Financial Modeling Consulting

I have been building Valuation Models for 18 years! Some key points that I have learnt along the way.. Try these to improve your analysis significantly! ✅ Always understand the business before approaching valuation. This will form the core of your valuation model assumptions. - What does the company do? - How does it make money - What is the value chain? - Are their any competitive advantages that it has? ✅ Check for the financials. - Pick out annual reports, Punch in the data for last 4-5 years, and calculate ratios - Look at all major numbers and find out what they are. Other expenses are large - check what they are. Provisions in Balance Sheet are large - check them. - Look for trends in numbers and ratios - Are margins increasing, decreasing, or constant. Is working capital cycle moving around? Any other trends worth noticing? - Why are these happening? Try and identify what are the key reasons ✅ Projections - Project financials. Build a completely linked valuation model. - Whether you are doing DCF, or Relative Valuation - it is good to get a sense of how projected financials would look like. - The Balance sheet should link completely and the model should be dynamic. The model is NOT COMPLETE without this. - If you change assumptions - it should flow cleanly and the Balance Sheet should remain balanced ✅ Valuation - This comes last. Choose the appropriate method - FCFE, FCFF, Relative Valuation, SOTP. If not sure, try more than one method - Focus on the business. You can do a sensitivity on the discount rate, but it is usually difficult to do that on the business. - It is ok to be uncertain about the final price you get. But you need to know the relationship between business parameters like volume / pricing / costs and the valuation Learning Valuation Modeling is a process. Key is to keep practising and reading on the business. Try this the next time you are building a valuation model. ----- I help people build a #career in #valuation and #investmentbanking through my writing and courses. Follow me (Peeyush Chitlangia, CFA) to stay tuned for future posts.

Understanding common FP&A Financial Models with the help of an example. Let's understand five common financial models using a simple example: a coffee cart business. 1. Budgeting Model You estimate that you'll sell 3,000 cups this month and make a profit of ₹30,000. This is your plan before the month begins. Question it answers: What do I expect to happen? 2. Forecasting Model Halfway through the month, you realize sales are lower than expected. You revise your estimates and project a lower profit. Unlike budgets, forecasts evolve with reality. Question it answers: Given current trends, what is likely to happen? 3. Sensitivity Analysis What happens if you increase the selling price from ₹30 to ₹35? Or if milk prices rise? Here, you change one variable at a time to understand its impact. Question it answers: Which factor affects my results the most? 4. Scenario Analysis Now consider three situations: Best Case: Higher footfall and strong sales Base Case: Business performs as expected Worst Case: Heavy rains reduce customer traffic and costs increase This helps prepare for multiple possible futures. Question it answers: What if the overall business environment changes? 5. Break-Even Analysis You calculate the minimum number of cups you must sell to cover all costs. Everything beyond that point contributes to profit. Question it answers: What's the minimum I need to achieve to avoid losses? The models may differ, but the objective remains the same: Making better decisions with numbers. Which financial model do you use most often in your work?

