Quant Hub

Investment banks like Morgan Stanley, Goldman and JP Morgan have specialized Quant desks which focus on specific asset classes or trading strategies Here's a glimpse into some of the key quant desks - Electronic Trading Desk: This is the high-speed, algorithm-driven hub. Quants here develop and implement complex trading models that execute trades in milliseconds, capitalizing on fleeting market opportunities. They work closely with electronic traders to ensure the algorithms are optimized and react swiftly to market changes. Delta-One Desk: Delta One products are essentially derivatives that have a delta close to or equal to one. This desk focuses on replicating the performance of underlying assets through a combination of various instruments. Quants here design strategies using futures, options, swaps, and other derivatives to achieve a "delta-one" exposure, meaning the price movement of the created product mirrors the underlying asset. Exotics Desk: This desk deals with complex, non-standard financial instruments like barrier options or knock-in puts. Quants here develop pricing models and risk management frameworks for these exotic derivatives, catering to clients with sophisticated investment needs. FX (Foreign Exchange) Desk: Here, quants build models for analyzing currency exchange rates and developing trading strategies. They incorporate factors like economic data, interest rate differentials, and political events to predict currency movements and advise traders on how to navigate the FX market. Rates Desk: This desk focuses on interest rate products like bonds, swaps, and futures. Quants here develop models to price these instruments, assess interest rate risk, and create strategies for clients looking to manage their exposure to interest rate fluctuations. Banks might have additional specialized desks focusing on commodities, credit derivatives, or even specific industries. The specific structure depends on the bank's overall strategy and client base.

Maverick Derivatives gave INR 1.5 crore at one of the IITs for their Amsterdam office during campus Placement Here is a question they asked in their test Consider a random permutation of the numbers 1,2,…,1000. Let A be the number of integers in the same position as before and B be the number of integers in a different position. What is the Variance of the quantity (B-A)? Solution - image attached below 100K Sale- 25% off on all our Products, Use the Coupon Code "100K25" to get 25% off ➡ Kickstart your Quant Interview Prep ↗ ‘Interview Byte’ contains 500+ Interview questions (https://lnkd.in/gkqcrrKf) ↗ Quant Insider Project Handbook has 15 industry-oriented projects, which include 10 industry-oriented projects based on challenges conducted by Top HFT's and Hedge Funds. (https://lnkd.in/gWBEn78U) ↗ Check out Quant Insider Stack - https://lnkd.in/gcfdUEfg A Bundle of Interview Byte and Project Handbook Quant Insider Career Catalyst is your guide to all interview prep tips, preparation roadmap and job application strategies (https://lnkd.in/gVhA4tNG) Resume Review - You will learn to make a tailored resume for the Job Description, effective use of keywords, and bullet points to create an impactful resume. We will be working on giving the best shape to your resume for your dream Quant role. https://lnkd.in/gi6yznXa ------------------------------------------------------------------------------------ Machine Learning for Finance course- Designed by Industry Veterans Hariom Tatsat, CQF, FRM with years of working at Wallstreet - https://lnkd.in/gtJDWcus Use Coupon code - "EARLYBIRD20" for 20% off on the ML for Finance course -------------------------------------------------------------------------------- Also check "𝐐𝐔𝐀𝐍𝐓𝐈𝐓𝐀𝐓𝐈𝐕𝐄 𝐀𝐧𝐚𝐥𝐲𝐬𝐭 𝐃𝐞𝐯𝐞𝐥𝐨𝐩𝐞𝐫 𝐒𝐭𝐫𝐚𝐭 -𝐓𝐇𝐄 𝐏𝐑𝐎𝐅𝐄𝐒𝐒𝐈𝐎𝐍" Course, taught by Andrey Chirikhin 𝐂𝐥𝐢𝐜𝐤 𝐨𝐧 𝐭𝐡𝐞 𝐋𝐢𝐧𝐤 𝐭𝐨 𝐄𝐧𝐫𝐨𝐥𝐥 𝐢𝐧 𝐭𝐡𝐞 𝐂𝐨𝐮𝐫𝐬𝐞 𝐍𝐨𝐰 - https://lnkd.in/gikuMSRg Use Coupon Code -"partner25" to get 25% off on the Course We have 𝐒𝐩𝐞𝐜𝐢𝐚𝐥 𝐩𝐫𝐢𝐜𝐞 𝐟𝐨𝐫 𝐭𝐡𝐞 𝐬𝐭𝐮𝐝𝐞𝐧𝐭𝐬- £𝟏𝟒𝟗 To get the course at £𝟏𝟒𝟗, you just need to register on the Website with your University Email ID and you will receive the Coupon For more quant finance memes follow us on Instagram - Quant Insider (https://lnkd.in/gfjc4hBu)

