2,837 reputation
1413
bio website
location
age
visits member for 3 years, 6 months
seen 12 hours ago

Jul
21
awarded  Nice Answer
Jul
7
answered How to calculate expected return based on historical data for Mean Variance Analysis
Jul
1
answered Reading recommendation on using statistical analysis in online fraud prevention
Jul
1
comment Optimality of Kelly criterion in non-normal environment
@bellamyj, The problem is, most people don't realize that when they look up the simple formula/applet for "Kelly Criterion", it is for a two-outcome event. They then apply that formula/applet to distributions that have nothing to do with two-outcome events, giving them "...a percentage of their stash..." to bet. That number is meaningless. It may or may not be in the ballpark with a reasonable "...optimal percentage of their stash...". The point is, if it is called "Growth Optimal Strategies", "Log Utility Strategies", or anything other than "Kelly Criterion", the confusion disappears.
Jun
30
comment Optimality of Kelly criterion in non-normal environment
@Joshua, Even though those slides use the term "Kelly Strategies", keep in mind that this is NOT what most people refer to when they discuss "Kelly" methods/strategies. The terminology in that paper would be less confusing if it used something like "Log Utility Strategies". "Kelly Criterion" as it has been known for decades is based on two-outcome events. Here's a typical example.... en.wikipedia.org/wiki/Kelly_criterion
Jun
30
answered Optimality of Kelly criterion in non-normal environment
Jun
30
answered Variable Selection in factor models
May
26
revised Mean Reversion Time Frame
deleted 1 characters in body
May
26
comment Mean Reversion Time Frame
@Abhay, I updated my answer to illustrate an example. Also, I'm sorry, but I don't use Excel for these types of calculations. You might consider R.
May
26
revised Mean Reversion Time Frame
added 1257 characters in body
May
26
answered Mean Reversion Time Frame
May
24
answered Tradable Volatility
May
24
comment If I have a model that gives 10% “probability edge” over random chance, how do I calculate the position size?
The Kelly Criterion is for a two outcome bet (not a continuous distribution of possible outcomes). At even money (50%/50%) and 1-to-1 odds, the amount to bet is 0% of your stash. At 1-to-1 odds and probabilities of 60%win/40%lose, the bet is 20% of your stash. At 1-to-1 odds and 80%win/20%lose, the bet is 60% of your stash. And at 1-to-1 odds and 100%win/0%lose, the bet is 100% of your stash.
May
24
awarded  Fanatic
May
20
revised Reliable Economic Data on China
added 607 characters in body
May
20
answered Reliable Economic Data on China
May
19
answered How are cryptography and speech recognition technology applied to forecasting financial markets?
May
19
comment How are dual class shares different from non dual class shares from a market makers' perspective?
@knorv: I added my response as an edit above.
May
19
revised How are dual class shares different from non dual class shares from a market makers' perspective?
added 1139 characters in body
May
18
answered How are dual class shares different from non dual class shares from a market makers' perspective?