Seeking Alpha

Experimental Trader

 
View as an RSS Feed
View Experimental Trader's Comments BY TICKER:
Latest  |  Highest rated
  • Fortune's Formula: The Untold Story Of The Scientific Betting System That Beat The Casinos And Wall Street [View article]
    satan2liberals,

    I think Chris is gracious on his answers to your questions. In my experience, it is to my advantage to understand and practice using the Kelly formula. There are a lot of opinions in the web; it is up to you to figure which is the truth and which is the voodoo science.

    (1) "The formula is particularly unsuited to markets where probabilities are unknown."
    If the probability is unknown, why do you trade it? Trading unknown is really "gambling" (in a bad sense).

    (2) "Too many traders try that silly trade a bunch of stocks thinking they are diversified."
    This has to do with diversification. In fact, the right way is that Kelly should be applied to a strategy, not to a stock. Overtrading highly correlated stocks is a violation to Kelly to start with.

    (3) "Everybody and their grandma thinks they have an exploitable "edge" which leads to inappropriate use of the formula in the first place"
    Yes, Kelly specifically tells you that if you don't have an edge, don't trade it.

    (4) "There's plenty of information out there on the web that demonstrates the Kelly formula is the sure way to lose all your money even in games that ARE appropriate for it."
    My experience is just the opposite. It is a sure way to lose money if a trader doesn't understand the danger of overtrading. The Kelly formula helps on that aspect.

    Just a bit of my background. I trade part time and mostly options. I am grateful to the Kelly formula as a wonderful tool for saving me from "blowing" up my account.
    Feb 8 10:35 AM | 1 Like Like |Link to Comment
  • An Investing Strategy That Is Half Warren Buffett And Half Nassim Taleb [View article]
    Not necessary, but we learn from survival instincts. If you play with fire everyday, you better have respect to fire.
    Feb 7 12:40 PM | Likes Like |Link to Comment
  • Asset Allocation In The Eyes Of The Kelly Strategy, Part II [View article]
    Ypa,

    Thanks for reading my articles and providing feedback.

    (1) I am in agreement with your remark on a small data sample does not make sense applying the Kelly formula. For monthly, this data series has approximately 300 data points. For yearly or multiple years, I am using monthly looking forward 1, 3, 5 years etc, so there are still approximately 250 data points. I agree that this data series does not have longer enough time span. I wish I have data date back say 1920's. It would be interesting to see how the calculation holds.

    (2) The main reason for me to choose Kelly is to answer the question how much asset I can allocate to a particular strategy. I mostly trade options, so the biggest fear is overleaveraging and "blow up" my account. I use Kelly to "curb my enthusiasm" both on the way up and on the way down. I usually calculate Kelly on my total trades (of the same strategy), and on my last 100 trades (to monitor the time drift effect), and choose the lesser of the two. It is interesting to see that if I have a series of bad trades, my Kelly ratio drops quite quickly and tell me to stop trading. So to a degree, I think my practice with Kelly is a way of anti-Martingale. On a longer timeframe, Kelly could act as Martingale (if I choose to believe I should take the bigger Kelly ratios of the two).

    (3) I also agree with your comments of modifying the formula; I haven't yet figured out how so inputs are very appreciated. Meanwhile, I fully acknowledge Kelly is an approximation, but in my opinion it is a good one and a convenient one.

    Lastly, my Kelly articles were inspired by my view that many investors are overleveraged without knowing they are overleveraged, even naively holding SPY. Hope I got my point crossed.
    Feb 7 10:49 AM | Likes Like |Link to Comment
  • An Investing Strategy That Is Half Warren Buffett And Half Nassim Taleb [View article]
    I like the LEAP JNJ ITM call for a different reason, and am bullish that it will come out ahead. But, Bik, I think you are correct, it is by no mean that it is an inexpensive trade to put in for the reason Brendan suggested, i.e., to achieve 1.46 leverage. Specifically, in order to put on this trade, it gives up the dividends ($.61*4*200 = $480) and the time premium ($1*200 = 200). So, it is roughly $680 out of pocket (which is 6.8% of the portfolio). It is the same as to borrow $5K from the bank and pay 13.6% interest, and put the entire bankroll in JNJ, and has an insurance of maximum 10% loss.

