Tuesday, January 01, 2008

Some thought on randomness

2007 has just ended, which also coincides with my first rotation in GDP with the Abu Dhabi project. While I suppose it's a good time to do a self-serving laundry list of the various things I've learnt or accomplished from my first posting, I find myself a bit uneasy and digging deeper to search for signs of any significant meaning of the things that has transpired in the past year (perhaps it's the sudden overdose of "free" time during the holiday season that gave me the luxury of some not-so-practical yet very thought-provoking "philosophical" self-reflection).

In any case, I am glad that I manage to fight off the temptation to plunge directly into the work for my next rotation and took the last week just catching up on a half-year inventory of Economists and various blogs and books. Particularly, I was really intrigued by Fooled by Randomness, a book written by Nassim Taleb. The book really crystallizes what has been at the back of my mind for a while now (regarding my life outcomes so far and into the future) but just couldn't articulate as well. Here're a few quotes from the book:

"I (author) do not dispute that arguments should be simplified to their maximum potential; but people often confuse complex ideas that cannot be simplified into a media-friendly statement as symptomatic of a confused mind."

"One cannot judge a performance in any given field by the results, but by the costs of the alternative."

"The frequency or probability of the loss, in and by itself, is totally irrelevant; it needs to be judged in connection with the magnitude of the outcome."

I see the above (trying to simplify an inherently complex idea) at my work all the time (I'm also as guilty as anyone); a temptation to summarize things into bite-sized pieces given the huge volume of signals in the background. The IRR calculation is a clear example. Everyone in the business world talks about IRR. When I work on the financial model, everyone usually asks for the IRR result first. Yet I'm actually always quite hesitant to "reveal" that number unless I know I'd have that person's attention for another few minutes to explain the assumptions. IRR is extremely sensitive to timing of the cashflow (especially if the profit margin is low), it does not have a mathematical real solution if the cashflow series experience multiple sign changes (ie. +-+-); even the concept of the rate of reinvestment at the calculated IRR is highly suspicious and of course not to mention the inherent difficulty in forecasting future cashflow, which IRR itself is based on.

Most finance textbooks will argue that NPV is a better way to do capital budgeting but making comparison with an arbitrary number with limitless range (NPV can potentially range from some negative to positive "large number", though never really infinity) is evidently more strenuous to the human mind than comparing using a number that has a limited range, like IRR. Of course we have "sensitivity analysis", but I seriously doubt its practicality in the way of influencing actual decisions.

The biggest loophole I see though is the absence of probability being factored into our calculations and to be really rigorous, one possibly will need to resort to some kind of Monte Carlo simulations, at which point I can already hear my teammates saying "your spreadsheet is already too huge! No one can read it but you!". Worse still, after all the hardwork, the simulations could yield a IRR of say 15%, but with a standard deviation of say 20%. Probably very few CEO would want to make a decision based on the statement "there is a 68% chance that the IRR will lie between -5% to 35% for this project", rather he/she'd want to hear "this project has a 15% IRR".

This all seems very academic, and I often feel that what I do on the spreadsheet has little relevance to real-life decisions (perhaps rightfully so since what I do is so incomplete without taking into account probability). But basing important decisions on "gut" feel or "business acumen" also seems a bit too fishy to me. Indeed, the most profitable trader at any one time could be the worst trader by most conventional measures. How do I know my "gut" feel, presumably based on past successful experiences, would work this time round? Karl Popper (philosopher/economist) says it best (as quoted in Nassim's book):

"There are only two types of theories…theories that are known to be wrong…(and) theories that have not yet been known to be wrong, not falsified yet, but are exposed to be proved wrong."

I guess that is why I always hold more than a normal dose of skepticism at what some well-respected and experienced men had to advise simply because I am very doubtful of how much of where I am now is really my own doing, instead of a streak of luck (not that I'm well-respected or experienced in the eyes of many, but I do consider myself a very very fortunate guy).

While all the ranting above seems irrelevant and not stemmed in reality, I think that randomness, or an estimate of its impact is actually highly appropriate when making major (business or life) decisions. One of my resolution for the coming year is, naturally, to integrate the concept of probability more rigorously into my judgment and perception. Fortunately, despite all the "one is not really in control of one's life" talk, we're not always victims of randomness due to its scaling property. So I looked for inspiration in someone who has proved to be successful through the years, and I found this RTHK (Hong Kong's answer to BBC) documentary on Li Ka Sing, the Hong Kong tycoon. After watching the show, I told myself "you really have not experienced enough hardship in your life Michael Cho. You really have to work SO MUCH harder."

All right, this post is getting way too long. I'd just like to end with this following phrase, which keeps me hopeful while reading this "you-are-where-you-are-because-you're-mostly-lucky" book:

"Fortune (Luck) favors the prepared mind." Happy New Year everyone :)

No comments: