I will appreciate any discussion on kelly ratio is sizing a portfoilo. A colleague of mine distrusts optimizers because results could be unstable and depend on too many things.

As far as I recall, it would rely on an optimizer. The Kelly Ratio, as Thorp described its usage, would first create the efficient frontier, than combine the Kelly ratio in the highest Sharpe portfolio with 1 - Kelly in the risk-free rate (assuming resultant portfolio had a positive geometric mean return). Thorp describes it here: http://www.bjmath.com/bjmath/thorp/paper.htmThe error of the traditional single period allocation model is not the idea of optimization (you can correct the model for hyper-sensitivity of the optimizer, see some of Michaud's work, for instance). It is the idea that return/risk ratio is constant regardless of leverage, and that return is linear in leverage. For a multi-period investor, this is patently false, as adding leverage will increase returns to a point, but will eventually result in a negative compounded return in expectation if leverage is increased too much. The traditional model doesn't understand this, because it is a single-period model.Who is a single-period investor? Poundstone's Fortune's Formula is basically all about Kelly, and it is written in a narrative format. Fun read.

Kelly is stupid in money management for many reasons, including,1) you always have to bet enough to compensate opportunity cost, so that 2) when you go broke you can work, and maintain your returns that way, or3) raise money by transferring expectation, up to the point where,3) you should always attempt to risk enough that - and move the price to where - risk-adjusted expected returns equal the next competing opportunity, as well as4) nobody ever controls someone's whole portfolio, among other things...

The Kelly criterion is quite risky in terms of drawdown risks. Fractional Kelly is very reasonable. The drawdown risk issue is discussed in Section 2.4 (freely available) of the new book Quantitavtive Strategies for Derivatives Trading.

QuoteOriginally posted by: dopemanThe Kelly criterion is quite risky in terms of drawdown risks.In other words, it is simply wrong to use it. The amount of risk you take relative to capital depends on one thing, the amount of risk an investor is taking in other parts of his portfolio, and his intended allocation percentage to what you are doing. If you have a higher rate of return, you must specifically solicit him to invest a great portion of his money.

why are constantly talking about imaginary investors that give you money to manage?oh, and btw, considering his other investments will not help you. if you lose a lot of money for clients but point out the rest of his portfolio is doing great so it's not so bad, what do you think he will say to you?

QuoteOriginally posted by: Lepperbeoh, and btw, considering his other investments will not help you. if you lose a lot of money for clients but point out the rest of his portfolio is doing great so it's not so bad, what do you think he will say to you?You are fukin retarded. You just compared your volatility to the volatility of his "other investments!" This is my point, you cannot use kelly to take risk greater than his other investments, or in fact to choose any level of risk unrelated to his other investments, for the simple reason that the risk of his other investments is a reflection of the investor's own choice of risk!If his other investments exhibit similar volatility, then you have calibrated your risk right. If they exhibit less relative to capital allocated, you have calibrated your risk wrong. Or, if you simply are losing money in the long run while his other investments make money, then no, neither kelly, nor fractional kelly, nor any reduction of capital risked approaching zero - subsituted for comparison to other investments - will help you. What kind of retarded nonsense are you thinking?Let me get this straight: You will perform flawed calculations, based on the flawed notion that if you lose all the client a money has currently allocated to you, he will be out of the game. When in reality, you are saying that only you will be out of his game. In fact, if the investor is using kelly over his entire portfolio, then he should allocate money away from you as his expectation for your returns goes down! And you would seek to frustrate this, and mislead him as to expectation, and claim that you are using kelly! Or if expectation has not gone down, and he allocates money away from you, it is to regain control of his volatility from you, you kelly fool. So you are saying you will put your own self interest ahead of the client. Which of course is accomplished in the real world not by using kelly, but by calibrating your risk competitively to match his other investments (while also trying to make money). And because of scumbags like you, is why money is allocated based on reputation, and why clients set their own volatility objectives, rather than allowing you to pick them.But of course no manager would choose to lose money anyway, this thread is about an actual real-world choice, the amount of capital risked. Based on this and your cryptic comments in the Olsen thread, I wonder if you are qualified for day work.