    This is where many option traders lose money. The JNJ LEAP is actually cheap; other trades are even more expensive.
    Feb 6 12:32 PM | Likes Like |Link to Comment
  • An Investing Strategy That Is Half Warren Buffett And Half Nassim Taleb [View article]
    There is nothing wrong to hold the cash! I am strong advocate for it. You may want to check out my (only) two seekingalpha write-ups on using the Kelly criterion and specifically talk about using cash as an optionality. I believe it is in alignment with what was published recently in marketfolly.com as interviewing Warren Buffett and other hedge fund managers.

    I think I get the "black swan" now. If it is just the outliner points, there is nothing to be concerned about. It is part of randomness. There are ways to manage it.
    Feb 6 01:32 AM | Likes Like |Link to Comment
  • An Investing Strategy That Is Half Warren Buffett And Half Nassim Taleb [View article]
    The LTCM folks do give a bad rep for the black-scholes model. IMHO, I think the black-scholes model is actually very VERY good. Yes, it has its limitations, but there is really no alternative (maybe small variations but nothing earth shattering). The problem with LTCM, however, is not that their black-scholes model is terribly wrong, but they don't quite understand and practice leveraging. The art and science of leveraging is a totally different ball-game. For most option traders, over leveraging is always a prominent danger, and we learn to live with it and manage it day-in and day-out.

    To answer your question if I hedge naked put, I don't hedge the trade itself per se, but I hedge with different strategies and more importantly, position sizing. To scale it, my average loss per "blow-up" is 2% of my portfolio. Once a while, I may have a 4% or 5% "blow-up", but certainly manageable. There is a remote possibility that if SPY goes to zero tomorrow, I will be completely wiped out. But, if SPY goes down 10% in one day, I will be ok and still have enough dry powder to fight another day.

    So, back to "black swan". If I understand your point correctly, the "black swan" is the outliner point outside the 99% confidence internal (as per Taleb?). Then, I think he managed to scare the readers rather than to educate the readers. A simple 90/10 indicates the "risky" strategy is not very effective. As a concrete example, I will only allocate 10% of my portfolio to a strategy that gives me 1:1 payout ratio with 55/45 odd. If there is a better strategy, should we increase the 10% allocation?
    Feb 6 01:05 AM | Likes Like |Link to Comment
  • An Investing Strategy That Is Half Warren Buffett And Half Nassim Taleb [View article]
    yeah, just checked and you are right. 4.3% is not expensive option and you can probably come out ahead next year. I agree that 70 is deep enough in the money.
    Feb 6 12:17 AM | Likes Like |Link to Comment
  • An Investing Strategy That Is Half Warren Buffett And Half Nassim Taleb [View article]
    Thanks for continuing this thread. First, let me say that I am here not to impose my view, but really would like to get help from someone to point out where I am wrong. This is a topic that I don't have enough knowledge, but also I have not yet hear a resounding counter-point that can change my view.

    I have some experiences in trading options; selling put is my favorite tactics. Based on my limited knowledge, I believe that if one to trade options randomly, regardless buy or sell options, at the end he will come up even (not counting small loss due to commissions and slippage).

    This is both the beauty and the Achilles' heel of the option pricing model. how does an option trader make money? While the aggregate market is efficient, there are special situations that are not efficiently priced. Exploring those situations adds the alpha for an option trader. As you suggested "preys on people's fear or uncertainty", what really happens is that people (or Mr. Market) willingly deviate from the random-walk model by discounting a future situation that may or may not happen. A "brave" (or "stupid") option trader picks up the other side of the trade and believe it will return to the random model.