A detailed breakdown of Discrete Dynamic Delta Hedging with Python code implementation A risk management strategy where an options trader periodically adjusts their hedge position to maintain a delta-neutral portfolio, mitigating the risk from movements in the underlying asset's price. Key Concepts: Delta (Δ): Measures the sensitivity of an option's price to changes in the price of the underlying asset. Delta-Neutral Portfolio: A portfolio where the net delta is zero, meaning small movements in the underlying asset's price have minimal effect on the portfolio's value. Implementation: Periodic Adjustments: Traders rebalance their hedge at discrete time intervals (e.g., daily or weekly) rather than continuously. Hedging Actions: Involves buying or selling the underlying asset to offset changes in the option's delta. Practical Considerations: Discrete Intervals: Continuous hedging is impractical due to transaction costs and liquidity constraints, so adjustments are made at specific times. Hedging Errors: Price movements between hedging intervals can cause the hedge to be imperfect, leading to residual risk. Frequency Trade-off: More Frequent Hedging: Reduces hedging errors but increases transaction costs. Less Frequent Hedging: Lowers transaction costs but increases exposure to risk. Effectiveness: Risk Reduction: Improves with increased hedging frequency but cannot eliminate risk entirely. Option Characteristics: Out-of-the-Money Options: Hedging errors are more significant due to higher sensitivity to price movements. Near Expiration: Options approaching maturity may exhibit greater volatility, increasing hedging challenges. 100K Sale- 25% off on all our Products, Use the Coupon Code "100K25" to get 25% off ➡ Kickstart your Quant Interview Prep ↗ ‘Interview Byte’ contains 500+ Interview questions (https://lnkd.in/gkqcrrKf) ↗ Quant Insider Project Handbook has 15 industry-oriented projects, which include 10 industry-oriented projects based on challenges conducted by Top HFT's and Hedge Funds. (https://lnkd.in/gWBEn78U) ↗ Check out Quant Insider Stack - https://lnkd.in/gcfdUEfg A Bundle of Interview Byte and Project Handbook Quant Insider Career Catalyst is your guide to all interview prep tips, preparation roadmap and job application strategies (https://lnkd.in/gVhA4tNG) Machine Learning for Finance course- Designed by Industry Veterans Hariom Tatsat, CQF, FRM with years of working at Wallstreet - https://lnkd.in/gtJDWcus Use Coupon code - "EARLYBIRD20" for 20% off on the ML for Finance course "𝐐𝐔𝐀𝐍𝐓𝐈𝐓𝐀𝐓𝐈𝐕𝐄 𝐀𝐧𝐚𝐥𝐲𝐬𝐭 𝐃𝐞𝐯𝐞𝐥𝐨𝐩𝐞𝐫 𝐒𝐭𝐫𝐚𝐭 -𝐓𝐇𝐄 𝐏𝐑𝐎𝐅𝐄𝐒𝐒𝐈𝐎𝐍" Course, taught by Andrey Chirikhin - https://lnkd.in/gikuMSRg We have 𝐒𝐩𝐞𝐜𝐢𝐚𝐥 𝐩𝐫𝐢𝐜𝐞 𝐟𝐨𝐫 𝐭𝐡𝐞 𝐬𝐭𝐮𝐝𝐞𝐧𝐭𝐬- £𝟏𝟒𝟗, you just need to register with your University Email ID to receive the Coupon

When a retail trader decides to move to algorithmic trading, they encounter a unique set of challenges that can make the transition difficult. Here are the key problems they face: 1. Limited Access to High-Quality Data Problem: Retail traders may rely on free or low-cost data that is often delayed, inaccurate, or insufficient for developing effective algorithms. They may struggle to access the same level of detailed and real-time data required for successful execution. 2. High Latency and Execution Speed Problem: Even small delays in order execution can result in significant losses, especially in fast-moving markets. 3. Cost of Infrastructure Building an execution platform requires substantial investment in technology infrastructure, including powerful computers, servers, and reliable internet connections. Problem: Without adequate hardware and software, retail traders may struggle to handle the computational demands of running complex algorithms, particularly in real-time environments. This can lead to performance system crashes, or inefficiencies in the execution process. 