Financial Engineering and Risk Management

Course Overview

Martin Haugh

Garud Iyengar

Columbia University Industrial Engineering and Operations Research

Why nancial markets?

Financial markets enable ecient allocation of resources across time across states of nature Young worker with a high salary. What should she do? Financial markets: invest in stocks and bonds to nance retirement, home ownership, education, etc. No nancial markets: consumption? what else? Farmer producing oranges. Financial markets: Hedge using futures markets, weather related derivatives. No nancial markets: only the spot market available.

More on markets and products

Role of Markets Gather information Aggregate liquidity i.e. demand and supply Promote eciency and fairness Products: satisfy needs Hedge risk Allow speculation Raise funds Fund liabilities

Modeling nancial markets

Two kinds of market models Discrete time models
Single period models Multi-period models

Continuous time models Discrete time models vs Continuous time models Pros: All important concepts with less sophisticated mathematics Cons: No closed form solutions ... have to resort to numerical calculations. Focus of this course: Discrete time multi-period models Caveat: Some continuous time concepts covered, e.g. the Black-Scholes formula.

Financial Economics vs Financial Engineering

Financial Economics: Use equilibrium arguments to Price equities, bonds and other assets Set interest rates Financial Engineering: Assume prices of equities and interest rates given Price derivatives using the no-arbitrage condition Not even close to being a complete dichotomy Capital Asset Pricing Model of interest to both

Central problems of FE
Security pricing Primary securities: stocks and bonds ... nancial economics Derivative securities: forwards, swaps, futures, options. Portfolio selection: choose a trading strategy to maximize the utility of consumption and nal wealth. Intimately related to security pricing Single-period models: Markowitz portfolio selection Real options Risk management: understand the risks inherent in a portfolio Tail risk: probability of large losses Value-at-risk and conditional value-at-risk Starting to become important for portfolio selection as well. Led to interesting applied math / operations research problems.
6

Introduction to no-arbitrage arguments

No free lunch ... every non-zero non-negative payo comes with a cost. Suppose c0 , c1 , . . ., cT cash (in)ow from a contract.
Weak No-Arbitrage: ck 0 for all k 1 c0 0 Strong No-Arbitrage: ck 0 for all k 1 and cl > 0 for some l c0 < 0

Interpretation c0 = cost of purchasing the cash ow (c1 , . . . , cT ) Suppose ck 0 for all k 1. Then cost c0 0 Suppose ck 0 for all k 1 and cl > 0 for some l . Then cost c0 > 0 Why? The cash ow has a seller who receives c0 and pays out ck for k 1 Since ck 0 for all k 1 the seller can increase the price c0 The seller can continue to increase the price until c0 = 0 Assumptions Price information is available to all buyers and sellers Enough buyers and sellers
7

Introduction to no-arbitrage arguments

No free lunch ... every non-zero non-negative payo comes with a cost. Suppose c0 , c1 , . . ., cT cash (in)ow from a contract.
Weak No-Arbitrage: ck 0 for all k 1 c0 0 Strong No-Arbitrage: ck 0 for all k 1 and cl > 0 for some l c0 < 0

Example. What is the value of a contract that pays A dollars in 1 year. Suppose one is able to borrow or lend unlimited amounts at r per year. Buy contract at price p and borrow A/(1 + r ) at interest rate r Cashow now A p + 1 + r Cashow in 1 year AA=0
A 1+r .

No-arbitrage: c1 0 implies c0 0, i.e. p

Sell contract at price p and lend A/(1 + r ) at interest rate r Same arguments imply that p
A 1+r