"Raj Rathnaram wa also a good predictor ." I think that he predicted that he wouldn't get caught. He was wrong and still made money (not to mention his lawyers).
Lucias, (meant respectfully): I am having trouble following your comments and the key points you are making because of the wordiness and apparent contradictions. Minds can easily hold multiple contradictions at the same time (as least I can LOL). You said "it is an axiom that is self demonstrated". This is a nonsensical statement to me, since an axiom is assumed to be true but can not (by definition) be proved (or demonstrated) with consistency in the system of rules in which it is assumed to be true. An axiom has to be accepted as true (a belief) to use the system. Zen and the Art of Motorcycle Maintenance has a classic example of this using one of Geometry's axioms. You said "The only advantage one can have in an active trading (or gambling) system is a predictive advantage." and "A prediction is a testable statement about the market" and then you proceed to give examples like arbitrage which makes no prediction. So, are there non-predicting systems or not according to you? Isn't one of your statements not true? "If I'm wrong then show me an example where one isn't required to predict something that can produce a profit (at a rate greater then chance)." Lets assume there is no slippage or costs for a theoretical example: How about a random entry daytrade system that takes a profit when it hits 1 dollar profit (or exits MOC) and takes a loss when it hits 20 cents. (I use theoretical since I don't want to disclose other hard-won information and I am interested in your rebuttle). Assuming that the delta is exactly 50% at the current price, wouldn't there be a consistent profit without any prediction involved. (Please don't try this system at home!!!!). If the market is random (which I don't really think it is) then creating a set of rules and executing them perfectly (which is not random) must produce either a winning or a losing record. How could it produce exactly zero over the long term? Lastly, so what if your statement is true? How does that help us make money or lower our costs as traders? Are we merely arguing semantics? I look forward to your answers.
I'm no expert on the subject but I love reading about this stuff. Think about this, when we look at the stars we are "seeing" what happened possibly millions of years ago. This is because the stars are light years away and it takes all of that time for the light to even reach us. We are constantly looking into the past by just looking at starts so the ability to look into the future doesn't seem that far out. How far could one possibly see into the future? Millions of years? That would contradict freewill and I don't like that idea but then again I don't know jack about the subject. Back to the topic of the thread ...I think there is a great difference between assessing probabilities of future outcomes and having 100% knowledge of future outcomes which may or may not be possible.
>You said "The only advantage one can have in an active trading (or >gambling) system is a predictive advantage." and "A prediction is a >testable statement about the market" and then you proceed to give >examples like arbitrage which makes no prediction. So, are there >non-predicting systems or not according to you? Isn't one of your >statements not true? No, these statements are not at odds. It is very doubtful that pure arbitrage exists in the market and even if it does then it is not going to be available to 99.99% of traders. So, it doesn't really conflict. This is the first law of gambling. Any gambling method that is sound is based on a predictive edge. This is how us dumb people (like me) can make quick decisions about whether a purported edge is worth reviewing. >Lets assume there is no slippage or costs for a theoretical >example: >How about a random entry daytrade system that takes a profit >when it hits 1 dollar profit (or exits MOC) and takes a loss when it >hits 20 cents. (I use theoretical since I don't want to disclose other >hard-won information and I am interested in your rebuttle). Expectancy would be zero. You'd get bigger winners less often but you'd get smaller losses more often. Your expectancy would be zero. >If the market is random (which I don't really think it is) then >creating a set of rules and executing them perfectly (which is not >random) must produce either a winning or a losing record. How >could it produce exactly zero over the long term? You are confusing a certain outcome with the expected outcome. A fair coin is expected to produce 50% heads/tails. There is a probability that one could get all heads out of 10 flips. It is not expected. We say such things are lower probability. The number of absolute heads/tail runs will deviate more and more from zero the more one flips the coin but as a proportion it will tend toward the true expected value. Also, it is nonsensical to talk about a perfect set of rules to be run over a random data. Random means all possibilities are equally likely. All rules would be equally good (or bad). No particular rule would be any more valuable then any other random rule. >>Lastly, so what if your statement is true? How does that help us >>make money or lower our costs as traders? It very important because it tells you what you are trying to do. A lot of traders don't know what they are trying to do. It tells you why you get payed or lose money. I can't think of anything more important. I recommend you read the first few chapters of Evidence Based Technical Analysis for a better introduction.
"No, these statements are not at odds. It is very doubtful that pure arbitrage exists in the market and even if it does then it is not going to be available to 99.99% of traders. So, it doesn't really conflict. This is the first law of gambling. Any gambling method that is sound is based on a predictive edge. This is how us dumb people (like me) can make quick decisions about whether a purported edge is worth reviewing." I actually asked about the logical truth of two seemingly contradictory statements and not about the odds of each statement's truth. In fact, I wasn't aware that logic involved odds. Is 99.99% True = logical True. And does "doesn't really conflict" mean logical True? So, by your reasoning, latency arbitrage doesn't exist and is not available to 99.99% of traders? Since there are likely way more than 10 million investors/traders in the markets, then there are at least 1000 traders who know something about the market that you don't know. Isn't that an edge for them? Arbitrage doesn't rely on prediction but other factors in an imperfect world. I am guessing that you think that trading is gambling. Would you also be comfortable with the opposite statement: gambling is trading? I would not. I would be interested in you sharing the other laws of gambling that you are aware of. Thanks.
I've stated my case, if one wants to profit from the market by active trading then one needs to predict it. I acknowledged there were possible exceptions: arbitrage and rebates. The theories of Modern Portfolio Theory might yield a different view. But, I'm not sure that falls under active trading. The results are also based on the ability to forecast future returns and risk: so, I don't think it is free of prediction. I've found focusing my strengths brings out my best performance.
I respect your case and so will leave you alone after this to present it. In my world: Random DOES NOT necessarily mean all possibilities are equally likely unless you are talking about particular distributions. There are other kinds of distributions and I think the market is likely one of those other kinds. Random is about predictability and not probability in my view. Implementation of MPT is mostly nonsense in my view. Add to that the efficient market theory. Each contains some truth along with some gotchas! OMG - I think we agree on something! Thanks for the replies.
I think there are factors that affect movements which, if known, can lead to something close to prediction. For example, it takes money to buy a stock, and it takes buying to move the price up. If I buy with my whole account, I can't buy more. And, again, no buying means no upward movement. So what happens when everyone has bought all they can buy and the crowd has no liquidity? The price must go down. This is why when more than 65% of the population becomes bullish, the price is ready to turn around. The people who are bullish have already bought a bulk of what they can buy, and the small portion that is bearish sells short with few buyers able to absorb it. This works on any scale. Day traders typically use half or more of their buying power on a single trade. So as a crowd they become non-liquid very fast, and stop being able to buy which allows for a small correction in the price. All you need to predict movements with a decent level of reliability is to have some indication of the crowd's liquidity and their sentiment (bullish/bearish). High bullish sentiment usually means low liquidity, and vice versa. And when I see that, it's time to sell. I see this well enough with only an MACD and V/OI. I'm sure that's how Paulson knew about the housing market crash. The truth is, the CEO at my old bank (Old National) knew of that as well back in 2004 and years before it happened, stopped doing housing loans. It wasn't hard to figure out people were over-extending themselves because of a belief that houses would always appreciate.