Experiences of top quants

-Book review of ‘How I became a quant’

This book tells stories of 25 famous quants, and their stories were written by quants themselves. So, you will see their different writing styles and experiences.

·      Most of them have a Doctor’s degree in STEM. It is reasonable because quants need to apply quantitative skills in solving financial problems, and advanced degrees are a proof of their strong academic problem solving skill.

·      25 quants have careers in different areas, including structured finance, portfolio management, risk management, derivatives, algorithmic trading, etc. It shows the broad application of quantitative tools in current finance industry.

·      Many of them worked for small companies focusing on a small area. The other many quants led quantitative groups within big and well known financial groups.

Why CLO may be a good investment opportunity?

Development of CLO

Collateralized loan obligation (CLO) is a form of structured products. It pools hundreds of business loans and passes cash flows to tranches of different priorities in receiving cash flow. Normally, these business loans are from below investment grade companies, which normally need to provide higher yield in their bond issuance. CLO is great at attracting all types of investors from risk averse to risk taking, to invest in corporate debts which were considered risky and not suitable for risk averse investors.

The first CLO was issued in 1987, and it provided incentives for depository banks to securitize these loans and remove these loans from their books by selling them to outside investors. The 2007-08 financial crisis is a big hit on the CLO market, although the key reason for the crisis was from subprime mortgage instead of corporate loans. Investors doubted the safety of CLOs since they have a similar structure with problematic subprime Collateralized debt obligation (CDO) and CDO². However, CLO quickly gained favor from investors, and had been the first type of secured products that recovered to the pre-crisis trading volume. 

Why do we care about Libor?

Background

The London Interbank Offered Rate (Libor) is the expected interest rate for leading banks in London to borrow money from other banks. Libor rates are calculated for five currencies and seven borrowing periods ranging from overnight to one year. It is considered to be the most important number in the financial world because at least $350 trillion in derivatives and other financial products are tied to it. (Article)

The Libor was invented by Minoz Zombanakis, who tried to let banks lend $80 million to the cash-strapped Iran in 1969. Banks hesitated to lend money at a fixed rate for long periods, because of the rising inflation in UK. Zombanakis offered a solution in which the interest rate would be recalculated every few months. He marketed the deal to a variety of local and foreign banks that could each take a slice, and the banks in the syndicate would report their funding costs just before a loan-rollover date. The weighted average plus a spread for profit became the price of the loan for the next period, and the rate was called London interbank offered rate (Article).

In 1984, the British Bankers’ Association (BBA) established the BBA standard for interest rate swaps including the fixing of BBA interest-settlement rate, which was later officially called BBA Libor in 1986.

C++ implementation of a simple order book

Please refer to my github for the code: https://github.com/DongliangLarryYi. 

1.     Data Structure

1.1  A basic order

Each order include ID, Side (buy or sell), Price, and Size.

//  Thanks to Daniel Cao who provided part of this code

#include <iostream>

#include <list>

#include <algorithm>

#include <iterator>

#include <fstream>

#define MAX_ID 100000 // at most 100000 orders

using namespace std;

// structure of an order which include id, side, price, size and related functions

class Order{

private:

    int id;

    char side;

    double price;

    int size;

Summary of some machine learning concepts and methods

Concepts

Cost function: It is used to measure the accuracy of a predictive model. It takes an average difference of all the results predicted by the model with inputs from x's (features or explanatory variables) and the actual output y's. (Week 1 of Andrew Ng’s class)

Regularization: It is added in the cost function to put some penalty on parameters of a model. It is used to prevent overfitting. (Week 3)

Gradient descent: It is a method to find the local minimum of a function with respect to some parameters. It is used in machine learning to find the best parameters in the model. (Week 1)