forecasting with datasets that lack conspicuous seasonal patterns. - Pretraining Advantage: The improvement in models like Variational Encoder and NBeats through pretraining on synthetic data suggests a promising avenue for enhancing neural networks' forecasting abilities. All models and datasets are open-source. For a detailed examination of models, datasets, and scores, visit https://lnkd.in/d6mMSudJ. Registration is free, requiring only your email. Our platform is open to anyone interested in benchmarking their models. Any feedback or questions are welcome. Let's raise the state of the art in forecasting!

Why I Think About Insurance Every Time I Look at My Portfolio You don't buy homeowner's insurance because you expect your house to burn down. You buy it because the cost of being wrong is catastrophic. Portfolio hedging works on exactly the same principle. Nassim Taleb, who literally wrote the book on extreme market events, advocates what's called "tail hedging." The concept is straightforward: spend a small, regular amount buying deep out-of-the-money put options to protect against rare but devastating crashes. Mark Spitznagel, who runs a hedge fund with Taleb as advisor, built a simple backtested strategy around this idea. Spend 0.5% of your equity portfolio per month on puts that are roughly 30% below the current market price. Roll them monthly. In normal months, they expire worthless. That's the insurance premium. In a crash, when SPY drops 20% and volatility spikes, those puts can increase 30-40x in value, offsetting most of your equity losses. The key insight: this strategy works best when the market is expensive. Spitznagel uses Tobin's Q ratio, total market value divided by replacement cost, as his gauge. When Tobin's Q is in its highest quartile, the tail-hedged portfolio outperforms a simple buy-and-hold approach by approximately 4% per year. Critics will say the cost is too high. A 6% annual drag from expired puts is real. But that cost only materializes in years when the market rallies continuously, precisely the years when your unhedged portfolio is performing well anyway. The question isn't whether you can afford the insurance premium. It's whether you can afford to be unprotected when the crash comes.