4. Backtesting and Strategy Development Problem: Retail traders may struggle to find affordable backtesting tools that simulate realistic market conditions. Many free platforms are either too limited in functionality or do not provide accurate data. 5. Data Storage and Management Algo trading platforms generate large volumes of data, including trade logs, order histories, and performance metrics. Storing and managing this data efficiently is a challenge, particularly for retail traders with limited resources. Problem: Retail traders may not have the tools or storage infrastructure to manage this vast amount of data. As a result, their ability to review past performance, troubleshoot errors, or refine strategies may be hindered. 6. Lack of Technical Knowledge Building an execution platform for algorithmic trading requires proficiency in both programming and understanding of financial markets. Retail traders may not have the coding skills needed to develop sophisticated algorithms, let alone implement them in a production environment. Problem: Retail traders without a background in programming (such as Python, C++, or other relevant languages) may struggle to write and optimize their own trading strategies. Even if they do find pre-built frameworks, customizing them to fit their own strategies or goals can be a daunting task. What's the solution? Quant Insider "Solution to Retail algo trading problem" is designed specifically for retail traders who want to take their trading to the next level. By providing high-quality data, low-latency execution, affordable infrastructure, powerful backtesting, seamless data management, and easy-to-use tools. To learn more about the product please fill out the Google form https://lnkd.in/gBWmCmkp

"Embracing Generative AI in Credit Risk" discusses how generative AI (gen AI) is being adopted in the credit risk industry. Adoption in Credit Risk: Generative AI's Rise: The rapid growth of generative AI, particularly after OpenAI's release of ChatGPT, has made its way into traditionally conservative sectors like credit risk. By 2023, major tech companies were already embedding AI capabilities into their products. Survey Results: A McKinsey survey revealed that 20% of credit risk organizations have already implemented generative AI, while 60% expect to do so within the year. Use Cases Across the Credit Life Cycle: Client Engagement: Gen AI is being used to offer hyperpersonalized products, draft communications for relationship managers, and assist with product identification. Credit Decision and Underwriting: AI tools can flag policy violations, draft communications for missing information, and help generate credit memos, thus streamlining decision-making. Portfolio Monitoring: AI automates report generation, summarizes optimization strategies, and can optimize early-warning systems to identify borrowers at risk. Customer Assistance: Gen AI assists in drafting personalized communications, guiding restructuring processes, and enhancing agent-customer interactions in real-time. Challenges in Scaling AI: Risk and Governance: Major risks identified include fairness in algorithms, data privacy issues, security vulnerabilities, and regulatory challenges. These risks must be managed carefully to avoid reputational and legal consequences. Internal Challenges: A lack of AI expertise, data quality issues, and decentralized project management are obstacles to successfully scaling gen AI. Only a small fraction of organizations have formalized support structures like centres of excellence (CoE) for AI initiatives. Implementation and Acceleration: Institutions that implement modular solution architectures and reuse open-source tools can reduce deployment time by 30-50%, making it possible to rapidly scale gen AI solutions in just a few weeks. Checkout our course Machine Learning for Finance -https://lnkd.in/gtJDWcus Use Coupon code - "EARLYBIRD20" for 20% off on the ML for Finance course 15+ Real-World Practical Applications Financial Applications Coverage - Algo Trading - Portfolio Management - Fraud detection - Leanding and Loand Default prediction - Sentiment Analysis - Derivatives Pricing and Hedging - Asset Price Prediction - and many more Course Description Supervised Learning Regression and Classification models 1. Linear and Logistic Regression 2. Random Forest and GBM 3. Deep Neural Network (including RNN and LSTM) Includes 6+ case studies Unsupervised Learning Clustering and Dimensionality Reduction 1. Principal Component Analysis 2. k-Means and hierarchical clustering Includes 5+ case studies Reinforcement Learning and NLP

