A little experiment...
Moderator: moderators
-
pineapples
- rank: <50 posts
- Posts: 45
- Joined: Fri May 04, 2007 11:49 am
- Reputation: 0
- Location: Belgium
- Gender:
Please add www.kreslik.com to your ad blocker white list.
Thank you for your support.
Thank you for your support.
Yeah, but I believe that there are four types of markets (and many markets sometimes change their "type" for some period of time):
1) Trend market. It is more probable that any longer move of price will not stop, at least for some time. Trend following strategies will earn there. Portfolios made of many markets sometimes are like that. Stocks sometimes are like that. Extreme example would be market that always move in one direction on long time scope (day chart or H4 chart etc.).
2) Noise market. It is more probable that any bigger/longer move of price will break and change direction than it will follow direction of trend. Counter-trend strategies will win here. Extreme example would be the market with price that fluctuates arround one level (looks like noise). Arbitrage trading is often a trying to make "synthetic" market that looks like noise market - by taking advantage of relationships between markets (some of them partially price same products).
3) Quasi-random-walk market. There is nearly the same or the same probability that trend will follow or that will end. Both strategies - counter-trend and trend-follow - will fail in longer period of time. They can be turned into martingale strategies with money management but after some finite time they will fail. Extreme example is marked made by random returns. But here some percent of traders will win - just by chance.
4) Markets that change between 1 (trend market) and 2 (noise market) in time on non-random basis. One can say that any random-walk market will change between 1 and 2, but it's change on random basis - it's not predictable. But hypothetically there can be markets that change between 1 and 2 and those changes can be predicted (and then strategy would be follow trend sometimes and counter-trend in other times - based on prediction). Extreme example would be a market that changes between 1 and 2 in regular cycles.
And I think that before you trade on some market (or some combination of markets) it's good to look what kind of market it is (if you choose between trend-follow or counter-trend strategies - because there are also strategies that are neither one nor another).
1) Trend market. It is more probable that any longer move of price will not stop, at least for some time. Trend following strategies will earn there. Portfolios made of many markets sometimes are like that. Stocks sometimes are like that. Extreme example would be market that always move in one direction on long time scope (day chart or H4 chart etc.).
2) Noise market. It is more probable that any bigger/longer move of price will break and change direction than it will follow direction of trend. Counter-trend strategies will win here. Extreme example would be the market with price that fluctuates arround one level (looks like noise). Arbitrage trading is often a trying to make "synthetic" market that looks like noise market - by taking advantage of relationships between markets (some of them partially price same products).
3) Quasi-random-walk market. There is nearly the same or the same probability that trend will follow or that will end. Both strategies - counter-trend and trend-follow - will fail in longer period of time. They can be turned into martingale strategies with money management but after some finite time they will fail. Extreme example is marked made by random returns. But here some percent of traders will win - just by chance.
4) Markets that change between 1 (trend market) and 2 (noise market) in time on non-random basis. One can say that any random-walk market will change between 1 and 2, but it's change on random basis - it's not predictable. But hypothetically there can be markets that change between 1 and 2 and those changes can be predicted (and then strategy would be follow trend sometimes and counter-trend in other times - based on prediction). Extreme example would be a market that changes between 1 and 2 in regular cycles.
And I think that before you trade on some market (or some combination of markets) it's good to look what kind of market it is (if you choose between trend-follow or counter-trend strategies - because there are also strategies that are neither one nor another).
-
BlackMage
- rank: <50 posts
- Posts: 7
- Joined: Sat Feb 23, 2008 1:22 pm
- Reputation: 0
- Location: Copenhagen
- Gender:
I would state it a bit differently. IMO there are three basic market states:
1) Unpredictable, untradeable
2) Predictable, untradeable
3) Predictable, tradeable
Also there is a fourth state which I currently believe in practice only applies for artificial markets (such as the examples charts in this thread):
4) Unpredictable, tradeable
The transitions between these states can in itself be predictable or unpredictable.
The notions of "trends" and "mean reversion" must at least be relative to some time scale. What is seen as a "trend" in one time scale could as easily be a "mean reversion" in another time scale (that is, if you can find a meaningful definition of a "trend" and a "mean reversion" at all). Also what is "noise" actually?
1) Unpredictable, untradeable
2) Predictable, untradeable
3) Predictable, tradeable
Also there is a fourth state which I currently believe in practice only applies for artificial markets (such as the examples charts in this thread):
4) Unpredictable, tradeable
The transitions between these states can in itself be predictable or unpredictable.
The notions of "trends" and "mean reversion" must at least be relative to some time scale. What is seen as a "trend" in one time scale could as easily be a "mean reversion" in another time scale (that is, if you can find a meaningful definition of a "trend" and a "mean reversion" at all). Also what is "noise" actually?
Financial Hacker
Please add www.kreslik.com to your ad blocker white list.
Thank you for your support.
Thank you for your support.