Yesterday, I made a post about my career change. It was about aspiration to move from "Financial Aspects" of Modelling/Valuations to "Tech Aspect" of Modelling/Valuation. Currently, most of the Financial Models are built using Excel and that too very basic Excel (Lookups, Sumifs etc.) As we pivoted from bullock cart to EVs and from steam engines to aeroplanes our "spreadsheets" & "models" also need an upgrade. So here is the detailed list of items I shall be exploring: 1. 𝗗𝘆𝗻𝗮𝗺𝗶𝗰 𝗔𝗿𝗿𝗮𝘆𝘀 based modelling using 𝗟𝗘𝗧, 𝗟𝗔𝗠𝗕𝗗𝗔, 𝗦𝗘𝗤𝗨𝗘𝗡𝗖𝗘 𝗳𝘂𝗻𝗰𝘁𝗶𝗼𝗻𝘀 2. Leveraging 𝗣𝗼𝘄𝗲𝗿 𝗤𝘂𝗲𝗿𝘆 (𝗠 - 𝗟𝗮𝗻𝗴𝘂𝗮𝗴𝗲) & 𝗣𝗼𝘄𝗲𝗿 𝗣𝗶𝘃𝗼𝘁𝘀 (𝗗𝗮𝗫) for data analysis especially Big Data which is now available due to technology-based tracking 3. 𝗗𝗿𝗮𝗳𝘁𝗶𝗻𝗴 𝗿𝗲𝗽𝗼𝗿𝘁𝘀 & 𝗰𝘂𝘁𝘀 on Power BI like tools as Spreadsheets have some limitations 4. Exploring 𝗣𝗼𝘄𝗲𝗿 𝗔𝘂𝘁𝗼𝗺𝗮𝘁𝗲 to automate workflows & manual tasks 5. 𝗘𝘅𝗽𝗹𝗼𝗿𝗶𝗻𝗴 𝗣𝘆𝘁𝗵𝗼𝗻 𝗶𝗻 𝗘𝘅𝗰𝗲𝗹. It is a new update in Office 365 if you are not aware 6. Building Practical use-cases & products for 𝗚𝗲𝗻 𝗔𝗜 𝗶𝗻 𝗙𝗶𝗻𝗮𝗻𝗰𝗲 & 𝗠𝗼𝗱𝗲𝗹𝘀. 𝗖𝗵𝗮𝘁 𝗚𝗣𝗧 𝗯𝗮𝘀𝗲𝗱 𝗔𝗱𝗱-𝗶𝗻𝘀, 𝗖𝗼-𝗽𝗶𝗹𝗼𝘁 𝗲𝘁𝗰 7. 𝗙𝘂𝘁𝘂𝗿𝗶𝘀𝘁𝗶𝗰 𝗠𝗼𝗱𝗲𝗹𝗶𝗻𝗴 (𝗣𝗿𝗼𝗯𝗮𝗯𝗶𝗹𝗶𝘁𝘆 𝗗𝗶𝘀𝘁𝗿𝗶𝗯𝘂𝘁𝗶𝗼𝗻𝘀) to model uncertainty and variability in future events. It often incorporates a range of possible outcomes (simulations) rather than a single predicted value Any references/thoughts/opinions/suggestions are most welcome. If you are a finance professional, what future trends you envisage? #ChinmayaAmteExcel

Messy data makes updating financial forecasts challenging. Here are various ways I import actuals into my models. 1. Account Mapping (SUMIFS Approach) Map the general ledger to the financial model using account code or ranges. Then use SUMIFS to dynamically pull totals into your cash flow model. This works best when accounts are consistently structured and categorized. Flexible ranges allow you to group GL codes (eg. 4000–4099 = Revenue) for direct mapping from the trial balance to a standardized model layout. I’ve historically used INDIRECT to get this to work seamlessly across fragmented files, but many people prefer avoiding this because of the volatility of the function. 2. Power Query (Data Staging Approach) If you’re working within a recurring process with consistent formats from a system like NetSuite, Power Query is your friend. It let’s you clean the data before it enters your model. My suggestion is to stage everything in a structured data table inside Excel driven by Power Query. If you’ve ever joined me live, you’ve heard me call this a data vessel or intermediary sheet. From there, you can use SUMIFS, XLOOKUP, or dynamic arrays to populate the model every month. 3. Triggering Actuals in a Rolling Model In many of my rolling TWCFs, I include a header row with an Actual/Forecast toggle. It’s not just a label, it drives logic. Actuals can be pulled in with a trigger as they become available. Or you can use dynamic arrays to automatically bring them in. There is no need to hardcode. If you want to add a 14th week after actuals fill in for Q1, just extend the week columns and let the formulas follow. The same technique can apply to rolling monthly forecasts. 4. Watch Out for This Cash Flow Trap One critical mistake is dropping transactions due to date roll-forwards. If a payment doesn’t happen when expected, and your model simply rolls to the next week, that payment may vanish. And your cash balance won’t reconcile. To prevent this: (a) Build a formulaic variance tracker (b) Compare the current forecast to the prior period’s forecast (c) Highlight timing shifts vs. permanent misses (d) Keep visibility on future anticipated misses, not just historical variances When cash is king, timing is everything. These small techniques add up to a more accurate, more trusted forecast.

Do I use a framework to create Financial models? I use my own. Of course, on the beginning I didn't call it TACTIC. It was just a collection of my preferred practices when it comes to creating financial models. But then I found myself repeating the same steps, using the same structure, relying on the same techniques. Isn't that what we call a framework? 🔹 𝗪𝗵𝘆 𝗧𝗔𝗖𝗧𝗜𝗖? Traditional models can be rigid and quickly outdated as business needs evolve. TACTIC models are designed to be dynamic and adaptable, enabling continuous improvement and enduring relevance. So let's break down its components: Ⓣ Target – Everything starts with clear, specific business questions. From budget planning to evaluating potential mergers, it's crucial that you know why you need that model. What is the business question you will answer? What is the Target? Ⓐ Assets – More than just data, assets include the contextual information and assumptions that deepen our understanding and enrich our models. Ⓒ Calculations – Here, we convert our assets into actionable calculations. This core processing stage is where our data becomes insights. Ⓣ Tools – This layer allows for the application of additional calculations and scenarios, giving us the flexibility to tailor our model to answer varied business questions without overhauling the base model. Ⓘ Insights – The apex of the TACTIC model where all analysis culminates into clear, actionable insights, answering our initial questions and guiding strategic decisions. Ⓒ Continuation or Correlations– TACTIC doesn’t stop at insights. It propels us forward, prompting new questions, strategies or correlated analysis, ensuring our models are as dynamic as the markets we operate in. But to me, the main advantages are: 🔄 The Iteration – By revisiting and refining each layer as new data and strategies emerge, TACTIC ensures my financial models remain precise, relevant, and aligned with evolving business objectives. 🧩 The Modular Design – With its distinct layers, TACTIC allows for quick adaptations—whether updating calculations or swapping analytical tools, flexibility is at its core.