Volatility Models Cheatsheet - Detailed overview of different volatility models 1. Constant Volatility Model Math: The volatility (σ) is assumed to be constant. The Black-Scholes formula and partial differential equations (PDEs) are used for pricing options under this assumption. Popularity: for pricing vanilla options (standard options with basic payoff structures). Key Characteristics: Despite being simple and tractable, it may not always accurately reflect market conditions, especially during periods of high volatility. 2. Deterministic Volatility Model Math: The volatility is modelled as a function of the underlying asset price (SSS) and time (t). The Black-Scholes PDE is extended to accommodate this deterministic but non-constant volatility. Popularity: for exotic options (complex options with path-dependent payoffs) where the constant volatility assumption doesn't hold. Key Characteristics: It accounts for changes in volatility over time but still lacks the flexibility to capture random shocks or sudden jumps in volatility. 3. Stochastic Volatility Model Math: The volatility is driven by its own stochastic process. The differential equation for volatility (dσ) includes stochastic terms, leading to higher-dimensional models. The use of advanced transforms is often required to handle this increased complexity. Popularity: for exotic options. Key Characteristics: More realistic compared to constant and deterministic models, but at the cost of increased mathematical and computational complexity. Often used in conjunction with Monte Carlo simulations or Fourier transforms for option pricing. 4. Jump Diffusion Model Math: Incorporates jumps in both the stock price and volatility. These jumps are modeled using Poisson processes, adding discontinuities to the otherwise smooth paths assumed by standard models. Popularity: Increasing in popularity due to its ability to capture sudden market movements that are not explained by continuous diffusion models. Key Characteristics: useful during events like earnings reports, political news, or financial crises when large, unexpected jumps in asset prices and volatility can occur. It is more flexible than purely diffusion-based models. Stochastic Volatility and Mean-Variance Model Math: Similar to the stochastic volatility model, but with an additional emphasis on mean-variance optimization, leading to higher-dimensional equations and more non-linearity in the pricing models. 100K Sale- 25% off on all our Products, Use the Coupon Code "100K25" to get 25% off ↗ ‘Interview Byte’ contains 500+ Interview questions (https://lnkd.in/gkqcrrKf) ↗ Quant Insider Project Handbook has 15 industry-oriented projects, which include 10 industry-oriented projects based on challenges conducted by Top HFT's and Hedge Funds. (https://lnkd.in/gWBEn78U) ↗ Check out Quant Insider Stack - https://lnkd.in/gcfdUEfg A Bundle of Interview Byte and Project Handbook

𝐖𝐞 𝐰𝐞𝐧𝐭 𝐟𝐫𝐨𝐦 𝟎 𝐭𝐨 𝟏𝟎𝟎𝐤+ 𝐢𝐧 𝟑𝟖𝟓 𝐝𝐚𝐲𝐬 We’re thrilled to announce that the Quant Insider LinkedIn community has officially grown to over 𝟏𝟎𝟎,𝟎𝟎𝟎 𝐟𝐨𝐥𝐥𝐨𝐰𝐞𝐫𝐬! To each one of our 100,000+ community members – thank you for being a part of Quant Insider! Your support, enthusiasm, and drive to learn and grow alongside us are what make this journey so rewarding. Together, we’re building a dynamic community that’s redefining what’s possible in quantitative finance. ‼ 𝗘𝘅𝗰𝗶𝘁𝗶𝗻𝗴 𝗧𝗵𝗶𝗻𝗴𝘀 𝗔𝗿𝗲 𝗖𝗼𝗺𝗶𝗻𝗴 ‼ We’ve been working on something groundbreaking – a powerful tool leveraging generative AI that will transform how you approach strategy creation. Imagine achieving new levels of ease and customization in your trading strategies. Curious? Stay tuned. 