Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Strategies & Market Trends : ahhaha's ahs -- Ignore unavailable to you. Want to Upgrade?

To: Bilow who wrote (451)	11/7/2000 10:48:21 PM
From: GraceZ	Read Replies (1) \| Respond to of 24758

At the micro level, prices are more predictable than at the longer term levels. They are both equally unpredictable. Isn't the long term built on what happens at the "micro level"? The term "micro level" is sufficiently vague so that you can't assert anything from it. The reason why you can't make money from it is because of the spreads. The spreads have nothing to do with whether or not price is random or predictable. But let me see if I get what you are saying, are you saying that price is predictable at the micro but its impossible to make money because of the spread between the bid/ask....and that you can't make money at the "longer level" because price isn't predictable enough. Is that what you are trying to say? The most obvious way to check for randomness is to observe the sequence of fills at the bid or ask. It isn't possible to establish whether any time series data is random especially not with measurement because a sample space is necessarily finite but you need the sample space to be infinite to establish randomness. The only thing you can establish is whether time series data contains information about where it's going, i.e., contains information about what value the series will have at a future time. Intuitively you can see that it's absurd to think there is. Not according to all these guys who think price contains information about future price. Endless testing by the academics has proven the validity of the intuitive. If they were random, you wouldn't see the long runs of executions all at either the bid or the ask. If an MM is working a large order they will take all at that level. When they are doing that the market mechanism is fixed by intervention. When they have finished working the order, then price reverts to the random noise of orders entering the market. It is easy to see this at work if you watch Island book after hours on a stock that is trading heavily. There you have no MM, just orders entering and filling with price moving every which way. If there is a large order that doesn't have a large off setting order on the other side, price will stand still while sufficient smaller orders are ticked off to fill that large order..... unless someone jumps in front of that big order with a better price. For the longer term, the best proof that prices are not random is to look at the tails of the distributions of price changes... The "tails" of a distribution only reflects the intrinsic effect of trend. Are you claiming that trend leads to trend? This is equivalent to price predicts price, but what happens when trend changes? And when do you know it changes? What is an effective criterion for deciding the trend has changed? This has baffled everyone in the business for a century. The solution is easy. There is no effective decision procedure to determine when a trend has changed. Lots of people use standard deviations and all kinds of related calculations, but the problem always boils down to predicting that next coin toss. Specifically, how far through a trend line must a stock go before you take action?