<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by JJ on Medium]]></title>
        <description><![CDATA[Stories by JJ on Medium]]></description>
        <link>https://medium.com/@captioned?source=rss-5086e44692d2------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*-qeBEv-I_B-GR_YEAbo_FQ.jpeg</url>
            <title>Stories by JJ on Medium</title>
            <link>https://medium.com/@captioned?source=rss-5086e44692d2------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 14:10:42 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@captioned/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Thoughts on Inflation]]></title>
            <link>https://medium.com/human-in-a-machine-world/thoughts-on-inflation-6c46a8806e6e?source=rss-5086e44692d2------2</link>
            <guid isPermaLink="false">https://medium.com/p/6c46a8806e6e</guid>
            <category><![CDATA[life]]></category>
            <dc:creator><![CDATA[JJ]]></dc:creator>
            <pubDate>Sun, 26 Jun 2022 15:05:25 GMT</pubDate>
            <atom:updated>2022-06-26T15:05:25.120Z</atom:updated>
            <content:encoded><![CDATA[<p>As a consumer, I’d like to know how much more I need to spend over the same period of time for the same type of goods. So, if I’m needing to buy my basket of goods more frequently, that adds to the money I need to set aside for spending.</p><p>Really, I’d like inflation calculations to capture all three of the below:</p><ul><li>The price for the same goods goes up (the price of milk increases)</li><li>The amount of the goods I get for the same price goes down (the size of individual toilet paper sheets shrinks)</li><li>Something happens to force me to use more of the goods than I needed to before (the lotion pump pushes out more lotion in a single pump)</li></ul><p>Both CPI and PCE miss the last two aspects.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6c46a8806e6e" width="1" height="1" alt=""><hr><p><a href="https://medium.com/human-in-a-machine-world/thoughts-on-inflation-6c46a8806e6e">Thoughts on Inflation</a> was originally published in <a href="https://medium.com/human-in-a-machine-world">Human in a Machine World</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Reflections of a New Manager]]></title>
            <link>https://medium.com/human-in-a-machine-world/reflections-of-a-new-manager-2ca65a38f1be?source=rss-5086e44692d2------2</link>
            <guid isPermaLink="false">https://medium.com/p/2ca65a38f1be</guid>
            <category><![CDATA[data-science-leadership]]></category>
            <category><![CDATA[imposter-syndrome]]></category>
            <category><![CDATA[management]]></category>
            <category><![CDATA[life]]></category>
            <dc:creator><![CDATA[JJ]]></dc:creator>
            <pubDate>Mon, 11 Apr 2022 04:07:04 GMT</pubDate>
            <atom:updated>2022-11-19T05:16:41.312Z</atom:updated>
            <content:encoded><![CDATA[<h4>One year in as a data science leader and first time official manager</h4><p>About a year ago, I interviewed for a fairly important sounding role. Since this was an internal position, I did my due diligence and solicited advice from my network of people-in-the-know. Some gave me helpful guidance, some gave me honest feedback on why they didn’t think it was a good role, and one told me that I’ll likely get the job because I check three diversity boxes.</p><p>Fast forward two months, I was offered the position. When I told the people I asked advice from, they reacted mostly with shock and one person suggested that it must be because I know the right people.</p><blockquote>Learning #1: Taking on more visible roles means that I have to deal with more people having and sharing opinions about me.</blockquote><p>So after almost 15 years of working as an individual contributor in the insurance industry, I became an official manager of a small team of data scientists in 2021. For the next several months, I did very little besides work.</p><p>I did the things that I knew had to be done before taking the role: creating a vision for what I want the team to be doing, building a good team culture, establishing best practices, liaising with business partners, managing ongoing projects and developing new projects.</p><p>And, I did the things that I hadn’t given much thought to before taking the role but needed to be done: addressing visa issues, addressing unprofessional behaviors, hiring to refill headcount from resignations, hiring and being ghosted by interns (it’s a tough job market), squabbling with HR on unreasonable salary offers, pushing back on a merit raise structure that I felt was unfair, dealing with employee concerns about returning to the office, learning IT systems and security considerations, learning cloud technologies and modern app development tools, politicking to obtain necessary approvals, and performing a lot of manual data cleaning work.</p><blockquote>Learning #2: For technical managers, HR and IT issues will take up a good chunk of time. Make friends where you can and choose your battles.</blockquote><p>In hindsight, perhaps my view of what this role entailed was too naïve. I did expect that some things would come up with HR and IT that I would need to deal with. I didn’t expect the extent of the dealings, nor did I expect how emotional I would get. Mostly, my emotions registered as feeling angry, which I then needed to try to present externally as just being serious because <a href="https://news.asu.edu/20151027-study-shows-angry-men-gain-influence-and-angry-women-lose-influence">angry women lose influence</a>. But to be honest, when I’m notified that I have to restart a 10-step approval process because someone added an extra whitespace at the end of an IT-managed identity name… I imagine flipping my table (╯°□°)╯︵ ┻━┻.</p><blockquote>Learning #3: For any manager, state of mind and emotions are much more consequential. Effort should be made to get them under control.</blockquote><p>There were a lot of things that I needed to learn quickly in the new role. And I needed to be able to apply those learnings fast. Since I was trying to lay the groundwork for my team, I felt that any misstep could have big undesirable consequences.</p><p>I also needed to keep up with insurance business news so that I could network effectively with our business partners and get their buy-in for our projects. In my department, from a data and analytics perspective, we are in the unfortunate situation of not being able to focus on any one specific domain and have to be interested in every major line of business.</p><p>On top of the things I felt that I had to do, other analytics professionals would send me information about new educational content and techniques in the open source or insurance community. Every day, I also receive a list of recommended articles from Medium that looks something like this:</p><ul><li>&lt;LARGE NUMBER&gt; Techniques all Data Scientists Should Know</li><li>&lt;LARGE NUMBER&gt; Killer Automation Scripts for Your Projects</li><li>&lt;LARGE NUMBER&gt; Ultimate Data Visualization Tools</li><li>&lt;LARGE NUMBER&gt; Python Operations for Your Data Analysis</li><li>&lt;LARGE NUMBER&gt; Features Your ML App Can’t Do Without</li><li>&lt;LARGE NUMBER&gt; Tools You Need to Add to Your Workflow</li><li>&lt;LARGE NUMBER&gt; Fantastic Packages</li><li>Working with &lt;PYTHON LIBRARY THAT SOUNDS LIKE GIBBERISH&gt;</li><li>Working with &lt;ANOTHER RIDICULOUSLY NAMED LIBRARY&gt;</li></ul><p>Enter imposter syndrome.</p><p>This isn’t new for me, but my feelings of inadequacy were more addressable when I was in junior roles. All I had to do was to work more hours, and since my field was so specialized, the extra hours were manageable. After moving into a management role, I had to conclude that there are not enough hours in the day for anyone to stay on top of the emerging trends in data science.</p><blockquote>Learning #4: Realize that having things to learn is an expected and good thing. Managers get to talk to a lot of people who each specialize in some area, but no one person can know everything.</blockquote><p>Now that I’m almost one year into the new job, I find myself still working much more than is probably healthy. Every day, more items get added to my to-do list than get checked off. And every day, I discover more things that I wish I had time to learn about. Prioritization might be the most important skill that I am developing.</p><p>I wonder what the next year will bring?</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=2ca65a38f1be" width="1" height="1" alt=""><hr><p><a href="https://medium.com/human-in-a-machine-world/reflections-of-a-new-manager-2ca65a38f1be">Reflections of a New Manager</a> was originally published in <a href="https://medium.com/human-in-a-machine-world">Human in a Machine World</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Scraping an Infinite Scroll Table]]></title>
            <link>https://medium.com/human-in-a-machine-world/scraping-an-infinite-scroll-table-f6ec94c6e2ae?source=rss-5086e44692d2------2</link>
            <guid isPermaLink="false">https://medium.com/p/f6ec94c6e2ae</guid>
            <category><![CDATA[code]]></category>
            <category><![CDATA[data]]></category>
            <dc:creator><![CDATA[JJ]]></dc:creator>
            <pubDate>Sat, 20 Feb 2021 21:49:16 GMT</pubDate>
            <atom:updated>2021-02-20T21:49:16.860Z</atom:updated>
            <content:encoded><![CDATA[<h4>Using Python / Selenium</h4><p>At the start of the new year, I joined this insurance auto pricing competition: <a href="https://www.aicrowd.com/challenges/insurance-pricing-game">https://www.aicrowd.com/challenges/insurance-pricing-game</a>. It’s quite similar to Kaggle, but the evaluation is way more complicated.</p><p>The competition has a model accuracy (RMSE) leaderboard that judges how close your model’s expected claims are to the actual claims in a test set (this part is similar to Kaggle), but the final evaluation is based on the profitability of your pricing model in “markets” against samplings of 9 “competitive” competitor models (there’s a game theory component here, I think).</p><p>After many nights and weekends of trying to improve my model but only achieving more and more unprofitability, I think it’s a fair question to ask if this game is actually winnable…</p><p>To get a better handle on this question, I wanted to scrape the RMSE and weekly profit leaderboards from the competition. Unfortunately, it was a bit of a pain because the leaderboard tables were displayed as infinite scroll tables.</p><p>Python and Selenium to the rescue! It took me a while to figure this out, so hope it can help someone else too :).</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/b5f3568e49afdef5c7968ac5f3bada7d/href">https://medium.com/media/b5f3568e49afdef5c7968ac5f3bada7d/href</a></iframe><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f6ec94c6e2ae" width="1" height="1" alt=""><hr><p><a href="https://medium.com/human-in-a-machine-world/scraping-an-infinite-scroll-table-f6ec94c6e2ae">Scraping an Infinite Scroll Table</a> was originally published in <a href="https://medium.com/human-in-a-machine-world">Human in a Machine World</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Women at Work]]></title>
            <link>https://medium.com/human-in-a-machine-world/women-at-work-5c58e887d354?source=rss-5086e44692d2------2</link>