    A "black swan" (or what I called uncertainty) is something can not be predicted in advance. In real life, I think occurence of natural disaster or man-made disaster may qualify as "black swan". I could not figure out how the bar-bell strategy can be successful in taking advantage of the occurence of the "black swan". Being sarcastic for a moment, taking the bar-bell strategy to an extreme, one should put 90% of his asset under the mattress and spend the 10% buying powerball. That's where I have trouble to understand how to trade a "black swan" situation.

    Feb 5 10:16 PM | Likes Like |Link to Comment
  • An Investing Strategy That Is Half Warren Buffett And Half Nassim Taleb [View article]
    That's randomness, not uncertainty. :) That's where it gets real muddy when we put the concepts in practice.
    Feb 5 03:23 PM | Likes Like |Link to Comment
  • An Investing Strategy That Is Half Warren Buffett And Half Nassim Taleb [View article]
    I concur with your notion on "the theory of relativity". Distinguishing uncertainty and randomness is not easy.

    IMO, uncertainty is not a tradable event. Since you can not predict or foresee, how can you bet on it? In contrast, randomness is where the money is generated. If the main design behind the bar-bell strategy is the "black swan", i.e., uncertain occurence, how can we be certain it is a winning strategy?

    Again, I am not against Taleb and his writing; I think he tried quite hard to distinguish them. But, I just have trouble to wrap my head around a strategy of trading uncertainty. Maybe you can enlighten me?

    Feb 5 02:48 PM | Likes Like |Link to Comment
  • An Investing Strategy That Is Half Warren Buffett And Half Nassim Taleb [View article]
    Read some of Talab's books, but I start to have more and more trouble with the notion of "black swan", especially how it is morphed into the common brief today. Specifically, there is a distinct difference between randomness and uncertainty. Randomness means while you can not predict the outcome of any individual occurence, but you can predict the aggregate outcome. In contrast, uncertainty means no one can predict and foresee. period.

    I think even Taleb tried to explain the difference, but it was lost in the context. Today, "black swan" has been used interchangably between uncertainty and randomness, which defeats the purpose of defining "black swan" to start with.

    Just MHO.
    Feb 5 02:14 PM | Likes Like |Link to Comment
  • An Investing Strategy That Is Half Warren Buffett And Half Nassim Taleb [View article]
    Interesting take but I think your strategy requires some refinement, especially in using LEAP options. Take your example of JNJ, what you are really betting is that JNJ is >$75 by Jan 2014, not >$70. It seems that the lost of the time premium has not yet accounted for.

    There are many ways to reduce the time premium. One possibility is to use deep ITM LEAP option.
    Feb 5 11:13 AM | Likes Like |Link to Comment
  • Fortune's Formula: The Untold Story Of The Scientific Betting System That Beat The Casinos And Wall Street [View article]
    Not really. One could figure it out if you keep a good and honest trading record of your strategy. For example, if I have a method that I think it has an edge, I will trade the same setup numerous times (>100 at least), at the end I will have good b, p and q estimates. Applying Kelly becomes very relevant.
    Feb 4 12:08 AM | 3 Likes Like |Link to Comment
  • Fortune's Formula: The Untold Story Of The Scientific Betting System That Beat The Casinos And Wall Street [View article]
    Shannon was actually a very good investor. He never invested other people's money, but the percentage in return of his personal account was at par with Warren Buffett's. And so was Ed Thorp, who ran hedge funds for many years.

    One other thing I want to point out is that the Kelly formula alone does not create the "edge"; it is basically an optimization solution if you have an edge. If you don't have an edge, the Kelly formula won't help.
    Feb 3 11:58 PM | 5 Likes Like |Link to Comment
  • Fortune's Formula: The Untold Story Of The Scientific Betting System That Beat The Casinos And Wall Street [View article]
    Exellent article.

    The essence of the Kelly formula is to balance the potential for gain with the reserve for loss, which is very applicable to both active and passive investing methods. This is an essential tool for any investors.
    Feb 3 11:45 PM | 2 Likes Like |Link to Comment
COMMENTS STATS
62 Comments
38 Likes