The recent market shocks have left a tremendous effect on investors’s mindmap. The volatility and the jump in the asset prices movements are extremely high. On a behavioural finance level, there is surely panic in the market leaving less headroom to ponder about the situations for normal retail investors. Thus, the implementation of mathematical models becomes a necessity not only to predict pricing value but considering volatility, jumps and high shocks. Although, the reason is different but at the end considering the dip in Japan stock market was lower than the Covid-19 pandemic. Using stochastic process mathematical models like Heston model could be used to predict both the asset price and its volatility, allowing for a mean-reverting volatility process while Hull White model for incorporating jumps in the asset prices. This way we get the volatility, jump and asset price. Also, if we consider multivariate volatility (time varying correlations with standardized returns) with correlation b/w the multiple assets, a great recommendation to opt for the extended GARCH model with dynamic conditional coorelation (DCC). Once you could predict the dynamic correlation with varying time portfolio optimization becomes more efficient with time-varying covariance matrix. No wonder, why maths with finance using tech makes such predictions better and high accuracy rates. #quantitativefinance #quant #finance #riskmanagement #japan

Portfolio optimization, grounded in Modern Portfolio Theory (MPT), is the foundational process of selecting the optimal distribution of assets to achieve maximum financial return while minimizing investment risk. Traditional financial methods like mean-variance optimization (MVO), uniform constant rebalanced portfolios (UCRP), and standard factor-based investment strategies are still widely adopted for asset allocation. In the last decade or so, quantitative finance has shifted toward machine-/deep-learning (ML/DL) and reinforcement learning (RL) to automate trading decision-making. However, current portfolio optimization approaches still face critical challenges. Traditional methods rely too heavily on rigid, historical data assumptions and struggle to adapt to volatile environments. Meanwhile, pure RL models suffer from a narrow focus; they primarily optimize for technical features like price signals or model architectures, completely ignoring macro market conditions and established economic theories (such as factor-based insights), leading to unstable performance during regime shifts. To bridge this research gap mentioned above, the authors of [1] introduce the Dynamic Factor Portfolio Model (DFPM), a hybrid framework that embeds financial domain expertise directly into a Deep Reinforcement Learning (DRL) structure. The DFPM addresses current shortcomings by utilizing a dual-module system: • Dynamic Factor Module (DFM): It tracks and dynamically scores five macroeconomically significant fundamental factors; Size, Value, Beta, Investment, and Quality. • Price Score Module (PSM): It analyzes real-time individual asset price data and inter-asset correlations. By integrating macroeconomic trends via the DFM with stock-level patterns from the PSM, the RL agent gains a comprehensive perspective. This enables the DFPM model to execute highly adaptive, interpretative, and stable asset weight adjustments as market environments shift. The DFPM was benchmarked against prominent baselines, including traditional strategies (like MVO, UCRP and conventional factor models) and state-of-the-art RL methods (such as PPO, A2C, and DDPG) across rigorous testing on the Nasdaq 100 and Dow Jones datasets. The experimental results demonstrate that the DFPM consistently and significantly outperforms all benchmarked baselines. It achieves superior risk-adjusted returns, as evidenced by its higher Sharpe ratios and Fractional Accumulated Portfolio Value (fAPV). The DFPM proves to be better precisely because it utilizes 'dynamic factor-informed knowledge' to recognize broad market contexts. This ensures it captures upward momentum during bull markets while aggressively reducing drawdowns and mitigating capital loss during periods of high volatility. The link to the paper [1] is posted in the comments.