            <guid isPermaLink="false">https://medium.com/p/5c58e887d354</guid>
            <category><![CDATA[women]]></category>
            <category><![CDATA[life]]></category>
            <category><![CDATA[diversity]]></category>
            <dc:creator><![CDATA[JJ]]></dc:creator>
            <pubDate>Sun, 10 Jan 2021 16:41:47 GMT</pubDate>
            <atom:updated>2021-01-11T12:55:00.312Z</atom:updated>
            <content:encoded><![CDATA[<h4>Issues in the hiring and promotion process</h4><h3>A problem of perceived ambition.</h3><h4>Entry level hiring</h4><p>In 2020, I interviewed a couple of people for entry level data analyst positions. Of the two top candidates, one was male and the other was female. When asked about their near-term career goals, the man said that he wanted to learn the company ins and outs enough to be promoted to a data scientist position in a year. The female said that she has so much to learn so she wanted to focus on learning and absorbing as much as she can.</p><p>If everything else about the two candidates’ background is the same, who would you hire? Does the man’s answer come across as more ambitious? Is this desirable? What if the female also answered this way? Would it come across as pushy or bragging? If you hired both candidates, would you actually expect their work product to be any different based on their answers?</p><h4>Leadership program</h4><p>In our company’s leadership program, all candidates have to pitch to a panel of senior leaders on why they are suitable for the program and what the program would allow them to do for the company in the future. A common part of male pitches include something along the lines of them wanting to do &lt;<em>fill in the blank great thing for the company</em>&gt; as a future C-level executive. Most female pitches often include the same &lt;<em>fill in the blank great thing for the company</em>&gt; but without attaching the “as a future C-level executive” portion.</p><p>A frequent feedback to the female candidates is that they don’t seem ambitious enough.</p><p><strong>My suggestion</strong>: <em>Name the position that you want to your manager and sponsors. Maybe not as immediately or directly as many men would, but make sure to work it into your conversations early on.</em></p><h3>Spurious correlations.</h3><p>One of our executives recently proudly commented that he promoted two females because they were the best for the job (they were in support staff roles). Another executive apologetically commented that he hired a white male executive because there were no qualified women or minorities.</p><p>Besides the facepalming comment made by the second executive, it made me wonder about what criteria they were using to judge candidates. “Best” and “Qualified” are obviously defined differently for different roles. But if the roles have traditionally been held by mainly one gender, then “Best” and “Qualified” likely also capture gender and gender traits.</p><p>Consider leadership. <a href="https://en.wikipedia.org/wiki/Masculinity#:~:text=Traits%20traditionally%20viewed%20as%20masculine,independence%2C%20leadership%2C%20and%20assertiveness">Wikipedia</a> says: “</p><blockquote>Traits traditionally viewed as masculine in Western society include strength, courage, independence, <strong>leadership</strong>, and assertiveness. ”</blockquote><p>Since leadership has been tagged as a male trait, it doesn’t seem a stretch to think that many hiring managers consciously or not use “being male” as a basis for qualification for leadership positions. (Promoting mostly females in female dominated roles is also a problem.)</p><p><strong>My suggestion</strong>: <em>When hiring, be very specific and explicit about the characteristics that you are hiring for in order to try to get away from some of the potential gender biases. For example, instead of hiring for a good “leader” (a male trait), hire for someone with “integrity”, “influence”, “empathy”, “courage”, and “respect” (traits that are more gender-neutral or at least have a mix of traits that better represents diversity).</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5c58e887d354" width="1" height="1" alt=""><hr><p><a href="https://medium.com/human-in-a-machine-world/women-at-work-5c58e887d354">Women at Work</a> was originally published in <a href="https://medium.com/human-in-a-machine-world">Human in a Machine World</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Homo Deus]]></title>
            <link>https://medium.com/human-in-a-machine-world/homo-deus-e5497d010cde?source=rss-5086e44692d2------2</link>
            <guid isPermaLink="false">https://medium.com/p/e5497d010cde</guid>
            <category><![CDATA[life]]></category>
            <dc:creator><![CDATA[JJ]]></dc:creator>
            <pubDate>Sun, 02 Sep 2018 04:42:57 GMT</pubDate>
            <atom:updated>2019-06-27T14:11:49.816Z</atom:updated>
            <content:encoded><![CDATA[<h4>Thoughts on Yuval Harari’s book</h4><p>A few weeks ago, my partner and I embarked on our annual summer vacation trip. This time to Utah, to visit all five of their national parks. While putting 1500+ miles on our rented Prius, we had time to finish two audiobooks.</p><p>Homo Deus was the second book we listened to, and… I have really mixed feelings about it. The first hour had me captivated. The author’s predictions of humankind’s three future goals were laid out, and thought provoking justifications were presented. After that, the rest of the book focused much more on history, and the ending really lost me. The author went on a lot of tangents that I did not think tied well to his main points, and by the end I wasn’t sure what the main points of the book were anymore.</p><p>It seemed that the author filled the pages of the book with ideas that fascinated him, with or without connection to bigger ideas in the book. Like, how did lucid dreaming make it into this book?</p><p>Because the ideas in the book seem disconnected to me, I enjoyed it more thinking of it as a collection of persuasive essays. The author’s biases come across pretty strong though. Some points were made ad nauseam. And some of the logic, I thought was lacking.</p><p>Harari went through a bunch of different arguments against why humans are not really better or greater than animals (except for being able to create complex societies). From an evolutionary standpoint, other humans are better/greater/more important to me than other animals just because humans are more similar to me. The humans versus other animals comparison is not objective because it’s a human making the comparison.</p><p>On the point that Harari made about humans not having souls because science has never been able to measure a “soul”, I think also falls short because not all things can be physically measured. Some people believe that a soul is a person’s relationship to god.</p><p>I don’t really get his claim that vaccinations for the masses are the result of needing the masses for war either. There are great humanitarians out there, even if I’m not one of them.</p><p>Two other parts that really irked me about the book are Harari’s use of the word religion and the portion on free will.</p><p>Equating social/political ideologies to religion was completely unnecessary. Harari could have just said that social/political ideologies and religions achieve a similar purpose to make his point. I think redefining liberalism (capitalism)/socialism/humanism (Nazism) as religions obfuscates his explanations because he’s not using the commonly understood definition of religion. On top of that, I imagine that his claim would piss off a lot of people. It bothers me when people seem to be contrarian just for the sake of being contrarian.</p><p>On the topic of free will, Harari’s gave a really poor argument for determinism. If arguing that researchers are able to predict what someone wants before the person knows it themselves, the unstated assumption is that awareness is required for a person to make a choice with their free will. This is not obviously true, so I’m not sold.</p><p>Besides sounding cool, I don’t understand what it means that between determinism and randomness, there is no room for freedom. From quantum mechanics (my knowledge is limited here, but thank you college), we live in a non-deterministic world with macro level restrictions and there is “freedom” in the form of unpredictability. Harari doesn’t discuss randomness, but the existence of randomness can be the basis for an argument for free will.</p><p>(I’ve always thought of science as very similar to religious doctrine. If science tells me something, I don’t really “know” it to be true since I didn’t take any part in the experiments and my knowledge of science if very limited. Believing in science means that I’d have to accept it on faith.)</p><p>My bigger issue with the author’s discussion on free will is, why does it matter? Take out religion, whether or not I have free will has no practical impact on anything. I don’t get how this has any relevancy to the rest of the book.</p><p>Imagine what life would be like for someone who actually doesn’t believe in self, choice and free will. Nothing would have any meaning. It wouldn’t matter if this person is alive or dead. No one actually lives like this, so talking about not having free will is purely a thought experiment. Logically then, if believing that there is free will is the belief that makes us the happiest (and have a stronger will to stay alive), evolution would have forced us to believe that free will exists and someone arguing against free will like Harari should not have existed or made it this far in life. There.</p><p>I’ll admit that I did fall asleep for a few portions of the audiobook, especially toward the end. I did enjoy the book immensely though because of how it made me think about stuff, and now I wrote about it to sort through my thoughts.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e5497d010cde" width="1" height="1" alt=""><hr><p><a href="https://medium.com/human-in-a-machine-world/homo-deus-e5497d010cde">Homo Deus</a> was originally published in <a href="https://medium.com/human-in-a-machine-world">Human in a Machine World</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[MOOCs]]></title>
            <link>https://medium.com/human-in-a-machine-world/moocs-1260241e2a3e?source=rss-5086e44692d2------2</link>
            <guid isPermaLink="false">https://medium.com/p/1260241e2a3e</guid>
            <category><![CDATA[life]]></category>
            <category><![CDATA[data-science]]></category>
            <dc:creator><![CDATA[JJ]]></dc:creator>
            <pubDate>Mon, 19 Feb 2018 21:17:40 GMT</pubDate>
            <atom:updated>2018-02-19T21:17:40.025Z</atom:updated>
