Quote from Thunderdog:
Two observations. First, with a coin flip, the odds per flip and the expectancy of outcome over many trials are mathematically calculable with a high level of confidence. And we are assuming that flips of a fair coin are indeed truly random in respect of one another. However, I don't think that we can legitimately assume that all price moves are entirely independent of one another. People tend to act in groups (or herds) and setups are normally based on the preceding price action. They may conflict with one another, but I would hardly call them independent (random). But that's neither here nor there because we're talking about coin tosses rather than the markets.
I am not making an assertion as to whether people act in herds or not, and whether you could gain information from noticing that they do. That was my whole point about asking the "wrong" question: it doesn't matter to many traders that are hugely profitable. In the hope of being more clear, I am making two statements:
1) Many if not most derivatives traders
assume that the underlying follows a [Geometric] Brownian Motion, or said another way that the underlying is a random variable like a coin flip. Whether that is the truth or not is a philosophical question
to derivatives traders. The way they make markets or compute theoretical prices completely ignores this philosophical question because the simplification of assuming a random variable and ignoring the philosophical question doesn't affect their profitability.
2) I stated that the question of trying to predict markets confused people into followig the wrong path to trading profitability. I tried to give an analogy by showing how you can make money playing games where the outcome of the event is random and therefore unpredictable by
definition. While I do not claim that markets are coin tossing games, the insight gained from reducing the complexities of markets to simple games like these should not be underestimated. In fact, if you understand the mathematics of coin flips and expentancies etc, it is amazing how close you get to the way options prices are derived mathematically.
My second observation is that your coin toss analogy seems to be alluding to the idea that your edge (for lack of a better word) is based not so much on the observation of non-random price phenomena, but rather on the observation of mispriced options based on the assumption of random price action of the underlying.
I had to read this a couple of times to understand what you say I am saying, but I think you got it. I am saying that an option's price does not predict where the underlying goes. It says that given a volatility expectation for that option, I can compute it's fair price by plugging that value along with other less complex parameters into the Black Scholes model. The BS model then outputs the theoretical price. This theoretical price plays the same role as the 50:50 probability that we realize is true for a coin flip. It is the average price when all possible paths are computed and weighted according to probabilities. Same with the coin flip.
We then make a market around that theo price, which is the edge we demand for making a market. If you fill where we make a market, it is like the coin toss game where we demand 60% if we win and we pay out 40% if we lose. Of course, on any one realization of this game, it may be 80:20 or 90:10 or whatever, but over many many trials and many many positions, that or somethig like it is our expectancy. Notice there is no prediction of the underlying whatsoever, and in fact we have no idea which trades will lose or win. Prediction is replaced by
computation of fair price, and our edge is the spread we make around fair price, which we demand for taking the risk and providing liquidity.
Personally, because I don't believe price action of the underlying is sufficiently random, I think that your premise is flawed. That is not to say that it won't necessarily work for you for a time. But I wouldn't be quite so confident in your options pricing model as I would be with your 60/40 coin toss analogy.
Nearly 100% of options market making firms do exactly this. It is hugely profitable and has been for decades.
nitro
