Alex Caffarini’s Post

Outliers in your data can greatly distort your analysis and results and quickly lead to bad decision making. Check out our latest post, "The Danger of Outliers: How to Find and Correct Them." https://lnkd.in/e9y5_s89

The FRED Dashboard tool is very cool if you want to have a quick glance of all the macro factors in the economy for a simple business cycle analysis. All the charts you like in one webpage. https://lnkd.in/d83UbTKg

Outliers in data is like filling water in a tank which has a whole in it. I say because when we apply correct logic on incorrect data, it won't yield any results. So, when data is correct logic works like magic. In my project of credit card analysis, everything was dependent on Annual Income because the credit score of customers was not available. So, the reduction of outliers in the annual income enhanced results and reduced skewness.

🔨 Addressing Data Censoring in Credit Analytics and Model Building #AddressingDataCensoring #CreditAnalyticsandModelBuilding

Starting an effective data practice at your bank or credit union doesn't have to be complicated. Or even budgeted. Literally, it can be you and three other teammates meeting every week... Aggregating a few sources. And discussing the data behind your decisions. Start small. Build momentum. And watch what happens :)

New Strategies for 2024 As we close out 2023 and look ahead into 2024 hopefully you are already thinking about your risk strategies and asking yourself what might need to change so that you can maintain acceptable levels of risk. Perhaps you are evaluating underwriting criteria, and or risk score performance, as you think about how best to manage the portfolio going forward. One of the typical exercises used to evaluate needed change is to conduct a retrospective analysis. During this process lenders often gather archive credit data and match that with your own performance data to understand what attributes would enhance segmentation or what data would improve score performance. If you are undergoing this type of typical process, I would urge you to consider Digital Matrix Systems. Our Credit Bureau data normalization and standardization process enables you to leverage tri- bureau summary attributes that can improve your segmentation and modeling schemes while eliminating the dependency on any one data source. I've personally been working with lenders for nearly 30 years in developing data sets for these analytic purposes and DMS far and away has the best set of summary attributes in the market today. If you would like to learn more reach out to set up a quick intro call. I promise it will be worth your time.

Cross-market analysis is a vital component of credit research. 📊 Direct comparisons between companies, markets and sectors are key to finding relative value - but it's a challenging process. Every analyst knows the pain of manually updating a ‘Comps sheet’, pulling financials from disparate models across multiple locations. Fortunately, we’ve solved this challenge. In the first of our series, we take a look at one of our most popular features - Comparables - and explain how it enables fast and convenient cross-market analysis. Read the post here: https://lnkd.in/eFmcxbuC #Credit #Bonds #CreditAnalysis #FinancialAnalysis

Doing a little research - love to hear your thoughts on this ! Let's say you have 10M unique data records for sale.  No matter which choice you make below, you'll sell them to a company who will package them with other data records then resell that new, cumulative data set to verified buyers. You're presented with 2 options - both of which are actively used in your market today. You won't be the only one doing whichever choice you make. (A) You can sell your data records to one of the leading companies in the market who will give you $3.25 per record. If you accept this, you can't sell them to anyone else.  (B) You can sell your data records to the three leading companies in the market who will each give you 35% of the revenue they generate from your records, with no cap. And if you find a 4th or 5th company in the market who is interested in the same deal, you can sell to them too. This is a financial analysis and risk choice.  Which one are you more comfortable with ?  

The best savings account for you 💰 Rand vs... In 2023, we thoroughly analyzed the offerings of various accounts in the sector and enhanced ours. ✅ We won't try to convince you of anything; we simply assert that the data doesn't lie. Examine the differences and decide with certainty. 🧠😉Try now ⬇ https://lnkd.in/dKT-cbSH

🔍 This is a short tutorial on inference procedures on categorical data. First we go over risks and odds ratios concerning bank and credit union ratings, then we go over market research data and work our way towards practicing chi-square tests on some medical data. 📊 #statistics #rprogramming #datascience https://lnkd.in/gFehsP25