            <content:encoded><![CDATA[<h4>My attempts with online learning resources.</h4><p>Inspired by Tirthajyoti Sarkar’s post on <a href="https://towardsdatascience.com/how-to-choose-effective-moocs-for-machine-learning-and-data-science-8681700ed83f">choosing effective courses for machine learning and data science</a>, I’m going to take stock of my progress so far and make an effort to figure out how I can better use the available educational resources.</p><h4>Pre-2018 Relevant Courses</h4><p>On evaluating the classes that I’ve taken, out of 21 audited MOOCs, I wouldn’t classify any of the contents covered at beyond a basic level of comprehension. This is my subjective view of course, but it gets at one of the biggest issues that I have with MOOCs: I can’t seem to move forward. MOOCs have been great for me to understand a breadth of material but I haven’t been able to get deeper into the subjects that interest me.</p><ul><li><strong>(Data Science | Basic) </strong>Data Visualization and Communication with Tableau, Duke University, Coursera</li><li><strong>(Data Science | Basic)</strong> The Data Scientist’s Toolbox, Johns Hopkins University, Coursera</li><li><strong>(Data Science | Basic) </strong>Getting and Cleaning Data, Johns Hopkins University, Coursera</li><li><strong>(Data Science | Basic)</strong> Regression Models, Johns Hopkins University, Coursera</li><li><strong>(Data Science | Basic) </strong>A Crash Course in Data Science, Johns Hopkins University, Coursera</li><li><strong>(Data Science | Basic) </strong>Developing Data Products, Johns Hopkins University, Coursera</li><li><strong>(Data Science | Basic) </strong>Building a Data Science Team, Johns Hopkins University, Coursera</li><li><strong>(Data Science | Basic) </strong>Data Science in Real Life, Johns Hopkins University, Coursera</li><li><strong>(Data Science | Basic) </strong>Managing Data Analysis, Johns Hopkins University, Coursera</li><li><strong>(Data Science | Basic) </strong>Computing for Data Analysis, Johns Hopkins University, Coursera</li><li><strong>(Data Science | Basic) </strong>Introduction to R for Data Science, Microsoft, edX</li><li><strong>(Data Science | Basic+)</strong> Introduction to Recommender Systems, University of Minnesota, Coursera</li><li><strong>(Data Science | Basic+) </strong>Practical Predictive Analytics: Models and Methods, University of Washington, Coursera</li><li><strong>(Finance | Basic)</strong> Competitive Strategy, Ludwig-Maximilians-Universität München, Coursera</li><li><strong>(Finance | Basic) </strong>Analyzing Global Trends for Business and Society, University of Pennsylvania Wharton School of the University of Pennsylvania, edX</li><li><strong>(Programming (py) | Basic) </strong>An Introduction to Interactive Programming in Python (Part 1), Rice University, Coursera</li><li><strong>(Programming (py) | Basic)</strong> Programming for Everybody (Python), University of Michigan, Coursera</li><li><strong>(Programming (Web) | Basic+)</strong> HTML, CSS, and Javascript for Web Developers, Johns Hopkins University, Coursera</li><li><strong>(Programming (Web) | Basic)</strong> HTML, CSS and JavaScript, The Hong Kong University of Science and Technology, Coursera</li><li><strong>(Programming (Web) | Basic) </strong>Responsive Website Basics: Code with HTML, CSS, and JavaScript, University of London International Programmes Goldsmiths, University of London, Coursera</li><li><strong>(Programming (Web) | Basic) </strong>HTML5 Coding Essentials and Best Practices, World Wide Web Consortium (W3C), edX</li></ul><p>After making this list (which took a while), I can see that at least part of the cause of my problem is that I keep picking basic courses to take. Reflecting upon how I picked these courses, I think the sequence of events unfolded something like the below:</p><ol><li>The MOOC platforms started to offer Specializations</li><li>I did not keep track of which Specialization the course I took belonged to or the future classes in the Specialization</li><li>When I finished a course and searched for another class on my topic of interest, I came upon more Specializations</li><li>I looked at the syllabus for the first course and there would always be at least one item that I didn’t feel like I had a good grasp of</li><li>The result is that I keep taking the first few courses in multiple Specializations</li></ol><p>It’s surprisingly hard to keep track of courses when only auditing (this is probably intentional). Also, the lack of structure in the self-paced classes hasn’t worked so well for me either. But since I’m getting all of this top rated education available for no cost, I can’t complain and it is really on me to keep track of my learning.</p><p>At this point now with so many people sharing their own experience of self learning data science, it shouldn’t be too hard to come up with a curriculum as a student. Colleges also publish their data science curriculums, and using that as a basis might be a good idea.</p><p>Another point that I should consider is that most of the classes I’ve taken have been through Coursera, and Coursera restricting the quizzes and assignments has been detrimental to my learning. I should probably explore more non-Coursera options too.</p><p>So this is my big takeaway: <strong>keep track of courses and specializations and create a curriculum of courses that can build upon one another.</strong></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1260241e2a3e" width="1" height="1" alt=""><hr><p><a href="https://medium.com/human-in-a-machine-world/moocs-1260241e2a3e">MOOCs</a> was originally published in <a href="https://medium.com/human-in-a-machine-world">Human in a Machine World</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Chickenshit Club]]></title>
            <link>https://medium.com/human-in-a-machine-world/the-chickenshit-club-f8b538e9b358?source=rss-5086e44692d2------2</link>
            <guid isPermaLink="false">https://medium.com/p/f8b538e9b358</guid>
            <category><![CDATA[finance]]></category>
            <category><![CDATA[life]]></category>
            <dc:creator><![CDATA[JJ]]></dc:creator>
            <pubDate>Wed, 08 Nov 2017 22:19:51 GMT</pubDate>
            <atom:updated>2017-11-08T22:21:45.430Z</atom:updated>
            <content:encoded><![CDATA[<p>After a hiatus to study for a UK actuarial exam (for questionable career prospects) and handle some odds and ends around the house, I’m about back to my normal timeline of quarterly existential crises.</p><p>I regretted signing up for the UK exam a little too late into the studying process (sunk cost, I know, but couldn’t help myself). Regardless of whether I pass or not, I do have a few important takeaways:</p><ol><li>I am genuinely interested in finance. Ten years of being in the financial services industry apparently wasn’t enough for me to know that. It’s amazing how self-reflection doesn’t happen even when I spent so much time looking at my navel. I enjoyed my psychology classes in college but wanted to do something in an applied context and without extensive human interaction, so that led nowhere. The finance world sort of provides a model for understanding and influencing psychology. I’m not sure why it took studying for this exam to complete my career self narrative…</li><li>I learned about copulas, extreme value theory and filled in some holes in my knowledge of statistical distributions.</li><li>I feel more strongly against the move to higher Bloom’s taxonomy test questions. It’s not that I don’t think evaluation and synthesis are important skills, but I think that the only way to be well prepared for a high Bloom’s question is through memorization (and having a good study guide). I had ~600 flashcards for this exam, way more than for any prior US lower-Bloom exams, plus I memorized a handful of mind maps to make sure I could cover enough points. If one of the big pros for moving to high Bloom’s is less memorization, I don’t think this is true in practice even though I can see why it could make sense in theory.</li></ol><p>I also realized how much I didn’t know about the financial crisis. So after the exam, this led me to read Jesse Eisinger’s new book, <a href="https://www.goodreads.com/book/show/34397551-the-chickenshit-club">The Chickenshit Club</a>.</p><p>The point of this book was to explain what a sad state of affairs the current financial regulatory system is in and how it got there (you get the gist just by reading the book cover). Through the multiple narratives that I had trouble keeping track of, the thing that stuck out to me most was how many passionate people tried and failed and gave up, mostly due to unintended consequences.</p><p>I wish the author would’ve closed with a lessons learned section. After finishing the book, I felt the same as when I watched Atonement — depressed that bad stuff happened and there’s nothing to look forward to. There’s a quote in the book that was in turn a quote from another book, American Pastoral, that sums up what I gained from reading this book:</p><blockquote>He had learned the worst lesson that life can teach — that it makes no sense.</blockquote><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f8b538e9b358" width="1" height="1" alt=""><hr><p><a href="https://medium.com/human-in-a-machine-world/the-chickenshit-club-f8b538e9b358">The Chickenshit Club</a> was originally published in <a href="https://medium.com/human-in-a-machine-world">Human in a Machine World</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[ERM Flashcards — Part 6 FINAL]]></title>
            <link>https://medium.com/@captioned/erm-flashcards-part-6-final-8e2b9e8f3303?source=rss-5086e44692d2------2</link>
            <guid isPermaLink="false">https://medium.com/p/8e2b9e8f3303</guid>
            <category><![CDATA[risk-management]]></category>
            <category><![CDATA[insurance]]></category>
            <category><![CDATA[finance]]></category>
            <dc:creator><![CDATA[JJ]]></dc:creator>
            <pubDate>Fri, 01 Sep 2017 13:55:41 GMT</pubDate>
            <atom:updated>2017-09-01T14:01:44.240Z</atom:updated>
            <content:encoded><![CDATA[<h4>ERM Implementation</h4><p>Final versions: [<a href="https://medium.com/@captioned/erm-flashcards-part-1-final-4f732e31f16"><strong>Pt.1</strong></a>] [<a href="https://medium.com/@captioned/erm-flashcards-part-2-final-2db77728cf31"><strong>Pt.2</strong></a>] [<a href="https://medium.com/@captioned/erm-flashcards-part-3-final-7ce95d747d1b"><strong>Pt.3</strong></a>] [<a href="https://medium.com/@captioned/erm-flashcards-part-4-final-cce3b7004b07"><strong>Pt.4</strong></a>] [<a href="https://medium.com/@captioned/erm-flashcards-part-5-final-7404de644fe4"><strong>Pt.5</strong></a>] [<strong>Pt.6</strong>]</p><ul><li>[30] Reasons to hold capital (3): 1. manage cashflow (working capital); 2. facilitate growth / new ventures (development capital); 3. cover unexpected losses (risk capital)</li><li>[30] Definition of capital (3): 1. amount meant to provide sufficient surplus to cover adverse outcomes; 2. with a given level of risk tolerance; 3. over a specified time horizon</li><li>[30] Process of capital management (3): 1. quantification; 2. allocation; 3. optimization</li><li>[30] Purpose of generic capital models: Used by capital providers and regulators to gain a consistent assessment of capital requirements across different firms</li><li>[30] Purposes of internal capital models (2): 1. used to simulate company specific view of the capital needed; 2. improve management’s understanding of business dynamics (thus improve decision making)</li><li>[30] Ancillary uses of internal capital models (11): 1. determine risk profile; 2. capital budgeting; 3. determine capital needed for M&amp;A; 4. insurance product pricing; 5. risk tolerances/constraints; 6. set investment strategy; 7. determine RAROC; 8. performance measurement; 9. incentive compensation; 10. alternative to regulatory/rating agency requirements; 11. disaster planning</li><li>[30] Forecast outputs of an internal capital model (3): 1. balance sheets; 2. profit and loss accounts; 3. cashflow</li><li>[30] Desirable features of an internal capital model (4): 1. asset model allows for asset correlations; 2. liability model considers reinsurance and liability correlations; 3. asset and liability models are integrated and allow for correlations between assets and liabilities; 4. model is dynamic</li><li>[30] Meaning of “dynamic” capital model: Asset and liability cashflows are linked by equivalent economic variables (e.g. inflation)</li><li>[30] Benefits of internal capital model being dynamic (8): 1. improved understanding of current strategy; 2. considers impacts of other strategies; 3. examines using other sources of capital; 4. useful for due diligence; 5. assesses risk adjusted performance of business units; 6. determines optimal asset mix; 7. helps understand impact of extreme events; 8. useful for Financial Condition Reports</li><li>[30] Factors discussed in FCR: Financial Condition Report is a formal assessment of the financial viability of the insurer and includes outputs from an internal capital model plus hard to quantify factors (reputational risk, ERM effectiveness)</li><li>[30] Differences between internal and generic capital models (): 1. objectives; 2. calibrations (1 in 500 vs. 1 in 200); 3. volatility views; 4. diversification allowances; 5. inclusion or treatment of different risks; 6. views on availability of certain assets as capital</li><li>[30] Stages in operating a successful capital model (6): 1. identify purpose; 2. identify and rank risks; 3. choose simulation approach for each risk; 4. define risk metrics; 5. select modeling criteria; 6. decide on method of implementation</li><li>[30] Ways capital management can increase shareholder value (4): 1. price competitively; 2. better reserving; 3. performance management; 4. risk management</li><li>[30] Bottom-up process for calculating capital requirements (4): 1. generate stand-alone distributions of changes in the enterprise’s value due to each risk source; 2. combine distributions (allow for diversification); 3. calculate total capital at desired standard for selected risk metric(s); 4. allocate capital to each activity based on amount of risk generated</li><li>[30] Analysis methods used to calculate capital (10): 1. probability of ruin; 2. economic cost of ruin; 3. full economic scenarios; 4. stress test method; 5. factor tables; 6. stochastic models; 7. statistical models; 8. credit risk methods; 9. operational risk methods; 10. option pricing theory</li><li>[30] Tiers of capital under Basel II (3): 1. bank’s equity and disclosed reserves; 2. other reserves and various debt instruments; 3. certain types of short dated capital</li><li>[30] Basel III changes (6): 1. tier 1 plus tier 2 capital at least 8% of RWAs; 2. tier 1 capital at least 6% of RWAs; 3. conservation buffer of 2.5% of RWAs; 4. additional deductions from common equity; 5. reduce pro-cyclicality by allowing capital requirements to fall in times of financial stress; 6. reduce systemic risk by limiting holdings in other banks</li><li>[30] Economic income created: EIC=(RAROC-[hurdle rate])*Capital</li><li>[30] Capital allocation methods (5): 1. use of risk measure (apply Euler principle); 2. marginal approach; 3. game theory approach; 4. pro-rata; 5. stand-alone (remaining capital retained in main corporate business line)</li><li>[31] Major processes to establish in ERM implementation (4): 1. corporate governance; 2. risk assessment and quantification; 3. risk management; 4. reporting and monitoring</li><li>[31] Business applications by stage of ERM maturity (3): 1. loss reduction; 2. uncertainty management; 3. performance optimization</li><li>[31] Key considerations in scoping an ERM implementation (3): 1. resourcing (internal v. external); 2. proportionality (to risks and size and sophistication of business); 3. top down or bottom up</li><li>[31] Key challenges in ERM implementation (2): 1. lack of risk awareness; 2. inappropriate risk culture</li><li>[31] Lam’s ERM maturity model (5): 1. definition and planning; 2. early development; 3. standard practice; 4. business integration; 5. business optimization</li><li>[31] Activities in Lam’s definition and planning stage (4): 1. identify internal and external requirements; 2. obtain Board and management support; 3. develop overall framework; 4. appoint key personnel</li><li>[31] Activities in Lam’s early development stage (4): 1. establish ERM policies and risk functions; 2. identify key risks; 3. coordinate risk and control processes; 4. education</li><li>[31] Activities in Lam’s standard practice stage (4): 1. establish risk database; 2. develop KRIs; 3. establish risk models; 4. measure risk adjusted performance</li><li>[31] Activities in Lam’s business integration stage (7): 1. evaluate business risks; 2. quantify cost of risk; 3. automate risk reporting; 4. allocate capital; 5. use risk triggers to prompt business decisions; 6. measure ERM effectiveness; 7. link to executive compensation</li><li>[31] Activities in Lam’s business optimization stage (3): 1. expand scope of ERM to include strategic risk; 2. integrate ERM into strategic planning; 3. allocate capital to optimize risk-adjusted performance</li><li>[31] McKinsey’s risk maturity model (4): 1. initial risk transparency; 2. systemic loss reduction; 3. risk-return management; 4. risk as competitive advantage</li><li>[31] Deloitte’s risk maturity model (5): 1. unaware / planning; 2. fragmented / specialist silos; 3. top-down; 4. systematic; 5. risk intelligent</li><li>[31] IAA stages of ERM maturity (3): 1. early; 2. intermediate (established); 3. advanced (established, consistent, understood)</li><li>[31] Areas to consider when assessing ERM maturity (4): 1. corporate governance; 2. risk language and culture; 3. competencies and performance management; 4. risk management processes and responsibilities</li><li>[32] Elements of organizational learning (4): 1. be open to discuss own past mistakes; 2. be able to learn from those mistakes; 3. be aware of mistakes of others; 4. adopt industry best practices</li><li>[32] Lam’s lessons learned (7): 1. know your business; 2. establish checks and balances; 3. set limits and boundaries; 4. keep your eye on the cash; 5. use the right yardstick; 6. pay for the performance you want; 7. balance the yin and the yang</li><li>[32] Mistakes leading to the 2008 financial crisis (8): 1. failure to understand and report risks inherent in business activities; 2. complex products were not well understood; 3. over-dependence on cheap debt; 4. remuneration encouraging short term goals; 5. unbundling and outsourcing led to poor sales practices; 6. deregulation of financial sector allowed risk to propagate unchecked; 7. poor corporate governance led to bad decision making at Board level; 8. credit rating agencies struggled to keep pace with complex products and banks that gamed the system</li><li>[32] Lessons learned from Barings case (4): 1. need to internal checks and balances; 2. need for proper supervision of employees with clear reporting lines; 3. auditors and top management should understand the business; 4. bonuses should be based on profits over a longer time horizon</li><li>[32] Examples of rogue traders (6): 1. Barings Bank (1995, Nick Leeson); 2. Morgan Stanley (2008, Matthew Piper); 3. MF Global (2008, unnamed); 4. Societe Generale (2008, Jerome Kerviel); 5. PVM Oil Future (2009, unnamed); 6. USB (2011, unnamed)</li><li>[32] Regulatory failures in Equitable Life case (4): 1. over tolerant attitude in light of GAD warning and complex reinsurance arrangements; 2. shortage of supervisory staff leading to arms-length monitoring; 3. over tolerant attitude by profession in regards to the Appointed Actuary also holding Chief Executive post; 4. failure of prior investigations to identify potential future related risks</li><li>[32] Lessons learned from LTCM case (3): 1. liquidity itself is a risk factor; 2. models must be stress tested and used to inform decisions rather than make them; 3. financial institutions should understand aggregate exposures to common risk factors</li><li>[32] Lessons learned from Madoff case (2): 1. if it looks too good to be true, it probably is; 2. do your due diligence</li><li>[32] Lessons learned from space shuttle Challenger (2): 1. ensure decision makers and leaders understand the risks that are being taken; 2. do not succumb to pressures to hit artificial targets at the cost of good risk management</li><li>[32] Lessons learned from Confederate Life (3): 1. apply checks and balances on the activities of those in positions of power; 2. concentration kills (set limits and establish a balance); 3. understand the business</li><li>[32] Lessons learned from Orange County (6): 1. beware of unconstrained star performers; 2. where there is excess return, there is risk; 3. powerful individuals can hide risk if company structure has gaps; 4. borrowing short and investing long leads to liquidity risk; 5. tie investment objectives to investment actions through framework of policies and oversight; 6. risk reporting should be complete and easily comprehensible to independent professionals</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8e2b9e8f3303" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[ERM Flashcards — Part 5 FINAL]]></title>
            <link>https://medium.com/@captioned/erm-flashcards-part-5-final-7404de644fe4?source=rss-5086e44692d2------2</link>
            <guid isPermaLink="false">https://medium.com/p/7404de644fe4</guid>
            <category><![CDATA[risk-management]]></category>
            <category><![CDATA[insurance]]></category>
            <category><![CDATA[finance]]></category>
            <dc:creator><![CDATA[JJ]]></dc:creator>
            <pubDate>Wed, 30 Aug 2017 23:14:40 GMT</pubDate>
            <atom:updated>2017-09-01T14:01:56.933Z</atom:updated>
