Calculations
October 18th, 2007 by MarkWork is going well. Not astoundingly well, but a solid improvement from a month and a half ago. My student numbers are up, my 3rd semester class has finished reading a 40 page Aladdin
The social life has been good, too. There are definitely some people who have left Taiwan that I miss, but things are good.
In terms of personal study, once again, things are good. I’ve had time to study a bit of Chinese, a bit of Japanese, some philosophy and some finance. I guess being a net-addict has its advantages.
What’s really on my mind though, is a choice. If I sell all my investments, I have an opportunity to start an incubator hedge fund. It would be expensive and it would be risky, but as the 62nd rule of Ferengi rule of acquisition states, “the riskier the road, the greater the profit.”
Since I’m a long-term value investor, running this sort of fund would take about the same amount of time I’m currently spending on researching investments (relatively little), but oh the risk! If my investing performance of the last 6 years is due to skill, then taking things to the next level is the thing to do. If I’ve just been lucky though, I could lose quite a bit by betting that luck will continue.
It’s time for some thinking.
October 19th, 2007 at 9:29 pm
I thought you’d like to know that thedailyinvestment.net/ is stealing your content and posting it without crediting you. He’s doing the same to me and to many others. I reported him to adsense abuse. The more of us who do that might make him stop getting paid for what we write.
google.com/adsense_dmca.html lists the steps to follow to file a complaint.
October 20th, 2007 at 3:33 am
Thanks for the heads up, Alex. Some people have copied my articles about language learning before, but only for what seemed to be altruistic purposes.
This is just ridiculous, though. Adsense really does seem to bring out the worst in people.
(their copy is at thedailyinvestment.net/calculations)
October 21st, 2007 at 4:15 am
Just a quick, unsolicited suggestion from someone who knows a lot of people in the financial industry - the hedge fund industry is extremely glutted right now. I would try something else if you’re looking for a new way to invest.
October 21st, 2007 at 12:15 pm
Thanks for the feedback, Nathan. Is there any alternative you can suggest that with similarly low barriers to entry?
I realize that the industry is glutted, but unfortunately I’m unlicensed and have no “pedigree”. As such, my options are very limited. Still, my investment performance over the last several years has soundly beaten over 99% of all managed funds. I’m not discounting the possibility that I’ve just been extremely lucky, but if I do have skills, then I want to make the most of them.
October 23rd, 2007 at 1:19 am
I was almost ready to suggest using Kelly Betting, but I quickly noticed that you’ve already included the Kelly Criterion equation at the end.
The first question to ask is how you define ‘ruin’ and how much you’re willing to risk. When using full Kelly Betting, the chance of being reduced to 1/n of your original bankroll is 1/n (for example there’s a 50% chance of being reduced to 50% at some point, a 10% chance to be reduce to 10%, and so on). Most people use a fraction of the Kelly Criterion to reduce the risk of ruin (for example, use 1/2 of what the Kelly Criterion recommends).
The next question is the amount of your bankroll that you should place on one bet (like a hedge fund). For example, let’s say that half the time you lose all you money, and the other half of the time you make 10x your money. We’d have this equation:
f = [bp - q]/b
(where f = percentage of bankroll to bet, b = odds of bet, p = percentage chance of winning bet, q = percentage chance of losing bet = 1-p)
f = [(10)(0.5) - (0.5)]/(10)
f = 4.5 / 10 = 45%
In fact, in the limiting case as b goes to infinity, the most you’d ever bet under the Kelly Criterion for a weighted coin flip is 50% of your bankroll. Taken a step further, the maximum bet for an infinite ‘b’ (payout) is the percentage chance of winning the bet.
Lim b->inf of f = [bp - q]/b is f = p
So, I would say that there isn’t any bet that’s so good that you’d ever bet your whole bankroll on it. When using half Kelly, the biggest bet you’d ever make is half your bankroll, and that would be for a sure bet only.
Of course, the hedge fund outcome is really the result of many small bets (not just one big bet), so it’s hard to apply Kelly Betting perfectly. My gut-feel, though, is that you should only start this with at most half of your net worth (and maybe get other investors to cover the difference)… Just a thought…
October 23rd, 2007 at 5:56 am
Matt, I appreciate the feedback. I’m willing to ride out extreme volatility, so I don’t have much interest in using half-Kelly betting. I’m trying to maximize long-term growth, so my IRA is off limits, but everything else is open for investing.
The real problem is trying to assign a value to p. If my performance up to this point is representative of my skills, then p would be very nearly 1. If I’ve just been lucky, then p might be 0.2 or less.
Getting other investors is pretty much right out of the question until I have a “usable track-record”.