Monday, January 19, 2009

Risk Management 101

It's been said that among traders there are those who quietly know and there are those who desperately hope. The trick is to equip yourself with the tools and mindset to enable you to join the ranks of the first group rather than the second. To that end I'm just finishing up Peter Bernstein's history of risk. Now, before you roll your eyes in utter boredom, I've got to say that this book is just as exciting as any Agatha Christie mystery or Charles Dickens classic. The cast of characters is equal to none. Along the way you'll learn about the quirks and foibles of such colorful notables as Fibonacci, Cardano, Bernoulli, Gauss, Galton, von Neumann (father of game theory), Markowitz and many, many more who were responsible for developments in mathematics that led to probability theory, betting, uncertainty principles, game theory, the insurance industry and, of course, risk management. The book covers a lot a ground but if trading is your thing and you want to develop a methodology to limit your losses, this book will certainly give you a lot of ideas to spring off from. This is not a trading book . . . there are no systems described or TS code for you to cut and paste. but you'll get a good perspective on how some of the best mathematical minds in history solved complex problems (sometimes with great success, sometimes with resounding failure). I bought a used copy through Amazon for $7.50, although it appeared new to me. A bargain at 10x the price IMHO.
Just a few things to ponder in closing:
Half the people you know are below average.
The 50-50 rule: anytime you have a 50-50 chance of making money in the market. .there is a 50-50 probability you'll lose money.
The early bird may get the worm, but the second mouse always gets the cheese.
And as my friend Steve Forbes says: "It's better to be lucky than smart".
Frankly, I tend to believe that you make much of your own luck and as such I'd prefer to be lucky and smart.

5 comments:

Tyro said...

Half the people you know are below average.

In what sense?

GS751 said...

I loved that book.

bzbtrader said...

tyro,
This post was about statistics, not about the people you know. An average is just that. . a midline. In a normal distribution, half the values will be above the midline and half below. This is not a comment on your friends, this is a statistical trusim. I could have just as easily said half the people you know are above average.
Hope that clarifies the inent of the pun.

Tyro said...

BZB,

I understand, I'm not taking it personally. I mean what is being averaged and what is the baseline for comparison? Our friends tend to be similar to us, so if you are above average in intelligence, education or income then your friends will tend to be above average as well (same for below average). The average income of the friends of a person living in a $100/mo apartment will be much, much less than someone in a penthouse condo. Real life tends to be clumpy which is why the average of your friends almost certainly does not reflect the average of a wider population. Friendship is non-random.

There's also the question of what it is you're averaging. Many traits are not normally distributed. In the US, there are many more people who have less than the mean net worth than have more. Even if you are just trying to say that half of your friends are above the average of your friends and not the average of your city or country this is still probably wrong. Say 8 of my friends have a MSc and 2 have a PHd, then 80% are below the average of my friends yet all are above average of the country as a whole. Look at CEO salaries or net worth - there are some who have 2-5 orders of magnitude more wealth than others. It would be an accident of selection if you picked a group where half were above, half below.

I was just thinking that any book about risk should talk about sampling errors, non-normal distributions and these other issues.

But it was probably an off the cuff comment, didn't mean much. Maybe just giving you a hard time :)

Anyway, thanks for the book recommendation, always looking for new things to read.

bzbtrader said...

tyro,
Just to be clear...the things to ponder at the end of the post was my feeble attempt at humor using statistics, and should in no way reflect on the content of the book, which is not a book explaining how various statistics are calculated, but a book about the history of risk management statistics and the men who went out on an intellectual limb (sometimes with serious political and religious overtones)in order to quantify all manner of phenomena that previously had no apparent rhyme or reason.