            <content:encoded><![CDATA[<h4>Risk Responses</h4><p>Final versions: [<a href="https://medium.com/@captioned/erm-flashcards-part-1-final-4f732e31f16"><strong>Pt.1</strong></a>] [<a href="https://medium.com/@captioned/erm-flashcards-part-2-final-2db77728cf31"><strong>Pt.2</strong></a>] [<a href="https://medium.com/@captioned/erm-flashcards-part-3-final-7ce95d747d1b"><strong>Pt.3</strong></a>] [<a href="https://medium.com/@captioned/erm-flashcards-part-4-final-cce3b7004b07"><strong>Pt.4</strong></a>] [<strong>Pt.5</strong>] [<a href="https://medium.com/@captioned/erm-flashcards-part-6-final-8e2b9e8f3303"><strong>Pt.6</strong></a>]</p><ul><li>[26] Ways risk management can optimize risk-return profile (4): 1. support selective growth of business; 2. support profitability through risk adjusted pricing; 3. use limit setting to control the size and probability of potential losses; 4. employ techniques to manage existing risks</li><li>[26] Benefit of active portfolio management and duration matching over risk transfer: Active portfolio management is often more cost effective and is a longer term solution (risk transfer tends to be quicker and easier to implement)</li><li>[26] Fundamental concepts in managing a portfolio of risks (5): 1. risk; 2. reward; 3. diversification; 4. leverage; 5. hedging</li><li>[26] Diversification: Reduce overall risk by investing in different projects or assets whose returns are not perfectly correlated</li><li>[26] Leverage: Borrow money and invest, increases potential risk and return profile</li><li>[26] Hedging: Enter into agreement that reduces risk (usually because hedge position is negatively correlated with existing position)</li><li>[26] Specific risk (MVPT): Risk specific to a company, can be eliminated through diversification</li><li>[26] Systematic risk (MVPT): Risk of being in the market, cannot be diversified away</li><li>[26] Efficient frontier: Portfolios with the highest returns for a given level of risk or lowest risk for a given level of return</li><li>[26] Separation theorem: The optimal portfolio of risky assets for an investor can be determined without any knowledge of their preferences toward risk and return (or liabilities)</li><li>[26] Benefits of portfolio management in ERM (4): 1. encourages companies to unbundle business into component projects; 2. provides mechanism for aggregating risks across the company; 3. provides framework for setting risk concentration limits and asset allocation targets; 4. influences investment, transfer pricing and capital allocation decision</li><li>[26] Active versus passive management of a portfolio of risks: There is no appropriate index to track so index-tracking is not an option, using risk concentration limits and asset allocation targets to apply active management then makes sense</li><li>[26] Responses to risk (4): 1. rejection (removal); 2. acceptance; 3. transfer; 4. management</li><li>[26] Process for implementing risk responses (4): 1. research possible responses and their costs; 2. determine response for each risk, with implementation deadlines; 3. assign risk manager responsible for implementation; 4. consider secondary risks</li><li>[26] Characteristics of a good risk response (5): 1. economical; 2. well matched to risk; 3. simple; 4. active; 5. flexible/dynamic</li><li>[26] Methods to reduce risk through transfer (8): 1. insurance; 2. reinsurance; 3. co-insurance; 4. sharing risk with policyholder via product design; 5. securitization; 6. purchase derivatives; 7. ART products; 8. outsourcing</li><li>[26] Securitization: Package risk into a marketable investment</li><li>[26] Elements of ERM that aid risk transfer (3): 1. provides a framework for recognizing net exposure to each type of risk and diversification of risks; 2. provides a framework to assess the cost of different risk transfer strategies; 3. establishes consistent risk transfer policies across company</li><li>[26] Considerations for risk transfer (5): 1. cost, 2. loss of upside potential; 3. counterparty risk; 4. regulatory restrictions; 5. market capacity</li><li>[26] Considerations for risk removal (3): 1. cost; 2. impact of removing risk on likelihood of project meeting original objectives; 3. whether opportunities will be lost as a result</li><li>[26] Situations where risk retention may be appropriate (3): 1. retaining particular risk is part of core business; 2. it is the most economical approach; 3. no alternatives</li><li>[26] Ways to reduce risk without transfer (8): 1. take on uncorrelated risks; 2. increase portfolio size; 3. better match assets and liabilities; 4. stronger internal controls and governance; 5. robust underwriting practices and analysis; 6. due diligence practices and tightly worded agreements; 7. remuneration and bonus systems that align interests; 8. increased capital or funding</li><li>[26] Circumstances giving rise to residual risks (4): 1. decision was made to retain risk; 2. secondary risk from risk response action; 3. imperfect hedge; 4. inability to transfer or fully mitigate risk</li><li>[26] Categories of ART products (2): 1. vehicles based on capital market instruments; 2. unconventional vehicles used to cover conventional risks</li><li>[26] Requirements for a company to develop ART products: 1. ability to quantify risk (likelihood and size); 2. ability to package, underwrite and sell securities; 3. capital if wanting to retain risk</li><li>[26] Advantages of ART products (5): 1. improved organizational focus; 2. customization of product and timing; 3. cost reduction and simplified administration; 4. earnings stability; 5. marking-to-market (establishes a market price for risks)</li><li>[26] Problems associated with ART (4): 1. may have higher initial costs than conventional products; 2. more complex; 3. company may need to change the way it assesses and manages risk in order to benefit; 4. staff may need to be educate</li><li>[26] Potential drivers of higher demand for ART (3): 1. conventional insurance becomes more expensive; 2. ERM becomes more widespread; 3. it becomes widely recognized that companies should focus on core business and eliminate other risks</li><li>[27] Market risk management strategies (3): 1. diversification; 2. investment strategy; 3. hedging</li><li>[27] Key activities of market risk management (5): 1. setting and monitoring policies; 2. setting and monitoring limits; 3. reporting; 4. capital management; 5. implementing risk-portfolio strategies (matching, hedging)</li><li>[27] Items covered by market risk policies (7): 1. roles and responsibilities; 2. delegation of authority and limits; 3. risk measurement and reporting; 4. valuation and back-testing; 5. hedging policy; 6. liquidity policy; 7. exception management</li><li>[27] Types of derivatives (4): 1. options; 2. forwards; 3. futures; 4. swaps</li><li>[27] Options: Gives the right, but not the obligation, for one party to exercise the contract in return for premium payment to the counterparty</li><li>[27] Forwards: Obligation for both counterparties to complete a transaction on a future date at a known price</li><li>[27] Futures: Like a forward, but a standardized contract traded on an exchange</li><li>[27] Swaps: Obligation for both counterparties to exchange a series of cashflows</li><li>[27] Characteristics of exchange traded contracts (6): 1. standardized; 2. trading is done through the exchange based on market prices; 3. deals are settled through a clearing-house; 4. clearing-house acts as counterparty to both buyer and seller; 5. counterparty risk is reduced for the clearing-house by pooling many contracts; 6. highly liquid with comparatively low transaction costs</li><li>[27] Characteristics of OTC derivatives (4): 1. trading is done at the convenience of the parties and no money changes hands until delivery date; 2. pricing is negotiable; 3. very flexible; 4. documentation is usually based on standard terms and conditions (published by ISDA)</li><li>[27] Types of OTC derivatives: Forwards, swaps, and some options</li><li>[27] Factors influencing the price of OTC contracts (6): 1. spot price of underlying; 2. time to delivery; 3. expectations of interest rates; 4. expected income yield on underlying; 5. residual counterparty risk; 6. cost of carry (of collateral and underlying)</li><li>[27] Marking-to-market process (3): 1. initial margin (cash) is deposited by counterparty into a margin account; 2. clearing house periodically determines if more or less is needed in the margin account; 3. counterparty must add variation margin if margin account drops below a specified maintenance margin</li><li>[27] Potential advantages of derivatives over underlying asset (3): 1. cheaper; 2. easier; 3. more flexible</li><li>[27] Risks associated with derivatives (9): 1. credit; 2. settlement; 3. aggregation; 4. operational; 5. liquidity; 6. legal; 7. reputation; 8. concentration; 9. basis</li><li>[27] Settlement risk: Risk that one counterparty does not deliver (a security or its value in cash) as agreed to when security was traded, after the other counterparty has already delivered</li><li>[27] Basis (futures contracts): Difference between spot price of asset and price of future, at a particular point in time</li><li>[27] Normal backwardation: Situation where the futures price is below the expected value of the future spot price (may occur if market expects income on asset to outweigh costs, or high demand for short positions in the future)</li><li>[27] Contango: Situation where the futures price exceeds the expected value of the future spot price (may occur due to high storage costs of underlying, or high demand for long positions in the future)</li><li>[27] Cashflow management techniques used for currency risk (2): 1. netting (foreign revenue match foreign expenses/liabilities); 2. leading and lagging (bring forward or delay foreign cashflows to exploit expected movement in exchange rates)</li><li>[27] Delta: Rate of change in portfolio value relative to price of underlying</li><li>[27] Gamma: Rate of change of delta relative to price of underlying</li><li>[27] Vega: Rate of change in portfolio value relative to the assumed level of volatility of the underlying</li><li>[27] FRA: Forward rate agreement is used to hedge direct exposure to interest rates, e.g. one fixed rate for one floating rate</li><li>[27] Techniques for managing direct interest rate risk (3): 1. FRAs; 2. caps; 3. floors</li><li>[27] Techniques for managing indirect interest rate risk (5): 1. cashflow matching; 2. interest rate swaps; 3. swaptions; 4. immunization (protect present value); 5. hedging using model points</li><li>[27] Practical problems with cashflow matching (3): 1. suitable assets may not be available; 2. future cashflows may not be known; 3. expected future cashflows may change frequently (costly)</li><li>[27] Best practices for managing credit risk (4): 1. avoid bad risks; 2. diversify credit exposure; 3. monitor exposure regularly; 4. take immediate action if and when default occurs</li><li>[27] Conditions required for immunization (3): 1. present value of asset and liability cashflows are equal; 2. discounted mean term of asset and liability cashflows are equal; 3. convexity of asset cashflows exceed that of liability cashflows</li><li>[28] Stages in credit risk management process (5): 1. policy and infrastructure; 2. credit granting; 3. exposure monitoring, management and reporting; 4. portfolio management; 5. credit review</li><li>[28] Elements of credit risk management policy and infrastructure stage (5): 1. establish appropriate credit environment; 2. adopt credit risk policies and procedures; 3. implement policies and procedures; 4. develop models; 5. define data standards and conventions</li><li>[28] Elements of credit risk management credit granting stage (4): 1. credit analysis and rating of counterparties; 2. credit approval; 3. pricing and setting terms and conditions for credit; 4. documentation</li><li>[28] Factors reflected in credit analysis and rating (4): 1. borrower’s repayment history; 2. analysis of borrower’s ability to pay; 3. their reputation; 4. availability and enforceability of guarantees or collateral</li><li>[28] Goals of the credit risk management monitoring stage (3): 1. prevent undue exposure to individual counterparty; 2. ensure appropriate portfolio diversification; 3. provide early warning of adverse credit events</li><li>[28] Uses of credit risk management exposure limits (4): 1. risk control; 2. allocation of risk-bearing capacity; 3. delegation of authority; 4. regulatory compliance</li><li>[28] Responsibilities of a credit review group (5): 1. review a sample of transactions and associated documentation to ensure data is correct; 2. test systems are working; 3. enforce underwriting standards; 4. check policies and procedures are being followed; 5. communicate to management</li><li>[28] Ways to transfer credit risk (3): 1. credit insurance; 2. credit derivatives; 3. securitization of assets</li><li>[28] Potential benefits of credit insurance (7): 1. protection against bad debts; 2. coverage for specific debtors; 3. coverage for specific region; 4. coverage for expenses; 