The book’s author, John Lanchester, explains the essence of risk management using the financial crisis of 2008 as the story’s background. Below are several key items I learned from this book:
What You Think May Not Be the Same as What It Truly Is
Do You Know the Data Behind the Data?
Have you ever heard of CDS (Collateralized Debt Obligation) and CDO(Credit Default Swap)? They are famous financial derivatives terms during the 2008 financial crisis. They were packaged as easy-to-understand financial products to sell to the public. Risky debts/loans disappear like magic when decomposed and mixed with other good financial products.
Where Did You Get Your Data Source? Who Is the Owner of the Data and Who Are the Key Influencers of the Data Owner?
Second, what was even worse was that the credit rating agencies misjudged the situation making the public believe what they had bought was highly safe. But if we clarify the relationship between the credit rating agencies and those who sell risky financial derivatives, we can get one question to answer: would you give a low score to your client’s products with the purpose of earning money from your clients?
The Data Used for a Prediction Model Could be Wrong
Does the Data We Collect Meet What We Aim to Predict
A risk prediction model to predict something we haven’t seen before in our database? Could this happen? Always, there is something that humans cannot measure. But what if we insist on the measure? This could lead to GIGO (Garbage In Garbage Out). The risk prediction model we built could not work but somehow people believe this is the model to protect us, therefore, enlarging an even greater risk impact in this case. So, before building a model, ask ourselves: does the data we collect meet what we aim to predict?
Variation and Correlation in Risk Management
When selecting a basket of financial investment targets, consider the variation of the targets in a specified timeline and better include the inverse correlation targets to reduce the risk caused by unknown economic factors.