100K Sale- 25% off on all our Products, Use the Coupon Code "100K25" to get 25% off ↗ ‘Interview Byte’ contains 500+ Interview questions (https://lnkd.in/gkqcrrKf) ↗ Quant Insider Project Handbook has 15 industry-oriented projects, which include 10 industry-oriented projects based on challenges conducted by Top HFT's and Hedge Funds. (https://lnkd.in/gWBEn78U) ↗ Check out Quant Insider Stack - https://lnkd.in/gcfdUEfg A Bundle of Interview Byte and Project Handbook Quant Insider Career Catalyst is your guide to all interview prep tips, preparation roadmap and job application strategies (https://lnkd.in/gVhA4tNG) Resume Review - You will learn to make a tailored resume for the Job Description, effective use of keywords, and bullet points to create an impactful resume. We will be working on giving the best shape to your resume for your dream Quant role. https://lnkd.in/gi6yznXa ------------------------------------------------------------------------------------ Machine Learning for Finance course- Designed by Industry Veterans Hariom Tatsat, CQF, FRM with years of working at Wallstreet - https://lnkd.in/gtJDWcus Use Coupon code - "EARLYBIRD20" for 20% off on the ML for Finance course -------------------------------------------------------------------------------- Also check "𝐐𝐔𝐀𝐍𝐓𝐈𝐓𝐀𝐓𝐈𝐕𝐄 𝐀𝐧𝐚𝐥𝐲𝐬𝐭 𝐃𝐞𝐯𝐞𝐥𝐨𝐩𝐞𝐫 𝐒𝐭𝐫𝐚𝐭 -𝐓𝐇𝐄 𝐏𝐑𝐎𝐅𝐄𝐒𝐒𝐈𝐎𝐍" Course, taught by Andrey Chirikhin 𝐂𝐥𝐢𝐜𝐤 𝐨𝐧 𝐭𝐡𝐞 𝐋𝐢𝐧𝐤 𝐭𝐨 𝐄𝐧𝐫𝐨𝐥𝐥 𝐢𝐧 𝐭𝐡𝐞 𝐂𝐨𝐮𝐫𝐬𝐞 𝐍𝐨𝐰 - https://lnkd.in/gikuMSRg Use Coupon Code -"partner25" to get 25% off on the Course We have 𝐒𝐩𝐞𝐜𝐢𝐚𝐥 𝐩𝐫𝐢𝐜𝐞 𝐟𝐨𝐫 𝐭𝐡𝐞 𝐬𝐭𝐮𝐝𝐞𝐧𝐭𝐬- £𝟏𝟒𝟗 To get the course at £𝟏𝟒𝟗, you just need to register on the Website with your University Email ID and you will receive the Coupon For more quant finance memes follow us on Instagram - Quant Insider (https://lnkd.in/gfjc4hBu)

 Here are the key takeaways from the paper "Statistical Predictions of Trading Strategies in Electronic Markets," Prediction Accuracy: Logistic regression models on Euronext Amsterdam data achieve strong out-of-sample accuracy: 70% for trade direction (buy/sell), 85% for price bucket (categorizing limit prices), and 62% for volume bucket (size categories). Prediction accuracy significantly drops when excluding algorithm and member identities, highlighting the value of these features. Key Predictive Features: Direction: Primary drivers include LOB imbalance for the algorithm’s orders, best bid/ask prices, and intraday accumulated inventory. Price: Influenced by bid-ask spread, volumes at the best bid/ask, and imbalance of algorithm-specific orders within the LOB. Volume: Predicted by intraday inventory and cash variables of both the algorithm and the trading member, reflecting inventory and liquidity considerations. Clusters of Algorithmic Behavior: Directional Traders: Exhibit low order-to-trade ratios, lean towards aggressive orders aligned with their inventory direction, and don’t maintain balanced liquidity provision. Opportunistic Traders: Primarily liquidity takers (73% of transactions), posting orders away from best quotes to target higher profits at a lower frequency. Market Makers: High order-to-trade ratios, balanced bid/ask liquidity, high likelihood of posting at the touch, and a tendency to revert inventories, aligning with classic market-making strategies. Application to Agent-Based Market Simulations: Models and behavioural clusters provide empirical foundations for constructing agent-based simulations that capture individual agent decisions for direction, price, and volume. These simulations can help firms and regulators assess the impact of new strategies or interventions under different market conditions. ➡ Kickstart your Quant Interview Prep ↗ ‘Interview Byte’ contains 500+ Interview questions (https://lnkd.in/gkqcrrKf) ↗ Quant Insider Project Handbook has 15 industry-oriented projects, which include 10 industry-oriented projects based on challenges conducted by Top HFT's and Hedge Funds. (https://lnkd.in/gWBEn78U) ↗ Check out Quant Insider Stack - https://lnkd.in/gcfdUEfg A Bundle of Interview Byte and Project Handbook Quant Insider Career Catalyst is your guide to all interview prep tips, preparation roadmap and job application strategies (https://lnkd.in/gVhA4tNG) Machine Learning for Finance course- Designed by Industry Veterans Hariom Tatsat, CQF, FRM with years of working at Wallstreet - https://lnkd.in/gtJDWcus Use Coupon code - "EARLYBIRD20" for 20% off on the ML for Finance course "𝐐𝐔𝐀𝐍𝐓𝐈𝐓𝐀𝐓𝐈𝐕𝐄 𝐀𝐧𝐚𝐥𝐲𝐬𝐭 𝐃𝐞𝐯𝐞𝐥𝐨𝐩𝐞𝐫 𝐒𝐭𝐫𝐚𝐭 -𝐓𝐇𝐄 𝐏𝐑𝐎𝐅𝐄𝐒𝐒𝐈𝐎𝐍" Course, taught by Andrey Chirikhin - https://lnkd.in/gikuMSRg We have 𝐒𝐩𝐞𝐜𝐢𝐚𝐥 𝐩𝐫𝐢𝐜𝐞 𝐟𝐨𝐫 𝐭𝐡𝐞 𝐬𝐭𝐮𝐝𝐞𝐧𝐭𝐬- £𝟏𝟒𝟗, you just need to register with your University Email ID to receive the Coupon

Yesterday Quant Insider concluded 2 day workshop at Birla Institute of Technology and Science, Pilani HYD campus On the Day 1 there was a detailed workshop on “Quant Finance industry” And one Day 2 we had a detailed discussion and presentation of “Statistical Arbitrage - Pair Trading” It was great experience interacting with the Students of BITS Pilani Thank you to ADITYA RANJAN and The Wall Street Club team for the amazing hospitality.

Important points on Delta Hedging that every Volatility trader should consider picture credit- Trading Volatility by Colin Bennett Delta Hedging and Matching Maturity To hedge an option's delta effectively, use a forward with the same maturity. Using different maturities or stock introduces dividend risk, as forwards exclude interim dividend benefits. This distinction helps prevent potential miscalculations in P&L. Dividend Risk in Forward Contracts A forward contract equates to holding stock but shorting dividends payable before maturity. This matters because a stock drops by the dividend value on the ex-date, leading to a dividend risk equal to the option's delta if you hedge with mismatched instruments. Borrow Cost & Option Pricing Borrow cost affects derivatives prices similarly to dividends: option owners pay it, but forward owners don’t. Banks usually price this cost asymmetrically, impacting volatility traders’ deltas. It’s minor for General Collateral (GC) but crucial for emerging market stocks. Zero Delta Straddles and Borrow Cost Zero delta straddles need borrow cost in one leg, calculated as two separate trades (call + put). For high borrow cost names, like in emerging markets, this can significantly impact pricing Myth: ATM Straddles and 50% Delta Contrary to popular belief, ATM options don’t guarantee a zero delta straddle. A true zero delta straddle is usually set above spot, as ATM straddles tend to carry a slightly negative delta. Why Delta Hedging Removes Equity Risk Delta hedging neutralizes an option’s equity exposure, allowing traders to profit solely from volatility. If you purchase an option at low implied volatility vs. realized volatility, you could profit – but only if equity exposure (delta) is fully hedged. Last Day of Diwali sale - Use the Coupon Code "Festive30OFF" to get 30% off on all our products and service ↗ ‘Interview Byte’ contains 500+ Interview questions (https://lnkd.in/gkqcrrKf) ↗ Quant Insider Project Handbook has 15 industry-oriented projects, which include 10 industry-oriented projects based on challenges conducted by Top HFT's and Hedge Funds. (https://lnkd.in/gWBEn78U) ↗ Check out Quant Insider Stack - https://lnkd.in/gcfdUEfg A Bundle of Interview Byte and Project Handbook Quant Insider Career Catalyst is your guide to all interview prep tips, preparation roadmap and job application strategies (https://lnkd.in/gVhA4tNG) Resume Review - You will learn to make a tailored resume for the Job Description, effective use of keywords, and bullet points to create an impactful resume. We will be working on giving the best shape to your resume for your dream Quant role. https://lnkd.in/gi6yznXa Machine Learning for Finance course- Designed by Industry Veterans Hariom Tatsat, CQF, FRM with years of working at Wallstreet - https://lnkd.in/gtJDWcus Use Coupon code - "EARLYBIRD20" for 20% off on the ML for Finance course