5. coverage for international risks; 6. specialist advice; 7. ability to secure better terms for financing</li><li>[28] Examples of credit events (5): 1. bankruptcy; 2. rating downgrade; 3. repudiation; 4. failure to pay coupon; 5. cross-default</li><li>[28] Uses of TRORS (2): 1. increase diversification; 2. manage specific risks</li><li>[28] Benefits of securitization to the originator (5): 1. convert a bundle of assets into a negotiable structured financial instrument; 2. raise money, that is linked to future expected cashflows; 3. provide alternative source of finance; 4. pass asset risk to third party; 5. sell assets that may otherwise be unmarketable</li><li>[28] Credit linked note: Collateralized vehicle consisting of a bond with an embedded credit default swap</li><li>[29] Key to managing operational risk: Have sufficient effective controls (combination of information, assessment and response)</li><li>[29] Desirable characteristics of controls (8): 1. focused on results; 2. accounts for both measurable and non-measurable events; 3. standardized for efficient communication; 4. high quality; 5. few, rather than many; 6. meaningful and appropriate; 7. timely; 8. simple</li><li>[29] Considerations for deciding to outsource (5): 1. regulatory environment; 2. financial standing of third party; 3. competency of third party; 4. legal agreements (right to terminate, sub-contract); 5. how to monitor third party</li><li>[29] Ways to mitigate business disruptions from operational risks (3): 1. business continuity plan for identified business interruption risks; 2. crisis management plan; 3. consequential loss insurance</li><li>[29] Actions to manage technology risks (7): 1. keep systems up to date; 2. routine maintenance; 3. thorough testing; 4. quick IT helpdesk response; 5. staff training; 6. restrictions on employee usage; 7. implement and test security software and routines</li><li>[29] Ways to manage people risk (4): 1. recruitment process; 2. training; 3. performance appraisals; 4. relationship management (with unions)</li><li>[29] Techniques to manage process risk (4): 1. pilot studies; 2. precise definition of requirements; 3. design systems that can be easily maintained; 4. careful deployment with user education</li><li>[29] Defenses against reputational risks (3): 1. sound ERM framework; 2. business continuity and crisis management plans; 3. strong relationship with key stakeholders</li><li>[29] Features of best practice operational risk management (5): 1. broad definition of operational risk; 2. internal and external early warning indicators; 3. qualitative and quantitative assessment tools; 4. capital is allocated to operational risk; 5. insurance function is fully integrated with operational risk function</li><li>[29] Actions aimed at managing operational risks (11): 1. outsourcing; 2. business continuity and crisis management plans; 3. horizon scanning; 4. maintenance; 5. security; 6. HR; 7. careful underwriting, product design and pricing; 8. education, checks and balances; 9. good change management; 10. strong relationship with key stakeholders; 11. sound ERM framework that is integrated into the business</li><li>[29] Enterprise-wide process for transferring operational risk (7): 1. identify exposure; 2. quantify; 3. integrate operational risk with credit and market risks to establish enterprise risk profile; 4. establish risk limits; 5. implement internal controls; 6. develop strategies for risk transfer and financing; 7. evaluate alternatives based on cost/benefit analysis</li><li>[29] Actions to manage retained operational risk (2): 1. establish reserves; 2. allow for operational risk when allocating capital (give management incentives to improve)</li><li>[29] Actions to manage liquidity risk (6): 1. active monitoring of liquidity requirements; 2. varying investment strategy; 3. using swaps; 4. maintenance of a contingency fund; 5. diversify sources of funding; 6. obtain contingent sources of funding</li><li>[29] Ways for a company to manage systemic risk (2): 1. ensure diverse range of counterparties; 2. trading via exchanges</li><li>[29] External regulatory actions to reduce feedback risk (5): 1. use of circuit breakers by exchanges; 2. government actions (propping up bank); 3. regulations requiring establishment of reserves; 4. avoid pro-cyclical regulations; 5. physically separating certain types of business</li><li>[29] Ways to manage insurance risk (5): 1. underwriting; 2. risk transfer; 3. reduce risk concentrations; 4. improve diversification; 5. hedging</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=7404de644fe4" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[ERM Flashcards — Part 4 FINAL]]></title>
            <link>https://medium.com/@captioned/erm-flashcards-part-4-final-cce3b7004b07?source=rss-5086e44692d2------2</link>
            <guid isPermaLink="false">https://medium.com/p/cce3b7004b07</guid>
            <category><![CDATA[insurance]]></category>
            <category><![CDATA[risk-management]]></category>
            <category><![CDATA[finance]]></category>
            <dc:creator><![CDATA[JJ]]></dc:creator>
            <pubDate>Wed, 30 Aug 2017 12:58:38 GMT</pubDate>
            <atom:updated>2017-09-01T14:02:54.348Z</atom:updated>
            <content:encoded><![CDATA[<h4>Risk Assessment</h4><p>Final versions: [<a href="https://medium.com/@captioned/erm-flashcards-part-1-final-4f732e31f16"><strong>Pt.1</strong></a>] [<a href="https://medium.com/@captioned/erm-flashcards-part-2-final-2db77728cf31"><strong>Pt.2</strong></a>] [<a href="https://medium.com/@captioned/erm-flashcards-part-3-final-7ce95d747d1b"><strong>Pt.3</strong></a>] [<strong>Pt.4</strong>] [<a href="https://medium.com/@captioned/erm-flashcards-part-5-final-7404de644fe4"><strong>Pt.5</strong></a>] [<a href="https://medium.com/@captioned/erm-flashcards-part-6-final-8e2b9e8f3303"><strong>Pt.6</strong></a>]</p><ul><li>[20] Black Monday: October 1987, S&amp;P 500 index fell by over 20%.</li><li>[20] Differences between actual financial data compared to a normal distribution (3): 1. more narrowly peaked; 2. fatter tails (leptokurtic); 3. returns are heteroscedastic</li><li>[20] Methods of defining Extreme Value (2): 1. max value in a set of n losses (block maximum); 2. all losses exceeding a certain threshold</li><li>[20] Main results of extreme value theory (2): 1. distribution of block maxima is approximately described by the GEV family of distributions if n is sufficiently large; 2. tail of the distribution above a threshold u can be approximated for large values of u by the GPD</li><li>[20] Parameters of the GEV family: 1. alpha (location); 2. beta (scale); 3. gamma (shape)</li><li>[20] Distributions in the GEV family (3): 1. Gumbel (gamma=0); 2. Weibull (gamma&lt;0); 3. Frechet (gamma&gt;0)</li><li>[20] Gumbel GEV characteristics: Tail falls exponentially</li><li>[20] Weibull GEV characteristics: Finite upper bound (absolute max), can fit reinsured losses and natural phenomenon such as temperature, wind speed, age of human population</li><li>[20] Frechet GEV characteristics: Have a lower bound and heavy tails (with infinite variance if gamma&gt;0.5) that follow a power law (this is typically the distribution used to model extreme financial loss events)</li><li>[20] Underlying distributions for Weibull GEV (3): Distributions with finite upper limits; 1. beta; 2. uniform; 3. triangular</li><li>[20] Underlying distributions for Gumbel GEV (6): Light tailed distribution with finite moments of all orders; 1. chi-squared; 2. exponential; 3. gamma; 4. log-normal; 5. normal; 6. Weibull</li><li>[20] Underlying distributions for Frechet GEV (5): Heavy tailed distributions whose higher moments can be infinite; 1. Burr; 2. F; 3. log-gamma; 4. Pareto; 5. t</li><li>[20] Return level approach for GEV: Select the max observation in each block</li><li>[20] Return period approach for GEV: Count the observations in each block that exceed some set level</li><li>[20] Trade-off considerations for block size in return level approach for GEV: Granularity versus variance of estimates</li><li>[20] Key limitations of the GEV approach (2): 1. a lot of data is ignored; 2. choice of block size is subjective</li><li>[20] Parameters for GPD: 1. beta (scale); 2. gamma (shape)</li><li>[20] Characteristics of the GPD (by different gammas): 1. lower bound when gamma&gt;0; 2. upper bound when gamma&lt;0; 3. becomes Pareto when gamma&gt;0</li><li>[20] Asymptotic property for GEV/GPD: When the maxima of a distribution converge to a GEV distribution, the excess distribution converges to a GPD with the same shape parameter</li><li>[20] Memory-less property of exponential distributions: Expected future waiting time for an event to occur is independent of the waiting time already elapsed</li><li>[20] Threshold selection for GPD: Determine the lowest threshold above which the mean excess function is linear in u (typically around 90–95th percentile of complete distribution)</li><li>[21] Stochastic uncertainty: Uncertainty arising from the randomness of a finite set of observations</li><li>[21] Ways to allow for parameter uncertainty through dynamic simulations (2): 1. use multivariate normal distribution (if covariance matrix is available); 2. use joint distribution for parameters (simulated by repeatedly re-fitting parameters to sets of data points each modeled using the original data set)</li><li>[21] Requirements for actuarial models (9): 1. model must be valid, sufficiently rigorous for its purposes; 2. adequately documented; 3. model points reflect the distribution of business being modeled; 4. components of model allow for all features of the business being modeled that could affect advice given; 5. input parameter values should be appropriate to the business; 6. workings of the model should be easy to appreciate and communicate; 7. outputs should be capable of independent verification; 8. capable of subsequent development; 9. not overly complex</li><li>[21] Requirements for actuarial models for ERM (6): Generic requirements plus; 1. amenable to an analysis of the impact of parameter uncertainty; 2. simulations should exhibit behaviors consistent with past events; 3. shortcomings of the model should be clearly stated; 4. reflect dynamics of the organization, now and in the future; 5. be comprehensive across all important well-defined risks; 6. produce outcomes that are balanced</li><li>[21] Sweeting’s model requirements (5): 1. be appropriate for purpose; 2. be robust; 3. be as simple as possible, while meeting its purpose; 4. be developed over time; 5. sometimes be avoided when more importance is attached to other activities</li><li>[21] Uses of ERM models (8): 1. price products or services; 2. assess economic value of company; 3. estimate possible volatility of future profits and earnings; 4. determine capital adequacy requirements; 5. project future capital or solvency position; 6. assess effect of risk management and mitigation techniques on profits and capital requirements; 7. assess effect of strategic decisions; 8. evaluate projects</li><li>[21] Steps in developing and applying model (10): 1. specify purpose; 2. collect data, group and modify as necessary; 3. choose model form, identify parameters; 4. select appropriate time period; 5. estimate parameters and correlations; 6. ensure acceptable goodness of fit; 7. ensure model is able to project all required outputs; 8. run model using estimated variables or stochastic