This paper introduces a novel component modeling approach that separates deposit decay into two distinct behaviors: account closure (customer stickiness) and average balance fluctuation (balance stickiness). While we often assume customer loyalty enhances deposit stability, reality is more nuanced, a loyal customer with multiple products may be less likely to leave the bank entirely, but still has economic incentives to seek better yields or withdraw uninsured balances during stress periods. The proposed framework allows banks to statistically identify true customer stickiness while using a combination of data and expert judgment to model balance sensitivity, particularly for larger, more sophisticated customers who respond more actively to economic incentives. The model is deliberately flexible so that institutions can customize variables (e.g. one can include explanatory variables for direct deposit, transaction frequency, geography, etc. into the closure rate model if so desired) and specify different interest rate spread definitions, credit spread measures (e.g. bank-specific CDS, industry benchmarks), and core/non-core classifications for the average balance model based on their specific needs. I encourage practitioners to adapt this framework for their institutions and share improvements as we work toward more standardized, dynamic deposit modeling approaches for both IRRBB and liquidity risk management. #BankTreasury #ALCO #IRRBB #LiquidityRisk #DepositModeling #BTRM

Direct Jump: Redefining Simulation in Stochastic Processes and Quantitative Finance 1. Evolution of Simulation Techniques: A Deeper Look → Early financial simulations relied on pathwise construction — tracing every small move an asset could make over time. → While intuitive, this approach faced growing challenges: • Computational Overload: Simulating each tiny step multiplied effort exponentially. • Error Accumulation: Small rounding errors compounded across steps, distorting outcomes. • Misaligned Focus: Most financial decisions depend on final outcomes, not the exact path taken. → Over time, it became clear that tracing every micro-movement added noise without enhancing insight. Finance needed a simulation philosophy focused on where the system ends, not every fluctuation along the way. → Direct Jump emerged from this realization: By simulating the final state directly from the start, it allowed models to bypass unnecessary complexity while staying true to the core uncertainty being measured. It marked a shift from microscopic noise-tracking to macroscopic risk-understanding. 2. What is Direct Jump: A Conceptual Revolution → Direct Jump simulates the future outcome of a stochastic process without calculating every intermediate step. → Instead of walking through each small movement, it samples directly from the known or approximated distribution at the desired horizon. → Many financial models — like Geometric Brownian Motion for stock prices — allow the final distribution to be expressed analytically, making direct sampling both feasible and precise. → Why Direct Jump Matters: • Purity of Simulation: Each scenario accurately reflects the model’s full intended uncertainty. • Efficiency: Dramatically reduces simulation time without sacrificing realism. • Focus on Relevance: Models what financial decisions are truly based on — the final result, not the path. → In stochastic modeling, Direct Jump became critical because it aligns perfectly with real-world financial objectives: Understanding future risks cleanly, efficiently, and without unnecessary artifacts. 3. Where Direct Jump is Critical → Risk Management: Simulating end-of-horizon outcomes (e.g., Value-at-Risk) becomes scalable and robust. → Long-Term Derivative Pricing: For contracts where only terminal values matter, Direct Jump ensures precise payoff modeling. → Strategic Planning: Portfolio forecasts, reserve calculations, and solvency assessments benefit from streamlined, outcome-focused simulations. → In each case, Direct Jump realigns simulation practices toward clarity about the future, rather than clutter about the past. #QuantitativeFinance #StochasticProcesses #DirectJump #MarketSimulation #FinancialModeling #RiskManagement #MonteCarloSimulation #QuantFinance

Working capital forecasting is where many financial models quietly break. Revenue gets projected carefully. Expenses are modeled in detail. But AR, AP, and inventory often get treated as static percentages. And that’s how cash forecasts drift away from reality. A better approach is to build a days-based working capital engine that dynamically converts your forecasted P&L into balance sheet movements. Here’s the simple framework I use in FP&A models. Instead of hardcoding balances, project them using operating drivers. Example structure: Accounts Receivable AR = (Revenue ÷ 365) × AR Days Inventory Inventory = (COGS ÷ 365) × Inventory Days Accounts Payable AP = (COGS ÷ 365) × AP Days Now your balance sheet automatically adjusts when revenue or cost forecasts change. More importantly, it allows FP&A teams to model operational improvements instead of just reporting them. For example: • Reducing AR days from 52 to 45 immediately shows the cash impact • Extending AP days by 5 days improves liquidity without touching the P&L • Improving inventory turns releases trapped cash • Leadership can instantly see how operational decisions affect the cash conversion cycle And the best part? You can run scenario analysis by adjusting only three inputs: AR days, AP days, and inventory days. Simple model. Powerful insight. That’s the difference between a financial model and a decision model. Curious how others handle this: Do you forecast working capital using days-based drivers or still rely on percent-of-revenue assumptions?