Engineering Corporate Finances
Portfolio engineering and investing strategy go hand-in-hand and are easily the most mathematically complicated subject in all of financial engineering. As Isaac Newton pointed out, modeling the madness of men is more difficult than modeling the movement of the planets. He was entirely correct, and portfolio engineering is an extreme example of how this is true.
Portfolio engineering is all about developing models and strategies that utilize combinations of assets to maintain a certain percentage return on investment. A wider variety of investments are typically utilized — including equities and debt, bundled and hybrid investments — but they’re almost always derivatives of some sort.
The intent is to ensure a certain amount of return on investment. Some of the early portfolio strategies including options have colorful names such as covered call, protective put, straddle, iron condor, collar, strangle, and ironfly. Many futures strategies focus greatly on generating revenue off the spread.
More and more corporations are beginning to move away from these simple strategies, however, and are moving toward the use of algorithms to determine financial transactions and strategies for the development of portfolios. These algorithms are often based in stochastic calculus, which, when applied to mathematical finance, sets out to estimate and predict time intervals of asset prices by treating them as a random variable. (See the nearby sidebar.)
On May 6, 2010, the stock market experienced something called a flash crash, where the Dow Jones Industrial Average (DJIA) stock index lost nearly 10 percent of its entire value almost instantly and then regained that value in just minutes. It was caused by the use of automated algorithms as a form of portfolio management.
Modern portfolio management is done, in very large part, automatically as managers preset computer algorithms that are designed to take specific actions if specific milestones are reached. For example, they may automatically buy or sell a certain amount of shares of any stock that changes in value to meet specific criteria, depending on the current moment’s value of other assets already in the portfolio.
Again, this strategy is all determined using mathematical models. So, when one algorithm triggered a sell-off of a particular quantity and type of asset, that triggered other algorithms to also sell certain things, and the entire thing became a chain reaction, like dominos being knocked over.
As prices dropped so low, those managers aware of what was occurring took the opportunity to buy up the undervalued assets, and at some point the algorithms eventually triggered a repurchase, driving up price again.
This example actually illustrates two trends in financial engineering. It’s true that the phenomenon was primarily caused by the use of automated mathematical modeling, making it an issue of portfolio engineering, but it could not have happened without the use of advances in computer engineering that allow for such things as high-frequency trading and automated responses to occur.