QuoteOriginally posted by: Lepperbewhy are constantly talking about imaginary investors that give you money to manage?I have the best statistical trading system in the world today.

Quotemr. farmer I believe - You are fukin retarded Get out of bed on the wrong side today did we farmer ? I suggest getting some more sleep .

QuoteOriginally posted by: farmerQuoteOriginally posted by: Lepperbewhy are constantly talking about imaginary investors that give you money to manage?I have the best statistical trading system in the world today.Yeah and no takers I guess. If you have the best system in the world why don't you trade your own money like a real man?And stop blabbing about risk versus other investments in the portfolio of an imaginary client.

QuoteOriginally posted by: farmerQuoteOriginally posted by: Lepperbeoh, and btw, considering his other investments will not help you. if you lose a lot of money for clients but point out the rest of his portfolio is doing great so it's not so bad, what do you think he will say to you?You are fukin retarded. You just compared your volatility to the volatility of his "other investments!" This is my point, you cannot use kelly to take risk greater than his other investments, or in fact to choose any level of risk unrelated to his other investments, for the simple reason that the risk of his other investments is a reflection of the investor's own choice of risk!If his other investments exhibit similar volatility, then you have calibrated your risk right. If they exhibit less relative to capital allocated, you have calibrated your risk wrong. Or, if you simply are losing money in the long run while his other investments make money, then no, neither kelly, nor fractional kelly, nor any reduction of capital risked approaching zero - subsituted for comparison to other investments - will help you. What kind of retarded nonsense are you thinking?Let me get this straight: You will perform flawed calculations, based on the flawed notion that if you lose all the client a money has currently allocated to you, he will be out of the game. When in reality, you are saying that only you will be out of his game. In fact, if the investor is using kelly over his entire portfolio, then he should allocate money away from you as his expectation for your returns goes down! And you would seek to frustrate this, and mislead him as to expectation, and claim that you are using kelly! Or if expectation has not gone down, and he allocates money away from you, it is to regain control of his volatility from you, you kelly fool. So you are saying you will put your own self interest ahead of the client. Which of course is accomplished in the real world not by using kelly, but by calibrating your risk competitively to match his other investments (while also trying to make money). And because of scumbags like you, is why money is allocated based on reputation, and why clients set their own volatility objectives, rather than allowing you to pick them.Interesting rant but I find it quite hard to follow your twisted logic.I actually don't think Kelly is of any use in the real world. Second, you can't adjust the volatility of your trading to your clients portfolio (different clients, changing portfolio etc.etc.). You just set the leverage to whatever leverage you are comfortable with yourself and find clients that have the same preference.QuoteBut of course no manager would choose to lose money anyway, this thread is about an actual real-world choice, the amount of capital risked. Based on this and your cryptic comments in the Olsen thread, I wonder if you are qualified for day work.Not sure what thread you're referring too. And yeah, you're correct ... I don't qualify for day work. Gave up trading for a boss a long time ago, didn't like having to deal with people like you that have no clue what they're talking about but are full of b*llsh*t anyway.

QuoteOriginally posted by: LepperbeYeah and no takers I guess. If you have the best system in the world why don't you trade your own money like a real man?"Like a real man," lol. Once again, as with "what do you think he will say to you," you are the one who seems to be more interested in play-acting than in reality."today i will trade. i will stop hiding and tell those investors what i am made of." -lepperbe

Don't spend another month in some lagging pop-culture investment, while I could be making you money. You can start building your wealth, and your reputation today.

QuoteOriginally posted by: LepperbeDon't spend another month in some lagging pop-culture investment, while I could be making you money. You can start building your wealth, and your reputation today. What's your point? I am sorry, there seems to be some non-investment related dynamic dominating your line of thought. Are you gay? Are you going to start lecturing me on how I dress?

QuoteOriginally posted by: LepperbeNo investor will ever take you seriously in those shoes!