simulations; 9. output results in appropriate format; 10. perform sensitivity tests</li><li>[21] Considerations in discussing model outputs (5): 1. corporate risk policy; 2. corporate risk appetite; 3. corporate risk preferences; 4. risk preferences and other investment opportunities of company owners; 5. qualitative factors, judgment and intuition</li><li>[22] Characteristics of actual returns on individual equities (4): 1. returns are rarely independent and identically distributed; 2. volatility appears to vary over time; 3. extreme returns appear in clusters; 4. return series are leptokurtic (heavy-tailed, non-symmetric)</li><li>[22] Characteristics of actual returns on equity portfolios (5): 1. correlations exist between returns of different equities at the same point in time; 2. correlations between different series vary over time; 3. multivariate returns show little evidence of cross-correlation (between time periods); 4. multivariate series of absolute or squared returns do show strong evidence of cross-correlation; 5. extreme returns in one series often coincide with extreme returns in several other series</li><li>[22] Characteristics of equity returns over a long time period (2): 1. volatility clustering is less marked; 2. returns appear to be more iid and less heavy-tailed (expected from the Central Limit Theorem)</li><li>[22] Excess kurtosis of daily financial data: Kurtosis measures “peakedness” of a distribution (kurtosis for daily financial data appears to be higher than that of a normal distribution, which means more of the variance is due to infrequent extreme variations)</li><li>[22] Steps to model portfolio returns using the multivariate normal distribution (6): 1. decide on frequency of calculation (daily, weekly); 2. decide on timeframe of historical data to use (volume versus relevance); 3. for each asset class, choose total return index; 4. calculate log-returns; 5. calculate average returns and variance of each asset class and co-variance between classes; 6. simulate series of returns with same characteristics based on a multivariate normal distribution</li><li>[22] Steps to model portfolio returns using PCA (10): 1.–5. same as using multivariate normal distribution; 6. derive matrix of deviations from average returns; 7. derive principal components that explain a sufficiently high proportion of the deviations; 8. project this number of independent normal distributed normal random variables using associated eigenvalues as variances; 9. weight projected series of deviations with eigenvectors; 10. add these weighted projected deviations to expected returns from each asset class</li><li>[22] Factors reflected by credit spread (4): 1. expected probability of default; 2. expected loss given default; 3. uncertainty around expected probability of default and LGD; 4. liquidity premium</li><li>[22] Common measures of credit spread (3): 1. nominal spread; 2. static spread; 3. option-adjusted spread</li><li>[22] Nominal measure of credit spread: Difference between the gross redemption yields of risky and risk-free bonds</li><li>[22] Static measure of credit spread: Addition to the risk-free rate at which discounted cashflows from a risky bond will equate to its price</li><li>[22] Option-adjusted measure of credit spread: Static measure plus stochastic modeling to allow for any options embedded in the bond</li><li>[22] Reasons why observed credit spreads tend to be higher than can be justified by historical bond defaults (6): 1. higher volatility of returns relative to risk-free asset (credit beta); 2. higher uncertainty of potential future returns; 3. greater skewness of potential future returns on corporate debt (more significant downside); 4. lower liquidity of corporate debt; 5. lower marketability of corporate debt (and higher costs of trade); 6. differences in taxation</li><li>[22] Features of a good benchmark for market risk (13): 1. unambiguous; 2. investable and trackable; 3. measurable on a reasonably frequent basis; 4. appropriate for investor’s objectives; 5. reflective of current investment opinion; 6. specified in advance; (7. contain high proportion of assets held in portfolio; 8. have a similar investment style to portfolio; 9. low turnover of constituents; 10. investable position sizes; 11. strong correlation with portfolio; 12. low correlation with market-benchmark; 13. variability of portfolio relative to benchmark is lower than portfolio relative to market)</li><li>[22] Market strategic risk: Risk of poor performance of benchmark used to judge manager’s performance, relative to liability-based benchmark</li><li>[22] Active risk: Risk of poor performance of manager’s actual portfolio relative to strategic benchmark</li><li>[22] Active return: Difference between return on actual portfolio and return on benchmark</li><li>[22] Brennan-Schwartz model assumptions (4): 1. changes in short term rates vary in line with steepness of the yield curve (differential between long and short term rates); 2. volatility of short term rates varies in proportion to the most recent level of short term rates; 3. changes in long term rates vary in proportion to the square of the level of long term rates and are also influenced by short term rates; 4. volatility of long term rates varies in proportion to the most recent level of long term rates</li><li>[22] Methods of modeling contagion risk (2): 1. consider contagion as a feedback risk (some serial correlation exists that can be modeled), though this in theory should be eliminated through arbitrage and thus ignored; 2. fit a t-copula with situation dependent correlations</li><li>[22] Methods of modeling market returns (5): 1. historical simulations (bootstrapping); 2. forward-looking data-based approaches (multivariate normal distribution); 3. forward-looking factor-based approaches (PCA); (If data is sufficient 4. multivariate distribution other than normal; 5. combining non-normal marginal distributions using a copula)</li><li>[22] Methods of modeling interest rate risk (3): 1. single factor models (for short term single interest rates); 2. two-factor models (Brennan-Schwartz); 3. PCA (for deviations from expected returns)</li><li>[23] Events that could be counted as default (4): 1. payment due is missed; 2. financial ratio falls above or below a certain level; 3. legal proceedings start against credit issuer; 4. present value of assets falls below liabilities</li><li>[23] Components of default risk assessment (3): 1. probability of default; 2. loss on default; 3. level and nature of interactions between various credit exposures and other risks in portfolio</li><li>[23] Methods of strengthening recoveries (2): 1. requiring collateral; 2. requiring third party guarantees</li><li>[23] Actions taken to minimize credit risk (3): 1. diversify credit exposure across a number of counterparties; 2. monitor exposures regularly; 3. take immediate action when default occurs</li><li>[23] Sources of information used to assess credit risk (4): 1. credit issuer (rating agencies); 2. counterparty; 3. publicly available data; 4. proprietary databases</li><li>[23] Factors considered in qualitative credit models (7): 1. nature of contractual obligation; 2. level and nature of any security; 3. nature of the borrower; 4. economic indicators; 5. financial ratios; 6. face-to-face meetings with credit issuer or counterparty; 7. how risks change over time</li><li>[23] Key advantage of qualitative credit models: A wide range of subjective factors can be incorporated into the assessment</li><li>[23] Disadvantages of qualitative credit models (4): 1. excessive subjectivity; 2. lack of consistency between ratings; 3. meaning of subjective ratings may change over the economic cycle; 4. ratings may fail to respond to changes in the economy or circumstances of the counterparty (reluctance to change rating)</li><li>[23] Examples of quantitative credit models (5): 1. credit-scoring; 2. structural (firm value); 3. reduced-form; 4. credit-portfolio; 5. credit-exposure</li><li>[23] Credit scoring models: Forecast the likelihood of a counterparty defaulting at a particular point in time given certain fundamental information about the counterparty (empirical models, expert models)</li><li>[23] Structural models (firm-value): Estimate the likelihood of default using market information such as the company’s share price and volatility of its share price (Merton, KMV)</li><li>[23] Reduced form models: Model default as a statistical process that typically depends on economic variables (credit migration models)</li><li>[23] Credit portfolio models: Estimate credit exposure across several counterparties, may allow for diversification effects of uncorrelated creditors (multivariate structural, multivariate credit migration)</li><li>[23] Credit exposure models: Estimate credit exposure for more complex circumstances, like when derivatives and guarantees are present (Monte Carlo)</li><li>[23] Common difficulties of quantitative credit models (4): 1. lack of publicly available data on default experience; 2. skewness of the distribution of credit losses; 3. correlation of defaults between counterparties; 4. model risk</li><li>[23] Merton model: Structural model using option pricing theory along with equity share price volatility to derive information on the value of company assets (and hence debt); Equity shares can be considered a call option on company assets, then the value of debt is equal to value of a risk-free bond less the value of a put option on company’s assets</li><li>[23] Key advantage of the Merton model for credit risk: Estimates an appropriate credit spread for a bond, even when the bond is unquoted</li><li>[23] Disadvantages of using the Merton model for credit risk (7): Unrealistic assumptions 1. markets are frictionless; 2. risk-free rate is deterministic; 3. asset value follows a log-normal random walk with fixed rate of growth and fixed volatility; 4. asset value is an observable traded security; 5. bond is a zero-coupon bond with only one default opportunity; 6. default results in liquidation; (7. Merton model results can be affect significantly by changes in market sentiment, absent any real changes to company prospects)</li><li>[23] Distance to default: Used in the KMV model, it is the number of standard deviations that the company assets have to fall in value before breaching a threshold</li><li>[23] Advantages of KMV model over Merton model (3): 1. coupon-paying bonds can be modeled; 2. more complex liability structures can be accommodated (KMV uses average coupon and overall gearing level); 3. starting asset values are not assumed to be observable and is derived from the value of the company’s equity shares</li><li>[23] Credit migration modeling process (3): 1. historical data is used to determine the probability that a company rated something at the start of the year will be rated something at the end of the year (recorded in rating probability matrices); 2. matrices are applied repeatedly to a counterparty’s current rating to estimate the likelihood of each possible rating in each future year; 3. using the probability of default for a company of a given rating, the model estimates the chance of default in each future year</li><li>[23] Advantages of credit migration models (2): 1. volatile equity markets should not overly impact results; 2. does not rely on publicly traded share information</li><li>[23] Assumptions of credit migration models (2): 1. credit migration process follows a time-homogeneous Markov chain; 2. there exists a credit rating that reflects the company’s default likelihood through the business cycle (rather than in the current environment)</li><li>[23] Disadvantages of credit migration models (7): 1. incorrect time-homogeneity assumption (recently downgraded company is more likely to be downgraded again); 2. assumes default probabilities for each rating in each future year can be estimated; 3. assumes likelihood of default can be determined solely by credit rating; 4. low number of distinct credit ratings results in a low level of granularity in default estimates; 5. rankings of organizations by different credit rating agencies do not always coincide; 6. not all companies have obtained a credit rating; 7. ratings are sometimes unavailable (withdrawn)</li><li>[23] CreditMetrics approach: Estimates variance of bond’s future value in one year’s time by combining default probabilities from credit migration models with estimated future values at each rating</li><li>[23] Assumptions of the multivariate CreditMetrics model (5): 1. each credit rating has an associated probability of default; 2. change in rating is a function of change in the company’s asset value and volatility of the asset value; 3. value of assets behaves log-normally; 4. correlation between asset values can be estimated from the correlation between corresponding equity values; 5. equity returns can be modeled using country-specific indices and independent firm-specific volatility</li><li>[23] Examples of credit portfolio models (5): 1. multivariate structural; 2. multivariate credit-migration (or financial) models; 3. econometric or actuarial models; 4. common shock models; 5. time-until-default (or survival) models</li><li>[23] Multivariate structural models: Multivariate KMV, model asset values by using correlation matrices or copulas</li><li>[23] Multivariate credit migration models: Multivariate CreditMetrics, assumes equity returns can be modeled using country-specific indices and firm specific volatility</li><li>[23] Econometric or actuarial models for credit risk: Estimates default rate of firms using external or empirical data</li><li>[23] Common shock models for credit risk: Determine probability of no defaults by assuming each bond defaults in line with a Poisson process, and considering shocks that cause the default of one or more of the bonds in the portfolio</li><li>[23] Time until default models for credit risk: Survival CDFs are linked by a suitably parameterized copula function to estimate the aggregate default rate for the bond portfolio</li><li>[23] Common measures of recovery (2): 1. price after default (short term measure); 2. ultimate recovery (larger, long term measure)</li><li>[23] Factors affecting the likely loss on default (6): 1. seniority of debt; 2. availability of collateral; 3. nature of industry; 4. point in economic cycle; 5. legal jurisdiction; 6. rights and actions of other creditors</li><li>[24] Reasons for increased interest in active management of operational risks (7): 1. advent of ERM; 2. new regulatory capital requirements; 3. increasing emphasis on sophisticated quantitative models for other types of risk; 4. operational risks have no inherent upside potential; 5. main driver behind many cases of major financial disasters; 6. inter-linked with credit and market risk and is particularly important to minimize during already stressed market conditions; 7. may otherwise be treated differently in different areas of the company</li><li>[24] Benefits of effective operational risk management (4): 1. minimizes impact of reputational damage (distinctive benefit); 2. minimizes day-to-day losses and reduces potential for more extreme incidents; 3. improves company’s ability to meet business objectives (reduces time spent on crisis management); 4. strengthens overall ERM process and framework</li><li>[24] Initial finding from analyzing publicly available data on operational risks (4): 1. distribution of operational losses is skewed to the right (large number of small losses, few large losses); 2. loss severities have a heavy tailed distribution; 3. losses occur randomly in time; 4. loss frequency may vary considerably over time</li><li>[24] Features of operational risk data that make applying statistical methods difficult (2): 1. low volume of data; 2. have a cyclical component and/or depend on current economic conditions</li><li>[24] Bottom up model for assessing operational risks: Estimates operational risk capital by starting the analysis at a low level of detail and then aggregating the results to obtain an overall assessment</li><li>[24] Top down model for assessing operational risks: Use readily-available data and fairly simple calculations to give a general picture of operational risks</li><li>[24] Steps in using scenario analysis to assess operational risks (5): 1. group risk exposures into broad categories; 2. for each group, develop a plausible adverse scenario; 3. for each scenario, calculate the consequences of the risk event; 4. calculate total costs for all scenarios; 5. assessments of frequency and severity can be displayed on a risk map</li><li>[24] Benefits of using scenario analysis to assess operational risk (5): 1. capture opinions, concerns and experience of risk managers; 2. not rely heavily on availability, accuracy and relevance of historical data; 3. provide opportunity to identify hard to predict, high impact events (black swans); 4. identify and improve understanding of cause and effect relationships; 5. reduce risk reward arbitrage opportunities</li><li>[24] Main advantage of bottom up models for operational risks: Give more robust picture of a company’s overall risk profile</li><li>[24] Limitations of bottom up models for operational risks (3): 1. difficult to break down reported aggregate losses into their constituent components; 2. there may be little robust internal historic data, especially for low probability and high severity events; 3. differences between companies mean that application of external data is difficult</li><li>[24] Basel AMA for operational risk assessment: Under the Basel advanced measurement approach, operational risk is assessed using internal models and scenario analysis, 1-year holding period and a 99.9% confidence interval</li><li>[24] Areas requiring credible data to use Basel AMA for operational risks (3): 1. internal data on repetitive, high frequency looses over a three to five year period; 2. external data on non-repetitive, low frequency losses; 3. suitable stress scenarios to consider</li><li>[24] Advantage of the implied capital model for operational risks: Simple and forward looking</li><li>[24] Limitations of the implied capital model for operational risks (3): 1. total risk capital needs to be estimated; 2. inter-relationships between different types of risk are ignored; 3. does not capture cause and effect scenarios</li><li>[24] Advantage of income volatility model over implied capital model for operational risks: There is better data availability for total income volatility than for total risk capital needed</li><li>[24] Limitations of the income volatility model for operational risks (2): 1. ignores rapid evolution of companies and industries; 2. does not capture the softer measures of risk (opportunity cost, reputation) by focusing on income rather than value</li><li>[24] Advantage of CAPM over the income volatility model for operational risks: Includes both the aggregate effect of specific risk events and the softer issues</li><li>[24] Limitations of CAPM for operational risks (4): 1. no information is provided on losses due to specific risks; 2. level of operational risk capital is unaffected by any controls put in place (little motivation to improve risk management process); 3. tail end risks are not thoroughly accounted for; 4. does not help anticipate incidents</li><li>[24] Stages in a comprehensive operational risk management process (5): 1. risk policy and organization; 2. risk identification and assessment; 3. capital allocation and performance measurement; 4. risk mitigation and control; 5. risk transfer and finance</li><li>[24] Components of a comprehensive operational risk management policy (6): 1. principles for operational risk management; 2. definitions and taxonomy; 3. objectives and goals; 4. processes and tools; 5. organizational structure as it applies to operational risk management; 6. roles and responsibilities of different business areas</li><li>[24] Purpose of loss incident database for operational risks: Help company learn lessons from past loss incidents, analyze trends and support analysis of root causes and mitigation strategies</li><li>[24] Controls self assessment: Internal analysis of key risks and their controls and management implications</li><li>[24] Examples of top down models for operational risks (4): 1. implied capital; 2. income volatility; 3. economic pricing models (CAPM); 4. analogue models (use data from similar companies)</li><li>[24] Factor based approach for assessing operational risks: Assume losses are related to the volume of transactions and apply a weighting to the actual or expected volume of transactions (Basel indicator and standardized approaches)</li><li>[25] Funding liquidity risk: Risk of money markets not being able to supply funding to a business when required</li><li>[25] Market liquidity risk: Lack of capacity in the market to handle asset transactions at a time when the deal is required</li><li>[25] Reasons why quantitative techniques are usually not applied to liquidity risk (2): 1. historic data on liquidity crisis is limited; 2. degree and nature of every organization’s exposure to liquidity risk is different</li><li>[25] Sources of cash inflows (3): 1. revenues/income generated by assets; 2. proceeds from sale of assets; 3. drawings upon sources of liquidity (issue of new debt or equity)</li><li>[25] Scenarios that should be considered when assessing liquidity risk (7): 1. rising interest rates; 2. ratings downgrade; 3. large operational loss; 4. large single insurance claim occurrence; 5. loss of control over a key distribution channel; 6. impaired capital markets; 7. sudden termination of a large reinsurance contract</li><li>[25] Level risk (or underwriting risk) in demographic risk: Risk that the particular underlying population’s loss costs are not as expected over the immediate future, due to shortcomings in the underwriting process</li><li>[25] Reserving risk in demographic risk (3): 1. volatility risk (finite policies); 2. catastrophe risk; 3. trend (or cycle) risk (future long term changes)</li><li>[25] Method of modeling volatility risk: Probabilistically or stochastically assuming some underlying statistical process</li><li>[25] Methods of modeling catastrophe risk (2): 1. scenario analysis; 2. more complex dependencies can be modeled by copulas</li><li>[25] Key distinctions between life and non-life insurance risk (3): 1. non-life policies may experience more than one claim; 2. move through different states over the lifetime of the policy; 3. severity may also need to be modeled</li><li>[25] Methods of assessing liquidity risk (2): 1. scenario analysis (when and why expected cash outflows might exceed inflows); 2. stress testing (examine effect on liquidity of an extreme event)</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=cce3b7004b07" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>